--- Log opened Mon Feb 26 00:00:21 2007 00:03 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)] 00:03 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2 00:12 -!- sepp [n=sepp@p213.54.58.134.tisdip.tiscali.de] has quit [Read error: 110 (Connection timed out)] 01:14 -!- Shingoshi [n=root@c-24-21-4-193.hsd1.mn.comcast.net] has joined #t2 01:16 < Shingoshi> Hello everyone! 01:17 < Shingoshi> ChanServ: I see that you also use or at least monitor Gentoo-amd64 as well. 01:19 < Shingoshi> ChanServ: Do you use t2 as a build environment for gentoo64? 01:22 < Shingoshi> Wait a minute. Is ChanServ a real person? 01:36 -!- mtr_ [n=Michael@kobz-590ca49f.pool.einsundeins.de] has joined #t2 01:52 -!- mtr [n=Michael@kobz-590cb18c.pool.einsundeins.de] has quit [Read error: 113 (No route to host)] 01:52 -!- mtr_ is now known as mtr 02:05 < sepp_> ChanServ is a bot 02:05 -!- sepp_ is now known as sepp 02:06 < sepp> hi Shingoshi :) 03:08 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)] 03:08 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2 03:12 -!- kensai [n=kensai@206.248.106.75] has joined #t2 03:25 -!- kensai [n=kensai@206.248.106.75] has quit ["Leaving"] 03:54 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2 03:56 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)] 05:41 -!- Shingoshi [n=root@c-24-21-4-193.hsd1.mn.comcast.net] has quit [Read error: 60 (Operation timed out)] 06:13 -!- Shingoshi [n=root@c-24-21-4-193.hsd1.or.comcast.net] has joined #t2 06:13 < Shingoshi> Hi sepp 06:29 < Shingoshi> sepp: My connection keeps breaking. So I hope if you write again, I will catch it. 07:37 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2 07:37 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)] 07:54 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has joined #t2 07:55 -!- emte__ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)] 08:17 < sepp> moin :) 08:35 < Shingoshi> So, how are you involved with this project? 08:36 < sepp> how, uhm, i maintain some packages 08:37 < Shingoshi> That's good to know. 08:37 < Shingoshi> Which pkgs are you maintaining? 08:39 < sepp> the website means "graphic packages, E17" 08:39 < sepp> :) 08:39 < CIA-8> susan * r22722 /trunk/package/develop/intltool/intltool.desc: * updated intltool (0.35.4 -> 0.35.5) 08:40 < CIA-8> susan * r22723 /trunk/package/network/pound/pound.desc: * updated pound (2.2.4 -> 2.2.5) 08:40 < Shingoshi> I am not familiar with E17. Is it gnome related? 08:40 < Shingoshi> Is that enlightenment? 08:41 < sepp> yup 08:41 < Shingoshi> ok. 08:41 < sepp> version 17 08:41 < Shingoshi> That's what I thought/ 08:41 < Shingoshi> I use KDE. Abondoned gnome a long time ago. 08:42 < sepp> i use e17 ... 08:42 < Shingoshi> I ran a couple of live cd's and dvd's. Most of them had gnome. 08:43 < Shingoshi> I did that today. 08:44 < Shingoshi> I really liked the Knoppix desktop. Very pretty. 08:44 < sepp> you can make your own livecd with anything you want on it - with t2 08:45 < Shingoshi> It is gnome. That's why I am installing t2. 08:46 < Shingoshi> I am still downloading opensde by svn. It looks like it is a derivate of t2. 08:46 < sepp> yup 08:46 < Shingoshi> Familiar? 08:46 < CIA-8> susan * r22724 /trunk/package/network/irssi/irssi.desc: * updated irssi (0.8.10a -> 0.8.11-rc1) 08:46 < Shingoshi> YOu beat me to it. 08:47 < sepp> :) 08:48 < Shingoshi> opensde is a very large download! 08:48 < Shingoshi> Just how large is the site. 08:50 < sepp> ? 08:50 < Shingoshi> It's been downloading for awhile. 08:51 < Shingoshi> Just curious how large opensde's svn repositoty is. 08:51 < rxr> re 08:51 < Shingoshi> re? 08:51 < sepp> moin rxr :) 08:52 < sepp> like in reply i guess 08:52 < Shingoshi> What language is primarily used here? 08:52 < rxr> in IRC? 08:52 < sepp> some sort of english :) 08:52 < rxr> en 08:53 [Users #t2] 08:53 [@ChanServ] [ CIA-8] [ idealm] [ mtr ] [ rxr ] [ Shingoshi] 08:53 [ Capey ] [ emte_] [ LMJ ] [ R4gnar0k] [ sepp] [ valentin ] 08:53 -!- Irssi: #t2: Total of 12 nicks [1 ops, 0 halfops, 0 voices, 11 normal] 08:53 < sepp> broken irc english :-D 08:53 < rxr> en_IRC 08:53 < sepp> lol 08:53 < Shingoshi> That's ok. Most people who speak english, barely do so! 08:54 < Shingoshi> At least in the US. 08:55 < Shingoshi> Finally. It seems to be done. 08:56 < Shingoshi> How long have each of you been with the project? 08:56 < sepp> with t2? 08:57 < Shingoshi> yes. 08:57 < sepp> maybe 2 years, i don't know exactly 08:58 * rxr from the beginning 08:58 [Users #t2] 08:58 [@ChanServ] [ CIA-8] [ idealm] [ mtr ] [ rxr ] [ Shingoshi] 08:58 [ Capey ] [ emte_] [ LMJ ] [ R4gnar0k] [ sepp] [ valentin ] 08:58 -!- Irssi: #t2: Total of 12 nicks [1 ops, 0 halfops, 0 voices, 11 normal] 08:58 < Shingoshi> Good. 08:59 < sepp> i dont like messy distros with kde or gnome and silly binary dependencies, so i use t2 to make my own with only the packages i want ... 09:00 < sepp> that is all :) 09:00 < rxr> Shingoshi: well, the copied ours and started munching it ... 09:01 < rxr> oh - sorry - was a bit up in the scrollbacked ... 09:04 < Shingoshi> Still running du to see what the size is. Must be very large though. 09:05 < Shingoshi> rxr: Did you mean opensde about copying? 09:07 < rxr> yes, it's a 1:1 copy with some munching which stalled recently anyway 09:07 < Shingoshi> opensde is 1.2Gbs. Damn that is large! 09:07 < rxr> that is I do not see much changes going on ... 09:07 < rxr> Shingoshi: you are definetly not running the correct command 09:07 < rxr> it's the same size as t2 09:07 < sepp> about 40mb 09:07 < rxr> -- some thousand revisions of fixes and updates ... 09:08 < CIA-8> rene * r22725 /trunk/package/printing/cups/cups.desc: * updated cups (1.2.7 -> 1.2.8) 09:08 < Shingoshi> I don't think so. I ran svn opensde* site. It downloaded the entire site. 09:10 < rxr> sepp: maybe 100MB 09:10 < Shingoshi> This is the exact command that I ran. svn checkout svn://svn.opensde.net/opensde site 09:10 < rxr> at least my local tree is about 100MB, including the .svn copy ... 09:10 < sepp> i exported my t2 tree and it has 43mb :) 09:10 < rxr> sepp: ah, exported, yes :-) 09:10 < sepp> without all the .svn noise 09:10 < rxr> Shingoshi: I think this checks out some branches and such 09:11 < Shingoshi> I guess. 1.2Gigs worth. 09:11 < rxr> yeah, copies of copies of copies ... 09:12 < rxr> wow - yesterday night an ebay got away on ebay - for 748 USD ... 09:12 < rxr> BeBox Dual PowerPC 603e 133Mhz, Coolest Computer Ever! 09:12 < rxr> Andere Artikel des Verkäufers 09:12 < rxr> US $748,23 09:12 < rxr> err 09:12 < rxr> http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&ih=002&sspagename=STRK%3AMEWA%3AIT&viewitem=&item=120088851732&rd=1&rd=1 09:13 < sepp> lol 09:13 < rxr> I also thought about doing a bit, but that is too much .. 09:13 < sepp> the one with the neat leds infront? 09:13 < rxr> yes, the geek box ... 09:13 < rxr> rather grab some alpha or ia64 then and improve that T2 areas 09:14 < rxr> got enough PPCs already .-) 09:14 < rxr> though it would definetly would have done a nice job in my computer museum :-) 09:15 < sepp> beos 5 was neat 09:18 < Shingoshi> rxr: There are only four dirs in the master dir. I have not gone that deep. But I don't see any copies. 09:19 < rxr> you need to checkout opensde/trunk 09:19 < rxr> AFAIR 09:20 < Shingoshi> I mean there are branches in most, but I don't think they are copies. 09:20 < rxr> in opensde/ there are some branches and thus multiple copies of the trunk - each spanning 40MB + 40MB .svn meta data 09:20 < rxr> a branch is always a copy of some other branch in the beginning ... 09:20 < Shingoshi> ok 09:21 < rxr> and in the opensde case most branches are dead ends with just 5 changes done to them and then leaving lingering around 09:21 < rxr> believe me, I monitor their T2 copy closely ... 09:21 < Shingoshi> So how can I clean this up? 09:21 < rxr> rm -rf :-) 09:23 < Shingoshi> yesh. But what would I remove? 09:24 < rxr> all ? 09:24 < rxr> well ok 09:24 < rxr> if you accidently have the opensde/trunk you can move that to opensde-trunk and rm -rf just the rest 09:25 < Shingoshi> Let me look. 09:26 < Shingoshi> opensde/opensde/trunk is 106Mbs. 09:29 < rxr> yes, as T2 ... 09:30 < rxr> quite funny to give opensde extraction instructions in #t2 ... 09:32 < Shingoshi> Well when I saw opensde, I thought it was associated with t2. So it is, indirectly. 09:32 < Shingoshi> I am running ls -R ./* > opensde.ls to see all of the dirs in opensde. 09:33 < rxr> Maybe I should document the othersde fork storry in my blog once and for all ... 09:33 < rxr> story it was .. 09:33 * rxr wondery why I always mistype story in the first try ... 09:34 < CIA-8> rene * r22726 /trunk/package/graphic/imagemagick/imagemagick.desc: * updated imagemagick (6.3.2-7 -> 6.3.2-9) 09:37 < Shingoshi> Don't feel bad. I do know how to type. And lately, I keep mistyping, directory. 09:38 < Shingoshi> Could be sleep deprivation! 09:39 < Shingoshi> Ok. This question pertains to t2. Well sort of. 09:40 < rxr> Shingoshi: hm? 09:40 < Shingoshi> I want to package all of my binaries as .tlz files. To be compatibale with tukaani's pkgtools. 09:40 < rxr> from the glibc list: 09:40 < rxr> "> I profiled make. It's spending around 60% of the time in 09:40 < rxr> > new_pattern_rule(), which does a linear search through the list of 09:40 < rxr> > pattern rules to check for duplicate rules. glibc generates ~2500 09:40 < rxr> > rules (in sysd-rules). 09:40 < rxr> Holy moly! How in the world do you get that many pattern rules?!?! The 09:40 < rxr> point of pattern rules is that they represent an entire class of 09:40 < rxr> targets, which means you typically would have orders of magnitude fewer 09:40 < rxr> pattern rules than you have targets. 09:40 < rxr> I don't have anything against making this more efficient, I'm just... 09:40 < rxr> surprised." 09:41 < rxr> .tlz == .tar.lzo ? 09:41 < rxr> yes, T2 can easily do that, it's even implemented, though not yet well tested 09:41 < rxr> that is we always just use .tar.bz2 and .tar.gz 09:42 < Shingoshi> Based on what I have read, .tlz seems to be the most advanced tech for compression. 09:43 < rxr> ah, maybe they us lza? 09:43 < rxr> because lzo is only fast, not compressing welll 09:43 < rxr> lzma ... 09:43 < Shingoshi> Here is my crazy idea. lzma. 09:43 < Shingoshi> Let me tell you what I intend to do. 09:43 < rxr> well, the most advancedness depends on multipl factors 09:44 < rxr> I thnk rzip (or what is was) even compresses a bit better 09:44 < Shingoshi> My dream is to create a clustered LTSP system. 09:45 < Shingoshi> I want only one application server in any location. Although, it can be replicated for security. 09:45 < Shingoshi> Every node in the cluster would get their apps from the master node. 09:46 < Shingoshi> And all of them would comtribute their processing power to the master node. 09:46 < Shingoshi> So I want to create the largest set of applications possible on the master node. 09:47 < Shingoshi> And then serve those apps throughout the entire network. 09:47 < Shingoshi> done 09:47 < CIA-8> rene * r22727 /trunk/package/security/vpnc/vpnc.desc: * updated vpnc (0.3.3 -> 0.4.0) 09:48 < rxr> sounds like a good plan 09:49 < Shingoshi> With this in mind, I also need to have as many architectures supported on the master server as well. 09:51 < Shingoshi> So, given these parameters, I think I should keep the opensde tree in it's entirety. 09:52 < rxr> Shingoshi: I would welcome you :-) 09:52 < rxr> do you already have an idea how to share the processing power ? 09:52 < Shingoshi> Exactly how? 09:53 < Shingoshi> Openmosix 09:53 < Shingoshi> or openssi 09:54 < Shingoshi> There are not enough developers on openmosix. 09:54 < rxr> valentin has some not so good experiene with an openmosix setup at his university 09:55 < Shingoshi> They also seem a bit disorganized. But I may be premature in my assessment. 09:55 < rxr> I think their setup locks up from time to time 09:55 < rxr> or at least did 09:55 < rxr> did not heared about new problems the last months++ 09:56 < Shingoshi> Yeah, and it crashes with 2.6 or something. 09:56 < rxr> nah, they are on 2.4 I think 09:56 < rxr> that cluster is in "production" for some years already 09:56 < Shingoshi> They have development branches of 2.6. But no releases yet. 09:57 < Shingoshi> 2.6 is only in cvs. 09:57 < Shingoshi> For OM. 09:58 < CIA-8> rene * r22728 /trunk/package/base/linux26/arch-ppc64-selected-deepnap.patch: * linux26 hotfix for powerpc64, PPC970MP to not DEEPNAP on rev1.0 09:59 < Shingoshi> What is CIA-8? 09:59 < rxr> it is open source informat project 09:59 < rxr> you can submit the SVN/CVS/... activity and either view HTML stats 09:59 < Shingoshi> Is it a bot also. 09:59 < rxr> and also select a IRC bot to join a channel and notify about changes there 10:00 < rxr> this is our SVN activity scrolling by ... :-) 10:00 < rxr> http://cia.navi.cx/stats/project/t2 10:02 < Shingoshi> What type of machine are you running? 10:02 < rxr> the one the patch is for :-) 10:02 < Shingoshi> Me. Opteron 2210 10:02 < rxr> well I run all kind of exotic stuff 10:03 < rxr> the patch is for the CPU my the dual-core G5 10:03 < Shingoshi> Damned you! LOL!!! 10:03 < rxr> I also have a Sgi Octane and UltraSPARC under my desk 10:03 < rxr> and a Core2Duo MacBook v2 ... 10:03 < rxr> and recently I got a AVR32 and Blackfin board on my desk ... 10:03 < rxr> you noticed that I'm the t2 founder? 10:04 < rxr> (and CTO of http://www.exactcode.de/) 10:04 < Shingoshi> And a Ferrari in the garage? 10:05 < rxr> no not yet 10:05 < Shingoshi> So I guess that's a yes for the Ferrari! 10:05 < rxr> nope :-( 10:06 < Shingoshi> I am still LOL!! 10:06 < rxr> if I would have that much money to "waste" I would rather pick something else I think ... 10:06 < Shingoshi> Murcielago? 10:06 < rxr> well, the cooperate backing is rather new, the last 6 years I most did OpenSource for the pure fun of it ... 10:06 < Shingoshi> Bugatti? 10:07 < rxr> lamborghini 10:07 < Shingoshi> That would be the Murcielago! 10:07 < rxr> I recently had been near the 1001PS Bugatti, though 10:07 < rxr> I know a desin person who did the PR material for the 1001PS Bugatti advertising ... 10:08 < Shingoshi> Haven't seen it. At least I don't think so. 10:08 < rxr> it's pretty new 10:08 < Shingoshi> 16clys like the other one? 10:08 < rxr> http://images.google.com/images?q=1001%20PS%20bugatti&ie=UTF-8&oe=UTF-8&um=1&sa=N&tab=wi 10:09 < Shingoshi> ok. I am gawking! 10:09 < rxr> IIRC the thing costs 1.000.000 EUR 10:09 < rxr> so 1.5xx.xxx US$ ... 10:09 < rxr> you are in the US, right ? 10:09 < sepp> gosh, it is just a car 10:09 < sepp> can it fly? 10:09 < rxr> no, but it has 1001 PS 10:09 < Shingoshi> Ok. I have seen it. Yes. 10:09 < rxr> maxe out around 400 km/h 10:10 < sepp> not on the street in front of my house 10:10 < Shingoshi> Ok, call me ignorant. What is 1001ps? 10:10 < sepp> there you can not even drive 30km/h on a bike :) 10:11 < rxr> PS 10:11 < Shingoshi> A measurement? 10:11 < rxr> horse powers ... 10:11 < rxr> was'nt that used in the US? I mean the new meassurement is kW ... 10:12 < Shingoshi> I think there is not a direct relationship between the two. 10:12 < Shingoshi> I am not sure if a kW is more or less than a HP. 10:13 < rxr> http://en.wikipedia.org/wiki/Horsepower 10:13 < rxr> "In scientific discourse the term "horsepower" is rarely used because of the various definitions and the existence of an SI unit for power, the watt (W)." 10:14 < Shingoshi> 400km/h @+ 250MPH? 10:14 < Shingoshi> 400km/h @= 250MPH? 10:15 < rxr> Shingoshi: this MPH thing is unintuitive for me, I would need to dig out a calculator widget to tell exactly ... 10:15 < rxr> 1 PS = 0.736 kW 10:15 < Shingoshi> Are you a mathematician? 10:16 < Shingoshi> I think it is .625Miles/Km. 10:17 < rxr> Over the last few days the Bugatti Veyron has achieved a top speed of over 400 Km/h (248.5 miles/h). 10:17 < rxr> ^- from google 10:17 < rxr> somehow I did not managed to get the conversion out of the google calculator ... 10:17 < Shingoshi> So I was right. 10:17 < rxr> but at least it offered a web hit below :-) 10:19 < CIA-8> rene * r22729 /trunk/package/x86/yasm/yasm.desc: * updated yasm (0.5.0 -> 0.6.0) 10:19 < Shingoshi> Wow! I see what you mean about HP. 10:19 < CIA-8> rene * r22730 /trunk/package/perl/perl-cairo/perl-cairo.desc: * updated perl-cairo (1.022 -> 1.023) 10:19 < CIA-8> rene * r22731 /trunk/package/perl/perl-dbi/perl-dbi.desc: * updated perl-dbi (1.53 -> 1.54) 10:20 < rxr> Shingoshi: Yeah, in germany its abbreviated PS, forgot that it's HP in the english language :-) 10:20 < rxr> well this /toy/ is quite powerful, costs a fortune and propably eats gasoline like crazzy :-) 10:21 < Shingoshi> Now, I know that you told me to rm -f most of the opensde. But now that I explained what I am up to, does that still seem like the right thing to do. I am not short on disk space. 10:21 < CIA-8> rene * r22732 /trunk/package/network/bittorrent/bittorrent.desc: * updated bittorrent (5.0.5 -> 5.0.6) 10:21 < rxr> well, opensde is not better for the task than t2 10:22 < rxr> and recently the commits in opensde rather scalled down to a bare minimum 10:22 < rxr> T2 is way more updates and fixes ahead 10:22 < Shingoshi> It wasn't a question of better. Only if the bloat may serve my purpose? 10:23 < rxr> and also they only support x86*, and kindly ask for hardware donation to take a look at the other, T2 just builds fine on all we list out, that is all the exotic hardware people are at T2 ... 10:23 < Shingoshi> ok. 10:23 < rxr> (+ av32 and blackfin support beeing added right now) 10:24 < Shingoshi> So t2 is pluralistic. And opensde is not? 10:24 < rxr> they want to support the other architectures as well, just noone has those hardware right now 10:24 < rxr> and currently there are few commits at opensde anyway 10:24 < Shingoshi> Why are you separate entities? 10:24 < rxr> where T2 is gaining new users and developers on a daily basis 10:25 < rxr> and we even signed a book contract so there will be a T2 book in the shelfs in Q3 about T2 10:25 < rxr> well - to be honest I do not know 10:25 < rxr> I can repeat the storry once more (I guess I really should blog it) 10:25 < Shingoshi> I am really tired of all the fragmentation. I want it to end! 10:25 < rxr> we had this nice contributor from chile, and when I needed a new employee we wanted to get into EU anyway 10:26 < rxr> so we gave him a try, payed the ticket, showed him the world largest computer fair (CeBIT) and then tried to work with him the next 3 weeks 10:26 < rxr> (that was in last summer= 10:26 < rxr> ) 10:26 < Shingoshi> So is this political? 10:27 < rxr> however in that 3 weeks hi did not contribute any noticeable line of code and did not finished any of the points TODO from his list 10:27 < Shingoshi> Well, I am not asking for a job. 10:28 -!- sparc-kly [n=mubex@adsl-72-50-101-246.prtc.net] has joined #t2 10:28 < rxr> so we politely told him that this does not work out and wanted to hand him the ticket back to chile, however he took that as a personal affront or whatever, stayed in germany and later forked t2 into opensde under the flag "technical reasons" though I can not find any 10:28 < Shingoshi> I am just a computer junkie who sits for hours in front of my computer compiling pkgs. 10:29 < rxr> yeah, most T2 people are not related to ExactCODE 10:29 < rxr> ExactCODE is just one company utilizing it for products, so are many others ... 10:29 < rxr> no problem with that 10:30 < Shingoshi> Ok. So that was your explanation about the forking of opensde. I can see where you might be a bit bitter! 10:30 < rxr> oh, forgot to mention that before we flight that person to germany we sent him money for two montly flat rents when he was short as well as a chritmas package with some Athlon board, cookies and stuff for his childs 10:30 < rxr> anyway sas storry end, opensde is not seeing many SVN commits anyway and T2 is happily growing and maturing every day 10:31 < rxr> sas => sad 10:31 < Shingoshi> I am single and live alone. I am 50 years old. I guess that explains why I am single and live alone. 10:31 -!- emte_ [n=emte@d64-180-45-14.bchsia.telus.net] has quit [Read error: 110 (Connection timed out)] 10:31 < Shingoshi> I spend all of my time in front of my computer. 10:32 < rxr> btw. here is the T2 ChangeLog: http://svn.exactcode.de/ChangeLog-t2 10:32 < rxr> just in case you want to scroll over the activity ... 10:33 < rxr> hi sparc-kly 10:33 < Shingoshi> ok. So if I use t2 exclusively. How much can I expect from it in terms of my goals? 10:33 < sparc-kly> hi rxr :) 10:33 < rxr> Shingoshi: building, including cross compiling some, packages, haveing most of the required packages ready 10:34 < rxr> (we miss openmosix and openssi so far I think, though they can be added easily) 10:34 < rxr> quite some extensive tech support here or on the list from other professional users 10:34 < rxr> and you get the package updates, e.g. of the kernel, X, and all the other stuff for free 10:35 < Shingoshi> Most of the pkgs can be as ports. I just need the base system ready to install and allow pkgs to compile on the fly. 10:35 < rxr> e.g. when you would use some own build script - you would have to hunt all the updates on your own 10:35 < Shingoshi> I was attracted to your automation. 10:35 < Shingoshi> I really like that. 10:36 < Shingoshi> I was also stressing the notion of an arch neutral source to Larhzu from the Tukaani project. 10:36 < Shingoshi> YOur's does that precisely. 10:37 < Shingoshi> I have been looking for a more automated build system, which doesn't have to be reconfigured for each arch. 10:38 < Shingoshi> I mean manually reconfigured. 10:38 < Shingoshi> I just want to run ./configure for each system and compile. 10:39 < Shingoshi> It seems that's what you do. 10:39 < rxr> yes, T2 really is a nice tool in this area 10:39 < rxr> what we really want is to be the definet solution in this area 10:39 < Shingoshi> Then I am likely here to stay! 10:39 < rxr> e.g. currently there is Gentoo for desktops or so, OpenEmbedded for embedded boards and so on 10:39 < rxr> T2 scales to all of them 10:40 < Shingoshi> Yeah. I was looking at Gentoo. 10:40 < rxr> that is it has (most, precentage increasing) of the packages you need for Gentoo style do-it-yourself desktops (like KDE, Gnome, ...) but can on the other hand cross compile most of what you usually need to cross compile for embedded use 10:40 < Shingoshi> Downloaded and chrooted a build environment of it on my system. 10:40 < rxr> ExactCODE is exactly heading in this are, cell phones, settop boxes, firewalls, etc. 10:41 < Shingoshi> Well you have my attention! 10:42 < rxr> currently T2 can cross build all you need for this stuff, the only stuff T2 can not yet cross build is all the sloppy written desktop stuff including KDE and OpenOffice and such 10:42 < Shingoshi> Going back to my original idea. How much does it interest you? 10:42 < rxr> but that is just because those packages are not written with cross compilation in mind, we could patch them to cross build properly - and we will do that at some point 10:42 < rxr> well - stronger clustering support is always nice to have 10:43 < Shingoshi> About patches, I have a question. 10:43 < rxr> we will definetly help and support you with this goal as free time permits here and on the list 10:43 < rxr> Shingoshi: yep? 10:43 < Shingoshi> Why can't patches be written like extensions for firefox, or modules for the kernel? 10:44 < Shingoshi> I am slow. 10:44 < rxr> Shingoshi: can you explain in depth what you imagine 10:44 < Shingoshi> Why can't the kernel be written as a browser of sorts? 10:44 < rxr> you mean the linux kerne ? 10:44 < rxr> +l 10:45 < Shingoshi> BSD has the ability to run linux as a kernel module. 10:45 < rxr> run linux? 10:45 < Shingoshi> If they can do that, why can't that idea be expanded upon for other purposes as well? 10:45 < rxr> you mean run linux applications ? 10:46 < rxr> the last time I looked BSDs can neither use linux driver nor run the kernel alongside of the BSD kernel 10:46 < Shingoshi> They have a psuedo linux kernel, which loads as a module. 10:46 < rxr> they ony can run linux applicatoins 10:46 < Shingoshi> FreeBSD. 10:46 < rxr> that is only the linux ABI implementation 10:46 < rxr> yes, but that is only a thin layer to allow running linux applications 10:47 < Shingoshi> Kind of like wine? 10:47 < rxr> yeah - kind of - just way smaller jsut converting "syscalls" 10:47 < rxr> linux has that too 10:47 < rxr> linux can run Solaris and sometimes other binaries on some architectures 10:47 < rxr> e.g. on Linux/SPARC you can run Solaris binaries ... 10:47 < rxr> (though I never tried, and never heared of someone using that) 10:48 < Shingoshi> Due to the underlying code of the processor? 10:48 < rxr> this has nothing to do with the processor 10:48 < rxr> would work for x86 as well 10:48 < rxr> (as FreeBSD has for Linux) 10:48 < Shingoshi> This is the kind of conversation I have been wanting to have for a very long time. THANK YOU!!! 10:48 < rxr> it is just translating the Solaris system calls from the user-space binary to the linux internal ones 10:49 < rxr> and on FreeBSD (or NetBSD) they are translating the Linux programs system calls to the BSD kernel internal ones 10:49 < Shingoshi> The reason why I mentioned processor, is because wine can't run on the PlayStation PS3 10:50 < rxr> I think the NetBSD people even has partial OS X / Darwin compatibility mostly going 10:50 < Shingoshi> It needs arch specific calls which aren't available to the PS3. 10:50 < rxr> ok - the PS3 is PowerPC(64) what do you want to run in that CPU ? 10:51 < Shingoshi> I have a friend who has one. I don't. He was asking about windows binaries. But as I suspected, wine won't run on PS3. 10:52 < rxr> ok, the thin layer of abstration I was talking about above (in FreeBSD, NetBSD and e.g. Linux/SPARC just works on the same CPU 10:52 < Shingoshi> I burned him a CD of Yellowdog. 10:52 < rxr> when you want to run binary code of another CPU you need to translate the whole Instruction Set Architecture (ISA) 10:53 < Shingoshi> Ok. Back to your question. 10:53 < rxr> that is what the Transmeta CodeMorphin (tm) or Qemu is doing 10:53 < Shingoshi> I have an old background in programming. 10:53 < rxr> so to run Windows or Linux/x86 binaries (such as the flash player for example) on the PS3 you need e.g. Qemu 10:54 < Shingoshi> I started right from the beginning 25 years ago using very modular code. 10:54 < Shingoshi> It was my belief that lines of code should not be duplicated, but instead called as macros or subroutines. 10:55 < Shingoshi> I used to play with my TI 85, and wrote a full ballistics program on it using modular code. 10:56 < Shingoshi> Including the graphics. 10:57 < Shingoshi> I tend to be a bit obsessive. Or at least focused in my thinking. As much as a 50 year can think! 10:58 < rxr> :-) 10:58 < Shingoshi> It seems that the kernel has become more modular. But I may be wrong. 10:59 < Shingoshi> One of the things I liked about my technique, was that I avoided lond compile times on my TI85 by writing smaller programs. 10:59 < Shingoshi> I used something like a directory or index structure in them. 11:00 < Shingoshi> It served as a traffic director. 11:00 < Shingoshi> It was basically the outline on the general intent of the program. 11:01 < Shingoshi> From it, everything would branch out and return their results to the main tree. 11:01 < Shingoshi> If something was wrong, I only had to recompile the subroutine. Not the entire code tree! 11:02 < Shingoshi> No part of the code tree was recompiled, unless it had specific alterations which required it. 11:03 < Shingoshi> Otherwise, once something was written, it would never be compiled again unless I found an error in my code. 11:04 < Shingoshi> If I wanted to add a new feature, I added it as a subroutine. 11:04 < Shingoshi> I would merely then call that module from the main tree, or the module which needed it. 11:05 < Shingoshi> A very clean system to maintain. 11:05 < Shingoshi> Done^^ 11:07 < Shingoshi> I guess I was writing the equivalent of a C library before I even knew what one was, in 1980-1. 11:08 < Shingoshi> There? 11:09 < Shingoshi> rxr: Are you gone? 11:10 < Shingoshi> rxr: ??? 11:11 < rxr> yeah - back 11:11 < rxr> got a call 11:11 < Shingoshi> ok 11:11 < Shingoshi> You can read through my long winded discourse. 11:12 < rxr> yep read it 11:12 < rxr> there is no question in it, is it ? :-) 11:12 < Shingoshi> That's what I was referring to as modules or extensions. 11:12 < rxr> do you want to know of modular I find the kernel or so? 11:13 < Shingoshi> You asked what did I mean. 11:13 < Shingoshi> It was the answer, not the question. 11:13 < Shingoshi> Yeah, I guess. 11:13 < Shingoshi> But not just specifically the kernel. 11:14 < Shingoshi> Just relating my mindset. 11:14 < rxr> you mean what you want to run on the non-native CPU or on your cluster? 11:14 * rxr upzzled 11:14 < rxr> puzzled even 11:14 < Shingoshi> More or less an overall paradigm. 11:15 < rxr> I completely agree :-) 11:15 < rxr> however I still find many programs not written too modular 11:15 < Shingoshi> I would like to see a system written in a strickly modular form. 11:16 < rxr> more like an overall mix like a spagetti plate 11:16 < rxr> the linux kernel definetly is not the best example of a true modular system, however 11:16 < rxr> though it becomes better 11:16 < rxr> but as a lot features can be switched off a complete rebuild id often needed 11:17 < rxr> like distribute IRQs to all CPUs or not, or NUMA scheduler or not and so on ... 11:17 < Shingoshi> You would have the basic nodes of operation (networking, graphics, etc) with extensions added to a single conceptual format. 11:17 < rxr> it just is not a microkernel but a extemely optimized and configurable monolyth 11:19 < Shingoshi> I probably am not as clear as I could be. 11:20 < rxr> :-( 11:21 < Shingoshi> I guess in my "world", only portions of the kernel would be recompiled. Not the entire thing. 11:21 < rxr> yes in yours .-) 11:21 < Shingoshi> LOL 11:21 < rxr> I just wanted to note due to the overal configurabilty and extreme opimizations the linux kernel often is 11:22 < rxr> e.g. I recompile the linux kernel for 3 das now on the dual-core G5 just to find out which damn config option is messing up the IRQ handling ... 11:22 < Shingoshi> There would probably be conditional checks in place in the main tree, to check if certain conditions were met. 11:22 < rxr> (^- of course not full time, just automated in the background with me hitting the reset button every now and then) 11:24 < Shingoshi> Yes. But most users would never need to do this. 11:24 < Shingoshi> I am speaking of general usage. 11:24 < Shingoshi> Debugging is an entirely different matter. 11:25 < Shingoshi> But I think even my system could help that. 11:25 < rxr> yes, do you think about writing a new kernel *g* :-? 11:26 < Shingoshi> If all of the portions of the kernel were separated, you would only need to compile the changes, not the entire structure. 11:26 < Shingoshi> They could simply be repackaged as a single entity as the user would see it in a tarball or .tlz. 11:27 < Shingoshi> I have to check my blood sugar. I am diabetic, and may need to eat. 11:27 < rxr> well, what part is true for the linux kernel, when you alter a driver just the driver will be rebul 11:27 < rxr> rebuild 11:28 < Shingoshi> bbl8r 11:31 < rxr> cu Shingoshi 11:34 < CIA-8> rene * r22733 /trunk/package/shells/pdksh/pdksh.desc: * added Url tag to the pdksh package 11:41 < Shingoshi> I'm back. 11:42 < Shingoshi> My original question was about patches being written like extensions to firefox or modules to the kernel. 11:43 < Shingoshi> I was asking that question because I read your comments about using as few patches as possible. 11:44 < Shingoshi> So I was wondering if patches could be written like extensions, to be added or removed as desired. 11:46 < Shingoshi> I have a particular need for extensions to mc. I have patched it for lzma. It can now use lzma with it's vfs. 11:47 < Shingoshi> But there is no addition to the F2 menu to allow lzma as a compression option. That's the kind of thing I was talking about. 11:48 < Shingoshi> rxr: ^^ 11:48 < rxr> ah !!! 11:48 < rxr> yes in my opionion we already do this 11:48 < Shingoshi> Now do you understand my convoluted questions? 11:48 < rxr> other packages or cpu architectures in T2 can already contribute patches 11:48 < rxr> and so can target 11:49 < rxr> targets is what you usually work on when you use T2 11:49 < Shingoshi> Well that is all that I am concerned about. 11:49 < rxr> so you start your target (in target/xyz) and define what packages to build and you can alter the package build for each of these packages including adding patches for those 11:49 < rxr> look into target/desktop or target/archivista for some examples 11:50 < Shingoshi> Yes. But I think you are speaking of groups of programs, rather than individual programs. 11:50 < rxr> so each "T2 based distribution (or system)" can come with it's own kind of extra patches without bothering the rest of T2 with at 11:50 < rxr> that 11:50 < rxr> well thoe packages are for individual packages 11:50 < Shingoshi> What I would like to see is a way to add menus to mc or any other program. 11:51 < rxr> and you can alter the way each individual source package by adding code to run to "hooks" in the build process 11:51 < rxr> like for firefox, hook_add postmake 5 "copy some custom extensions from there to here" 11:51 < Shingoshi> To create patches to the main program which don't require recompilation of the original prog. 11:52 < rxr> that hook content can be anything, function, direct shell code, ... 11:52 < rxr> hm, ok the patches are patching the source, you need to build the patched program then 11:52 < rxr> altering a packages without recompiling it of course requires that the program to modify already has some sort of plugin mechanism 11:53 < Shingoshi> I know that zsh is supposed to be modular. I just don't know to what extent. 11:53 < rxr> bash likewise 11:53 < rxr> or e.g. the KDE desktop 11:53 < Shingoshi> Yes! Precisely. But once that is provided, it should never need to be done again. 11:54 < rxr> the KDE desktop is very modular and you can plug in all sorts of extensions without recompiling anything of the big KDE thing 11:54 < rxr> well - when you use your working copy of T2 this is tree 11:55 < Shingoshi> Yes. But you are talking about adding programs to the tree. I am talking about adding functions to the program. 11:55 < rxr> for the whole T2 project as it this is not the case people might use different cpu optimizations or optimize for size and not for speed and so on 11:56 < Shingoshi> Yes. But in my case, I want the most of everything there is to offer. 11:56 < Shingoshi> Kind of like the local library. Everyone comes to it to check out the books they want. 11:57 < Shingoshi> That's my idea of the application server I spoke of. 11:57 < Shingoshi> It could be distributed though. 11:57 < rxr> yes, definetly in your use case you can pre-build the whole "application library" and then only recompile parts when e.g. security updates show up for a single package 11:58 < Shingoshi> Applications specific to certain functions could be kept on different machines. 11:58 < Shingoshi> Precisely. 11:58 < Shingoshi> But I am also speaking of the individual programs. 11:59 < Shingoshi> What I need are books. 11:59 < Shingoshi> And I need to study them when I have them. 12:00 < Shingoshi> I also tend to be a bit lazy at times. 12:00 < Shingoshi> Ok, Most of the time! 12:00 < Shingoshi> LOL 12:00 < Shingoshi> Where in Germany do you live? 12:00 < rxr> Berlin 12:01 < rxr> (though grown up around Hannover) 12:01 < Shingoshi> What is the city like. I have only seen it in movies. 12:01 < rxr> ouh, that's such a difficult questions ... 12:01 < Shingoshi> Yes. I am sorry. 12:01 < rxr> multi culti ? 12:02 < rxr> http://images.google.com/images?q=potsdamer%20platz&ie=UTF-8&oe=UTF-8&um=1& 12:02 < Shingoshi> Cities, no matter how small have personlities within them. 12:02 < Shingoshi> Yes. Culturally and Architecture. 12:02 < Shingoshi> I am a photographer. Or at least I used to be. 12:03 < Shingoshi> I grew up in New York, but not the city. 12:03 < Shingoshi> I love photographing buildings. The older the better. 12:04 < rxr> well, as Germany's largest city and 4 million citicens it's not easy to stuff it into a few sentences ... 12:04 < Shingoshi> I really like old farm houses and barns, especially the broken ones. 12:04 < Shingoshi> That's why I apologized. 12:04 < rxr> of course berlin is still recovering from the times of the cold war with the dividing wall in between ... 12:04 < rxr> http://images.google.com/images?svnum=10&hl=en&ie=UTF-8&oe=UTF-8&q=berliner+mauer&btnG=Search 12:05 < Shingoshi> Division is something I am all too familiar with. 12:05 < Shingoshi> It is why I am so devoted to global equality. 12:05 < rxr> definetly 12:06 < rxr> our office is just 10 foot minutes away from the potsdamer platz quotesd earlier and thus where the wall was about nearly 20 years ago 12:06 < Shingoshi> I have this one core belief. The Immediate Equalization of All Knowledge Among All Beings. 12:06 < Shingoshi> If you search for that line on the internet, you will have likely found me. 12:07 < Shingoshi> My name is Xavian-Anderson Macpherson. 12:08 < Shingoshi> In my later years, I have developed a deeper sense of community than I even had as a child. 12:09 < Shingoshi> And mine was pretty well developed even then. 12:10 < Shingoshi> My perception of strife is far more critical now than it was before. 12:11 < Shingoshi> Am I spouting on too much? 12:12 < rxr> nope, quite an interesting conversation 12:13 < Shingoshi> I struggle with comments comedians make. Realizing the power they have on the perceptions of the public. 12:13 < rxr> comedians and news agencies ... 12:14 < Shingoshi> I was very impressed with Craig Ferguson last week on his Latenight Show on CBS. 12:15 < Shingoshi> The problem with comedians, is that they tend to teach us to find the suffering of others as cause for laughter. 12:15 < Shingoshi> I disagree with that. 12:16 < Shingoshi> I believe we all take something out of ourselves when we take pleasure in the suffering of another. 12:17 < Shingoshi> For me right now, I would like to use software as a means of bridging gaps between groups of people. 12:18 < Shingoshi> Kind of idealistic, huh? 12:18 < rxr> yes, as most of us, Open Source society, are, hm? 12:19 < rxr> btw. I can not follow references to US letenight shows as we do not receive too many US stations 12:19 < rxr> CNN is the nearly only one I remeber on the cable line 12:19 < Shingoshi> I know. 12:20 < Shingoshi> But you can find it on YouTube. 12:20 < rxr> :-) 12:20 < Shingoshi> It made the news here. 12:20 < Shingoshi> It was that big! 12:21 < Shingoshi> He basically apologized for his participation in ridiculing Brittney Spears. 12:21 < rxr> oh, will try to remeber to look it up after the workday ... 12:21 < rxr> (/me wonders if I have a youtube compatible video player on my box ...) 12:21 -!- idealm [n=ideal@203.45.99.98] has quit [Remote closed the connection] 12:21 < Capey> good god, there has been lot of activity since my sleep 12:21 < rxr> hi Capey :-) 12:21 < Capey> wont scroll and read.... 12:21 < Capey> hi rxr 12:22 < Shingoshi> You can see it with firefox. 12:22 < rxr> does it need the flash player? (that is not yet available for the x86-64 box I'm on ...) 12:22 < rxr> anyway I'll take a look later 12:22 < rxr> I'm busy tracking some run-time linux regression on the multi-core G5 ... 12:25 < Shingoshi> Back to modularity. 12:26 < Shingoshi> Take browsers. It seems like every time someone wants something new, they write a new broswer. 12:27 < Shingoshi> I am curious. Could links be modified with extensions to function like firefox? 12:28 < Shingoshi> I mean you take the simplest form of a browser, add that functionality we spoke of for extensions, and the go from there. 12:29 < rxr> links? that is a rather minimalistic browser anyway 12:29 < rxr> I think they advertise links specifically as "monolythic" on their homepage 12:29 < Shingoshi> Precisely. 12:29 < rxr> so I guess it does not support any form of extensions 12:29 < rxr> at least run-time loadable 12:30 < rxr> when you rebuild it with some add-on source you can of course modify it :-) 12:30 < Shingoshi> But what I am speaking about is taking a code base like it and modifying it to your own needs. 12:30 < Shingoshi> It doesn't have to be links. 12:30 < Shingoshi> But it should be small to begin with. 12:31 < rxr> there are some other browser like dillo and co 12:31 < Shingoshi> I don't know how large the source is. But I would take it, and rewrite it to be fully modular. 12:31 < rxr> but i'm not sure that you become happy with their limitted feature set and stanard conformance ... 12:32 < Shingoshi> It is better in my technique to start with the smallest one. 12:32 < rxr> hm - ok, firefox is indeed a bit huge to rewrite 12:32 < Shingoshi> You then have less code to change from the beginning. 12:32 < rxr> kHTML is quite good code quality from what I saw so far 12:32 < rxr> that is also why Apple used it as the base for webkit 12:33 < rxr> maybe it is worth buiding your browser up on kHTML / webkit 12:33 < CIA-8> rene * r22734 /trunk/package/perl/perl-test-manifest/perl-test-manifest.desc: * updated perl-test-manifest (1.14 -> 1.17) 12:33 < Shingoshi> How old are you? I am asking for chronological context only. 12:33 < Shingoshi> Historical context. 12:36 < Shingoshi> You mentioned Apple. They had something years ago called HyperText(?) or something like it. 12:36 < Shingoshi> It was a macro program. 12:37 < rxr> I'm preogressing to the 30 ... :-) 12:37 < rxr> but I learned to code when I was 10 or so, the first box was a XT followed by a 286 12:37 < Shingoshi> One of the things that bothers me about firefox, is the massive amount of duplication in extensions. 12:38 < Shingoshi> A system like mine would virtually eliminate that. 12:43 < rxr> what bothers me most is the sheer amount of code the motilla project wrote to implement their browser 12:43 < rxr> I also use KDE's konqueror for daily browsing 12:43 < rxr> it's also way faster and such 12:43 < rxr> I mean on a average machine (2GHz Athlon or so) firefox needs nearly one hour to just build ... 12:44 < rxr> though in project most compatibility is required we still supply Firefox 12:44 < Shingoshi> Yes. I use both. 12:44 < rxr> as kHTML does not honor some typos and such and thus some sites do not look perfect 12:45 < rxr> however it is not nice to see how firefox performs on sub GHz machines 12:45 < rxr> e.g. 200-700MHz embedded boxes / boards ... 12:45 < rxr> quite a bit sluggish 12:45 < Shingoshi> Too much code! 12:46 < Shingoshi> A modular code base runs faster. 12:46 < Shingoshi> Not just on the compile. 12:46 -!- Stelz [i=stelz@unaffiliated/stelz] has joined #t2 12:47 < Shingoshi> Only those portions of code that are needed are actually loaded into memory. 12:47 < Shingoshi> By modular, again I mean subroutines. 12:48 < Shingoshi> Only one portion of the program remains in memory at all times, It's the operational directive. 12:48 < Shingoshi> The portion that calls all of the other parts as they are needed. 12:50 < Shingoshi> I need to get some books, specifically on programming. 12:50 < Shingoshi> I have yet to settle on a language yet though. 12:50 < Shingoshi> My hobby is math. For it, I prefer APL. 12:51 < rxr> :_) 12:51 < Shingoshi> I don't know between python perl etc, which is better for what I want. 12:52 < rxr> I guess your browser will rather be written in something like C++, ObjC, D or so 12:53 < rxr> neither perl or python 12:53 < Shingoshi> I guess if I were to take the same approach to languages, I would want the simplest one that can be extended as well. 12:53 < rxr> YMMV 12:53 < Shingoshi> ?? 12:53 < rxr> your millage may vary 12:53 < rxr> mileage 12:55 < Shingoshi> It would need to be the most cross compileable of all. I guess that means C? 12:56 < rxr> nope depends what cross compiler you have 12:56 < rxr> you can cross compiler C++, ObjC, or Ada just fine 12:57 < Shingoshi> Which lang besides assembler can create all of the rest? 12:57 < rxr> in theory any langauge you get a cross compiler for 12:59 < Shingoshi> Is there any one of them that is the most reduced in instruction set, and not specific to an arch? 12:59 < Shingoshi> APL has very few operatives, but is amazingly strong. 13:00 < Shingoshi> Most don't like it because it is considered too cryptic. 13:00 < rxr> neither of this high languages, such as C, C++, ObjC, D does rely in any way on the later target CPU instruction set 13:01 < rxr> when you do not inject assembly or do other tricky stuff but just write normal code you can let the compiler create the resulting machine assembly for any source you inject into it ... 13:01 < rxr> personaliy I would use a object orient language with and higher level ob abstraction than C, such as C++ or D for a new, modular project 13:02 -!- sparc-kly_ [n=mubex@adsl-72-50-101-246.prtc.net] has joined #t2 13:02 < Shingoshi> Symbolism eludes most people. 13:02 < Shingoshi> So I guess I want a fairly low level language which isn't machine specific. 13:02 < Shingoshi> Be nice if fewer than 100 instructions. 13:02 < Shingoshi> The less the better. 13:02 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has joined #t2 13:03 < rxr> wb misl 13:03 * rxr phone - back a bit later 13:05 -!- Netsplit calvino.freenode.net <-> irc.freenode.net quits: LMJ, misl, mtr, R4gnar0k, Shingoshi, sepp, valentin, Capey 13:05 -!- Netsplit calvino.freenode.net <-> irc.freenode.net quits: CIA-6 13:06 -!- sparc-kly_ [n=mubex@adsl-72-50-101-246.prtc.net] has quit [Nick collision] 13:06 -!- Netsplit over, joins: CIA-6 13:06 -!- sparc-kly_ [n=mubex@adsl-72-50-101-246.prtc.net] has joined #t2 13:06 -!- Netsplit over, joins: misl, Shingoshi, mtr, sepp, LMJ, Capey, R4gnar0k, valentin 13:17 < rxr> re 13:18 -!- sparc-kly [n=mubex@adsl-72-50-101-246.prtc.net] has quit [Connection timed out] 13:20 [Users #t2] 13:20 [@ChanServ] [ LMJ ] [ R4gnar0k] [ Shingoshi ] [ valentin] 13:20 [ Capey ] [ misl] [ rxr ] [ sparc-kly_] 13:20 [ CIA-8 ] [ mtr ] [ sepp ] [ Stelz ] 13:20 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal] 13:22 < Shingoshi> rxr: How many commands does C have? 13:22 < Shingoshi> Average/ 13:22 < Shingoshi> Approx. 13:23 < Shingoshi> Over 100, less than 200 13:23 < Shingoshi> ??? 13:25 < sepp> 32 13:27 < Shingoshi> Wow!! That's all? 13:28 < Shingoshi> What about C++ 13:29 < rxr> some more keywords 13:30 < rxr> but in fact C++ is way more complex than pure C 13:30 < sepp> some more like 'class' 13:30 < rxr> and namespace 13:30 < rxr> though you might want the object abstraction for your modularitly ... 13:30 < rxr> so - lunch - cu 13:30 < Shingoshi> How much does it help? 13:31 < Shingoshi> Shingoshi@comcast.net 13:31 < Shingoshi> Write me! 13:32 -!- Enqlave [i=stelz@unaffiliated/stelz] has joined #t2 13:32 -!- Stelz [i=stelz@unaffiliated/stelz] has quit [Nick collision from services.] 13:32 -!- Enqlave is now known as Stelz 14:14 < rxr> re 14:14 < rxr> Shingoshi: In m opinion object orientation does quite help to keep the code modular 14:14 < rxr> and extendable 14:15 < rxr> however Gnome coder would say they archive simillar result with low level hackery ... 14:15 < rxr> I suggest you read some introduction yourself 14:19 < Shingoshi> I am doing that now. But I should be in bed. 5:19 in the morning here. You can always write me. My email is above. 14:19 < Shingoshi> bye. 14:20 < Shingoshi> Shingoshi@comcast.net 14:23 < rxr> cu Shingoshi have a good sleep 14:50 -!- Netsplit calvino.freenode.net <-> irc.freenode.net quits: @ChanServ, Stelz 14:51 -!- Netsplit over, joins: @ChanServ 14:51 -!- Netsplit over, joins: Stelz 14:54 -!- sepp_ [n=sepp@p85.212.65.245.tisdip.tiscali.de] has joined #t2 15:02 -!- sepp [n=sepp@p85.212.18.61.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)] 15:58 < CIA-8> rene * r22735 /trunk/package/graphic/xsane/xsane.desc: * updated xsane (0.992 -> 0.993) 16:14 -!- sepp [n=sepp@p85.212.65.245.tisdip.tiscali.de] has joined #t2 16:27 -!- sepp_ [n=sepp@p85.212.65.245.tisdip.tiscali.de] has quit [Read error: 110 (Connection timed out)] 16:47 < CIA-8> rene * r22736 /trunk/package/shells/pdksh/initrd-usability.patch: 16:47 < CIA-8> * slightly tweaked pdksh for default usability, especially due to 16:47 < CIA-8> initrd use without configuration files available 18:17 < Capey> rxr: when converting wrt2 to pkgsel format, do i still check if libc was dietlibc or uclibc? 18:17 < Capey> in config.in 18:21 < rxr> nope, I would choose a sane one (uclibc for now?) 18:21 < Capey> yep 18:21 < rxr> this option come from the embedded target which offers the choice to show various pathes 18:21 < rxr> I would not use this on well defined products 18:22 < rxr> I would choose a sane set of stuff needed and make sure this all integrates well together 18:29 < rxr> http://www.t2-project.org/targets/wrt2.html 18:29 < rxr> ^- injected your name there :-) 18:29 [Users #t2] 18:29 [@ChanServ] [ LMJ ] [ R4gnar0k] [ Shingoshi ] [ valentin] 18:29 [ Capey ] [ misl] [ rxr ] [ sparc-kly_] 18:29 [ CIA-8 ] [ mtr ] [ sepp ] [ Stelz ] 18:29 -!- Irssi: #t2: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal] 18:30 < Capey> :) 18:37 < rxr> == 18:36:27 =[0]=> Building base/gcc [4.0.2-atmel.0.99.2 7.0-trunk]. 18:37 < rxr> I really like the new architecture/target package overloading facility sooo much :-) 18:39 < Capey> oh, we have that already in trunk? 18:40 < rxr> that avr32 bit or package overloading ? 18:40 < Capey> package overloading 18:40 < rxr> yes, but there is some work still missing 18:40 < rxr> basic parts work - announced some weeks ago on the mailing lsit 19:02 < CIA-8> rene * r22737 /trunk/scripts/ (Build-Pkg functions.in): 19:02 < CIA-8> * removed some bitotting (rev 1!) pkgdir fragments that are unused 19:02 < CIA-8> and undocumented 19:06 < CIA-8> rene * r22738 /trunk/scripts/Create-ErrList: * removed unused variable pkgdir in Create-ErrList 20:01 < LMJ> moin moin 20:04 < rxr> moin moin LMJ 20:04 < LMJ> s'up rxr 20:13 < rxr> working on new, exotic T2 features (and rebooting the G5 in a row to find which .config option makes the IDE interrupt ooops ...) 20:17 < rxr> LMJ: register on your new site, today 20:17 < rxr> ah - damn 20:17 < rxr> wasted 10 minutes debugging a thing that actually worked as intended ... 20:19 < R4gnar0k> lol 20:24 < LMJ> i saw 20:25 < LMJ> pascal & Rene rxr ;) 20:26 < LMJ> should i switch one to admin ? 20:27 < rxr> pascal is my brother 20:27 < rxr> you can switch rene to admin if you like 20:27 < LMJ> ok 20:28 < rxr> my brother is not a coder, he is more assigned as exactcode pr person 20:28 < LMJ> oki 20:28 < LMJ> brb 30min, i will do it just after 20:30 < rxr> LMJ: I'm at home then 20:31 < rxr> but that's ok 20:31 < rxr> will just have evening dinner and fall into bed that 20:31 < rxr> I'll take a look tomorrow morning then 20:34 < CIA-8> rene * r22739 /trunk/scripts/ (Build-Pkg functions.in): 20:34 < CIA-8> * implemented .conf inheritance and .patch dropping for redefined 20:34 < CIA-8> architecture/target packages as outlines on the mailing-list 20:36 < CIA-8> rene * r22740 /trunk/scripts/Build-Pkg: * removed a line of debug code slipped in in r22739 20:37 < rxr> Capey: now the package dedefine code is more complete 20:38 < rxr> Capey: I just implemented that part, because I did not need that so far 20:38 < rxr> Capey: I think now it is behaves exactly as outlined on the list 20:38 < rxr> Capey: the only part missing is Download-ing the redefined [D] tags 20:38 < rxr> Capey: but for that I have to rewrite some parts of Download 20:38 < rxr> Capey: I'll do that tonight or tomorrow morning 20:39 < rxr> Capey: but at least I think that will improve Download readability by reusing function.in bits ... 21:01 < Shingoshi> Hi guys! 21:02 < LMJ> rxr : you are super admin 21:05 < rxr> LMJ: thanks I think I have more edit this pages knobs now 21:05 < LMJ> ;) 21:09 < Shingoshi> rxr: By the way. Yesterday, when I was checking out those live cd's, I almost installed yours on my system. But because it didn't recognize RAID, I couldn't do so. 21:09 < Shingoshi> My /usr and /pub are on RAID5. 21:10 < Shingoshi> They are also both xfs filesystems. 21:14 < LMJ> hardware raid Shingoshi ? 21:14 < Shingoshi> Upon investigation, it seems that your system layout is compatible with my slamd64. That is, it looks like you put things in the same place. 21:14 < Shingoshi> Software 21:16 < Shingoshi> I think my card supports hardware RAID. But I configured it beforehand, and don't know how using hardware RAID would affect it. 21:16 < LMJ> If i understand well, how a live cd could determine /usr & /pub are in fact /dev/md* stuff who are RAID software from a couple of partitions of differents disks ? 21:16 < Shingoshi> I have a Promise tx4 sata2 21:17 < LMJ> ok 21:17 -!- sparc-kly_ [n=mubex@adsl-72-50-101-246.prtc.net] has quit [Read error: 104 (Connection reset by peer)] 21:17 < Shingoshi> There appeared to be no RAID configuration possible with the installation. 21:17 < LMJ> i think you are right 21:17 < LMJ> should be done later i guess 21:19 < Shingoshi> Hopefully, soon! 21:19 < Shingoshi> But no loss. I can do it from my existing system. 21:20 < Shingoshi> Does t2 install anything to /usr/local? 21:20 -!- sepp [n=sepp@p85.212.65.245.tisdip.tiscali.de] has quit [Read error: 110 (Connection timed out)] 21:21 < LMJ> should in /usr/bin by default 21:21 < Shingoshi> I looked into the packages for the 64bit system, and it seems everything goes into either /lib64 or /usr/lib64. Is that right? 21:22 < R4gnar0k> mostly 21:22 < Shingoshi> I am sorry. I meant libraries. 21:22 < Shingoshi> What are the exceptions that you can think of? 21:22 < Shingoshi> I mean for /usr/local. 21:23 < Shingoshi> I tend to avoid /usr/local like the plague. 21:23 < R4gnar0k> driconf installs something there 21:23 < rxr> sorry folks gotta go home, grab evening dinner and the like 21:23 < R4gnar0k> should be fixed I guess 21:24 < Shingoshi> hi rxr! 21:24 < LMJ> Shingoshi : grep -ir "/usr/local" /var/adm/flists/ 21:24 < LMJ> see you rxr 21:24 < Shingoshi> Ok. I can do that even if I haven't installed the system yet, right? 21:26 < LMJ> ha 21:26 < LMJ> i did here, i didn't find anything 21:26 < LMJ> do you R4gnar0k ? 21:26 < LMJ> 36K /usr/local/ 21:26 < LMJ> quite empty here ;) 21:28 < R4gnar0k> 122325 < R4gnar0k> driconf installs something there 21:28 < LMJ> k 21:33 < Shingoshi> Gues I can install t2 over Slackware/Slamd64 then. 21:34 < Shingoshi> I haven't yet installed t2. Maybe I didn't make that clear. 21:34 < Shingoshi> I was looking in the packages to see if I could find any that used /usr/local. 21:35 < Shingoshi> Anyway, I turned that line you wrote into a script for future use against any package or the like. 21:36 < Shingoshi> I can now use it in the process of converting packages. 21:37 < Shingoshi> If I do find a package that uses /usr/local, I can simply change it during the installation to /usr. 21:44 < Shingoshi> For all of the scripts in t2-trunk/scripts, where and if, are they installed on the system? I am thinking of making a Slackware pkg from them so as to be installed. 21:44 < Shingoshi> If they are installed, where on the system? 21:47 < R4gnar0k> they are not 21:50 < Shingoshi> So are they only used during the build process? 21:50 < R4gnar0k> yea 21:51 < Shingoshi> So they have to be called by /dir/foo? 21:51 < Shingoshi> Or in this case ../scripts/foo 21:51 -!- d-marc [n=d-marc@HSI-KBW-091-089-001-120.hsi2.kabelbw.de] has joined #t2 21:53 < Shingoshi> Do you manually call these scripts, or are they called by the installation process itself? 21:53 < R4gnar0k> hm? o_o 21:53 < Shingoshi> When are they used, and how? 21:54 < R4gnar0k> Build-Target builds your target of choice ... which you created with Config 21:55 < R4gnar0k> Emerge-Pkg builds a package into the running system 21:55 < Shingoshi> Gotcha! 21:55 < Shingoshi> Can you build more than one type of target? 21:56 < Shingoshi> One overlaying another? 21:56 < R4gnar0k> that doesn't make much sense? o_o 21:56 < Shingoshi> Let me look more closely. 21:56 < Shingoshi> bb 22:05 -!- Stelz [i=stelz@unaffiliated/stelz] has quit ["I wanna live my life like thunder!"] 22:18 -!- Stelz [i=stelz@unaffiliated/stelz] has joined #t2 22:24 -!- d-marc [n=d-marc@HSI-KBW-091-089-001-120.hsi2.kabelbw.de] has quit ["KVIrc 3.2.4 Anomalies http://www.kvirc.net/"] 22:26 < Shingoshi> I just compressed the entire t2-trunk to bz2. It's less than 5Mbs. 22:26 * R4gnar0k nods 22:26 < Shingoshi> If I convert that to lzma, it will even smaller. 22:27 < R4gnar0k> :p 22:28 < Shingoshi> Ok. This will sound crazy, I know. How would it work to create a@Live target to be installed on a hard drive? 22:29 < R4gnar0k> hm? o_o 22:29 < R4gnar0k> define live target 22:30 < Shingoshi> I want to create the system to fit on a dvd using squasfs.lzma. On the t2 site, one of the targets is @Live for a live cd. 22:30 < R4gnar0k> ah ok 22:31 < Shingoshi> I want to compile as much software as will fit on that dvd. 22:31 < R4gnar0k> yea you can do that ... it'll be slow tho ... 22:32 < Shingoshi> I know. But the intended end result is to have a system which can be immediately installed to disk. 22:32 < Shingoshi> Basically, a Live installer. 22:32 < Shingoshi> @Live/installer 22:33 < R4gnar0k> I see 22:33 < Shingoshi> How much software do you think I could get on it> 22:33 < R4gnar0k> but a livecd has different requirements ... like filesystem being mostly readonly and such 22:34 < R4gnar0k> well my whole pkgs dir just fits on one dvd 22:34 < Shingoshi> I have strange intentions for this. One thing is security. Unionfs could be used as a means to prevent security issues. 22:35 < R4gnar0k> if I wanted to install that, I'd just burn it to a minimal livecd and then fdisk/create fs/mount and then mine -i pkgs/* 22:35 -!- misl [n=chatzill@84-104-172-187.cable.quicknet.nl] has quit ["Chatzilla 0.9.75 [Firefox 1.5.0.9/2006120612]"] 22:35 < rxr> re 22:36 < Shingoshi> Another strange question. Is it possible to have more than one bootable kernel on a live cd? 22:37 < R4gnar0k> sure 22:38 < CIA-8> rene * r22741 /trunk/package/base/uclibc/uclibc.conf: 22:38 < CIA-8> * converted uclibc to not use confdir, but $base/... so it's 22:38 < CIA-8> .conf can be reused from redefined packages 22:38 < Shingoshi> So I could create only one cd/vd with the compilation tools for every supported arch? 22:39 < R4gnar0k> in theory I guess ... 22:39 < R4gnar0k> but you won't be able to boot a sparc cd in a i386 system 22:39 < R4gnar0k> or vice versa 22:39 < Shingoshi> How much of the build system would not need to be replicated? 22:40 < Shingoshi> That's what I was asking 22:40 < Shingoshi> Is it an issue of formatting or the kernel itself. 22:40 < R4gnar0k> everything basically has to be replicated. you'll have different binaries 22:40 < R4gnar0k> both. and of the boot loader. 22:40 < Shingoshi> rxr: Are you back? 22:41 -!- kensai [n=kensai@206.248.106.75] has joined #t2 22:42 < Shingoshi> A sparc won't read the cd format of a 386 and reverse? 22:42 < Shingoshi> I mean the cd. 22:42 < R4gnar0k> it will read it. but it won't boot from it. 22:43 < Shingoshi> So the bootloader must be arch specific? 22:43 < R4gnar0k> of course 22:44 < Shingoshi> Has there been any attempt to create a noarch bootloader? 22:44 < R4gnar0k> how about: not possible? 22:44 < Shingoshi> Just wondering. 22:45 < Shingoshi> What if you had one cd for each arch, and a separate dvd for all of the sources. Every source imaginable. 22:46 < Shingoshi> The arch cd's would have the compilation tools, and the dvd the sources. 22:46 < R4gnar0k> possible 22:47 < Shingoshi> This is where my discussion with rxr about modularity comes in. 22:47 < R4gnar0k> of course, my download dir is 17180750K ... so you might need a blu-ray for that 22:47 < Shingoshi> 1.7Gbs? 22:48 < R4gnar0k> no. 17Gbs 22:48 < rxr> Shingoshi: nope just at home looking something up via a cellular link 22:48 < R4gnar0k> but that's probably highly redundant 22:49 < Shingoshi> That's not a problem. I think with lzma, much of it would fit on a single dvd. 22:49 < Shingoshi> hi rxr! 22:49 < Shingoshi> Maybe all of it. 22:49 < rxr> R4gnar0k: Shingoshi: it is possible to place the boot loaders of some architectures on the same cd 22:49 < R4gnar0k> these are bz2 compressed already 22:49 < R4gnar0k> lol 22:50 < R4gnar0k> some yea ... but not all supported 22:50 < Shingoshi> But would they run, no? 22:51 < Shingoshi> So then, two dvd's. 22:51 < Shingoshi> And cd's for each arch. 22:52 < Shingoshi> Did you peruse my earlier discussion with rxr this morning? 22:52 < R4gnar0k> no 22:52 < Shingoshi> Did you read it? 22:52 < Shingoshi> ok 22:53 < Shingoshi> I told him that I am interested in creating a clustered LTSP system. 22:53 < R4gnar0k> ok my cleaned up downloads has 4.9 gig ... you might be able to fit that on a dvd if you leave out some stuff 22:53 < R4gnar0k> LTSP? 22:54 < Shingoshi> I want all of my applications to be served from a single server. 22:54 < R4gnar0k> ah ok 22:54 < Shingoshi> For an entire network. 22:54 < R4gnar0k> that's possible 22:54 < Shingoshi> With the ability of replication for speed and security. 22:55 < Shingoshi> The servers themselves would appear as RAID. 22:55 < R4gnar0k> that doesn't really matter tho 22:55 < Shingoshi> The replicated ones. 22:56 < Shingoshi> But mostly I don't want to be required to install software in a network more than one time. 22:56 < R4gnar0k> you could do it very simply by having just a small root on the clients, and then mounting /usr and /opt by nfs 22:57 < Shingoshi> Yes, but the conf is already done by LTSP. Linux Terminal Server Project. 22:58 < R4gnar0k> ok o_o 22:58 < Shingoshi> I don't want to reinvent any "new" wheels! 22:58 < Shingoshi> What was done once, doesn't need to be done again. 23:01 < R4gnar0k> yea 23:03 < Shingoshi> How much of your cleaned up download would actually need to be on the arch-spec cd? 23:03 < R4gnar0k> depends on what you want ... bbl tho 23:03 < R4gnar0k> lunch 23:04 < Shingoshi> Great, now you decide to leave! 23:04 < Shingoshi> I just invited Larhzu from tukaani here! 23:11 < Shingoshi> rxr do you know Larhzu from tukaani? 23:25 -!- kensai [n=kensai@206.248.106.75] has quit ["Leaving"] 23:54 -!- idealm [n=ideal@CPE-203-45-99-98.nsw.bigpond.net.au] has joined #t2 --- Log closed Tue Feb 27 00:00:22 2007