00:00:06 <ehird> ^add a .; I typed it but my client stripped it
00:00:36 <AnMaster> ehird, two chars stripped for me and fizzie, remember that
00:00:43 <AnMaster> and seriously fix that bug in your client
00:00:45 <ehird> Not me, it's my client
00:01:02 <oerjan> ais523: KeyMaker missed a spam page today: Talk:Main Page/index.php
00:01:16 <ehird> Hmm, the T43's battery life isn't all that good; 5 hours 23 minutes
00:01:39 <ehird> But no, not really; I'm looking for 6 hours, minimum.
00:01:47 <ehird> The X200 gets up to 9 hours.
00:01:48 <AnMaster> ehird, and since you won't use cd probably just get an ultrabay battery as well
00:01:57 <AnMaster> note that both can't charge at once
00:01:58 <ehird> Can you do that? The T43 is from 2005.
00:02:11 <ehird> It still has IBM on it.
00:02:27 <AnMaster> ehird, oh that guy I met at uni with the 9 cell battery had an ultrabay too, said he got close to 14 hours or something like that
00:02:40 <ehird> How much did he get without? I guess it was X-series.
00:02:53 <ehird> But the ultrabay battery is presumably heavier than the optical drive, no?
00:02:59 <AnMaster> ehird, didn't look like that X series pic you showed me
00:03:16 <ehird> That's a new X-series.
00:03:16 <AnMaster> rather more like mine, and as I said, it was IBM era
00:03:42 <ehird> Also, since you can choose different WiFi cards and stuff, I suppose I could replace it with a draft-n one.
00:03:49 <ehird> when I replace the HD with an SSD
00:03:50 <AnMaster> ehird, and I didn't ask about it without ultrabay
00:04:09 <AnMaster> ehird, not sure you can get at the wifi card easily
00:04:12 <ehird> Apparently, the T42 was lighter and faster than the T43... but at expense of battery
00:04:50 <AnMaster> I mean... how fast is your internet connection?
00:04:59 <ehird> All I know is that this WiFi is slow.
00:05:08 <ehird> 200kiB/s vs 800kiB/s.
00:05:13 <AnMaster> ehird, how many bars or whatever do you get there?
00:05:23 <ehird> It could just be a shitty antenna.
00:05:33 <ehird> I'm afraid to consider things older than T60, anyway; they're all single-core Pentium M affairs.
00:05:39 <ehird> Seems like it'd be upsettingly slow.
00:05:51 <AnMaster> ehird, oh and what speed is it? On AP I mean
00:06:20 <AnMaster> ehird, 11a, 11b or 11g, or even pre-n?
00:06:33 <ehird> It's, uh, the one before n.
00:06:38 <ehird> It's a shitty ISP-provided "Livebox".
00:06:45 <ehird> http://www.thinkwiki.org/wiki/Category:300 ;; want!
00:07:31 <ehird> Who cares? It probably runs DOS
00:07:34 <AnMaster> ehird, well then I bet the issue is either latency or crappy AP
00:07:48 <ehird> The AP is across the room.
00:07:51 <ehird> I doubt it is latency.
00:08:02 <AnMaster> ehird, latency is always larger than for ethernet IME
00:08:22 <ehird> Well 200kiB/s just isn't acceptable; I can get 800kiB/s, dammit, and I want to use a good portion of that.
00:08:22 <AnMaster> ehird, especially if AP doesn't support short preamble
00:08:36 <AnMaster> ehird, I bet you will need to replace your AP then
00:08:54 <ehird> I have one of those Linuxable blue/green Linksys routers with the sticky-up antennae.
00:09:12 <ehird> And also a similarly-styled Linksys ADSL modem; just a question of linking those together sometime.
00:09:22 <AnMaster> ehird, because pre-N isn't going to be any better than g at 200 vs 800
00:09:32 <ehird> Surely it speeds up the non-internet stuff, though?
00:09:35 <ehird> Like, the overhead.
00:09:35 <AnMaster> since 11g should manage 800 kbps without issue
00:09:50 <AnMaster> ehird, well it speeds up LAN yes
00:10:07 <ehird> I guess I could get very large antennae for the Linksys AP.
00:10:14 <ehird> I'm sure you can buy gigantic ones.
00:10:27 <fizzie> Oh, I get ads for gigantic ones every day in my inbox.
00:10:46 -!- coppro has joined.
00:10:51 <ehird> Why aren't there houses with antennae in the walls, I wonder?
00:10:59 <AnMaster> ehird, anyway latency will always be higher with wlan than switched ethernet. Always
00:11:04 <fizzie> But really, with a 802.11g AP in the same room you indeed should get reasonable speeds; it goes theoretically up to 54 Mbps == 6.75 MB/s.
00:11:15 <AnMaster> probably also true for routered ethernet
00:13:08 -!- oklopol has quit (No route to host).
00:13:45 <ehird> Is switched ethernet, like, direct computer-to-computer?
00:15:07 <ais523> no, switching's like networking without any intelligence in the routers
00:15:15 <ais523> so instead, the computers at each end have to work out the routes for themselves
00:15:36 <ais523> it works well for things like relatively small company networks, though
00:16:00 <fizzie> There's quite a bit of intelligence in a modern switch, though; for things like sensible multicast forwarding only to listening ports and such.
00:16:03 <ehird> seems like it'd be less of a pain to just buy an AP
00:16:30 <fizzie> The student housing network (which I assume had few thousand hosts) used to be a single gigantic Ethernet segment, and it worked reasonably well.
00:16:41 <fizzie> (Except when people misconfigured things.)
00:16:45 <AnMaster> ehird, *shrug* anyway latency will always be less with ethernet. A lot less if AP doesn't support short preambles
00:16:45 <ehird> that's a ton of wasted CPU time :P
00:16:56 <ehird> AnMaster: *a lot more
00:17:09 <ais523> as part of my final-year project, I had to design a hardware implementation of a networking switch
00:17:29 <ais523> (nothing quite as complex as ethernet, though, just simple packets made up of address and data)
00:17:40 <AnMaster> ehird, ethernet will have a LOT less latency if AP doesn't support short preamble
00:17:43 <fizzie> Nowadays it's about "one network segment per 50 apartments" and DHCP.
00:17:49 <ais523> and I managed to make it work for arbitrary sizes despite VHDL's amazingly annoying lack of recursion
00:18:18 <ehird> Silicon recursion: proving space is continuous since 2010
00:18:29 <ais523> ehird: it bottomed out, and it's unrolled by the compiler
00:18:31 <ehird> The circuit keeps getting smaller!
00:18:34 <ehird> ais523: Shush you :P
00:18:48 <ais523> but still, it's quite difficult to design the code so that the code doesn't get bigger when the number of inputs does
00:19:23 <ais523> it was ridiculous, I had one array which held all the temp vars, and lots of polynomials to index it
00:19:31 <fizzie> Some of those "broadband routers" have been made with the assumption that Internet access is slow, which means that if you plug a direct 100M/100M Internet connection in the "WAN" port, and then connect a computer with 100M Ethernet directly to the "router" (I hesitate to call them that), it's unable to transfer full-speed, since it's built on some really crappy general-purpose-CPU embedded platform and runs some equally silly software on it.
00:19:34 <ais523> (it reminded me a lot of esoprogramming, in fact)
00:20:01 <fizzie> This was a often-noticed thing in the student networking newsgroup.
00:20:17 <AnMaster> fizzie, well for me the wan port looks like a special modem port
00:20:20 <ais523> fizzie: heh, I see the reasoning there
00:20:29 <ehird> The only real routers are pentium 3 boxes.
00:21:05 <fizzie> Yes, well, for routers that include an ADSL modem, it sort-of makes sense to make some assumptions about the maximum speed.
00:21:06 <ehird> A bunch of people use old computers as routers because, you know, who cares about noise, space or power consumption, let's be cool and use this as a router.
00:22:10 <ehird> Incidentally, it seems that Amazon has old ThinkPads, although unfortunately at low CPU speeds.
00:22:14 <fizzie> But there are WLAN dealies with "single Ethernet interface for an external internet, like a separate ADSL box, and then a four-port switch for the LAN" and still design the hardware so that it can't route and NAT packets from LAN to WAN at full speed.
00:22:38 <fizzie> Oh, and a wlan ap in all that; that's why I said "WLAN dealies".
00:22:54 <ehird> But seriously, I won't survive on a single-core Pentium M, will I?
00:23:44 <ais523> ehird: I probably would, but somehow I suspect you wouldn't
00:25:07 <ehird> I'm basically comfortable with this early-model 2.16ghz Core 2 Duo and 2.5GB of RAM, but I'll be putting an SSD in the notebook, so that gives it quite a performance advantage. Still, single-core is kind of bad.
00:25:08 <AnMaster> ehird, how did you manage 5 years ago?
00:25:28 <ehird> Five years ago I was only starting to program shitty PHP craps, and my system had so much crap on it I don't even know.
00:25:35 <ehird> And it crashed So. Much.
00:25:57 <ehird> It was that frankenstein a-bit-here-a-bit-there cobbled PC that existed in so many various configurations since 1998.
00:26:00 <AnMaster> well XP is quite stable for being windows
00:26:17 <ehird> The only way I can describe that machine is a constant war between crappy hardware.
00:26:20 <ehird> And man, it was so loud.
00:26:28 <ehird> I slept a meter away from it, and it took me ages to adjust.
00:26:30 <ehird> WHIIIIIIIIIIIIIIIIIIIR
00:26:34 <fizzie> And speaking of old computers as a router, I admit I'm one of those people.
00:26:35 <ehird> Such a loud, deep noise.
00:26:41 <AnMaster> ehird, you didn't shut it down at night?
00:26:54 <ehird> No, because starting it up took about three to five minutes.
00:26:58 <fizzie> (Except that I did put the Atom box in to work as the router to eat a bit less electricity.)
00:27:12 <ehird> AnMaster: Whoa! Too advanced to work.
00:27:24 <ehird> No, that did not work.
00:27:28 <ehird> AnMaster: The case, optical drives and disks survive today
00:27:42 <ehird> But everything else was replaced recently. It is serving as my mother's computer
00:27:58 <ehird> It's now an AMD Athlon X2 thing. Still quite noisy.
00:28:09 <ehird> AnMaster: The 80GB is still an installation of Windows since at least 2006.
00:28:18 <ehird> The 500GB, which is the one that is used, is Ubuntu 9.04.
00:28:41 <ehird> She used the 80GB until late last year, when I installed Ubuntu, and it's been about 10x less sluggish ever since.
00:29:08 <ehird> I think that particular one probably dated back to 2005.
00:29:39 <ehird> But yeah, I haven't really "coped" with any computer before this one.
00:30:24 <AnMaster> ehird, sure does. My desktop is noisy. Couldn't sleep in same room
00:30:30 <ehird> Oh, you mean the bed
00:30:45 <ehird> This computer makes a barely audible hum, but I'm sensitive enough that I still put it to sleep.
00:31:01 <ehird> Well, it clicks a bit when the disk is working.
00:31:06 <AnMaster> ehird, I have above normal hearing too so... yeah
00:31:13 <ehird> But it's something like 23dBA.
00:31:18 <ehird> measured by silentpcreview about that
00:31:24 <AnMaster> zalman tnn... would be so cool
00:31:33 <ehird> And around 20dBA thinks become completely inaudible
00:31:39 <ehird> AnMaster: Actually, it can handle a quad-core CPU
00:31:47 <ehird> just not a good graphics card with it
00:32:01 <ehird> So use a dual-core CPU
00:32:12 <ehird> Also, the current TNN sold is an (iirc, less beefy) Mini-ATX variant; you'd have to get it from a retailer that has some stockpiled.
00:32:24 <ehird> AnMaster: You wouldn't be able to get _high_ end.
00:33:12 <AnMaster> ehird, actually, I'd just get http://helmer3.sfe.se/ and place it on another floor. Wait... Make that "another building" even
00:33:21 <ehird> I'd say the max you could get would be 9600 GT + 2.6GHz dual-core
00:33:36 <ehird> Which isn't a powerhouse by any measure, but the SSD should make it very snappy.
00:33:41 <ehird> And also, you know, more silent.
00:34:52 <ehird> AnMaster: But, er, "The completely fanless case, power supply and heatpipe-based cooling system was massive, heavy, and very expensive, well over US$1,000."
00:35:14 <ehird> Introduced 2003; that sentence is from a "background" section of an early 2006 review of a pre-assembled computer based on it.
00:35:15 <AnMaster> ehird, yeah I was thinking about the heat
00:35:24 <ehird> AnMaster: Think about the damn price!
00:35:44 <AnMaster> and you know radiation. It would be like having a dual screen laptop right on your lap when standing a few meters away!
00:35:49 <ehird> TNN-500AF + X25-M is already nearing $2,000
00:36:19 <ehird> Add the rest of the computer components, totalling say around $700
00:36:31 <ehird> and that makes $2,700 for a mid-end system
00:36:36 <ehird> well, better since it's SSD, but
00:37:07 <ehird> AnMaster: Also, er, http://www.silentpcreview.com/files/images/epcn-tnn500/epcn1.jpg
00:37:22 <ehird> Yes, those wheels are needed; it's something like 30kg.
00:37:47 <ehird> I want a TNN notebook!
00:37:59 <ehird> Bonus: Sell it as a cheap method of birth control
00:38:13 <ehird> Well, okay. Not cheap.
00:42:19 -!- Pthing has joined.
00:53:37 -!- calamari has quit (Read error: 54 (Connection reset by peer)).
01:10:08 -!- calamari has joined.
01:17:31 <ehird> http://porg.es/blog/wp-content/uploads/2008/11/unicodedata-520d2txt.mp3 Unicode makes passable chiptune noise music (warning: loud).
01:17:55 <ais523> is that mp3 that's also valid unicode?
01:18:09 <ehird> It's `lame -r -m m -s 16 --bitwidth 8 ~/Downloads/UnicodeData-5.2.0d2.txt` where that file is http://unicode.org/Public/5.2.0/ucd/UnicodeData-5.2.0d12.txt.
01:18:12 <ehird> From http://porg.es/blog/what-unicode-sounds-like.
01:18:16 <ehird> It actually sounds good.
01:18:25 <ehird> I legitimately like it, musically.
01:21:42 <ais523> hmm... I just had the ridiculous idea of writing a brainfuck misinterpreter
01:21:52 <ais523> as in, it reads in a BF program, then runs it incorrectly
01:22:02 <ehird> that is indeed ridiculous. forget it quickly
01:22:33 <ais523> hmm... although I still think Befunge2K is an interesting idea
01:22:45 <ais523> (befunge, all commands have a constant chance of being no-ops rather than doing what they should do)
01:22:54 <ais523> I wonder if it's possible to write a program that's 100% guaranteed to work?
01:23:12 <ais523> (also: should # and ; be affected? they aren't really commands...)
01:24:29 <ehird> every command could fail
01:25:44 <ais523> ehird: you could set it up so that it's in an infinite loop until the commands start working, though
01:25:58 <ais523> for instance, if going vertically, _ has a different failure behaviour to success behaviour, entirely
01:26:14 <ais523> and say you arranged it so that the TOS was 0 unless the previous command had run correctly...
01:26:16 <ehird> ais523: probability, though
01:26:20 <ehird> you could get fails, all the time
01:26:23 <ehird> even though it has p=0
01:26:31 <ais523> ehird: that's still 100% guaranteed to work
01:26:58 <ais523> if something has p=1 of working, it's 100% likely to work
01:26:58 <ehird> well, sure, it is certain to happen, it's just that it's perfectly possible not to
01:29:13 <ehird> http://www.thinkwiki.org/wiki/PC110
01:32:58 <ehird> lol, I like how people are claiming that IBM wanted the PC to be cloneable
01:33:13 <ehird> being that that's, you know, the opposite of reality
01:33:20 <ehird> they desperately wanted a monopoly
01:40:29 <Sgeo> Why was the IBM PC so clonable, as opposed to other machines?
01:40:46 <ehird> apart from being quite off-the-shelf
01:40:59 <ehird> but heck, the IBM PC *sucked*
01:41:36 <ehird> I mean, the Apple Lisa came out two years later and was massively better, albeit more expensive
01:41:42 <ehird> Next year the Macintosh, which was like half the price
01:41:51 <ehird> Whereas the PC was, uhh, sucking.
01:58:57 <ehird> HOW MANY PEOPLE ARE.
01:58:58 -!- calamari has quit (Read error: 104 (Connection reset by peer)).
01:59:22 <oerjan> `wolfram How many people are
01:59:28 <ehird> I think it's a good question,
01:59:48 <HackEgo> ? \ addquote \ calc \ commands \ creatures \ define \ esolang \ etymology \ fortune \ google \ helpme \ imdb \ minifind \ paste \ quote \ runfor \ sayhi \ strfile \ toutf8 \ translate \ translatefromto \ translateto \ unstr \ url \ wolfram
01:59:55 <HackEgo> How many people are \ \ Input interpretation: \ \ all countries \ Summary: \ \ population \ \ total highest lowest \ \ 6.68 billion people 1.31 billion people China 48 people Pitcairn Islands \ 2006, 2007, 2008 and 2009 estimates \ \ Generated by Wolfram|Alpha (www.wolframalpha.com) on September 5, 2009 from Champaign,
02:00:01 <HackEgo> fuck \ \ Input interpretation: \ \ fuck English word \ Definitions: \ \ fuck \ \ slang for sexual intercourse \ \ Word origins: \ \ Old English Middle English Classical Latin first recorded use: 1916 93 years ago \ Pronunciation: \ \ f \ \ k \ \ f' uhk \ \ Synonyms: \ \ fuck \ \ nookie shtup \ \ nooky \ \ piece of
02:00:08 <ehird> 6.68 billion people are.
02:00:17 <ehird> ...also, nookie shtup.
02:02:33 <Sgeo> `wolfram Scientology
02:02:40 <HackEgo> Scientology \ \ Input interpretation: \ \ Scientology English word \ Definition: \ \ Scientology \ \ a new religion founded by L. Ron Hubbard in 1955 and characterized by a belief in the power of a person' s spirit to clear itself of past painful experiences through self knowledge and spiritual fulfillment \ \ Word origins:
02:02:45 <Sgeo> `Wolfram I love you
02:03:02 <ehird> `wolfram sexual intercourse
02:03:10 <HackEgo> sexual intercourse \ \ Input interpretation: \ \ sexual intercourse English phrase \ Definition: \ \ sexual intercourse \ \ the act of sexual procreation between a man and a woman; the man' s penis is inserted into the woman' s vagina and excited until orgasm and ejaculation occur \ \ Synonyms: \ \ sexual intercourse
02:03:21 <ehird> So THAT'S why you have to be 18!
02:03:23 <ais523> `wolfram infinity squared
02:03:24 <ehird> It's practically erotica.
02:03:28 <HackEgo> infinity squared \ \ Input: \ 2 \ \ Result: \ \ Generated by Wolfram|Alpha (www.wolframalpha.com) on September 5, 2009 from Champaign, IL. © Wolfram Alpha LLC—A Wolfram Research Company \ \ 1 \ \
02:03:52 <ehird> fun fact: HackEgo's Wolfram Alpha interface is the only thing in the universe more buggy than Wolfram Alpha itself
02:03:57 <ehird> fun fact: HackEgo's Wolfram Alpha interface is the only thing in the universe more buggy than Wolfram Alpha
02:03:59 <ehird> (reads better that way)
02:04:06 <ehird> http://www.wolframalpha.com/input/?i=infinity%5E2
02:04:10 <ehird> Result = sideways 8
02:04:29 <Sgeo> Does HackEgo screen scrape, or does W|A provide an API?
02:04:38 <ehird> It screen scrapes the PDF.
02:04:38 <ais523> you have to pay for the API
02:05:05 <ehird> It'd be better to scrape the Mathematica notebook, but
02:05:48 <ehird> ais523: does Mathematica have a button to turn a notebook into code?
02:06:06 <ais523> ehird: the code's embedded in there somehow
02:06:23 <ehird> it contains base-64, presumably images
02:06:26 <ehird> and a shitload of boxes
02:08:19 <Sgeo> Who came up with the term "screen scraping"? It's not like a screen scraping tool is taking a screenshot and interpreting it
02:16:34 <Sgeo> HackEgo violates W|A's terms of use
02:16:57 <ehird> Meanwhile, nobody gives a shit
02:16:59 -!- calamari has joined.
02:17:01 <ehird> I violate it by being under 18.
02:18:06 <oerjan> i violate it just out of solidarity
02:29:05 <ehird> Advantage of ball mice: move more smoothly. Gah this mouse+pad.
02:29:24 <ais523> I thought ball mice got less smooth, when they were dirty
02:29:25 -!- calamari has quit (Read error: 131 (Connection reset by peer)).
02:39:12 <ehird> ais523: oh, no doubt
02:39:16 <ehird> but that beats scratchiness
02:39:29 <ehird> don't say felt pads; that's just like mud for a laser mouse
02:39:39 <ehird> actually i'm not sure why we obsoleted ball mice
02:39:43 <ehird> well, i guess cleaning
02:39:47 <ehird> when they were clean they worked perfectly
02:40:05 <ehird> I wonder if there are any bluetooth ball mice...
02:41:19 <ehird> Every time I look at a trackball I cringe
02:41:27 <ehird> It just reminds me of popping your thumb out of its socket...
02:43:40 <ehird> *optical mouse; I'm not using a laser mouse atm
02:47:49 -!- calamari has joined.
02:48:12 -!- Gracenotes has joined.
02:49:34 <oerjan> next: the aureal mouse
02:52:14 <ehird> Hmm, apparently the T42 had good battery life
02:52:30 <ehird> I seem to be travelling further and further back in the ThinkPad line, and am scared at the performance it will lead me to...
03:10:55 <ehird> "Why Psychologists Are Infinitely More Dangerous Than Conspiracy Theorists" from infowars.com.
03:11:03 <ehird> So much textbook crazy concentrated in one place.
03:11:19 <ehird> They want to destrooooooy your mind! Obamacare is eugenics NWO 9/11!
03:16:57 * Sgeo wishes Twitter was decentralized. I think I've said that before, and someone mentioned identi.ca
03:22:00 <ehird> identi.ca is not decentralised.
03:22:04 <ehird> also, nobody gives a shit.
03:22:15 <ehird> if twitter went down, nothing much would be lost; it is inherently transient
03:22:21 <ehird> and seriously, distributed this, distributed that
03:22:24 <ehird> it's mostly pointless.
03:24:14 <oerjan> distribution using pointless topology
03:37:38 <coppro> anyone know a way to scan for a given process being started?
03:37:51 <coppro> preferably without ptracing the invoker
03:38:05 <ehird> while sleep 1; do pgrep
03:38:09 <ais523> coppro: if you don't need it accurately, you could poll
03:38:28 <coppro> I need to figure out a command line being invoked
03:38:35 <ehird> grep can print at ree.
03:41:13 <coppro> augh, pgrep is annoying, but it should work
03:41:22 <coppro> I hate piping in to read
03:45:48 -!- coppro has quit ("The only thing I know is that I know nothing").
03:46:56 -!- coppro has joined.
04:02:27 -!- ehird has quit.
04:18:35 -!- Gracenotes has left (?).
04:25:12 -!- coppro has quit (Read error: 110 (Connection timed out)).
04:40:08 * Sgeo realizes that he can malloc from Python, kind of
04:40:35 <ais523> Sgeo: why would you need to?
04:40:44 <Sgeo> ais523, for fun? Being silly?
04:40:58 <ais523> I'm not sure what you'd mean by that, though
04:42:46 <Sgeo> WTF? ??2@YAPAXI@Z
04:42:48 <Sgeo> WTF is ??2@YAPAXI@Z
04:55:08 -!- coppro has joined.
05:44:57 <ais523> wow, I just realised how terribly awful http://windows7sins.org/#1 is from a graphics design point of view
05:45:08 <ais523> and knowing the FSF, the content isn't much better either
05:46:49 <ais523> coppro: first I've seen the page
05:47:01 <ais523> I haven't actually read much of the content, I'm just guessing it's bad
05:47:07 <ais523> the graphic design is obvious, though
05:47:17 <coppro> I thought you meant you'd been there several times and just realized
05:47:22 <ais523> (the real reason open source isn't really taking off is that the FSF is so awful at marketing)
06:32:05 <coppro> w00t clean Ubuntu install is nice :)
06:32:22 <coppro> I should seriously consider doing this every 6 months, rather than upgrading
06:58:22 -!- Sgeo has quit (Read error: 60 (Operation timed out)).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:29:03 -!- KingOfKarlsruhe has joined.
08:31:08 -!- pikhq_ has joined.
08:42:40 -!- oerjan has quit ("leaving").
08:43:02 -!- pikhq has quit (Read error: 110 (Connection timed out)).
08:53:11 <ais523> the shops will be opening so I can buy breakfast
08:53:32 <Deewiant> I managed to do http://ars.userfriendly.org/cartoons/?id=20090906 with only two Google searches; should I be worried or something?
08:54:02 <AnMaster> (so that you remember uf had such a crossword)
08:54:27 <AnMaster> I thought you meant "googled to find the crossword"
08:54:39 <Deewiant> No, I meant I managed to fill it in :-P
08:54:58 <ais523> I like the way they avoided 4e references
08:55:16 <Deewiant> There's even an error in it; it says "twenty-sided solid" where it means "twelve-sided solid"
08:55:38 <AnMaster> anyway I couldn't manage it, since I never played D&D
08:55:41 <ais523> AnMaster: the crossword stops at third edition, which is probably a wise choice
08:55:57 <ais523> because fourth edition is a more or less completely different game, just with the same name for marketing purposes
08:56:02 <ais523> and a vaguely similar theme
08:56:54 -!- coppro has quit (Read error: 54 (Connection reset by peer)).
08:57:29 -!- calamari has quit ("Leaving").
08:57:30 <AnMaster> <ais523> wow, I just realised how terribly awful http://windows7sins.org/#1 is from a graphics design point of view <-- um yeah that the frame scrolls sideways makes it look terrible for a start
08:57:45 -!- coppro has joined.
08:57:49 <AnMaster> that it then doesn't use said frame all the way...
08:58:07 <ais523> I don't even get why it does the start of a page in a frame, then leaves it for the rest of the page
08:58:52 <AnMaster> ais523, any idea how to connect to internet using your phone in ubuntu?
08:59:06 <AnMaster> I mean, I have 3G internet thingy on my phone
08:59:16 <ais523> I don't see why people would assume i do
08:59:23 <ais523> my family own a landline, I can use taht
08:59:27 <ais523> and I use payphones sometimes too
08:59:45 <AnMaster> ais523, I mean, I would have assumed at least an old GSM one with black and white screen or so
08:59:56 <AnMaster> (or black and grey rather, you know, like old calculators)
09:00:43 <AnMaster> (or possibly modern cheap calculators too, but I haven't used a non-graph-capable one for years, so no clue)
09:00:53 <ais523> modern cheap calculators still do that, yes
09:01:01 <ais523> and exams is a pretty good reason to use them
09:01:12 <ais523> often calculators that do much more than arithmetic are banned
09:01:57 <AnMaster> ais523, for this course graphing ones are allowed, but ones that can do CAS aren't
09:02:10 <AnMaster> like simplifying expressions and such
09:02:20 <AnMaster> mine is just below what is allowed.
09:03:41 <AnMaster> ais523, also I think mine doesn't follow the physical laws... I haven't replaced the batteries for years, and yes I used it a lot.
09:03:52 <AnMaster> four of those small AAAA or whatever they are called
09:06:04 <AnMaster> considering that I replaced batteries in one of those cheap calculators a lot more often...
09:07:02 <AnMaster> ais523, Shouldn't a graphing calculator use more power than a cheap one in general?
09:09:20 <fizzie> Is this "using your phone over bluetooth" or what?
09:09:54 <AnMaster> fizzie, um that was before, but last was "calculator"
09:10:15 <AnMaster> fizzie, but about phone for connecting to internet over bluetooth, if you know anything about it, please tell
09:11:30 <fizzie> I've done it "manually" in Debian once, but it was quite a kludge; using that USB 3G modem in Ubuntu was impressively easy, though. My guess for phone-and-bluetooth would be the Bluetooth Device Wizard, maybe.
09:11:34 <AnMaster> fizzie, I mean, opera mini works fine on the phone, yet I can't see to share it with the laptop over bluethooth.
09:11:49 <fizzie> I should perhaps actually try this, since I have the laptop and Ubuntu open and all.
09:12:28 <ais523> AnMaster: would depend on how much calculation it was doing, I suppose
09:12:35 <AnMaster> fizzie, hm... same wizard as where you pair the devices up? Because I don't remember seeing any option for that there. I can mount the phone and browse it's file system and such
09:12:49 <ais523> actually, maybe it wouldn't; it's more likely to have a more expensive processor, which therefore probably has better power-save features
09:12:55 <AnMaster> ais523, well the graphing one has been used a LOT more.
09:13:23 <ais523> the graphing one may also have batteries which hold more charge
09:13:37 <ais523> (and ofc if it's solar-powered, the batteries may never run out)
09:14:07 <AnMaster> ais523, well 4 AAAA vs 2 AAA? Not sure
09:14:26 <ais523> most AAAs hardly hold any charge, they're mass-produced for consumers
09:14:36 <AnMaster> ais523, but AAAAs are even smaller?
09:14:37 <ais523> AAAA, I don't know; but likely they'd be higher-quality due to being harder to get hold of
09:14:49 <ais523> small doesn't necessarily equal low charge, it depends on what they're made of
09:15:00 <AnMaster> ais523, both are Duracell thingies. So not exactly "cheap no-name"
09:16:06 <fizzie> Curious, there's no immediately-obvious thing. What with Ubuntu's "just works" motto, I'd have assumed they have a user-friendly no-config-file-editing-involved solution for this.
09:16:19 <ais523> it just works for Palms
09:16:35 <ais523> but I imagine phone-tethering would be hard to develop, as most networks ban it
09:16:54 <AnMaster> anyway I got as far as using rfcomm to bind the DUN interface for the phone, but there I got stuck....
09:16:55 <fizzie> Nobody bans it around these parts. And what do you mean "expensive"?
09:17:02 <ais523> call charges while testing
09:17:16 <fizzie> Er, it's a fixed 10 eur/month for unlimited data.
09:17:31 <ais523> wow, your mobile operators are much less rip-offy than the ones over here
09:17:32 <AnMaster> I did manage to use minicom to seriously mess up the phone with AT commands (had to power cycle it... heh), but not to connect to internet
09:17:51 <ais523> here, most operators are pre-pay, but the rates go up astronomically high if you pay less than £15 a month
09:18:09 <fizzie> Come to think of it, I don't really know how to set up a traditional dial-up networking connection over Ubuntu either.
09:18:21 <AnMaster> iirc it is per data here, but max 9 kr per 24 hours
09:18:38 <ais523> how much is a kr in euros or pounds?
09:19:12 <HackEgo> 9 Swedish kronor = 0.878224893 Euros
09:19:32 <ais523> so pretty much the same as our £15 a month effective rate
09:19:56 <ais523> and that wouldn't even be unlimited
09:19:58 <AnMaster> ais523, yeah, I have one of those SIM cards that you load with money, rather than pay per month thingy
09:20:08 <AnMaster> not sure what the English term for it is
09:20:25 <ais523> although "pay as you go" is the usual advertising phrase
09:21:33 <HackEgo> 9 SEK in EUR \ \ Input interpretation: \ \ convert kr 9 Swedish kronor to euros \ Results: \ \ € 0.88 euros \ \ at current quoted rate \ \ Minimal currency form: \ 50c 20c 10c 5c 2c \ \ 0.72 c \ \ Local currency conversion: \ \ $ 1.25 US dollars \ \ at current quoted rate \ \ Generated by Wolfram|Alpha (www.wolframalpha.com)
09:21:43 <AnMaster> ais523, also I meant that above 9 SEK you aren't charged any more for the surfing
09:25:20 <fizzie> I don't think any Finnish operators offer that sort of deal, which is a bit of a shame, since I only use the mobile interwebs quite rarely, but a lot when it happens.
09:26:50 <fizzie> Hm, Network Manager's "add mobile broadband connection" wizard (which was used for the 3G modem stuff) says the connection can be made "using a 3G modem or a phone", but it looks a bit like no-one's bothered to add over-Bluetooth support to it yet.
09:27:45 <AnMaster> fizzie, indeed my conclusion too
09:28:39 <fizzie> There's that "Blueman" GTK bluetooth manager app, which according to some sources should make it easy to have network manager see the phone.
09:29:12 <AnMaster> fizzie, and with the rfcomm stuff, it seems I'm supposed to know stuff about establishing the connection I have no clue about, I mean... it just works on the phone itself, why shouldn't it just be able to share that connection to the computer...
09:30:32 <AnMaster> like: <computer> please establish data 3G connection, you take care of the details <phone> ok, please stand by... (a few seconds pass) <phone> done, lets talk with some simple data stream protocol now <computer> great!
09:30:35 <fizzie> Oh, that won't ever work. I mean, unless your phone specifically has support for Internet-sharing thing. (Some PDA's at least support that sort of "connection setup on the phone, then Internet sharing over the PAN bluetooth protocol thing")
09:30:38 -!- M0ny has joined.
09:31:06 <AnMaster> fizzie, yeah why does it have to act as a clueless modem when it *can* do this sort of thing itself already
09:31:24 <fizzie> Anyway, that network-manager's "mobile broadband" wizard just has you select your provider from the list, with no messing-around with settings.
09:31:45 <AnMaster> fizzie, yep, there doesn't even seem to be anywhere to select what device you want to use for it
09:32:13 <AnMaster> and yes I tried it, but quickly found it wouldn't work in any way whatosever
09:33:06 <AnMaster> fizzie, so what was the complex way under debian that you did?
09:33:31 <fizzie> Yes, network-manager expects to find the modem-to-use over HAL. That blueman app is supposed to provide the necessary glue there.
09:33:50 <fizzie> It's not in the official Ubuntu repositories, though. (But there is a APT source for it.)
09:34:05 <AnMaster> also I wouldn't trust random repos.
09:34:08 <ais523> fizzie: is it in the contrib repos?
09:34:22 <AnMaster> ais523, not in universe or multiverse at least.
09:35:07 <fizzie> Nnno. It's at https://launchpad.net/blueman, won't comment on trustworthiness.
09:36:11 <ais523> hmm... it seems that bit.ly have got themselves an even shorter URL
09:36:47 <ais523> it's basically a mirror
09:37:40 <fizzie> Libya, and Northern Mariana islands.
09:40:01 <fizzie> chi.mp seems to be giving out "free" (haven't read the details) .mp domains.
09:41:13 <ais523> oh, quite a few TLDs have been taken over commercially, now
09:41:22 <ais523> which IMO is making a mockery of the Internet
09:41:54 <fizzie> Well, you know how the saying goes. You can't make a mockery without breaking a few... mocks.
09:42:04 <AnMaster> nice seems gcc has an __attribute__ that allows you to control -O* and -f* on a per-function basis
09:44:34 -!- BeholdMyGlory has joined.
09:44:50 <fizzie> It's funny how they speak constantly about you owning your own identity; then when you look at the small print term-definitions: "Owner the term used by Saipan DataCom to describe you as the subscriber to the CHI.MP Service and owner of 'Customer Content.'"
09:48:22 <fizzie> Most places (at least .fi) frown on single-character domains, but I guess not everyone.
09:51:23 <fizzie> There's a relatively recent network-manager blog-post which shows how easy the internet-over-bluetooth-over-phone is, assuming you have a bleeding-edge network-manager: http://blogs.gnome.org/dcbw/category/networkmanager/
09:51:31 <fizzie> So I guess someone's actually working on it.
09:52:41 <fizzie> And of course also: "DUN support needs a bit more work (its more complicated but also more capable); but tons of phones have PAN so lets get it out there."
09:53:09 <AnMaster> fizzie, how do I check if mine has PAN?
09:53:48 <fizzie> Presumably it should have some sort of menu for Internet sharing. (Personally I haven't ran into any such phone yet.)
09:55:21 <fizzie> The PAN approach is that more sensible "phone knows actual connection details like AP names" one.
09:56:37 <AnMaster> fizzie, sdptool to list services yields one that seems slightly relevant: Service Name: Network Access Point Service
09:58:03 <fizzie> That does actually sound relevant.
09:58:44 <AnMaster> http://pastebin.ca/1555921 is the full list btw
09:59:38 <AnMaster> (that excludes the channel info and so on for brevity)
10:00:18 <AnMaster> fizzie, it doesn't seem to be the PAN one though
10:00:38 <AnMaster> because I found some website mentioning "Service Name: PAN Network Access Point" from that command.
10:00:52 <AnMaster> Service Description: Personal Ad-hoc Network Service which provides access to a network
10:01:28 <fizzie> Well, it does sound good. But that PAN support in the network manager was (at the time of the blog post) only in the git master. Probably will take some time for it to dribble down to Ubuntu.
10:02:45 <AnMaster> fizzie, yes and this is separate from PAN it seems
10:02:53 <AnMaster> plus there is no menu for this on the phone afair
10:05:05 <fizzie> My list of services is a lot shorter: http://pastebin.ca/1555928
10:05:10 <fizzie> Okay, so the phone is quite ancient too.
10:06:52 <AnMaster> fizzie, wait, what does "PAN" stand for?
10:07:19 <AnMaster> if so that is the description indeed of that. So it is PAN but isn't called PAN
10:07:39 <AnMaster> fizzie, also how would one go about using PAN without network manager?
10:08:01 <fizzie> "personal area network" is the usual expansion. And no clue, really. There was some tutorial about it though.
10:10:17 <fizzie> What do you have in the sdptool Service Class ID List field for that? I think those should be quite unambiguous, unlike the service name field which is a bit free-form.
10:10:34 <AnMaster> "Network Access Point" (0x1116)
10:12:11 <fizzie> It should be the PAN-AP indeed.
10:12:43 * AnMaster is looking at http://bluez.sourceforge.net/contrib/HOWTO-PAN and wondering if bluez is new enough in ubuntu 9.04
10:12:52 <fizzie> Symbian sources: "return 0x1116; // PAN NAP
10:12:56 <AnMaster> Programmet "pand" är för närvarande inte installerat. Du kan installera det genom att ange:
10:13:46 <fizzie> Yes, sounds obsoletey. (Though really, if you got a working rfcomm serial line and can speak to the phone, you should be able to set up a traditional dial-up networking connection over it; it's just some AT commands and PPP.)
10:14:06 <AnMaster> fizzie, I didn't manage to get the correct AT commands. Somehow
10:14:57 <AnMaster> okay it seems I have a pan0 interface now, though ifconfig says it is down
10:15:12 * AnMaster wonders what now, since the guide he looked at is so outdated it seems to refer to 2.4 kernels
10:15:56 <fizzie> If it's somehow "connected" to the network the phone is on already, you should be able to just bring it up and run dhclient or some-such on it.
10:16:14 <AnMaster> fizzie, it isn't connected as such
10:16:24 <AnMaster> well with the pan0 interface at least
10:16:33 <fizzie> Ah, well, I don't know how that stuff works. :p
10:16:59 <fizzie> I do have an OS X modem script for the GPRS dial-up networking for "Nokia phones" in general; it's got one set of AT commands that could work.
10:18:37 <fizzie> I'll try to extract the relevant stuff; there's a lot of modem-script-programmatic nonsense.
10:19:31 <fizzie> All kind of test-if-it's-a-Nokia-phone and control flow strangeness.
10:20:35 <AnMaster> fizzie, got a link to that PAN in network manager thing? it might help to look at the relevant commits
10:20:45 <AnMaster> and google seems unable to find it here
10:21:39 <fizzie> Anyway, "AT&FE0V1&D2&S0&C1S0=0+IFC=3,1" should be the magical init string for the modem, followed by >> AT+CGDCONT=1,"IP","[access point name here]" << for more initialization, and finally "ATD*99***1#" to dial.
10:22:03 <fizzie> (Had to use a different quotation notation for that middle string, since it has embedded quotes.)
10:22:17 <AnMaster> what would access point name be?
10:22:29 <fizzie> That one should be somewhere in the phone menus.
10:22:42 <fizzie> For me it's "saunalahti.internet".
10:22:50 <AnMaster> fizzie, this phone has swedish menus, it is a bit hard to guess what might be intended
10:23:41 <AnMaster> which seems to be "Telia GPRS" (though it is 3G actually -_-)
10:23:57 <fizzie> Hmm. Well, that might either be the real name or the human-readable name.
10:24:00 <AnMaster> actually iirc it is "3G if available, otherwise GPRS"
10:24:11 <fizzie> Is your provider listed in the network-manager mobile-broadband thing?
10:24:35 <AnMaster> fizzie, well yes, but I didn't manage to actually make that thing use my phone
10:24:36 <fizzie> If so, you can add a dummy "connection" there, then do "Edit" to it and look at the "APN" field it set up.
10:25:40 <fizzie> That sounds like a reasonable format. Providers seem to like to make it resemble a DNS name, even though it's not really related.
10:26:02 <fizzie> The "saunalahti.internet" is just [ISP name].[fictional .internet TLD].
10:26:04 <AnMaster> fizzie, seriously PAN sounds a lot saner
10:26:26 <fizzie> And these AT commands might not work; they do work with the N-Gage, though.
10:26:58 <AnMaster> fizzie, so where did you read that about PAN in network-manager, because if I can read the patch I can possibly use it with command line tools myself or such
10:27:44 <fizzie> http://blogs.gnome.org/dcbw/category/networkmanager/ was the blog-post about it.
10:28:36 <fizzie> It has a date, which might be useful maybe.
10:29:42 <fizzie> http://git.gnome.org/cgit/network-manager-applet/commit/?id=3bec17daa718354e22ca9b5116a91a4cf594eeee possibly.
10:30:47 <AnMaster> um that seems less useful than I had hoped
10:30:51 <fizzie> And, well, the files the code refers to; that actual patch looks pretty short and network-manager-infrastructury.
10:32:49 <AnMaster> "You’ll need gnome-bluetooth >= 2.27.7.1, bluez >= 4.42" <-- right, both are way outdated here
10:33:15 <fizzie> Still, I wouldn't want to have to grok the whole network-manager-applet mess just for this.
10:34:11 <AnMaster> fizzie, ok, about dialup thingy, what sort of place does one enter those commands in?
10:34:42 <fizzie> That sort of depends on whether you want to use Ubuntu's network-settings-thingies or some separate PPP dialer (like wvdial).
10:35:21 <fizzie> If you have a working rfcomm channel (with the device file present) it is possible that the device shows up in Ubuntu/Gnome's System > Administration > Network.
10:36:14 <AnMaster> but yes I have the rfcomm0 device file
10:36:15 <fizzie> Is there a point-to-point connection, though?
10:36:42 <AnMaster> fizzie, hm? where whould that be? can't see one at least
10:37:06 <fizzie> I have one in that "Network Settings" connections tab.
10:37:49 <fizzie> (Don't know where it comes from, really.)
10:38:04 <AnMaster> fizzie, wait, do you mean in the "network tool" thingy with ping, netstat and so on? or in the network manager thing?
10:38:35 <fizzie> Neither. There are three things; the network manager applet (from tray), the "network tool" thing, and a "Network Settings" dialog.
10:38:57 <fizzie> Typically pretty much everything in "Network Settings" is set to "roaming mode", in which case the network manager applet takes care of them.
10:39:00 <AnMaster> fizzie, can't find the final one
10:39:47 <fizzie> Well, the command name is "network-admin"; maybe they've cleaned it away from the menus because it was confusing. :p
10:40:06 <AnMaster> tells me to do: sudo apt-get install gnome-network-admin
10:40:44 <fizzie> Ah. Well, then... if you want to try that, you can. It's either that, some other graphical dialer thing, or just /etc/network/interfaces magic, but I don't quite know how that last one works.
10:41:11 <AnMaster> but I can't see if it is for the built in modem in computer thing or not
10:41:19 <AnMaster> also it reports pan0 as a wired network XD
10:41:50 <fizzie> Yes, same here too. (I have a useless pan0 too.)
10:42:06 <AnMaster> anyway I suspect it may be for the built in modem
10:42:28 <fizzie> Just Properties > Enable it, then see the Modem tab for "modem port".
10:42:33 <fizzie> Hopefully you'll have your rfcomm device there.
10:42:49 <AnMaster> modem port is empty, lets see...
10:43:05 <AnMaster> nop, lists /dev/modem and some ttyS[0-3]
10:43:11 <AnMaster> and this doesn't even have a serial port
10:43:41 <fizzie> Hrm. Well, you need a serial-port-like connection to the phone, and a /dev node which looks like one -- something you can write AT commands to with minicom.
10:43:44 <AnMaster> oh indeed /dev/modem doesn't exist even
10:44:00 <fizzie> Maybe it's a static list of suggestions, then -- same list here.
10:44:07 <AnMaster> fizzie, I could write AT commands with minicom to the rfcomm device, but that made the phone lock up
10:44:16 <AnMaster> possibly I entered the wrong ones
10:44:25 <fizzie> Well, if you want to try it, you can put that rfcomm device just manually there in modem port.
10:44:47 <fizzie> Also it's possible that you can do "Connection type" > "GPRS/UMTS" in the "General" tab and just enter the "Access point name" to the field which should appear.
10:44:58 <fizzie> Possibly it might then do a reasonable generic initialization/dialing thing.
10:45:22 <AnMaster> fizzie, I don't know what to enter for user/password there...
10:45:30 <fizzie> Just leave them empty, it works for me. :p
10:45:37 <fizzie> (Well, in OS X, at least.)
10:45:54 <fizzie> Then you probably want the "Options" tab default-route and use-ISP's-DNS settings too.
10:46:18 <fizzie> Oh, and I actually don't know how to make it *dial* the configured connection, heh.
10:46:43 <fizzie> Though I think that network-settings should configure it so that you can "ifup ppp0" manually. Maybe.
10:48:05 <fizzie> (Applying settings will probably reload the network-manager-applet which will temporarily drop your wlan, it seems.)
10:48:23 <AnMaster> no luck, it just made my computer keep cycling connection to ethernet and phone crash again
10:49:10 <fizzie> Did your phone reply to plain "AT" with "OK" in minicom?
10:50:58 <fizzie> Hrm, that's pretty strange.
10:51:30 <fizzie> Maybe you could try with those nokia-specific initialization/dial commands. I'll try to see if I can figure out where you can put that stuff in the network-settings place.
10:51:45 <fizzie> No, I mean, it should do that and then work properly. :p
10:52:25 <AnMaster> <fizzie> Anyway, "AT&FE0V1&D2&S0&C1S0=0+IFC=3,1" should be the magical init string for the modem, followed by >> AT+CGDCONT=1,"IP","[access point name here]" << for more initialization, and finally "ATD*99***1#" to dial.
10:52:36 <AnMaster> if I try that first one it seems minicom stops accepting any more input
10:52:57 <AnMaster> but then any key press does nothing, Ctrl-A Z opens the minicom menu thing still though
10:53:23 <AnMaster> oh wait, it just stopped echoing
10:53:37 <fizzie> That's the E0 possibly.
10:53:53 <fizzie> Well, you can indeed try those commands in minicom.
10:54:11 <fizzie> It should spew out some binary stuff for PPP connection initialization after the ATD... command, if it works.
10:54:13 <AnMaster> well phone locked up at "ATD*99***1#"
10:54:23 <fizzie> Rrright. That's not good.
10:54:26 <AnMaster> but it did spew some binary stuff
10:55:05 <fizzie> I don't quite know what's making it lock up all the time.
10:55:09 <AnMaster> fizzie, also on phone there is an icon for "data connection established", it is an E when connecting, and an E with a box around once the connection is established
10:55:16 <AnMaster> and it got to the E but never to the box
10:55:53 <AnMaster> I think around where I am atm there is only GPRS, no 3G, and the phone falls back on GPRS if it can't find 3G
10:56:39 <AnMaster> because that E thing is only for GPRS, while there is some kind of "3" symbol when there is full 3G around
10:56:58 <fizzie> It's "G" for the N-gage (always, since it's GPRS only).
10:57:04 <fizzie> Actually it could be E for EDGE?
10:57:35 <fizzie> EDGE's a slight speed upgrade to GPRS.
10:57:42 <fizzie> Pretty much the same technology, I guess.
10:57:42 <AnMaster> fizzie, it is a "nokia 3210 classic" btw
10:57:49 <AnMaster> fizzie, well possibly, I wouldn't know
10:58:26 <fizzie> I guess I could try to see what the N-gage says if I minicom those strings into it, to see if it works. I'll have to do the rfcomm setup first, though.
10:58:58 <AnMaster> like: rfcomm bind 0 <MAC-of-phone> <channel number>
10:59:02 <AnMaster> where channel number is the DUN channel
10:59:24 <AnMaster> and 0 is what will be at the end of the device node
10:59:34 <AnMaster> rfcomm unbind 0 <mac of phone>
10:59:49 <AnMaster> where MAC is the bluetooth one
11:00:05 <AnMaster> if phone is set to visible you can use hcitool scan to find it
11:00:29 <fizzie> Well, it did create rfcomm0 there. Will have to apt-get minicom first, though. :p
11:03:50 <fizzie> Eh, it picked up /dev/tty8 as the "serial port" by default, or something.
11:04:08 <AnMaster> and select to edit the relevant line
11:05:34 <fizzie> Hm, I pointed it at rfcomm0 and started; the phone did the "bluetooth active and talking" icon for a second, then it went away and said "minicom: cannot open /dev/rfcomm0: No such file or directory". Maybe I typoed the address or something.
11:05:52 <AnMaster> fizzie, possibly you selected wrong channel?
11:06:41 <AnMaster> sdptool search --baddr <MAC of phone> DUN
11:06:52 <AnMaster> you can drop the baddr stuff if you make the phone visible
11:08:33 <fizzie> That looks like it should be correct:
11:08:38 <fizzie> http://pastebin.ca/1555961
11:11:45 <AnMaster> it says it is closed here too until I try to use it
11:11:58 <fizzie> "rfcomm connect 0 00:60:57:B8:DF:55 1" gives the "accept connection request?" dialog on the phone and connects properly, though.
11:12:27 <AnMaster> fizzie, hm I have my phone set to auto accept connections from my laptop (but of course not from anywhere else)
11:12:31 <fizzie> Then I can give AT commands to the phone with minicom; but I have to keep that "rfcomm connect" running.
11:12:59 <AnMaster> didn't need to do that at all hm
11:13:12 <fizzie> Well, I'll check the commands. Though minicom did send some sort of automatic init string -- "AT S7=45 S0=0 L1 V1 X4 &c1 E1 Q0" -- which hopefully won't break anything.
11:13:26 <AnMaster> rfcomm0: 00:21:AA:FC:F2:F4 channel 1 connected [tty-attached]
11:13:51 <AnMaster> not sure what that command does
11:16:18 <fizzie> Hmmn. Well, I got some binary stuff (which looked reasonably familiar, I think it's the PPP setup) but just the "G", not the "boxed G". Which sort-of makes sense; the PPP server is the phone, after all, and presumably it's going to wait until the PPP connection is formed before actually doing the GPRS call.
11:16:45 <AnMaster> fizzie, did the phone lock up though?
11:17:07 <fizzie> Nope. I don't know what's up with that.
11:17:48 <fizzie> Though the "G" symbol seems to be stuck there now. (Normally it's displaying just an "antenna" icon when nothing's connected.)
11:18:32 <AnMaster> fizzie, try power cycling the phone
11:19:23 <fizzie> Well, it went away after I browsed (with the phone browser) to google.com and then closed the browser. Silly stuff, though.
11:19:54 <fizzie> I think I'll try that full dial-up networking connection too, just in case.
11:27:06 <fizzie> Yes, not just minicom commands.
11:27:19 <fizzie> And, well, it didn't seem to want to work. But it didn't seem to want to lock up either.
11:27:24 <AnMaster> fizzie, as for wvdial, it refuses to work without user/pass it seems
11:27:48 <fizzie> Heh, that's a bit silly. Though you can try entering "x" and "x" for both, it really shouldn't matter.
11:29:26 <AnMaster> but it didn't crash phone, and it connected (boxed E)
11:29:42 <fizzie> It should get IP and DNS from the PPP-specific protocol, I've forgotten the name.
11:30:12 <fizzie> No, just some "part of PPP connection setup" thing.
11:30:37 <fizzie> pppd is probably writing rather verbose logs (it has a habit) somewhere in /var/log.
11:31:40 <AnMaster> well good to know it works, not that I will be using it a lot, just good to know I can if needed
11:31:48 <fizzie> The lockups are strange. Did you do any setup with wvdial, or was it just "defaults for everything except /dev/rfcomm0 as the modem port"?
11:32:06 <AnMaster> fizzie, I used the relevant setup strings
11:32:23 <AnMaster> Init1 = AT&FE0V1&D2&S0&C1S0=0+IFC=3,1
11:32:23 <AnMaster> Init2 = AT+CGDCONT=1,"IP","online.telia.se"
11:32:54 <fizzie> Ah, those. Hmm. Well, maybe I'll try to set that vwdial up too later, just in case I've forgotten the 3G stick home or something.
11:33:10 <AnMaster> fizzie, well 3G stick would be cheaper I bet
11:33:35 <AnMaster> fizzie, as for lockup I think the phone locked up between wvdial and ppp was done
11:34:00 <AnMaster> otherwise wvdial was waiting for a non-existing login prompt
11:34:15 <fizzie> It's the same 10 eur/month for both the 3G stick as well as any data over the phone; they're part of the same contract.
11:35:06 <fizzie> Even the second SIM card (for the 3G stick) was free, though I had to sign up my soul to them for the next 24 months for that.
11:35:36 <AnMaster> fizzie, can't find any ppp log files, possibly it was the stuff that went to the terminal I ran wvdial in
11:36:13 <fizzie> Yes, that sounds low-level enough.
11:36:50 <fizzie> That MultiSIM thing is really strange, though; if I were to stick that second SIM into a phone, any incoming calls would cause both phones to ring.
11:37:31 <fizzie> Answering one makes the others stop ringing. I wonder if there's some sort of race condition there.
11:37:52 <AnMaster> what second sim? one for the 3G stick?
11:38:32 <fizzie> Though you can use it in another phone if you want, it's this generic "do you have multiple phones? buy a 'secondary' SIM card" deal.
11:39:35 <fizzie> The second SIM card has a some sort of "technical" phone number, but logically they work under the "primary" SIM card number. (I.e. calls to that number go to both, when dialing out the receiver sees the primary number always, they share the same voicemail box, and so on.)
11:39:52 <fizzie> Oh, and if one of the phones is busy, incoming calls still ring in the other phones.
11:40:09 <fizzie> SMS messages only go to the phone with the primary SIM card.
11:40:16 <fizzie> I don't really know what sort of use case this is for.
11:41:41 <fizzie> From the ad-blurb (translated): "Do you have two mobile phones? For example a work phone and a party phone? Would you like to take a small and light phone with you when you go out to exercise, rather than your new multimedia phone? Or does your car perhaps have a fixed phone installation?"
11:43:04 <fizzie> They all sound a bit far-fetched to me; the only one that makes sense is this "want to get online from both the phone and the computer without messing around with internet-over-phone stuff" case.
11:45:39 <AnMaster> wait hm, the max cost is 9 kr / dygn. And it says that the cost per MB is 20 kr. So that means you hit the max cost / dygn quite fast
11:45:44 <fizzie> "Note! The MultiSIM service is personal, and you may not distribute the secondary SIM cards to a third person." Otherwise I think quite a lot of people would take the "5 mobile internet connections at a price of 1" deal.
11:45:52 <AnMaster> (dygn = Swedish word for 24 hour period)
11:46:31 <fizzie> That's just "day" (overloaded a bit) in English.
11:46:57 <HackEgo> May 9, 2009 ... 2 EUR, = 3.11 CAD, 100 EUR, = 155.59 CAD. 5 EUR, = 7.78 CAD, 200 EUR, = 311.18 CAD. 10 EUR, = 15.56 CAD, 500 EUR, = 777.95 CAD. 20 EUR ...
11:46:58 <AnMaster> well day in English could be a 12 hour period as well
11:47:15 <HackEgo> 20 Swedish kronor = 1.95161087 Euros
11:47:25 <AnMaster> fizzie, what is the SEC -> EUR thingy
11:47:32 <AnMaster> it doesn't make a lot of sense
11:47:52 <fizzie> I guess it just picked some popular currencies when it didn't recognize SEC.
11:48:18 <fizzie> Anyway, it's 1.5 EUR/MB here without a specific data plan thing.
11:49:15 <fizzie> Hmm, I guess that `calc is from Google?
11:49:40 <fizzie> It seems to miss the "Did you mean: 20 SEK in EUR" correction, and pick up the summary for a "Convert Euros to Canadian Dollars" page.
11:50:10 <fizzie> Maybe it just looks for the first "X = Y" line on the page.
11:53:14 <AnMaster> prices really are a lot more reasonable nowdays than just half a year ago
11:54:53 <fizzie> The 10 eur/month (well, 9.80/month to be exact) was pretty much the upper limit for me, since the Nebula ADSL we have at home is so awfully expensive.
11:55:22 <AnMaster> fizzie, btw the built in modem thingy doesn't seem to show up in linux at all. No great loss though
11:56:25 <fizzie> If it's one of those winmodems, it's not much of a surprise. I think the iBook modem showed up as a serial port in Linux, though not sure about that.
11:56:50 <fizzie> (And "upper limit" in the "is willing to pay" sense, maybe not quite in the "can afford" sense.)
11:57:40 <AnMaster> fizzie, how much do you use it btw?
11:59:10 <AnMaster> 00:03.0 Communication controller: Intel Corporation Mobile 4 Series Chipset MEI Controller (rev 07)
12:00:46 <fizzie> It seems to be related to http://en.wikipedia.org/wiki/Intel_Active_Management_Technology maybe. No clue.
12:01:50 <fizzie> And I probably don't use it enough to justify paying every month. I just got it as a preparation for an Android phone some day; those are a bit internet-dependent.
12:02:01 <fizzie> Google Maps at 1.5 eur/MB doesn't sound like a win.
12:02:54 <fizzie> Well, except with that sort of "maximum price per day" thing, but we don't have that.
12:03:23 <fizzie> At least I don't think any Finnish operator does it; many do "you pay only this much for telephone calls per day", but they've cleverly not included data in that.
12:05:36 <AnMaster> fizzie, there is no such thing for the calls, only for the data
12:06:02 <AnMaster> fizzie, so maybe Sonera (sp?) has it?
12:06:33 <fizzie> I doubt that; Sonera is (or at least has traditionally been) one of the most expensive options here.
12:06:58 <AnMaster> fizzie, same for Telia here. Reliable, good coverage, but expensive.
12:07:13 <AnMaster> but that is going down now, still I guess the other ones are less expensive
12:08:16 <fizzie> Sonera has recently removed all "named" deals; nowadays they just have that pre-paid thing, and then for "pay monthly" there's this "build your own GSM connection" thing, where you pick all kinds of options and it computes the final price.
12:09:15 <fizzie> Actually yes, they indeed do have one max-price-per-day data thing.
12:09:56 <fizzie> "Data Tunti" (where fi:tunti == en:hour) -- you pay 0.90 EUR/hour for data, with a maximum of 6.90 EUR for a 24 hour period.
12:10:33 <fizzie> Though the 6.90 EUR/day is a lot more than your 20 SEK -> ~1.95 EUR.
12:10:55 <AnMaster> fizzie, yep. 20 SEK was per MB. But 9 SEK max per 24 hour period
12:10:56 <fizzie> Well, then it's really a lot more than that.
12:11:57 <AnMaster> fizzie, though, calls are pretty expensive. More so than when I first got this it seems.
12:12:02 <fizzie> Hey, they have (and this must be a new thing, wasn't there less than a year ago) also a "Data Päivä" (fi:päivä == en:day), where you pay 2.90 EUR at the start of each 24-hour data-use period.
12:12:29 <AnMaster> fizzie, http://www.telia.se/files/iSurfMiniMidiMaxi_TSP1543_priser_07aug.pdf
12:12:37 <AnMaster> http://www.telia.se/files/TeliaRefill_1139_3april.pdf
12:13:29 <fizzie> Since Sonera only does this custom-option-pickery, I don't have a good price sheet to point at. The web-based customizer looks a bit Finnish-only.
12:13:31 <AnMaster> fizzie, it seems that data is the only cheap thing on it
12:13:52 <AnMaster> which is quite the reverse of when I first got it
12:14:01 <AnMaster> weird they didn't tell their customers about that
12:14:41 <HackEgo> 0.69 Swedish kronor = 0.0673305752 Euros
12:14:43 <AnMaster> well... data isn't exactly cheap either. What was it you paid
12:15:09 <HackEgo> 9.80 Euros = 100.429856 Swedish kronor
12:15:21 <AnMaster> much cheaper than mine would be if I used it every day
12:15:32 <AnMaster> not that I'm likely to use it very often for that
12:15:35 <fizzie> Yes, well, that's obviously optimized for sporadic use.
12:15:59 <fizzie> Like Sonera's 2.90 EUR/24 hours. They have an about-10 EUR/month option too.
12:16:01 <AnMaster> fizzie, oh and the 0.69 thing is "within a month after you refilled the account"
12:16:11 <AnMaster> otherwise it is the cost inside the parentheses
12:16:32 <HackEgo> 1.99 Swedish kronor = 0.194185282 Euros
12:16:39 <fizzie> Oh, and this 9.80 EUR is with a 384 kbps speed limit, though; the faster options cost more.
12:18:08 <AnMaster> fizzie, mine is for 1 mbps limit iirc
12:19:06 <fizzie> I don't quite remember how much I pay for calls, since they no longer sell this limit.
12:19:49 <fizzie> The calls are reasonably expensive (something like 0.069 eur/minute, which is pretty close to that 0.69 SEK actually) but there's no monthly charge at all. (Except for this data add-on service.)
12:19:49 -!- M0ny has quit (Read error: 60 (Operation timed out)).
12:20:10 <AnMaster> fizzie, oh and they seem to do it for whole minutes. As in... anything less than one minute still cost one minute
12:20:19 <AnMaster> and between 1 and 2 minutes cost 2 minutes
12:20:38 <AnMaster> fizzie, no monthly charge at all here either
12:21:04 <fizzie> Pre-paid ones tend not to have one.
12:21:13 <fizzie> Heh, it seems that in August (when I got this mobile Internet) I used it for 96 separate connections, totaling about 27 hours, and moved 66488 kilobytes of data.
12:21:41 <fizzie> With the 1.5 EUR/MB cost, that's 97.39 EUR.
12:22:00 <AnMaster> fizzie, didn't you say it was fixed to 10 EUR?
12:22:16 <fizzie> Yes, well, s/'s/ would be/
12:23:32 <AnMaster> my phone claims I so far received 5024540 bytes
12:23:52 <fizzie> I don't think the N-Gage keeps a running count, unfortunately.
12:24:31 <AnMaster> fizzie, total connected time: two hours, 58 minutes and 32 seconds
12:24:41 <AnMaster> it is slightly less than a year old
12:26:12 <fizzie> Nope, can't seem to find such statistics in this phone. Aw.
12:26:38 <AnMaster> fizzie, it is in the same menu as called numbers, missed calls and such
12:26:44 <fizzie> There's a "GPRS data counter" in the "missed calls" menu, indeed.
12:27:01 <fizzie> "All sent data: 6.17 MB", "All receiv. data: 19.31 MB".
12:27:19 <fizzie> I've had this for at least five years now.
12:27:40 <AnMaster> fizzie, was that the same plan as your 3G usb stick thing or?
12:28:02 <fizzie> Well, same connection, but the phone can't count that of course.
12:28:41 <fizzie> I think data was something like 10 EUR/MB when I bought the phone, didn't really use it back then.
12:30:52 <fizzie> All calls: 41 hours, 52 minutes, 23 seconds.
12:31:02 <fizzie> The firmware upgrade probably reset these counters anyway.
12:32:03 <AnMaster> btw I wonder if you can get internet over bluetooth the other way around
12:32:10 <AnMaster> I mean, phone connecting to laptop
12:32:24 <AnMaster> would be useful for expensive firmware upgrades
12:32:30 <fizzie> Though to a desktop, not laptop.
12:32:33 <AnMaster> expensive as in: will easily hit the 9 kr
12:32:41 <AnMaster> fizzie, well, my desktop lacks bluetooth
12:33:03 <fizzie> So does mine, but I had borrowed an USB stick for it.
12:33:04 -!- M0ny has joined.
12:34:05 <fizzie> Gah, I don't remember the details. There was a *really* strange kludge involving an imaginary ".mrouter" TLD; Nokia phones use DNS queries to that to see if they're running under the PC Suite thing.
12:35:04 <AnMaster> well probably not worth the trouble then
12:35:09 <fizzie> http://www.bwestermann.privat.t-online.de/3650linux_en.html seems to be about it; probably very obsoleted, and bad English.
12:36:26 <fizzie> It might also not work on newer phones: http://discussion.forum.nokia.com/forum/showthread.php?t=35332
12:36:54 <fizzie> And of course it could be completely different for different models.
12:41:49 <fizzie> For Symbian phones I think there was some app. The PPP connection between the computer and the phone of course doesn't care which way the bits flow, but some sort of kludge is needed to (a) make the phone not close the connection when you're not making it connect to the Internet over GPRS/etc., and (b) to set up a default route with the computer as a gateway.
12:42:05 <AnMaster> fizzie, I don't think this is a symbian one
12:42:38 <fizzie> If it doesn't constantly reboot, it probably isn't. :p
12:42:45 <fizzie> (Okay, so that's more of a N-Gage problem.)
12:43:55 <fizzie> I can't seem to find a "3210 classic" from Nokia's device specs. There is a 3120 classic, though.
12:44:09 <fizzie> (S40 5th edition FP1.)
12:46:17 <fizzie> Incidentally, the PAN-AP thing is not listed in the "Bluetooth Profiles" list (A2DP, AVRCP, DUN, FTP, GAP, GAVDP, GOEP, HFP, HSP, OPP, SAP, SDAP, SPP; I don't know what most of those mean) for 3120 classic, so if that's your phone, it probably doesn't support the PAN approach for interweb sharing at least officially.
12:48:52 <AnMaster> fizzie, err 3120 classic was what I meant
12:49:17 <AnMaster> fizzie, and what does that mean exactly?
12:49:53 <fizzie> It's a Nokia term; most of their phones are S40 (non-Symbian, custom proprietary OS) or S60 (Symbian).
12:50:02 <fizzie> "Series 40" and "Series 60" officially.
12:50:30 <fizzie> And the S40 OS is called "Nokia OS" in the specs; I think it had some silly internal codename too.
12:50:49 <fizzie> http://www.forum.nokia.com/devices/3120_classic
12:53:26 <fizzie> The only Symbian phones in the 3xxx series seem to be 3230, 3250, 3600, 3620, 3650 and 3660, all pretty old (2003-2005).
12:54:31 <AnMaster> fizzie, anyway you say it is more advanced than the S60 one?
12:54:53 <AnMaster> also is symbian not propietary?
12:55:41 <fizzie> I guess so, maybe "internal" or something like that would be a better word. You can buy a Symbian license; you can't buy a "Nokia OS" license.
12:56:36 <fizzie> And I guess the first question depends on the precise definition of "advanced". This N-gage might be a bit more "programmable" (you can write native-code apps for it; it's just J2ME on S40 phones) but certainly yours has a lot of features that this doesn't. (Like 3G.)
12:57:06 <AnMaster> fizzie, and 30 MB flash built in or so
12:57:13 <AnMaster> no I don't have a micro sd card for it
12:57:41 <fizzie> Besides, fat load of good the programmability does me: *no-one* supports the "1st Edition" S60 devices any more, only from 3rd edition upwards.
12:58:50 <AnMaster> fizzie, how would one write J2ME apps without ending up killing one self though?
12:58:57 <fizzie> Not even open-source folks bother to support it any longer. I have to use an outdated copy of PuTTY because the couple of newest versions no longer support it.
12:59:12 <AnMaster> fizzie, can't you update the OS on it?
12:59:15 <fizzie> I don't know, I don't think J2ME is that bad. I mean, compared to Symbian development it's pure pleasure.
12:59:28 <AnMaster> oh symbian development is worse?
12:59:40 <fizzie> Symbian development is legendary in painfulness.
13:00:29 <fizzie> And you can't even flash the firmware on the N-gage without specialized hardware, if you could magically get a N-gage firmware with a newer Symbian version, which you can't.
13:00:36 <AnMaster> fizzie, is there any putty for my phone though? would be quite nice to have, though rather useless
13:00:53 <fizzie> There's no PuTTY port, but there's a J2ME SSH client, sure.
13:01:03 <fizzie> http://www.xk72.com/midpssh/
13:01:10 <fizzie> Should work just fine in your phone.
13:02:05 <fizzie> I can't seem to find the source link right now though.
13:02:50 <fizzie> Oh, it's at the end of the download page, I'm just selectively blind.
13:03:31 <fizzie> http://www.xk72.com/midpssh/downloadversion.php?version=1.7.3 and at the end, for example.
13:03:54 <fizzie> For the devel version. Of course reading the source doesn't really help unless you build it yourself, so you'd need the J2ME SDK stuff.
13:04:23 <AnMaster> so should one select devel or stable hm
13:04:43 <AnMaster> oh and, jar or jad file? for cost reasons I hope I can transfer this over bluetooth
13:04:50 <AnMaster> using the file protocol thingy (which works)
13:05:37 <fizzie> At least on my phone you can just file-transfer a .jar and it'll launch the app-installer.
13:06:10 <AnMaster> fizzie, is that file transfer ftp like or file transfer "mount phone as a directory" like?
13:06:32 <AnMaster> fizzie, also is there no ssh2 only build ?
13:06:58 <fizzie> Er, I've just used the "send file to device" OS X bluetooth thing; it speaks OBEX internally.
13:07:36 <fizzie> The file arrives as a "message" then. If your phone has a file browser app, though, you can probably just transfer it however you like and "launch" it.
13:07:48 <AnMaster> fizzie, aren't jar just zip files?
13:08:23 <AnMaster> because when I try to open it I get: "Invalid crc on data descriptor 0/134695760 on midpssh-full.jar"
13:08:32 <fizzie> Zip files with some specially handled file names (in the MANIFEST directory), yes.
13:08:47 <fizzie> Which version was this?
13:09:42 <fizzie> wget http://www.xk72.com/midpssh/v1.6.0/midpssh-full.jar + "unzip -l midpssh-full.jar" looks fine to me.
13:10:18 <AnMaster> well not with unzip, I was trying advzip to compress it better, which tends to work for every other jar file I tried it on
13:11:28 <fizzie> Technically in a .jar file the META-INF directory needs to be first, but I guess advzip wouldn't go change the file order anyway.
13:11:47 <AnMaster> fizzie, hm, well java -jar foo.jar after always worked
13:11:58 <AnMaster> 495 Defl:N 500 867670159% 02-27-07 20:32 171efe4a small.png <-- that looks a bit wrong
13:12:07 <AnMaster> there are a few more like that
13:12:07 <fizzie> Sure, I don't think apps really care about it.
13:12:50 <fizzie> 495 Defl:N 500 -1% 2007-02-27 20:32 171efe4a small.png
13:12:54 <fizzie> That's what my "unzip" says.
13:13:45 <fizzie> The .png files indeed have a bit larger compressed size. Optimally they wouldn't be deflate-compressed at all.
13:14:08 <AnMaster> indeed, and the pngs would first be optimially compressed with optipng and advpng
13:14:29 <AnMaster> though some apps tend to throw up on 4 bit gray scale pngs and such
13:14:34 <AnMaster> so might not be a good idea always ;)
13:14:54 <AnMaster> (I love how complex png actually is)
13:15:04 <fizzie> With 30 MB of flash (and something like a megabyte-or-two Java heap) I don't think you should care *that* much about a 121164-byte .jar.
13:15:27 <fizzie> You haven't put a gigabyte-or-a-couple microSD card in the phone?
13:16:55 <AnMaster> fizzie, I remember having problems running non-signed jars in my phone hm
13:22:22 <AnMaster> fizzie, wait, can it do public key auth?
13:26:15 <AnMaster> fizzie, I seem unable to specify a non-standard port too
13:26:37 <fizzie> I haven't really used it, just heard of it.
13:27:33 <fizzie> The S60 port of PuTTY does public-key auth, fortunately. I'm not sure about MidpSSH; J2ME apps can't really access the phone filesystem freely, IIRC, so it'd have to have some sort of own key-store thing.
13:28:23 <fizzie> Not listed on the "Features" page.
13:28:53 <fizzie> "Use Public Key - for SSH 2 connections you can opt to use public key / private key authentication rather than a password."
13:28:58 <fizzie> Yes, it should support that.
13:29:08 <fizzie> http://www.xk72.com/midpssh/wiki/public-key
13:29:55 <AnMaster> fizzie, the non-standard port bit doesn't seem to work though
13:30:10 <fizzie> "now available", doesn't say which version it refers to. And doesn't support passphrases in the public key yet. :p
13:30:31 -!- oerjan has joined.
13:30:52 <fizzie> And you should be able to use just "host:1234" as the hostname to connect to port 1234.
13:31:26 <AnMaster> fizzie, just timed out, and yes I know it works because I done it from uni with my laptop
13:31:36 <fizzie> Well, based on http://www.xk72.com/midpssh/wiki/using-sessions it's supported. Feel free to check sources whether that feature too is only in the devel version. :p
13:31:50 <fizzie> Must prepare some food. →
13:32:28 <AnMaster> fizzie, well ssh keys was in stable it seemed
13:35:18 * oerjan wonders if that is genuine Quenya
13:35:50 <oerjan> and if it has anything to do with balrogs
13:37:32 <oerjan> hm Valarauko does mean that
13:38:55 <oerjan> ah but the phrase itself only gives iwc
13:39:29 <oerjan> hm just the word hirdinya in fact
13:40:08 <AnMaster> ah the dev version works better
13:41:43 <oerjan> in fact all the words i tested _expect_ valarauko only exists on that page, so i take it DMM just made something to sound right
13:44:48 <oerjan> oh and iirc quenya transliteration doesn't use ë except at the end of words
13:45:18 <oerjan> (tolkien used it only so english-speakers wouldn't make it silent iirc)
14:02:49 <AnMaster> oerjan, didn't the annotation say it was Tengwar or whatever
14:03:14 <oerjan> tengwar is the script, which he _didn't_ use
14:03:28 <oerjan> it definitely resembles quenya
14:03:59 -!- M0ny has quit.
14:04:01 <AnMaster> oerjan, you realise saying that makes you look incredibly nerdy?
14:04:12 -!- M0ny has joined.
14:04:21 <AnMaster> (though not as much as if you was able to speak it fluently)
14:04:36 <oerjan> i even used to subscribe to the conlang list
14:05:20 <oerjan> fortunately i'm much too lazy to actually _learn_ it ;)
14:15:15 <oerjan> makes me slightly less nerdy? :D
14:37:20 <oerjan> incidentally there's a thread on the iwc forum where they are discussing who is geekiest. sample: "I mentioned to a friend at an SCA event the other day that I'd translated lines from Dr. Horrible's Sing-Along Blog into the Black Speech for use by my half-orc D&D character."
14:39:34 <oerjan> (and that was just the beginning of his post)
14:42:52 -!- Pthing has quit (Remote closed the connection).
14:56:47 -!- fizzieg has joined.
14:57:34 <fizzieg> AnMaster: Incidentally, one of the reasons why this net-over-phone might not have worked was that I was trying to use the APN "saunalahti.internet" when the correct would have been "internet.saunalahti".
14:58:38 -!- fizzieg has quit (Client Quit).
14:59:55 <oerjan> those pesky finns have got internet in their saunas!
15:01:08 <fizzie> fi:lahti is also en:bay, in the geographical sense. So, "the Sauna Bay".
15:01:55 <fizzie> It is a very common place-name, actually, what with all the saunas around and all. I think there are at least four of them, probably many more.
15:04:07 <AnMaster> there, secure auth by first public key and then enforced password. (all thus using a restricted user)
15:10:56 -!- ehird has joined.
15:12:10 <ehird> 21:44:57 <ais523> wow, I just realised how terribly awful http://windows7sins.org/#1 is from a graphics design point of view
15:12:11 <ehird> fucking hell that's ugly
15:12:20 <ehird> looks like one of those one-page sites where they get you to buy a get-rich book
15:12:37 <ais523> and to top it, the content isn't particularly good either
15:12:58 <ehird> I wonder if you can anti-donate from the FSF
15:13:05 <ehird> like, transfer $10 from the FSF to the EFF
15:13:25 <ais523> do you think that the FSF are paid by Microsoft to spread FUD so obvious it backfires?
15:13:42 <ehird> ais523: that would explain the free software song
15:13:48 <ehird> JOOOOOOOIN US NOW AND SHARE THE SOFTWARE
15:13:59 <ais523> or that awful Stephen Fry documentary?
15:14:23 <ehird> I love how it was Freedom Fry when he's british
15:14:36 <ehird> 22:32:05 <coppro> w00t clean Ubuntu install is nice :)
15:14:36 <ehird> 22:32:22 <coppro> I should seriously consider doing this every 6 months, rather than upgrading
15:14:36 <ehird> you're doing something severely fucked up to your upgradees...
15:15:50 <ehird> 00:58:52 <AnMaster> ais523, any idea how to connect to internet using your phone in ubuntu?
15:16:09 <ehird> 00:59:16 <ais523> I don't see why people would assume i do
15:16:09 <ehird> 00:59:18 <ais523> *I do
15:16:10 <ehird> because everyone does
15:17:01 <ehird> AnMaster: on ubuntu it's easy
15:17:11 <ehird> connect via bluetooth, then connect to → bluetooth
15:17:44 <AnMaster> ehird, read the whole discussion with fizzie below before commenting on that :P
15:18:07 <ehird> 01:16:35 <ais523> but I imagine phone-tethering would be hard to develop, as most networks ban it
15:18:08 <ehird> 01:16:37 <ais523> rather expensive, too
15:18:19 <ehird> the iphone 3g s has tethering, but not in the us/AT&T
15:18:41 <ehird> 01:17:16 <fizzie> Er, it's a fixed 10 eur/month for unlimited data.
15:18:41 <ehird> 01:17:31 <ais523> wow, your mobile operators are much less rip-offy than the ones over here
15:18:57 <ehird> you can get a practically-unlimited-for-light-stuff contract for £25/mo
15:19:01 <ehird> with unlimited 3g interwebs
15:19:57 <ehird> ais523: also you can have a pay-as-you-go + £10/mo unlim 3g internet
15:21:34 <ehird> 01:40:01 <fizzie> chi.mp seems to be giving out "free" (haven't read the details) .mp domains.
15:21:34 <ehird> you can only host an openid thingy on them tho
15:21:51 <ehird> 01:41:13 <ais523> oh, quite a few TLDs have been taken over commercially, now
15:21:51 <ehird> 01:41:22 <ais523> which IMO is making a mockery of the Internet
15:21:51 <ehird> like everything else involving money and the internet since the 70s, right?
15:22:00 <ehird> it must be a pretty damn big mock by now for you
15:22:55 <fizzie> There was some talk about a website in the .mp, but I really didn't bother to read the details.
15:24:26 * oerjan is disappointed hum.an doesn't exist
15:27:40 <fizzie> "Netherlands Antilles"
15:27:52 <ehird> ais523: so stop using it and stop whining
15:28:07 <AnMaster> fizzie, eh.. netherlands is .nl iirc, so what is this "Netherlands Antilles"?
15:28:15 <oerjan> apparently you have to be from there to register, but it's strange no one there thought of it...
15:28:16 <ais523> ehird: I mostly don't use most of the bits I don't like
15:28:24 <ais523> the reason I whine is because I rather enjoy it
15:28:26 <oerjan> AnMaster: ever heard of colonialism?
15:28:31 <ehird> ais523: it's annoying.
15:28:48 <fizzie> It's two islands in the Caribbean; that seems to often be the case.
15:29:04 <fizzie> "The Netherlands Antilles was scheduled to be dissolved as a unified political entity on December 15, 2008, so that the five constituent islands would attain new constitutional statuses within the Kingdom of the Netherlands,[3] but this dissolution has been postponed to an indefinite future date which will at the latest be October 2010." -- maybe that's why people don't like their TLD.
15:29:11 <AnMaster> oerjan, I would have assumed they dropped the name of the parasite country when they became a separate country. And would have used the same TLD if not a separate country
15:29:36 <ehird> country TLDs are so stupid
15:30:14 <fizzie> Autonomous regions of one country often have their own ccTLDs. land has .ax, to pick a nearby example.
15:30:15 <ehird> the only thing they can usefully classify non-arbitrarily is single-country organisations (and you never know when an organisation will spread), and they break every single domain in them if the country disappears, unless they're persuaded to not drop it
15:31:47 <ehird> I wonder if there's a 3g internet service with free roaming, dammit :P
15:31:49 -!- kar8nga has joined.
15:32:00 -!- FireFly has joined.
15:32:56 <fizzie> They don't sell .ax domains to any non-land-related companies/individuals, though. I don't think I've happened across an .ax site ever.
15:33:22 <ehird> people who use country TLDs for personal stuff amuse me
15:33:30 <ehird> "I am never ever ever going to move!"
15:34:05 <ehird> grr, I wish people would stop linking to http://www.gnu.org/philosophy/selling.html
15:34:14 <ehird> the business model it outlines is utterly impractical and useless...
15:35:12 <fizzie> I think there was some talk in engadget-mobile referring to some US operator with data plans that included roaming (but 3G only in the States, only GPRS when roaming, and I don't know how unlimited-data the roaming parts were).
15:35:38 <ehird> my crapbook thing has that awful gprs/ie thing
15:35:47 <ehird> but I wonder if the gprs can be used for non-IE-server stuff?
15:35:51 <fizzie> For Finnish operators roaming-GPRS tends to cost some really absurd amount of money, like well over 10 eur/MB.
15:35:59 <ehird> who cares, I'd rather pay £10/mo for 3g
15:36:10 <ehird> but yeah, roaming, grumble
15:36:24 <ehird> the whole point of mobile internet is to make location irrelevant imo
15:36:43 <fizzie> As long as you move within your little box, yes.
15:36:45 <ehird> AnMaster: http://www.engadget.com/2009/09/06/new-thinkpad-keyboard-features-crowdsourced-design-lower-price/
15:36:51 <ehird> New ThinkPad desktop keyboard thing.
15:36:56 <ehird> Funny how I was talking about it yesterday.
15:37:04 <AnMaster> ehird, so you are saying that ccTLDs only make sense for stuff like governments and such?
15:37:20 <AnMaster> (at least they usually don't spread)
15:37:22 <ehird> AnMaster: pretty much. and local branches of international metaorganisations, maybe
15:37:34 <ehird> like "Fishermen International" has a group "Fishermen Ireland
15:37:38 <AnMaster> ehird, what about something like BBC?
15:37:40 <ehird> then fishermen.ie would make sense
15:37:49 <ehird> AnMaster: that's a governmental organisation, more or less
15:37:58 <ehird> with a few degrees of seperation
15:39:20 <ehird> but the trackpoint is a boon when typing ofc
15:39:36 <AnMaster> yeah, and I prefer a keyboard that is less shallow
15:40:10 <ehird> who likes desktops anyway :P
15:41:31 <AnMaster> ehird, for long term use I prefer a computer I don't have to look down at to see the screen. Maybe a laptop with a monitor on some sort of extendable hydralic arm or such?
15:41:42 <ehird> You can plug displays into notebooks
15:42:08 <ehird> and flatten down the screen to use the kb/nub/pad
15:42:09 <AnMaster> ehird, well yes sure, but also a full sized keyboard is nice when typing for long periods.
15:42:19 <ehird> thinkpad keyboards are full sized
15:42:26 <ehird> the keys are the same size as on a desktop keyboard
15:43:12 <AnMaster> ehird, yes mostly, Ctrl/Alt and space are not full sized for example
15:43:21 <ehird> not on the older ones
15:43:27 <AnMaster> because they needed to fit the arrow keys in on the same row kind of
15:43:53 <AnMaster> and they dropped one windows key to be able to do it
15:44:09 <ehird> AnMaster: http://www.notebookreview.com/assets/3072.jpg
15:44:46 <AnMaster> that one drops both windows keys and the menu key to manage it
15:44:56 <ehird> AnMaster: use fn as an extra modifier
15:45:04 <ehird> after all, most of the keys have no fn action
15:45:08 <ehird> so you can use it for them
15:45:18 <ehird> I don't have any menu key on this keyboard and I've had zero problems, too
15:45:22 <AnMaster> ehird, iirc impossible, fn generates no key event to computer I think... let me check
15:45:30 <ehird> maybe it does on the older ones :P
15:45:57 <ehird> http://www.thinkwiki.org/images/b/b4/ThinkPadX60.jpg ;; this was the machine the uni battery guy had, right?
15:46:02 <ehird> it's one of the most famous ones, I'm pretty sure
15:46:13 <ehird> and prolly the only IBM one to hit 14hrs...
15:46:18 <AnMaster> oh wait it does, but BIOS intercepts some, such that Fn-PgUp generates no key event apart from Fn
15:46:38 <ehird> well, sure, but it's a minority of keys that have a fn action
15:46:41 <ehird> so you can use it for all the oters
15:48:11 <AnMaster> ehird, anyway, what about escape key relative F keys... that is one thing I often typo, hitting F2 when I mean F1, because F1 is where Esc is usually
15:48:36 <ehird> I'm sure I'll adjust
15:48:51 <ehird> I just need the main block of keys, enter, backspace, delete and modifiers most of the time
15:49:00 <ehird> (I'll probably map Fn-Backspace to delete for easiness)
15:49:10 <AnMaster> ehird, and it still doesn't solve the issue with feeling I mentioned above. I much prefer keys that goes down further (like traditional desktop keyboards)
15:49:21 <ehird> yes, but you're crazy,
15:49:23 <AnMaster> sure, thinkpads have great keyboards compared to many other laptops
15:49:31 <ehird> no, they have great keyboards full stop.
15:49:51 <AnMaster> ehird, depends on if you prefer full depth or shallow ones.
15:50:42 <fizzie> The logitech illuminated Fn key doesn't generate anything visible to the computer, which is a bit of a shame.
15:50:59 <ehird> fizzie: scissor-switch right? buy the thinkpad keynub! :P
15:51:01 <ehird> [15:45] ehird: http://www.thinkwiki.org/images/b/b4/ThinkPadX60.jpg ;; this was the machine the uni battery guy had, right?
15:51:33 <AnMaster> ehird, mmh nop don't think so. Hard to say from the pic
15:51:47 <fizzie> I just switched keyboards, I think I'm going to have to use that at least until I break its spacebar too.
15:52:05 <ehird> prolly T60, or T61 if it was widescreen
15:52:11 <AnMaster> ehird, anyway I can tell you on friday, same course then.
15:52:25 <ehird> were the point's buttons red, i wonder
15:52:39 <ehird> if so then it'd be T43
15:53:39 <AnMaster> ehird, if you could show me the left side of it closer I could tell better I think.
15:53:45 <AnMaster> because I saw that most of the time
15:53:59 <AnMaster> oh and it definitely didn't have a floppy, not sure if that one there does
15:54:16 <ehird> I don't think they've ever had floppies
15:54:24 <ehird> AnMaster: I just think that any ultrabay battery must be really heavy if it changes 5 1/2 hours into 14
15:54:31 <ehird> so I doubt it's a T-series, really
15:54:47 <ehird> 9 hours with the 9-cell?
15:54:52 <ehird> It's definitely an X series.
15:54:56 <AnMaster> ehird, well maybe it was 10 cell then
15:55:19 <AnMaster> ehird, it was sticking out quite a bit, so possibly more than 9 cell
15:55:34 <ehird> The T43 gets 5 1/2 hours on the 9 cell
15:55:39 <AnMaster> ehird, one card reader slot above a pc card slot on the right side
15:55:41 <ehird> Similar for the T60
15:55:44 <ehird> So it can't possibly be a T series
15:56:03 <AnMaster> ehird, well it wasn't http://www.thinkwiki.org/images/b/b4/ThinkPadX60.jpg because it did have a touchpad.
15:56:24 <ehird> Then he was lying to you about the battery life :
15:56:47 <AnMaster> ehird, is there any step above 9 cell?
15:57:11 <ehird> Sure, like, up to 12-cell. Regardless, 5 1/2 hours on 9 cell != 9 hours on 12 cell
15:57:38 <AnMaster> ehird, might have been 12-cell then,
15:57:49 <AnMaster> I might have misremembered that bit
15:57:50 <ehird> It wasn't a T series.
15:57:59 <ehird> Only the X series, which has no touchpads, gets that much battery.
15:58:05 -!- xfire35 has joined.
15:58:34 <AnMaster> ehird, design looked very similar to my R500, the whole thing was slightly smaller (looked like maybe 14" instead of 15.4") and the battery was sticking out quite a bit
15:58:48 <ehird> It may have been 15" normalscreen.
15:59:19 <ehird> AnMaster: http://www.notebookreview.com/assets/9343.jpg
15:59:25 <ehird> http://macbitz.files.wordpress.com/2009/06/thinkpad-t601.jpg
15:59:26 <AnMaster> ehird, possibly, but definitely not 4:3
15:59:34 <ehird> http://www.thinkwiki.org/images/4/4b/Thinkpad.jpg
15:59:38 <ehird> AnMaster: Are you sure
15:59:54 <ehird> http://www.landbiksen.dk/catalog/images/5153-IMG7415s.jpg
16:00:02 <ehird> http://www.vernontech.ca/uploads/lenovo_thinkpad_t61.jpg
16:00:19 <AnMaster> http://www.notebookreview.com/assets/9343.jpg <-- not too much off, but don't remember the bit that is sticking out on the right side of the screen
16:00:28 <ehird> See links after "If so, then"
16:02:02 <AnMaster> http://www.thinkwiki.org/images/4/4b/Thinkpad.jpg <-- looks quite similar, but less widescreen? http://www.landbiksen.dk/catalog/images/5153-IMG7415s.jpg <-- PC card slots are off, it had one pc card slot and above that a card reader which was not part of a pc-card. Also that seems to have a thicker boarder for the screen?
16:02:34 <AnMaster> but maybe the thicker boarder is just doe to the angle of the picture
16:02:50 <ehird> http://www.thinkwiki.org/images/4/4b/Thinkpad.jpg is not after "If so, then".
16:02:54 <ehird> http://www.thinkwiki.org/images/4/4b/Thinkpad.jpg is the 4:3 T60.
16:03:13 -!- xfire35 has quit ("Leaving").
16:03:29 <AnMaster> ehird, well, can you wait until friday, I will try to see what the model number is then ;P
16:03:44 <ehird> I'm convinced it either doesn't exist or you're misremembering.
16:04:48 <AnMaster> ehird, I could be misremembering that it was a ibm and not a lenovo (and didn't some early lenovo ones still have the ibm logo on them?)
16:05:14 <fizzie> Maybe it was actually a live goat and not a laptop at all?
16:05:15 <ehird> Only the latest two models or so don't have the IBM logo, whereas Lenovo has owned it since 2004/2005.
16:05:24 <ehird> So there are many Lenovo-manufactured models that have the IBM logo on them.
16:05:40 <AnMaster> ehird, right. Mine doesn't have the IBM logo btw
16:06:15 <AnMaster> ehird, anyway, show me pics of 9 cell and 12 cell and I'll tell you which of those it was
16:06:48 <ehird> Unless that ultrabay battery back weighs like a kilo, no way a trackpad-having ThinkPad gets that much battery, though.
16:07:07 <AnMaster> ehird, the trackpad uses a lot of energy?
16:07:36 <ehird> No, it's just that the X series is the only one that could handle that sort of battery life.
16:07:41 <ehird> And they don't have trackpads.
16:07:56 <ehird> AnMaster: Have you seen http://www.thinkwiki.org/wiki/PC110?
16:08:12 <ehird> Pic with scale: http://www-cs-students.stanford.edu/~linma/gif/ibm_palmtop_pc110.gif
16:08:22 <ehird> Actual photo: http://www.mobile-diary.com/pc110/pc110_5.jpg
16:08:36 <ehird> AnMaster: No idea, but jeez!
16:08:45 <ehird> Imagine that keyboard. Ouch.
16:08:50 <ehird> Gotta be worse than the iPhone's.
16:09:07 <ehird> I like how http://www-cs-students.stanford.edu/~linma/gif/ibm_palmtop_pc110.gif is Droste.
16:09:37 <ehird> Also really pointless.
16:09:44 <AnMaster> what is the meny system thingy on it
16:09:57 <ehird> Some proprietary bootup thing, I think. It also runs Windows.
16:10:03 <ehird> And DOS, and Linux and BSD.
16:10:11 <ehird> It's also very thick.
16:10:32 <fizzie> "So's your..." and so on.
16:11:15 <ehird> There are two completion to that that come to mind and only one is an insult. Although the other is mildly disturbing.
16:13:26 -!- KingOfKarlsruhe has quit (Remote closed the connection).
16:13:27 <AnMaster> aah I know what the other one is! "goldfish"
16:13:54 <ehird> Yo goldfish so fat, she doesn't fit in a standard-sized goldfish bowl.
16:14:51 <ehird> "Yo momma so fat, she's slightly above the standard BMI!"
16:14:58 <fizzie> Wife's family have this Toshiba T1000 PC "laptop", it's so very portable: http://www.myoldcomputers.com/museum/comp/museumpics/toshibat1000a.jpg
16:15:11 <ehird> fizzie: What, really?
16:15:22 <ehird> Please tell me they actually use it.
16:15:41 <fizzie> 640x200 screen and all. And a 256-kilobyte ROM chip (with ms-dos 2.something) in place of a HD; but you can use floppies for storage.
16:16:07 <fizzie> There was something wrong with it last I heard, but it wasn't very many years ago when we played a bit of Tetris on it.
16:16:15 <ehird> Wait, it has a graphical screen? Amazing.
16:16:17 <fizzie> Can't exactly say it'd be in *active* use right now though.
16:16:27 <ehird> That keyboard is probably buckling-spring.
16:16:54 <fizzie> CGA-compatible monochrome screen, but graphical, yes.
16:16:56 <ehird> Deewiant: That old?
16:17:03 <ehird> Anyway, awesome people used this: http://www.computercloset.org/Apple_Macintosh_Portable.jpg
16:17:11 <ehird> It's got a trackball and button and all.
16:17:18 <AnMaster> also what do you mean... non-graphical screen?
16:17:42 <ehird> The Macintosh Portable was also stylish! All whit eand all.
16:17:49 <ehird> Also it ran system 6 up to system 7.5.
16:17:50 <fizzie> AnMaster: There's a serial port.
16:17:55 <ehird> And had a 16Mhz processor!
16:17:59 <AnMaster> fizzie, hm that's a bit too slow
16:18:01 <ehird> Plus you had a whole 1MiB of RAM.
16:18:02 <fizzie> "Alongside the earlier T1100 and T1200 systems, the Toshiba T1000 helped to change the image of portable computing. Before the T1100, previous supposedly portable computers were large and unwieldy." Fortunately that's not unwieldy at all.
16:18:09 <ehird> Even expandable to 9MiB or 8MiB if it's the backlight version.
16:18:11 <Deewiant> AnMaster: You think the CPU could handle anything faster? :-P
16:18:14 <ehird> It's so high-spec, isn't it?
16:18:18 <ehird> It even had a hard disk!
16:18:20 <AnMaster> ehird, but wasn't the CRT itself technically able to render graphics, just the hardware wasn't able to be told to do so?
16:18:23 <ehird> And it had a modem inside.
16:18:28 <AnMaster> (not that the difference matters)
16:18:37 <fizzie> That T1000 had the standard 4.77 MHz.. well, it's a 80C86, but compatible anyway. And 512 kB of RAM.
16:18:38 <ehird> AnMaster: True, I guess.
16:18:53 <ehird> fizzie: See? The Macintosh Portable was way superior. And came out in 1989. :-P
16:18:54 <AnMaster> ehird, of course it could be non-graphical like a cheap calculator
16:18:58 <ehird> Too bad it cost $6,500 for the base model.
16:19:12 <fizzie> ehird: The T1000 came out in 1987, so it's understandable it's a bit less powerful.
16:19:24 <ehird> Also, it had a good display.
16:19:25 <AnMaster> ehird, wasn't the macintosh portable WAY heavier?
16:19:34 <ehird> AnMaster: 16 pounds
16:19:40 <ehird> `calc 16 pounds in kg
16:19:41 <HackEgo> 16 pounds = 7.25747792 kilograms
16:19:50 <AnMaster> ehird, that Toshiba looks lighter
16:19:55 <ehird> I doubt the other one was much less luggable.
16:20:13 <AnMaster> ehird, maybe I misjudged the scale...
16:20:18 <ehird> It doesn't really matter at that point, considering you can't really carry it.
16:20:22 <ehird> AnMaster: They're about the same size.
16:20:37 <AnMaster> ehird, one has a tracball on the side too
16:20:43 <ehird> But its keyboard is smaller.
16:20:44 <AnMaster> possibly smaller keyboard though
16:20:45 <fizzie> http://pc-museum.com/gallery/rcm-061-big-Macintosh-portable.jpg -- it does look a lot kewler.
16:20:57 <fizzie> Ah, there was a picture already.
16:21:19 <AnMaster> anyway my plan when asking for network was:
16:21:43 <AnMaster> but with serial it wouldn't be usable even as that
16:21:46 <fizzie> It's usable as a serial terminal.
16:21:53 <ehird> You know, it'd be easier to plug in a monitor and keyboard to the server. :P
16:22:13 <ehird> fizzie: Does that Toshiba have a fan in it?
16:22:15 <AnMaster> fizzie, well yes, but I was thinking something that was low end graphical, something that 10 mbps Ethernet could have handled
16:22:18 <ehird> I bet the Macintosh Portable didn't!
16:22:34 <ehird> Since Steve Jobs hated fans. Which is why the Mac kept frying.
16:22:37 <fizzie> I don't really know, could be either way.
16:22:54 <fizzie> I used to have a VT510, but it burned out. :/
16:23:23 <fizzie> An awesome piece of hardware, did some sort of 132x50 tiny-font display and everything.
16:23:27 <ehird> http://dec.vecmar.com/products/productpage.asp?pid=2660-DEC_VT_510_Terminal
16:23:31 <ehird> so not exactly a rarity :P
16:23:32 <AnMaster> I wonder why there is a useless indention in the "screen" (technically not the screen, but the area around it) on that mac... I mean one is for the trackball
16:24:01 <ehird> It looks nicer symmetrical.
16:24:11 <ehird> I want a GreenArrays chip. I'd write a little OS thingy for it. :P
16:24:24 <fizzie> Sure it's not rare, but I don't think I can justify getting another.
16:24:38 <fizzie> I don't think it did color, though, unlike the VT525. (Might be remembering wrongly too.)
16:24:44 <ehird> "Here's a comparison of GA32 and 2 GA4 packages. We're testing chips and making mock-ups for possible applications. Hopefully mock-ups will help raise funds for real boards. Testing, of course, goes on forever."
16:24:44 <ehird> http://colorforth.com/G32-G4.jpg
16:24:46 <ehird> Wow they're small.
16:24:58 <ehird> (That's what she said to the FSM.)
16:25:24 <AnMaster> anyway, how long would a 9 cell battery last with those?
16:25:49 <ehird> Flying Spaghetti Monster
16:25:55 <AnMaster> (of course they aren't intended for the same purpose)
16:26:00 <ehird> And not very on their own, you assemble them into a supercomputer.
16:26:02 <ehird> A bunch of them in one.
16:26:09 <ehird> They're unclocked, so they all run at different speed.
16:26:17 <ehird> (native silicon speed, pretty much)
16:26:30 <ehird> http://colorforth.com/haypress2.jpg is an assembled littleputer thingy.
16:26:57 <AnMaster> ehird, don't you need clock for some stuff? Like making sure avoiding metastable state and such.
16:27:04 <AnMaster> (I'm no expert on this though)
16:27:13 <ehird> Chuck Moore is clever. he'll have fixed it
16:27:18 <ehird> He knows that kind of stuff.
16:28:36 <AnMaster> thought you needed some kind of speed limiting to make some components work by making sure they didn't settle into a meta stable state or whatever. "flip-flops"?
16:29:18 <ehird> Does anyone know ... whatever I was going to ask
16:30:03 -!- oerjan has quit ("Good night").
16:31:56 <fizzie> No, I think ehird's question is still an unsolved problem.
16:36:33 -!- sebbu2 has joined.
16:37:23 <ehird> ais523: i'm curious as to an answer to AnMaster's q too
16:38:38 <fizzie> At least it's not the first clockless CPU design; there have been other working-but-experimental models.
16:39:05 <ais523> ehird: sorry, I've been afk for a while
16:39:05 <ehird> I wonder what native silicon speed is, anyway
16:39:12 <ais523> AnMaster: what's the issue?
16:39:45 <AnMaster> <ehird> They're unclocked, so they all run at different speed.
16:39:45 <AnMaster> <ehird> (native silicon speed, pretty much)
16:39:47 <AnMaster> <AnMaster> ehird, don't you need clock for some stuff? Like making sure avoiding metastable state and such.
16:39:49 <AnMaster> <AnMaster> thought you needed some kind of speed limiting to make some components work by making sure they didn't settle into a meta stable state or whatever. "flip-flops"?
16:40:03 <ais523> yes, you do need speed limiting, due to slew rates
16:40:20 <ais523> the metastable state only really happens with self-references, unless you're really unlucky
16:40:29 <AnMaster> ais523, so how could that colourforth chip work as unclocked?
16:40:36 <ais523> but if you try to run a circuit too fast, there just won't be time to change 1s to 0s
16:41:01 <ais523> and asynchronous design doesn't imply that it's running max speed with nothing limiting; just that there isn't a global clock
16:41:16 <ehird> is silicon speed quantifiable in hz?
16:41:22 <ais523> there'll almost certainly be local synchronization or control circuits, either implemented clock-style or some other way
16:41:40 <ais523> it's massively temperature-dependent, for one thing
16:41:40 <ehird> I assume it's like 50ghz normally or sth?
16:42:09 <ais523> but in one of our lab experiments, we chained 7 NOT gates in a loop, which for that sort of logic gave oscillation rather than a metastable state
16:42:28 <ais523> and the resulting oscillation was so fast that we could only just notice it on the student oscilloscopes, even setting them to the max frequency range
16:42:40 <ais523> I can't remember offhand what value they go up to, though
16:42:56 <fizzie> "Caltech designed and manufactured the world's first fully Quasi Delay Insensitive processor. During demonstrations, the researchers amazed viewers by loading a simple program which ran in a tight loop, pulsing one of the output lines after each instruction. This output line was connected to an oscilloscope. When a cup of hot coffee was placed on the chip, the pulse rate (the effective "clock rate") naturally slowed down to adapt to the worsening performance
16:42:56 <fizzie> of the heated transistors. When liquid nitrogen was poured on the chip, the instruction rate shot up with no additional intervention."
16:43:43 <ais523> ofc, the massive slowdown in silicon-based circuits when they're heated is one of the more interesting demonstrations
16:43:57 <ehird> no overclocking required
16:43:59 <ais523> it basically makes all the components electronically leakier
16:44:04 <ehird> 1. Beef up the cooling
16:44:06 <ehird> 2. There is no step two!
16:44:11 <ais523> you get more stray electrons doing random things, rather than what you want them to
16:44:45 <ehird> so what kind of range does it get at usual CPU temperature
16:44:51 <ehird> double digit ghz? triple?
16:45:13 <ais523> it depends a huge amount on the fabrication techniques used
16:45:43 <ais523> you have to pay quite a lot extra to get a circuit going faster than a few tens to hundreds of megahertz, but ofc modern computers are much faster than that
16:46:06 <ais523> (which is one of the reasons that modern processors are quite expensive as silicon chips go)
16:46:38 <fizzie> One would assume the "instruction rate" would depend also on how the chip does whatever it does, i.e. how much logic there is to go through.
16:46:51 <ehird> ais523: let's say AMD produced a clockless chip using typical desktop cooling
16:47:02 <ehird> with as good as process as their x86 chips
16:47:09 <ais523> ehird: the answer is I don't know
16:47:21 <ais523> my guess is, though, that the circuitry would be more complex
16:47:30 <ais523> and the circuit would have to run more slowly as a result
16:47:40 <ais523> OTOH, you'd have a considerably simpler motherboard
16:47:58 <ais523> half of what a motherboard does is trying to match the wildly varying clock speeds of different bits of the system
16:54:09 -!- sebbu has quit (Read error: 110 (Connection timed out)).
16:54:12 -!- Asztal has quit (Read error: 110 (Connection timed out)).
17:27:33 -!- MigoMipo has joined.
17:35:09 <ehird> http://www.reddit.com/r/programming/comments/9hudh/the_most_complex_computer_science_operating/ ;; BSD astroturf? :-P
17:36:51 <ehird> http://www.losethos.com/godsongs.html
17:43:39 <ehird> "I consult God regularly. One design decision He made was no child windows -- policy of one window per task. Another decision He made was 640x480 resolution. I posted a promise of sticking to this resolution. Then, God said He didn't mean permanantly, but I had already promised and God expects you to keep your word. Those were the main influences by God on my operating system. He's made little things, too."
17:43:56 <ehird> ("God's court is the supremest court. It's funny -- I hear these punks saying piracy is not theft. I'd like to see them tell God. I thought so!")
17:44:03 <ehird> you never fail to make me laugh, losethos
17:44:12 <ehird> ("Wouldn't you believe in God if He talked to you? What's nutty about that?")
17:44:28 <ehird> "I downloaded opensound with a BSD license, but couldn't uncompress it. I even bought a new copy of winzip -- no luck."
17:47:22 <ais523> what format was the archive in? tgz?
17:53:07 <ehird> he's just so crazy :)
18:30:12 <fizzie> If "opensound" means OSS, the download form lets you decide between .rpm, .deb and .tar.bz2.
18:31:26 <fizzie> A reasonably new winzip should be able to do bzip2, I think.
18:31:33 <ehird> He decided not to because it's evil BSD license
18:31:39 <Deewiant> Stupid valgrind, segfaulting on an unstripped binary :-(
18:31:45 <ehird> And it allows no commercial use therefore it's against people making money
18:32:15 <fizzie> Source downloads are .tar.bz2 only.
18:36:59 <AnMaster> Deewiant, also what the hell did you do to make it do that
18:37:19 <Deewiant> strip -s ccbi && valgrind ccbi -- OK
18:37:36 <Deewiant> Also happened with 3.4.something last week
18:37:38 <AnMaster> becuase I guess this is some issue with reading debug info
18:37:58 <AnMaster> Deewiant, I hope your valgrind is built with debug info so you can report a bug about it
18:38:26 <Deewiant> They'll want a copy of the binary anyway so I won't bother
18:38:53 <AnMaster> Deewiant, oh? I reported bugs to valgrind before. All but one fixed so far
18:39:06 <AnMaster> Deewiant, really good devs there who tries to figure out each bug
18:39:17 <AnMaster> of course, a good test case is always a good idea
18:39:39 <AnMaster> but failing that they seem to accept binaries that cause the issue too
18:39:52 <Deewiant> Yeah, but I don't want to give them a copy even of the binary :-P
18:40:15 <AnMaster> also, make a simple test case in D
18:40:34 <AnMaster> Deewiant, ccbi2? you made the hg repo public iirc.
18:40:42 <AnMaster> because I'm pretty sure I have a copy somewhere
18:41:10 <Deewiant> Hmm yeah, I could try a non-secret copy
18:41:39 <ehird> <AnMaster> HOW DARE YOU BE A CATHEDRAL
18:42:59 <Deewiant> Yeah, evidently the non-secret one fails too
18:43:32 <ehird> Deewiant: how's it secret :P
18:44:06 <Deewiant> It's not a /problem/ for me though, I can just run valgrind on the stripped copy and then ask gdb unstripped_copy what the symbol at a given address is
18:45:42 <Deewiant> strip --only-keep-debug makes it unexecutable, nice
18:47:42 <Deewiant> Ah, it could be http://bugs.kde.org/show_bug.cgi?id=197988
18:47:42 <fizzie> Isn't that what it should do, anyway?
18:48:10 <Deewiant> Is it? Isn't there non-debug not-needed-for-running stuff
18:48:13 <fizzie> "The intention is that this option will be used -- to create a two part executable. One a stripped binary which will occupy less space in RAM and in a distribution and the second a debugging information file [which is not supposed to be executed anyway]."
18:49:31 <fizzie> I guess there's no "strip everything not necessary for running, but do keep debug info" flag.
18:50:42 <Deewiant> Great, they had a fix for it (2M stack) but backed it out
18:51:06 <Deewiant> I guess it's time to build my own valgrind and see if that helps
18:52:11 <AnMaster> <Deewiant> Yeah, evidently the non-secret one fails too <-- what is the secret one?
18:59:42 <Deewiant> http://bugs.kde.org/show_bug.cgi?id=197988 presumably wasn't it after all, I made the stack size about 300 times bigger and it still doesn't work.
19:00:48 <AnMaster> Deewiant, don't use such insanely long symbols
19:00:59 <Deewiant> I doubt they are that insanely long
19:01:01 <AnMaster> Deewiant, tell me what your long symbol is
19:01:13 <Deewiant> I don't know whether that even is the problem
19:01:15 <ehird> Don't you compile your binaries manually
19:01:29 <AnMaster> Deewiant, nm ./ccbi-unstrripped <-- look for the longest line
19:01:33 <Deewiant> dobelx64 is manually compiled though
19:01:42 <Deewiant> Can I do a sort by length somehow easily
19:01:55 <AnMaster> Deewiant, hm... I was thinking about that myself
19:02:00 <ehird> awk print length | sort numeric
19:02:16 <Deewiant> How does one get length in awk, is the question
19:02:30 <Deewiant> It is, evidently, length. How obvious.
19:02:57 <AnMaster> Deewiant, include the symbol too
19:03:01 <AnMaster> Deewiant, also that is insanely long
19:03:06 <ehird> [[awk '{print length" "$0}' | sort -n]], FWIW
19:03:22 <Deewiant> [[awk '{print length, $0} | sort -n]] is what I did
19:03:36 <Deewiant> The symbol is apparently full of MD5
19:03:42 <Deewiant> Meaning it's been compressed quite a bit already.
19:03:53 <AnMaster> wait, isn't md5 supposed to be one way?
19:03:56 <Deewiant> Some-end of the compiler does that to compress them.
19:04:11 <Deewiant> Yes, it is, so it's not fully clear where the symbol came from.
19:04:25 <Deewiant> It does have the initial part though, so I can guess.
19:04:45 <AnMaster> also if you get symbol names you need to compress to fit you are doing something wrong elsewhere ;P
19:04:45 <Deewiant> AnMaster: Think about it. If this were reversible, it'd be another 10-100x bigger.
19:05:01 <Deewiant> The compiler is doing something wrong by emitting this.
19:05:09 <AnMaster> Deewiant, yep that is what I meant
19:05:10 <Deewiant> This is a known bug, of which the cause is not known.
19:05:28 <Deewiant> It just shouldn't be emitted because it's compile-time only
19:05:48 <AnMaster> I meant that the compiler is doing something wrong if it manages to construct a symbol that long
19:05:49 <Deewiant> Or I mean, there is something that long there, it just should be discarded.
19:05:56 <Deewiant> Exactly, and you're wrong there.
19:06:23 <AnMaster> Deewiant, or possibly the user is doing something wrong,
19:06:25 <Deewiant> This is presumably the concatenation of all fingerprints' instructions' identifiers' names
19:06:33 <ehird> have you considered that YOU fail
19:06:54 <AnMaster> ehird, no I meant the common case where C++ generates long ones due to lots and lots of templates
19:07:00 <Deewiant> And since they're all "ccbi.fingerprints.rcfunge98.trds.dosomethingcool" or the like, it's no surprise that this grows a bit.
19:07:16 <ehird> Objective C++: "All the high level expressivity of C, with the lean and slim design of C++ together with the speed of Smalltalk."
19:07:34 <ehird> (http://www.reddit.com/r/programming/comments/9hqbk/drag_mouse_to_rotate_look_at_tree_through_portals/c0cu74a)
19:08:08 <AnMaster> ehird, unable to view that flash?
19:08:21 <Deewiant> No, I don't think he is unable.
19:08:35 <ehird> I was linking to a comment, and no.
19:08:42 <ehird> You chose not to install flash, so you don't see flash content.
19:08:49 <ehird> That is a decision you have made and it is only because of you.
19:09:29 <Deewiant> The sum of all symbol lengths is 4895700
19:09:36 <Deewiant> Note, this is in the non-"-g" version
19:10:27 <Deewiant> The -g version goes up to 4974975; not that much more
19:11:35 <Deewiant> Anyway, this is what happens when you do compile-time functional programming with a crap compiler frontend ;-P
19:12:09 -!- augur has joined.
19:12:31 <ehird> augur: I DON'T LOVE YO
19:12:38 <ehird> i have hyperactive enterkey syndrome
19:13:06 <Deewiant> It might be a linker fail in some way too, I guess, since they shouldn't be used from anywhere
19:14:18 <augur> ehird: well one would hope not
19:14:28 <augur> youre WAY too young
19:14:32 <ehird> YOU SAID THAT THE ONE OF THE TWO OF THE writing lines is boring
19:14:38 <ehird> anyway you would know, pedo
19:14:44 <ehird> HAVING CLAIMED TO LOVE ME THE LAST TIME YOU WERE IN HERE
19:14:47 <ehird> what am i even talking about?
19:15:09 <AnMaster> Deewiant, you could try nm on each *.o
19:15:22 <AnMaster> I mean, if it was linker only or not
19:15:49 <AnMaster> Deewiant, try non-whole-program-insanely-fast mode?
19:16:17 <AnMaster> so you have one *.o per compilation unit
19:18:55 -!- coppro has quit (Remote closed the connection).
19:19:46 <Deewiant> 13M ccbi.instructions.templates.o
19:20:16 <Deewiant> A 69-line file with four workarounds for compiler bugs
19:20:38 <Deewiant> Vim doesn't recognize it as D and thinks it's dtrace
19:24:15 -!- coppro has joined.
19:24:35 -!- Asztal has joined.
19:24:36 <ehird> http://blogs.telegraph.co.uk/technology/shanerichmond/100003402/google-ufo-logo-mystery-solved/
19:25:12 <AnMaster> Deewiant, also what about emacs?
19:25:25 <Deewiant> I don't know, I don't have emacs.
19:26:02 <AnMaster> Deewiant, this just proves how crappy a) current C implementations and b) vim are
19:26:23 <ehird> being unable to detect a file type = VIM SUX EMACS RULZ
19:26:28 <ehird> VIM PERHAPS EVEN DROOLZ
19:26:37 <AnMaster> ehird, yes, it is horrible isn't it
19:28:25 <Deewiant> The 87.97-MB emacs doesn't seem to have D support at all.
19:28:57 <AnMaster> 87 MB? what emacs variant is that?
19:29:12 <Deewiant> "emacs-23.1-3" from Arch's repositories.
19:29:26 <fizzie> Emacs is a bit lacking in the different-file-types front; of course there's a major mode for just about anything, but they don't ship everything by default.
19:29:55 <Deewiant> I like how they ship me a /var/games/emacs/snake-scores but not filetype support.
19:30:20 <AnMaster> Deewiant, I don't install that part of emacs usually
19:30:21 <fizzie> "ls /usr/share/emacs/23.1.50/lisp/progmodes/*.elc | wc -l" => 85, vs. "ls /usr/share/vim/vimcurrent/syntax/*.vim | wc -l" => 522.
19:30:51 <AnMaster> $ ls /usr/share/emacs/22.3/lisp/progmodes/*.elc | wc -l
19:31:14 <fizzie> Deewiant: If you want to know the test:
19:31:16 <fizzie> if match(lines, '^#!\S\+dtrace\|#pragma\s\+D\s\+option\|:\S\{-}:\S\{-}:') > -1
19:31:39 <Deewiant> Haven't bothered to modify it.
19:31:45 <fizzie> What part of the test is confusing it?
19:32:56 <Deewiant> Equivalent to * but non-greedy, evidently
19:33:42 <Deewiant> Yes, the line const C = "'" ~ (i=='\''?r"\'":i=='\\'?r"\\":i=='"'?"\"":""~i) ~ "'"; matches that last bit.
19:34:03 <fizzie> Apparently so; sounds like the Perl *?
19:34:19 <Deewiant> Beats me, I don't know the details of Perl regexp that well.
19:34:42 <AnMaster> <Deewiant> Yes, the line const C = "'" ~ (i=='\''?r"\'":i=='\\'?r"\\":i=='"'?"\"":""~i) ~ "'"; matches that last bit.
19:34:54 <fizzie> It has *?, +? and ?? as non-greedy variants of plain *, + and ?.
19:35:13 <fizzie> Oh, and {...}? for {...}.
19:35:23 <fizzie> I think Java borrowed thom too.
19:35:43 <AnMaster> Deewiant, what the hell is that D line
19:35:47 <fizzie> No, "thom". It sounds more impressive that way.
19:35:59 <Deewiant> ~ is concatenate, the rest is C-compatible.
19:36:23 <Deewiant> Oh, and r"" is a string without escape chars
19:36:24 <AnMaster> Deewiant, ok... why the hell then :P
19:36:42 <fizzie> I think I was trying to write "those" and "them" simultaneously.
19:37:04 <AnMaster> Deewiant, why do it as a series of nested ?: expressions?
19:37:38 <AnMaster> Deewiant as opposed to if else?
19:37:48 <Deewiant> if-else isn't a constant expression, is it now
19:37:53 <AnMaster> Deewiant, but that isn't constant either
19:38:15 <AnMaster> and if i is, then what is the point?
19:38:18 <Deewiant> i is a template parameter, so it is indeed constant.
19:38:40 <AnMaster> Deewiant, tell me, what the hell is it that you are doing :P
19:38:49 <Deewiant> That line is a workaround for http://d.puremagic.com/issues/show_bug.cgi?id=1059
19:39:48 <Deewiant> This code is probably in your copy of CCBI2, too.
19:41:33 <AnMaster> how many open D compiler bugs are there?
19:42:01 <Deewiant> That includes frontend and backend
19:42:12 <AnMaster> Deewiant, and in the ldc backend?
19:42:34 <Deewiant> 53, but some of those might also be for their frontend fork
19:42:44 <AnMaster> Deewiant, oh and how many compiler bug workarounds does CCBI2 contain?
19:43:35 <AnMaster> Deewiant, what is so secret about it ;P
19:44:01 <Deewiant> 7 of those are for bug 810, which was fixed in the penultimate release but hasn't been merged into LDC
19:44:13 <Deewiant> Because that release breaks backwards compatibility with something else
19:44:48 <AnMaster> Deewiant, how long since D reached stable version?
19:45:03 <AnMaster> as defined by what upstream consider stable
19:45:06 <Deewiant> Oh, and 8 of them are for a workaround in the build tool
19:45:18 <Deewiant> Which could be removed if I switched to xfBuild, which I haven't bothered to do
19:45:50 <Deewiant> Well, 1.0 was called stable, although it actually wasn't (GRR 1.017, breaking several parts of my code)
19:46:02 <Deewiant> Anyway, that was beginning of 2007
19:46:12 <Deewiant> When the D1 - D2 fork happened
19:46:58 <AnMaster> I'm a bit surprised that there are so many bugs still left in D after that time. Few devs working on it?
19:47:02 <fizzie> Debian has 77742 active bugs in 11741 packages; what a rat's nest of bugs.
19:47:23 <AnMaster> fizzie, about 7 bugs per package?
19:47:55 <AnMaster> fizzie, any stats over which package has most?
19:48:08 <AnMaster> it would be interesting to see how evenly distributed it is
19:48:41 <fizzie> The median number of bugs is 2, however.
19:49:41 <Deewiant> Workaround stats for the DMD frontend in ccbi: 7 for 810, 5 for 1059, one each for 1640,1715,2288,2339,2991
19:49:41 <fizzie> 4804 packages with one bug, 2024 with two. Oh, and this is from the active-bugs index, so it's completely ignoring all packages with no bugs at all, which is not quite fair. :p
19:50:19 <AnMaster> Deewiant, I don't think I have any in cfunge, and I had one in efunge. Yay for stable software :P
19:50:38 <Deewiant> Yay for compilers with more than one maintainer :-P
19:50:41 <fizzie> Oh, and the top-three active bugs:
19:50:44 <fizzie> * apt (maintainer: APT Development Team <deity@lists.debian.org>) has 828 bugs
19:50:44 <fizzie> * installation-reports (maintainer: Debian Install Team <debian-boot@lists.debian.org>) has 2113 bugs
19:50:44 <fizzie> * wnpp (maintainer: wnpp@debian.org) has 3528 bugs
19:50:52 <AnMaster> of course there are some I more or less consider workarounds, like the double nested stringification macro
19:51:09 <AnMaster> so possibly a stupid standard workaround
19:51:17 <Deewiant> Just like installation-reports presumably is
19:51:30 <AnMaster> fizzie, what is the wnpp meta-package for?
19:51:38 <Deewiant> http://www.google.com/search?q=wnpp
19:54:15 <fizzie> Yes, "installation-reports" bugs are I think general "problems installing Debian" issues.
19:54:48 <fizzie> The apt bugs seem reasonably real, though.
19:55:18 <AnMaster> I guess it is just a software everyone use
19:55:29 <fizzie> Of course it counts minor and wishlist and other such things.
19:56:01 <fizzie> "Outstanding bugs -- Wishlist items; Unclassified (243 bugs)" is the largest single category for apt.
19:56:34 -!- sebbu2 has changed nick to sebbu.
19:56:40 <fizzie> The 7 other packages in the top-10 bugs list are aptitude, debbugs, dpkg, emacs21, evolution, iceweasel and linux-2.6. Pretty popular stuff.
19:57:01 <Deewiant> I'd've expected xorg in that list
19:57:22 <fizzie> Deewiant: It's split in too many parts, I think. :p
19:58:20 <fizzie> 952 bugs in everything matching /xorg/, that would make it replace apt in the top-3.
19:58:35 <fizzie> Though not if counting apt+aptitude+dpkg together.
19:59:50 <fizzie> 227 bugs in xserver-xorg-core, 178 bugs in xserver-xorg-video-intel; those are the only two three-digit numbers there.
20:00:37 <fizzie> There's a funky graph at http://bugs.debian.org/release-critical/
20:18:01 <pikhq_> Fried rice you made yourself is quite delicious.
20:18:28 <ehird> Not making it yourself inherently reduces its flavour. True story.
20:18:39 -!- pikhq_ has changed nick to pikhq.
20:19:03 <pikhq> Restaurants like covering it in soy sauce. Far too much salt, glutamates.
20:19:15 <ehird> What. I love soy sauce.
20:19:31 <pikhq> Some soy sauce adds nice flavor.
20:19:44 -!- M0ny has quit.
20:19:48 <pikhq> Adding cups of soy sauce makes for a dish saltier than potato chips.
20:22:03 <Deewiant> http://asset.soup.io/asset/0426/5459_3af7.jpeg
20:23:52 <ehird> Deewiant: that was... almost funny :\
20:24:16 <Deewiant> Just not funny, or something more
20:24:33 <ehird> theoretically it should have been funny
20:36:51 -!- boily has joined.
20:38:27 -!- sebbu2 has joined.
20:38:31 <Ilari> E621... a.k.a. MSG... Wasn't it causative for overweight in animal experiments and associated with overweight in humans? :-)
20:38:31 -!- ehird has left (?).
20:38:36 -!- ehird has joined.
20:39:06 <ehird> Ilari: what's your point
20:40:02 <Ilari> To that "Far too much salt, glutamates". And who cares about salt (apart from bad taste) unless one is exceptionally sensitive to sodium.
20:40:40 <Ilari> Statistically signficant != actually significant.
20:41:44 <ehird> Ilari: of course he means taste...
20:41:48 <ehird> unless you can't taste salt?
20:42:01 <pikhq> Ilari: "Bad taste" is the primary complaint.
20:42:38 <Ilari> Yeah, except that many have attitude: Eek... Salt (same type as eek... fat).
20:42:55 <pikhq> And I'm complaining about glutamates in general, not specifically MSG. (soy sauce has naturally-occuring glutamates. They overpower everything else when too much is used.)
20:43:58 <Ilari> Soy sauce proabably also has other nasty stuff in it (that screws with things like nutrient intake)...
20:44:25 <pikhq> Likely. I'm just complaining about how it tastes bad when you use tons of it, though.
20:44:33 <ehird> soy sauce is delicious though, who cares
20:45:24 <Ilari> That "assocated with"... As some research reports would say, MSG consumption is a risk factor for being overweight.
20:45:58 <pikhq> Most of the things that have MSG added have many other reasons for being bad for you.
20:46:08 -!- boily has quit ("leaving").
20:46:25 <ehird> i seem to be able to survive on just about anything
20:47:40 <Ilari> Yeah, but that MSG-weight association seems independent of that...
20:48:42 <Ilari> Yeah, most stuff in stores that contains MSG contains quite nasty stuff.
20:49:05 <ehird> i'm not sure I can physically get overweight
20:49:13 <ehird> I've been underweight my whole life no matter how much I eat
20:50:55 <Ilari> (ever wondered why GI of fructose is so low compared to glucose even if both are very closely related)?
20:51:11 <ehird> Never in my life, Ilari.
20:54:58 -!- sebbu has quit (Read error: 110 (Connection timed out)).
20:55:22 <Ilari> Glucose GI: 100 (defintion), Fructose GI: ~20.
20:55:41 <ehird> I have a feeling Ilari is involved in some sort of nutritionist field
20:58:57 <ehird> Wolfram Alpha doesn't help you care about stuff
20:59:33 <Ilari> It just seems that most of current "nutrion science" is total garbage...
21:01:50 <Ilari> Badly designed research, conclusions that do not follow from data, bad media reporting, etc...
21:11:46 <Ilari> Standard trick: Send embragoed (keep secret until publication of the actual study) press release about some study to press (of course, distorting even further from already distorted conclusions). The press then distort it even further and then all give press without reading what the actual study says.
21:17:54 <Ilari> And of course this might come on top of study design so bad that one wonders wheither incompetence is sufficient to explain it.
21:19:45 <ehird> Ilari: it's a conspiracy!
21:19:53 -!- MigoMipo has quit ("QuitIRCServerException: MigoMipo disconnected from IRC Server").
21:22:11 -!- augur has quit (Read error: 54 (Connection reset by peer)).
21:26:59 <Ilari> One of the funniest was rating various diets based on model which compared them to modern food guidelines... And one of the diets there was Atkins... One can guess how well it fared...
21:30:22 <Ilari> Oh, and then there was effects on cholesterol computed for compared diets. And let's just say that those values were not sanity-checked against real-world values...
21:31:14 -!- GregorR has quit (Remote closed the connection).
21:31:50 <ehird> Ilari: do you like research this much into absolutely everything :D
21:33:05 -!- GregorR has joined.
21:34:18 <Ilari> What I come across... I have seen lots of quite wild stuff. Including stuff that would have serious HSQ if the math would check out (I usually skip the heavy math on first reading).
21:34:46 <Ilari> Holy Shit Quotent.
21:37:16 <Ilari> Stuff like deriving Special Relativity, something close to General Relativity and some cosmological stuff (without having to invoke dark energy) out of model with *absolute* time.
21:39:19 <ehird> I need to replace the batteriiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiies eseseseseseses in this keyboard, because it keeps repeating the same key which just closed all my browser windows and did ← that.
21:39:46 <ehird> But that was amusing.
21:40:26 <Ilari> Why does that remind me of Linux doublecore keyboard bug? :-)
21:41:12 <ehird> Every key repeated or something?
21:41:33 <Ilari> Every keypress appeared twice.
21:42:29 <ehird> Deewiant: run the keyboard code on both cores
21:42:41 <Deewiant> How do you manage that accidentally
21:42:43 <ehird> so that it notifies each core about it, they both send it to the rest of the system
21:42:52 <ehird> Deewiant: corespawn(kb);corespawn(kb);
21:42:59 <ehird> or maybe it runs everything on every core by default in that section
21:43:02 <Deewiant> // OOPS, didn't mean to do that
21:43:04 <ehird> like, it's a general-stuff-for-the-cpu
21:43:10 <ehird> which each cpu core needs
21:43:15 <ehird> and they had the kb in that for single core
21:43:20 <ehird> without realising it would break with more cores
21:44:03 <Ilari> Something OS-visible that's shared between cores but not been processors in SMP configuration...
21:45:39 <fizzie> For a multi-core system you can program the APICs to sensibly route interrupts; I could believe they can be configured to deliver keyboard interrupts to both cores accidentally.
21:46:10 <Ilari> That wouldn't expain it.
21:46:28 <Ilari> (assuming the keyboard interface is similar to what PS/2 used).
21:47:03 <ehird> dual pentium III, fuck yeah
21:47:35 <Ilari> Double-IRQ would mess with it for sure, but at least PS/2 interface suppiled the real scancode data from I/O port).
21:48:17 <Ilari> The first IRQ would read the key and the second would then get nothing. Or they both would get parts and really mess things up.
21:50:01 <fizzie> There are interrupt-based USB endpoints, but I have no clue how those work really.
21:56:49 <Ilari> From /proc/interrupts: LOC: 657 698 369 (Timer), IRQ17: 367 783 122 (NIC), IRQ14: 35 914 225 (IDE), IRQ1: 8 306 061 (Keyboard), IRQ15: 170 134 (IDE), IRQ23: 1 007 (USB), IRQ16: 587 (USB), IRQ12: 323 (Mouse), IRQ0: 44 (Timer), IRQ4: 7 (Serial), IRQ6: 4 (floppy), IRQ8: 3 (RTC).
21:56:51 <ehird> in your end point.
21:59:34 <Ilari> Hmm... Racked 72 more IRQ6's backing up some source code. :-)
22:05:10 <Ilari> Hmm... Esolang that has 94 no-arguments operators (!-~) and every previous character affects decoding of the next... Would make code totally unreadable & unmaintainable.
22:06:35 <Ilari> Except that malbolge also scrambles code when running...
22:09:32 <Ilari> Probably best would be language that could be proven to be TC with non-constructive proof but that wouldn't have any easy constructive proof that it is TC.
22:11:15 <Ilari> Or where the size of programs really explode when they get more complex (but still its TC).
22:13:40 <fizzie> Even /proc/interrupts is so complicated nowadays: http://pastebin.com/m2c5bc2b4
22:14:28 <fizzie> There used to be just IRQ lines from 0 to 15 on a PC, and that was enough! Uphill both ways!
22:14:55 <Deewiant> Mine is even more complicated: http://pastebin.com/f7f0bb0ca
22:15:13 <Ilari> PCI-MSI-edge? HPET-MSI-edge?
22:15:20 <fizzie> More cores, more numbers.
22:15:36 <Deewiant> Instead I have "machine check" stuff
22:16:50 <fizzie> MSI is that PCI/PCIE message-signaled-interrupt thing, with the device sending a message over the bus to cause an interrupt, instead of having physical lines for that.
22:16:59 <Ilari> Those are just polls. Actual machine check events are worse.
22:17:33 -!- kar8nga has quit (Read error: 110 (Connection timed out)).
22:18:34 <fizzie> Oh, and "lspci"; at one point there was this one PCI bus, but now lspci lists five.
22:19:20 <Deewiant> I have three vacant PCI Express ports and one vacant PCI bridge
22:19:36 <Deewiant> Oh, and a PCI Express root port, whatever that is.
22:20:58 -!- ehird has quit.
22:21:02 <fizzie> http://pastebin.com/m18086ab0 -- I would like to, for example, know why the bt878 card has a different bus number there.
22:21:10 <Ilari> This computer has 4 USB ports. And according to lspci, 3 USB1 UCHI controllers plus one USB2 ECHI controller...
22:21:38 <fizzie> According to "lspci -t" that bus 04 is connected via that 00:14.4 "PCI to PCI Bridge".
22:22:27 <fizzie> And buses 01, 02 and 03 go through the more reasonable PCI bridges 00:01/02/06 up there; they even have some sort of labels there.
22:23:54 <Ilari> PCI bridge on card itself?
22:24:16 <fizzie> I doubt it'd be "ATI Technologies Inc SBx00" then.
22:24:53 <Ilari> What that SBx00 even is?
22:25:10 <fizzie> Probably "SB600 and SB700", product numbers.
22:25:14 <fizzie> South Bridge, I guess.
22:25:39 <Deewiant> Ah yes, it's an ATI motherboard isn't it
22:25:46 -!- ehird has joined.
22:28:42 <fizzie> That Intel Atom box has a significantly shorter lspci output; much less cruft there.
22:31:11 <ehird> So what window managers are you darned kids running these days
22:32:01 <ehird> icewm would be wonderful if it had just a lil more polish
22:35:31 <Ilari> And some bugs fixed, like that "lose input focus" bug.
22:36:00 <ehird> openbox is so strange. why do you have to keep some space free to get a menu to do anything? that's a ridiculous UI
22:36:03 <Ilari> (sometimes when focused window is closed, no window gets the focus).
22:37:05 <ehird> Fact is that you have to hunt for a bit of desktop
22:37:16 <Deewiant> For the menu? Just bind it to a key
22:37:19 <ehird> Which is ridiculous: Desktop is space that should be filled with useful things
22:37:36 <ehird> Deewiant: Pressing a key then directly taskswitching to the mouse seems inefficient/
22:37:40 <Deewiant> I have it on Win-esc myself, although I do always have some desktop free in the corners next to my toolbar as well
22:37:58 <Deewiant> See, your problem is that you use the mouse ;-)
22:38:08 <Ilari> The "losing input focus" is annoying because then "next window" and "previous window" don't work. To give something focus from keyboard, one has to use window list.
22:38:27 <ehird> it's just always seemed to me like openbox is a WM designed to make pretty translucent screenshots with small fonts /shrug
22:38:48 <Deewiant> Maybe it is, but it works for me with 0% translucency. :-P
22:38:56 <Ilari> Main WM menu is Capslock+M here. :-)
22:39:29 <ehird> i wish the automatic window managers actually made an attempt at intelligently managing them
22:39:35 <ehird> instead of "hurr, one to the left rest to the right"
22:39:39 <ehird> or "hurr, fibonacci spiral"
22:42:18 <Ilari> At least the windows don't spin by themselves...
22:42:45 <Ilari> One WM had key combo which made current window start to spin...
22:44:12 <Ilari> The same WM also had stuff like window rotation with three degrees of freedom, window zooming (seperate from resizing) and that...
22:44:37 <fizzie> At some point Enlightenment was the WM that seemed most focused on the pretty tricks.
22:45:11 <ehird> enlightenment has always seemed crappy to me
22:45:35 <ehird> i wonder if E17 is out yet :P
22:45:47 <fizzie> "Enlightenment allows you to have a grid of workspaces called virtual desktops. Switching between them is achieved by hurling the mouse cursor to the edge of the screen, at which the desktop appears to slide across to reveal the next. The maximum grid size is currently 8 by 8 desktops, and you can have 32 grids (each with a different background), making 2048 total possible desktop spaces."
22:46:07 <ehird> fizzie: 2048 desktops, fuck yeah
22:46:30 <ehird> (Users can enable a map of the desktops, in case they get lost, which is called the pager.)
22:46:37 <ehird> Enlightenment: So you can get lost in your desktops.
22:46:54 <Ilari> 2048 virtual desktops? I only have 11. :-(
22:47:17 <fizzie> "DR16 has been the choice of power users and artists due to its low overhead, highly graphical, widely theme-able, extremely configurable, yet unobtrusive interface."
22:47:27 <Ilari> (named "F2"-"F12").
22:48:00 <Ilari> Guess why there isn't "F1".
22:49:25 <Ilari> Nope, CAPSLOCK+SHIFT+Fx hits three-key same-row/column restriction with x = 1.
22:49:58 <ehird> on what keyboard is capslock+shift+fx comfortable
22:50:02 <Ilari> That is, one can't hit Capslock+Shift+F1.
22:50:19 <ehird> sure you can, with difficulty :)
22:50:33 <fizzie> From a suitably skewed perspective I could claim to have 216 virtual desktops, since there's six independently selectable tags on all three monitors, and 6^3 = 216. Though from an even more wonky perspective I guess I could claim 262144, since you can select any combination of tags, and (2^6)^3 = 262144.
22:50:45 <Ilari> Capslock+Shift+F1 just doesn't work.
22:51:49 <fizzie> I guess he means with a "on this keyboard" qualifier.
22:51:53 <ehird> i mean 'pataphysically?
22:52:14 <Ilari> If one hits three keys at once, the keys must either be in distinct rows in scan matrix or in distinct columns in scan matrix.
22:52:24 <Ilari> Otherwise the third key won't register.
22:52:33 <ehird> i don't think that is true with my keyboard, at least...
22:53:54 <Ilari> How the keys are located in scan matrix may vary depending on keyboard. But for any keyboard that uses scan matrix, that three-key restriction appiles.
22:54:13 <ehird> ó_ó I've never heard about this before
22:54:18 <fizzie> They do make keyboards completely without that problem; and I think quite many keyboards put the modifier keys "independently" recognizable.
22:54:37 <ehird> i guess the more expensive keyboards don't do that.
22:55:41 <fizzie> I think it depends on what sort of "expensive" it is; if it's just expensive-because-it-has-stylish-design they might not care. It's not like it hurts "normal use", especially if the modifier keys always work.
22:56:12 <fizzie> But the key-jamming problem is not a new one; Star Control came with a keyboard testing thing you could use to pick up keys so that two people can play the melee game without problems.
22:56:14 <Ilari> Pretty much the only use where the scan matrix layout really matters is hotseat multiplayer games.
22:56:31 <ehird> I mean like mechanical keyswitches
22:57:16 <fizzie> I don't think full n-key rollover is very common even in expensive-like-that keyboards, but I don't have any empirical data on this.
22:57:22 <fizzie> http://en.wikipedia.org/wiki/Rollover_%28key%29 is about it.
22:57:26 <ehird> It's not related to n-key rollover, is it?
22:57:37 <ehird> That's just being able to hit a bunch of keys at once, not neccessarily in the same row/column.
22:57:44 <Ilari> At least on this keyboard, left shift is part of normal scan matrix. Capslock+F1 is possible, Capslock+Shift+F1 isn't.
22:59:20 <fizzie> The reason you can't hit multiple keys is because of the matrix-style design. And it makes it sometimes rather nontrivial to find out which keys you can do.
23:00:01 <ehird> I've never had a rollover problem
23:00:05 <fizzie> On this logi-illu one, I can hit 'q', 'w', 'e' and 'r' successfully, but if I keep those four pressed 't' won't register any more.
23:00:11 <ehird> I type at like 100wpm, but only ever hit one or two keys at a time.
23:00:51 <Ilari> On this keyboard, two key columns in main block (alphabet, numbers & co) are on same scan matrix column.
23:01:11 <fizzie> It's not like it's an issue when typing. But when you have one person using wasd and a couple of "fire" and "special" buttons controlling one car, and another one with the cursor keys and another couple of buttons, it's trickier.
23:02:00 <fizzie> On the second row I apparently can do 'a', 's', 'd', 'f' and 'g', but with all those down 'h' will no longer register.
23:02:16 <ehird> fizzie: 3-key rollover is considered essential for English touch typing. [3]
23:02:27 <ehird> which doesn't make any sense to me
23:02:51 <Ilari> Here second row registers only up to 'f'. 'g' won't register anymore.
23:03:35 <fizzie> I'm sure the mechanical design does come to play too, when it comes to the speed how fast keys rise up and so on. Can't say I've had trouble *typing* with even the cheapest keyboards I've used.
23:03:58 <Ilari> Typing so fast that previous key doesn't have time to register as relased before next is pressed?
23:05:17 <fizzie> This is certainly no gaming keyboard, I can't seem to find any two wasd-shaped sets of keys where you could successfully record all eight simultaneous keys; it just goes up to six.
23:06:10 <ehird> I have never tried to press eight keys in a game at once.
23:06:16 <Ilari> If keys register on positive edge, and there is no ghosting/jamming, then release delay (if not enough to cause false autorepeat), won't usually be problematic.
23:06:32 <fizzie> ehird: Maybe you haven't been multiplayering much with a single keyboard, then?
23:06:59 <ehird> I generally use crazy invention like : Inter - net!
23:07:36 <ais523> fizzie: meh, hjkl-shaped > wasd-shaped
23:08:16 <fizzie> ais523: I'm not sure that's true for gaming, especially if the game is usually played with cursor keys and the second player just needs a suitable replacement.
23:08:30 <fizzie> Inter net! But it's so expensive, and my parents don't want me to keep the phone line busy all the time, and my 2400 bps modem gets so bad latency, and... well, no game supports that sort of thing anyway.
23:08:41 <ehird> wasd is perfectly intuitive, hjkl is just being contrarian for the fuck of it.
23:09:04 <ehird> fizzie: just hook up a modem and dial a friend running a Doom game
23:09:13 <ehird> you can use ordinary telephone numbers!
23:09:23 <Ilari> The second in main block is usually ijkl, with actions being something like 'b'.
23:09:45 <fizzie> I did that, but it's still expensive, and has the phone-line-busy problem, and anyway only a tiny fraction of games supported modem-based multiplayer.
23:09:48 <Ilari> (first being wsad, with actions being tabs, and various left modifiers).
23:10:26 <ehird> fizzie: just run a looooooong ps/2 cable to your friend
23:10:29 <ehird> to another keyboard
23:11:46 <fizzie> Warcraft 2 did, and ROTT, and Descent, and obviously Doom; but none of the cave-flying games like kops, v-wing, luola; eh, I guess that genre is not so widespread outside Finland, really.
23:12:27 <fizzie> Slicks'n'Slide, too... actually I think that's a Finnish invention too.
23:12:35 <ehird> what were the game mechanics of these... cave-flying ames
23:12:42 <ehird> http://www.luolamies.org/software/luola/screens/screen6.png ;; looks great
23:12:45 <ehird> did you just lik... fly
23:13:03 <fizzie> Yes, and shoot your friends with a multitude of weapons.
23:13:30 <Ilari> (at least if you still have flyable ship)...
23:13:41 -!- FireFly has quit ("Later").
23:13:50 <ehird> fizzie: oh, and if you hit things you die?
23:13:52 <fizzie> Yes; v-wing lets you continue with just your "guy" when the ship goes boom.
23:14:11 <fizzie> Depending on settings, but usually bumping into things does take up some of your health.
23:14:20 <fizzie> And the levels are usually bitmap files (or a few of them) so there's quite a variety of levels available.
23:14:37 <coppro> we should just all play falling sand
23:14:38 <Ilari> The ship from Luola has ejector seat...
23:14:57 <ehird> falling sand kind of doesn't have any game mechanics
23:15:26 <coppro> but there are other versions
23:15:35 -!- BeholdMyGlory has quit (Remote closed the connection).
23:15:42 <Ilari> And the standard equipment for player moving on foot includes parachute and ninja rope...
23:15:53 <Ilari> Plus some kind of peashooter.
23:15:59 <ehird> Ilari: ninja rope stolen from Worms or the other way around?
23:16:21 <coppro> stupid make is slowing my computer down
23:16:33 <ehird> What you need, when you need it
23:16:51 <Ilari> Stolen poorly from Worms. The one in Worms is much more useful.
23:17:07 <ehird> Ilari: well, the whole flying thing is basically identical to the super sheep in worms ofc
23:17:15 <ehird> except... shooting
23:17:23 <ehird> but i doubt that was stolen
23:17:27 <ehird> since it's pretty obvious
23:18:05 <ehird> it has no game mechanics.
23:19:03 <Ilari> Also in Luola if ship got destroyed and the player got to docking pad on foot, one got new ship...
23:19:39 * ehird downloads this luola thing in the hope it has single-player AI
23:19:57 <Ilari> There is also modes where players start on foot and level has N (or N-1 in another) ships scattered around.
23:21:19 <fizzie> There's also that "N players run around with Worms-style ninja-rope and shoot the others" game; or I think at least two of them. Don't remember the names.
23:22:28 <ehird> that's called "Worms"
23:23:13 <Ilari> Liero was ~first such game with 2-players. I have also played one such game over the net (back when I had 56K, which was actually 33.6k modem).
23:23:38 <ehird> Oh, Liero is ridiculously fun
23:24:38 <Ilari> I have also played Liero on Xbox (using the standard Xbox controllers). Didn't run very well.
23:24:54 <ehird> Liero's ninja rope is a bitch, though
23:25:11 <ehird> in fact all ninja ropes apart from Worms 2 onward seem to really suck
23:25:48 <Ilari> The one in Luola is probably the worst I have seen.
23:26:51 <fizzie> Right, Liero and some clone of it.
23:27:28 <Ilari> Aren't there at least 3 OSS one that allow N players? Soldat, NiL and Bankiz...
23:27:46 <fizzie> Oh, right MoleZ; and it was in fact the one that came before Liero.
23:28:36 <fizzie> http://en.wikipedia.org/wiki/Liero -- wikipedia knows all this important information.
23:29:46 <Ilari> And sometimes programming course has project subject to make (simple) Liero clone... Some years ago one such course produced two more liero clones...
23:30:21 <Ilari> (probably not found anywhere anymore)...
23:30:29 <ehird> Aw, Luola has no AI.
23:30:35 <ehird> Or networked multiplayer.
23:31:02 <fizzie> What that means: you need some real friends. :p
23:31:39 <fizzie> Kops is open source nowadays, too: http://jet.ro/kops/ -- it's a bit different from the other cave-fliers, though, may be an acquired taste.
23:31:44 <ehird> fizzie: (as opposed to all the chatbots I talk to in here?)
23:32:28 <fizzie> (But Kops won't have an AI either, or networked multiplayer for that matter.)
23:33:26 <fizzie> Kops is also the reason why I wrote that "VNC client with keyboard only", since we needed to play it with two keyboards to avoid the multiple-keys issues.
23:35:03 <Ilari> What, Xinput won't let you treat another keyboard as just another input device? :-)
23:35:31 <fizzie> There was just the DOS version back then.
23:35:56 <fizzie> But it ran under Windows, even with a VNC server running.
23:36:00 <fizzie> Had some palette issues, though.
23:37:42 <fizzie> It's easy nowadays since you can just stick in more and more USB keyboards.
23:37:53 <ehird> INFINITE DAISY-CHAINED USB KEYBOARDS
23:38:15 <fizzie> Not that I know whether Windows handles multiple USB keyboards, but one can always hope.
23:38:47 <Ilari> For those that like to do the same with mice, there's MPX... :-)
23:39:54 <fizzie> You can also just have multiple single-pointer-controlling devices and have the players fight with it.
23:40:03 <ehird> MPX + pointer warping = Trippy, man.