| _nemesis_ | hallo, i found a bug, uptime: 02:04:31 up 8588 days, 23:09, 4 users, load average: 0.41, 0.42, 0.35 | 01:11 |
|---|---|---|
| _nemesis_ | uname -a: Linux 8101 2.6.24-22-generic #1 SMP Mon Nov 24 18:32:42 UTC 2008 i686 GNU/Linux | 01:12 |
| _nemesis_ | what can i also search for, to prove that this uptime is real ;) | 01:13 |
| phix | hey | 02:45 |
| phix | installing on an encrypted volume, LVM, on dual boot system, when defining /boot partition grub is setup to boot off hd0,0 instead of hd0,1 and kernel for /boot/$kernel_version instead of /$kernel_version | 02:45 |
| phix | bug, some one file it, I CBF, my day off :P Lol | 02:45 |
| ScottK | lool: If you have any ideas about how to get that spec approved, I'm all ears. | 03:17 |
| phix | ScottK: what specs? | 03:32 |
| * Hobbsee scratches head at kees | 03:54 | |
| === jscinoz_ is now known as jscinoz | ||
| NCommander | hey ScottK | 04:21 |
| === freeflyi1g is now known as freeflying | ||
| karthik_ | hey i need to switch netween workspaces.. how does it happen internally in the code? | 08:31 |
| === pwnguin_ is now known as pwnguin | ||
| teddyb | hi | 09:04 |
| teddyb | http://packages.ubuntu.com/search?keywords=spice&searchon=names&suite=intrepid§ion=all : doesn't have spice | 09:05 |
| teddyb | ubuntu has gnucap, but no spice, which is practically the basis for all circuit simulations | 09:05 |
| teddyb | gentoo pulls from: http://www.ibiblio.org/pub/Linux/apps/circuits/spice3f5sfix.tar.gz and creates a package. it'd be nice if a similar package was created for ubuntu | 09:06 |
| teddyb | my only worry is that the spice on that link is from 1999. not to say that old is bad, but if there hasn't been _any_ activity since, it's a bit concerning | 09:06 |
| teddyb | also gnucap isn't really fully compatible with spice. there are lots of annoying little fixes that have to be introduced to patch the differences | 09:07 |
| persia | karthik_, Look at libwnck | 09:08 |
| teddyb | this might help http://osmirrors.cerias.purdue.edu/pub/gentoo/portage/sci-electronics/spice/spice-3.5.5.ebuild : the ebuild has a few good tips to clean | 09:10 |
| teddyb | for example, spice doesn't like any more optimization than -O1 | 09:10 |
| teddyb | also, this patch: http://osmirrors.cerias.purdue.edu/pub/gentoo/portage/sci-electronics/spice/files/spice-3.5.5-gcc-4.1.patch seems to be relevant | 09:12 |
| teddyb | i don't have ubuntu on me right now (i use ubuntu at uni, and my current computer is my old gentoo box). i'll see if i can get a vm running soon of ubuntu and i'll try to crank out a spice package | 09:13 |
| teddyb | that is if someone else isn't working on something similar | 09:13 |
| persia | teddyb, spice isn't very free: you might look at easyspice with ngspice or gEDA. Alternately, if I'm wrong, you're welcome to submit a package. | 09:14 |
| teddyb | the other package i'd been longing for on ubuntu (and even on gentoo it's pretty unstable) is ngspice | 09:14 |
| teddyb | i checked out gEDA, but it needs a spice backend | 09:14 |
| persia | teddyb, See https://wiki.ubuntu.com/UbuntuDevelopment/NewPackages | 09:14 |
| teddyb | easyspice is a gui as well i believe | 09:14 |
| teddyb | will do. i have a slow connection, so it'll take a while :( | 09:15 |
| persia | teddyb, Actually, as it turns out, there are license issues even with ngspice: see Debian bug #299748 | 09:16 |
| ubottu | Debian bug 299748 in ngspice "ngspice: legal issue" [Normal,Closed] http://bugs.debian.org/299748 | 09:16 |
| teddyb | i'm also not an expert by any means. i'm just a student who's trying to learn this on the side, but are you sure about the spice isn't very free line? | 09:16 |
| teddyb | hmm, well gentoo has a sci-electronics/ng-spice-rework ... i wonder what that is. | 09:18 |
| pwnguin | wha'ts wrong with ngspice? | 09:18 |
| teddyb | yeah, it's bsd and gpl-2? | 09:18 |
| teddyb | http://ngspice.sourceforge.net | 09:18 |
| persia | pwnguin, Dunno: the bug cites license issues, and it was removed. | 09:19 |
| teddyb | i'm still looking through the debian bug (downloading an ubuntu cd takes a lot of bandwidth, so pages load slowly :( ) | 09:19 |
| persia | teddyb, Looking at http://embedded.eecs.berkeley.edu/pubs/downloads/spice/index.htm, it looks like it should have become free in 2007 (which is more recent than the 2005 date for the removal). If you'd like to investigate, and get it back in, that wouldn't be bad, but be sure to only use the post-free sources. | 09:20 |
| teddyb | check this gentoo ebuild as well btw: http://osmirrors.cerias.purdue.edu/pub/gentoo/portage/sci-electronics/ng-spice-rework/ng-spice-rework-15.ebuild | 09:20 |
| teddyb | sure, i wouldn't mind. | 09:20 |
| pwnguin | odd. the guy who filed the legal issue bug had an ieee email | 09:20 |
| pwnguin | do they hand those out to all members? | 09:20 |
| teddyb | i guess it's good to keep on multiple distros. keeps everyone on their toes :) | 09:21 |
| teddyb | i think so | 09:21 |
| teddyb | have yet to get a membership (can't justify the cost yet) | 09:21 |
| persia | All full members, yes.] | 09:21 |
| pwnguin | have you seen oregano? | 09:22 |
| persia | Hrm. The 1994 Berkeley "Research Software license Agreement" doesn't look all that free: it permits redistribution, but only gratis (or with charge for supporting materials), and has the advertising clause. | 09:23 |
| teddyb | oregano needs spice backend | 09:23 |
| teddyb | they are all frontends in ubuntu's repositories. the only exception is gnucap | 09:23 |
| teddyb | and it's good, but i've been running into a lot of problems rectifying the differences between what the ibiblio circuits books say to use for circuit simulations on spice and what gnucap will work with | 09:23 |
| persia | Well, it looks like spice got free in 2007, so with a bit of work, and some license review, it ought be suitable for inclusion. | 09:24 |
| teddyb | wonderful | 09:24 |
| pwnguin | "got free" | 09:24 |
| persia | pwnguin, Well, it was made more free. See the homepage. | 09:24 |
| pwnguin | fyi, the guy who does the fedora electronics lab spin also pays attention to ubuntu | 09:24 |
| teddyb | lab spin? | 09:25 |
| pwnguin | a remix | 09:25 |
| pwnguin | comes with a bunch of things installed by default like gEDA | 09:26 |
| teddyb | oh, i thought you were being negative or something | 09:26 |
| teddyb | oh yeah, there's a fatty DVD of electronic programs available through gEDA iirc. ngspice is one of the things found on it | 09:26 |
| persia | Are Fedora remixes conventionally called "spin"s? | 09:26 |
| pwnguin | persia: i think so. | 09:27 |
| pwnguin | they only recently tried the idea | 09:27 |
| pwnguin | http://spins.fedoraproject.org/ | 09:29 |
| teddyb | so it seems as though the spice3f5sfix is antiquated. would it make more sense to package ngspice then? | 09:32 |
| teddyb | especially considering ngspice is based on the berkeley spice3f5? | 09:33 |
| persia | teddyb, If ngspice inherited the freeness of spice, yes, but you'll want to check the licenses carefully. | 09:33 |
| teddyb | i don't know if i'm qualified to do that | 09:34 |
| emgent | HNY people. | 09:34 |
| teddyb | i'm not a lawyer and i don't want to taint ubuntu with legally dubious packages | 09:34 |
| teddyb | does ubuntu have any sort of legal consulation for package maintainers? | 09:35 |
| persia | teddyb, You can ask for input on the freeness of a license in #ubuntu-motu | 09:35 |
| persia | The ultimate decision lies with the archive admins, but they don't tend to have time to speculatively consult. | 09:36 |
| persia | If you can get a couple of the MOTU to agree that the license is probably free, the archive admins will review and check. | 09:36 |
| pwnguin | teddyb: you could always just ask debian about it ;) | 09:36 |
| persia | And you could always package it for Debian, and let Ubuntu inherit it :) | 09:36 |
| teddyb | oh true | 09:37 |
| teddyb | it'll hit main by 2035 ;) | 09:37 |
| pwnguin | in fact, i bet there's a rocket enthusiast who might appreciate ng-spice | 09:37 |
| persia | heh | 09:38 |
| teddyb | oh, just googling for ngspice debian gets a few packages already made | 09:38 |
| pwnguin | http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=489768 | 09:39 |
| ubottu | Debian bug 489768 in wnpp "ITP: ngspice -- A Spice circuit simulator" [Wishlist,Open] | 09:39 |
| teddyb | particularly annoying: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=489768#15 | 09:40 |
| ubottu | Debian bug 489768 in wnpp "ITP: ngspice -- A Spice circuit simulator" [Wishlist,Open] | 09:40 |
| persia | Why is that annoying? | 09:40 |
| teddyb | that was more than a month ago, but then again, i guess thanksgiving, winter break, etc. | 09:40 |
| pwnguin | they may have forgotten to follow up | 09:41 |
| persia | Also, it means that someone else is chasing the licensing situation, which can only be a good thing :) | 09:41 |
| teddyb | well, i remember being tangentially interested in this in the summer and not getting much feedback from anyone related to ubuntu, so i quickly gave up and got back to other stuff | 09:41 |
| teddyb | but i really have nothing better to do, so might as well make it worth my while. i'll see if i can send an email out to either "Gudjon I. Gudjonsson" <gudjon@gudjon.org> | 09:42 |
| teddyb | or to the ngspice people | 09:42 |
| pwnguin | get some debian people interested in it ;) | 09:42 |
| persia | I'd recommend sending a follow-up to the bug first, asking about the status, just so nobody else does that: those working on it only need to be poked once. | 09:43 |
| teddyb | you keep ;)... but i'm not too sure why. are they difficult people to work with or something? | 09:43 |
| pwnguin | i donno | 09:43 |
| pwnguin | too lazy to use : i guess | 09:43 |
| teddyb | i know a lot of ubuntu people (developers especially) are extremely friendly and willing to help. colin watson definitely springs to mind | 09:44 |
| teddyb | and these two people i met on irc.. persia and pwnguin :) | 09:45 |
| pwnguin | well, if ngspice is good enough for ubuntu, its usually good enough for debian | 09:45 |
| pwnguin | in which case, you'd want someone in debian to take up debian's end of the deal | 09:46 |
| teddyb | what's ubuntu's policy with old packages that are no longer getting support by the way? | 09:46 |
| teddyb | like spice3f5 | 09:46 |
| teddyb | is it better to have crufty old packages in the repositories just in case someone wants them? | 09:46 |
| pwnguin | well, if it's a case of "never works" it's probably a waste of compile time and just a gordian knot | 09:47 |
| persia | teddyb, Opinion is split. There are several folk who argue they are potentially useful, and others that argue they should be gone. | 09:47 |
| pwnguin | if it's old and functional, I don't see the pain | 09:48 |
| teddyb | gordian knot.. hmm, wiki'ing. also, it works. i just compiled it earlier today and it works | 09:48 |
| pwnguin | i mean, by that criteria we would probably lose our only font editor | 09:48 |
| persia | Practical result is that packages typically get removed when either 1) they get removed from Debian, or 2) there are persistent unfixable compilation, runtime, or legal issues that attract the attention of a developer | 09:48 |
| teddyb | although, it was gentoo patched, which could mean the original source was a dog and i ended up with a stallion | 09:48 |
| pwnguin | well, upstream is active | 09:49 |
| pwnguin | in the case of ngspice | 09:49 |
| persia | teddyb, If the patches are good, there's no reason not to use them :) | 09:49 |
| pwnguin | so as persia says, no reason they can't be pushed upstream or sideways | 09:49 |
| persia | Generally, I prefer both :) | 09:50 |
| teddyb | yeah, but spice3f5 isn't. i'll definitely look more into ngspice, but since spice works as well... why not? | 09:50 |
| teddyb | anyway, i'm really tired, should head to bed. it's been a pleasure discussing this with the two of you. thanks for all your input | 09:51 |
| teddyb | goodnight | 09:51 |
| pwnguin | im super confused about where the tarball for ngspice-rework018 is | 09:53 |
| pwnguin | oh, apparently they don't know how to seperate packages correctly =/ | 09:54 |
| pwnguin | so the mingw version supercedes -18 | 09:55 |
| === asac_ is now known as asac | ||
| === cjwatson_ is now known as cjwatson | ||
| veloc1ty | moin | 12:33 |
| crid | Hi all. And happy new year! | 13:17 |
| crid | If I have some project I would like to see in repsitories of Ubuntu. What I should do? | 13:18 |
| crid | Project which I'm talking about is here: http://ait.berlios.de | 13:18 |
| sebner | !revu | crid | 13:21 |
| ubottu | crid: REVU is a web-based tool to give people who have worked on Ubuntu packages a chance to "put their packages out there" for other people to look at and comment on in a structured manner. See https://wiki.ubuntu.com/MOTU/Packages/REVU | 13:21 |
| crid | thanks | 13:23 |
| anilg | Hi.. is there a bounty system at launchpad? | 15:04 |
| persia | anilg, There used to be, but it was mostly delinked or removed because it didn't turn out to work so well. | 15:05 |
| persia | anilg, For more detail, you might ask in #launchpad next week. | 15:06 |
| anilg | Thanks. Why next week?.. | 15:06 |
| persia | Because it seems that most of the #launchpad folk are on vacation this week. | 15:08 |
| savvas | what's a bounty system? :p forgive my ignorance | 15:21 |
| persia | savvas, A place where one posts work one wants done, and remuneration on offer for the work. | 15:22 |
| persia | Others can add to the remuneration, or volunteer for the work. | 15:22 |
| savvas | like revu for packagers? | 15:24 |
| savvas | looks interesting, thanks | 15:24 |
| persia | Not really. It's about money. | 15:24 |
| savvas | ah :P | 15:24 |
| savvas | now it makes sense :) rent-a-coder hehe | 15:25 |
| anilg | savvas: yes. | 15:26 |
| anilg | A bounty system makes sense on a palce like launchpad.. I wonder what didnt work out | 15:26 |
| savvas | anilg: What would happen if people place money on something they want done that doesn't get implemented? Would they allow team cooperation or single-person updates when using such a system and how would the team share the money, by length of bytes? | 15:31 |
| anilg | I would assume folks would collaborate, and start off with what everyone would do and how they would share the bounty | 15:32 |
| anilg | usually bounties are small tasks that individuals can take up | 15:33 |
| anilg | so its more of a do this and get this | 15:33 |
| savvas | then there's the possibility of abuse, a lot of people taking up a lot of tasks and never doing anything :p I hope that's not the case in the real world though | 15:35 |
| Mithrandir | not all bounty systems allow you to lock a bounty, it's just a race to completion. | 15:36 |
| savvas | and quality :) | 15:36 |
| savvas | but then people with money would actually run the development with "do this and do that" tasks | 15:37 |
| Mithrandir | savvas: yes, and? It's not like you can't effectively do that today, if you have enough money. | 15:37 |
| Mithrandir | usually, bounties have been tiny compared to how much it'd cost to hire somebody to do the work | 15:38 |
| savvas | hm.. you mean making custom versions of programs? | 15:38 |
| savvas | without actually touching the official releases? if it wouldn't be dependant to the projects' trunk, then I'm all for it! | 15:39 |
| savvas | er.. * if the projects' trunk wouldn't be dependant to the tasks | 15:40 |
| Mithrandir | integration into mainline is usually a condition on most bounties. | 15:40 |
| savvas | yes, that :) | 15:40 |
| === DktrKranz2 is now known as DktrKranz | ||
| === giskard_ is now known as giskard | ||
| === bluesmoke_ is now known as bluesmoke | ||
| === lan3y is now known as Laney | ||
| === popey_ is now known as popey | ||
| bbechdol | Hello everyone | 19:52 |
| bbechdol | Happy New Year :) | 19:53 |
| pwnguin | Mithrandir: i dont think the problem with the lp bounties was in the race to finish part; it's probably a function of a) tiny bounties funded by individuals and b) failure to aggregate bounty across individuals | 19:59 |
| pwnguin | so one guy says "i'll pay 20 dollars for this" | 20:00 |
| pwnguin | and another guy says "i'll pay 40" | 20:00 |
| pwnguin | lp didnt add up those bids; and nobody will work for a 20 dollar project | 20:01 |
| rootard | Hi all (and hny), I'm having a strange issue with dpkg unpacking a symlink that overwrites a file (in the same package) | 20:23 |
| rootard | This is on Nexenta (OpenSolaris) so it's not strictly an Ubuntu issue but I'm looking for some pointers on where to dive into | 20:23 |
| rootard | there is a file "/usr/bin/tcsh" and a link "/bin/tcsh -> /usr/bin/tcsh" and when the package is unpacked (via dpkg -X tcsh_*deb /) then /usr/bin/tcsh becomes a link to /usr/bin/tcsh | 20:24 |
| rootard | however, if I unpack to another directory (dpkg -X tcsh_*deb tmp) then the directory structure is fine | 20:25 |
| andersk | Is there a reason you are unpacking the package to / rather than installing it (dpkg -i tcsh_*.deb)? | 20:26 |
| rootard | to try and debug the issue | 20:26 |
| rootard | dpkg -i does the same thing | 20:26 |
| andersk | This sounds like what might happen if you somehow had /bin a symlink to /usr/bin (or vice versa), though that would be really weird. | 20:27 |
| * rootard smacks self | 20:28 | |
| rootard | crap | 20:28 |
| pwnguin | heh | 20:28 |
| pwnguin | why would you do that? | 20:28 |
| rootard | good question... I don't know what that is done but it seems to be a Solaris'ism | 20:29 |
| rootard | :-/ | 20:29 |
| rootard | *why | 20:29 |
| rootard | well, thanks. The patch to the package is simple then. | 20:30 |
| rootard | Hmm, if I were to make the patch a little more robust than just commenting out the "ln -s" in debian/rules, how might I approach this? I think hurd (last time I used it...) had some kind of crazy /usr -> / setup. | 20:35 |
| rootard | so just detecting /bin to be a symlink isn't enough. | 20:35 |
| rootard | is there some (simple/portable) way to see if /usr/bin is the same as /bin ? | 20:36 |
| andersk | Checking whether /bin is a symlink with [ -L /bin ] is probably close enough. | 20:39 |
| andersk | It's dangerous for a package to install anything into a symlinked directory. | 20:40 |
| pwnguin | if the symlink is usually there in nexenta | 20:40 |
| pwnguin | it's probably the case that the nexenta developers have solved this before | 20:40 |
| rootard | I think it will always be there. I just checked a Solaris 10 host and it's the same | 20:40 |
| andersk | You could try checking whether [ "$(stat -Lc %D%i /usr/bin)" = "$(stat -Lc %D%i /bin)" ], I guess. | 20:45 |
| rootard | bash seems to support if [ /bin/ -ef /usr/bin/ ] | 20:48 |
| andersk | Oh, cool. (So does dash.) | 20:49 |
| === bluesmoke is now known as Amaranth | ||
| === lool- is now known as lool | ||
| kolby_ | are any developers here paid by canonical? | 23:49 |
| Hobbsee | a fair few of them | 23:52 |
| Hobbsee | likely not here right now, though | 23:52 |
| kolby_ | I don't think I have a good enough resume to apply. | 23:53 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!