[03:46] <madrazr> Hii all, yesterday there was a session on Building Upstream kernels at #ubuntu-classroom. I just had this question but unfortunately could not ask.
[03:47] <madrazr> Suppose I now have the rpm-src of the Fedora 8 kernel and I want to build it for Ubuntu, how to do it?
[03:49] <madrazr> I badly want this because, there is a tool called NCTUNS, which must be mandatorily used by students as per the University for Networks Simulation and this runs only on Fedora
[03:55] <cradek> maybe you could extract the source from that rpm with rpm2cpio and build it.  but I'm pretty sure this is not the right channel for rpm questions.
[04:14] <pwnguin> madrazr: is the source to nctuns available?
[04:15] <madrazr> pwnguin: yes
[04:16] <madrazr> pwnguin: can some one build it for Ubuntu, I am ready to take the initiative but I am not well versed in Ubuntu package maintainance
[04:16] <madrazr> So If someone helps me, we can bring it to Ubuntu and I can pursuade my college authorities to use Ubuntu from nowonwards
[04:17] <pwnguin> im grabbing a tar.gz right now, but im not clear what on earth it is
[04:17] <madrazr> We are actually not happy with Fedora, and we want to shift to Ubuntu, but NCTUNS is standing in between our college and Ubuntu
[04:17] <pwnguin> it has linux 2.6.24 as part of then ame
[04:17] <madrazr> pwnguin: can you link me to the page frm where you are downloading?
[04:19] <pwnguin> no
[04:19] <pwnguin> stupid frames
[04:19] <pwnguin> my first guess is something related to tun
[04:20] <madrazr> pwnguin: I did not get you
[04:20] <pwnguin> http://nsl10.csie.nctu.edu.tw/
[04:20] <pwnguin> that place uses frames well enough that i cant figure out what link to give you
[04:20] <madrazr> http://nsl10.csie.nctu.edu.tw/products/nctuns/download/download.php
[04:21] <madrazr> this is the link to download page
[04:21] <pwnguin> like i said, im downloading now
[04:21] <madrazr> oh sorry, fine then
[04:22] <pwnguin> anyways, you might consider attending a session by dholbach tomorrow on making packages
[04:22] <madrazr> pwnguin: ok
[04:23] <pwnguin> but really, i dont see this happening if it involves patching the kernel
[04:24] <madrazr> pwnguin: why so?
[04:25] <madrazr> pwnguin: how can end users use this tool then?
[04:26] <madrazr> pwnguin: cant we have a NCTUNS supporting kernel in repos, so those who want to use NCTUNS can install this kernel, actually thats what happens when we install it on Fedora
[04:26] <madrazr> Grub gives a separate option at boot time
[04:26] <pwnguin> from what i can tell, this NCTUNS provides a patched kernel for fedora and the application uses it to attach to a special network emulator forked from tun
[04:29] <madrazr> pwnguin: yes it does so
[04:29] <pwnguin> who would maintain this kernel?
[04:29] <madrazr> this patched kernel??
[04:29] <pwnguin> yes
[04:29] <madrazr> its maintained by those developers only
[04:29] <pwnguin> indeed
[04:30] <madrazr> pwnguin: so do you mean to say, until those people provide patches to Ubuntu kernel, no one can help us?
[04:31] <pwnguin> What I mean to say is that ubuntu is fairly burdened already, especially on the kernel. If someone new were to step up and meet the obligations, that would be one thing
[04:31] <pwnguin> if the NCTUNS had put their patches into the kernel tree, that'd be another thing
[04:31] <pwnguin> but asking ubuntu to pick up a fedora package that even fedora hasn't picked up seems a bit silly
[04:33] <madrazr> pwnguin: ok fine, suppose I am ready to maintain the packages and I am ready to take the initiative for making it work on, what kind of support will I get from Ubuntu community and kernel team?
[04:33] <pwnguin> well, dholbach will love you
[04:34] <madrazr> pwnguin: sooper, where can I find him??
[04:34] <madrazr> which channel?
[04:34] <pwnguin> #ubuntu-motu
[04:34] <madrazr> pwnguin: fine, thanks a lot
[04:34] <pwnguin> you should also write the kernel mailing list
[04:34] <madrazr> pwnguin: ubuntu kernel mailing list?
[04:34] <pwnguin> yea
[04:35] <pwnguin> about whether the patches are okay or not
[04:35] <madrazr> fine
[04:35] <madrazr> pwnguin: actually I dont know where to get started, thats the whole problem here now, because I am not very well versed with Kernel development
[04:35] <pwnguin> well, in your case
[04:35] <pwnguin> the first thing to do is make an actual patch
[04:36] <pwnguin> they provide a set of files, but they're not in patch format
[04:37] <pwnguin> you might also write the NCTUNS folk and ask them if they have plans to get the nctun interface into the kernel
[04:37] <pwnguin> linus' kernel
[04:37] <madrazr> ok
[04:40] <pwnguin> the patch would make it easier to evaluate whether the modifications to say, syscall_table is dangerous etc
[04:40] <pwnguin> im not even sure regular tun is still in the kernel
[04:41] <madrazr> ok
[04:42] <pwnguin> anyways, if the kernel team were to decide this was worthy of inclusion in ubuntu mainline, you'd basically provide a patch and they'd include it, then the rest of the work would be in MOTU -- but the idea of providing a special nctun build of the kernel I don't think will fly
[04:43] <pwnguin> maybe I'm wrong
[04:44] <madrazr> pwnguin: ok
[06:24] <ackrcv1> Hi, how do I go about installing 2.6.25 for 8.04?
[10:27] <Saied> is there a 2.6.25 kernel available for ubuntu?
[17:08] <BenC> Ok people...there's supposed to be a kernel meeting today, but there's not much to discuss on account of work is starting for intrepid
[17:08] <BenC> But if anyone has any questions, we can answer them now
[17:09] <laga> (i've got a question about a bug report, but i'm not sure if it's appropriate for the meeting)
[17:09] <BenC> Closer to UDS we will review the topics for the kernel track, and we invite people to add to the list of topics
[17:09] <BenC> laga: It's usually not, but for lack of anything else to discuss, feel free to bug drop :)
[17:10] <laga> sure. i'll just need a minute to make thunderbird behave again..
[17:11] <laga> https://bugs.launchpad.net/bugs/223381
[17:11] <laga> that guy basically suggests a change to CONFIG_HZ to get his remote working, and i'm not sure if that's the right way ;) any advice?
[17:24] <smb> laga: To throw in my opinion, I agree with your initial answer. A driver that requires a certain HZ number is broken and should be fixed instead of changing HZ.
[17:30] <BenC> Yeah, that's my answer too
[17:32] <laga> good, thanks
[17:49] <alex_joni> BenC: mind if I jump in with a question?
[17:49] <BenC> alex_joni: sure
[17:50] <alex_joni> abogani asked me if I plan to push the custom flavour I did for hardy (IPIPE patch in the kernel, and RTAI packages) back to Ubuntu
[17:51] <alex_joni> so far I haven't considered it (until he approached me), mainly because I didn't think there's enough interest.. also maintaining might be a bit of a problem
[17:51] <alex_joni> are there any thoughts/pages about processes like these?
[18:58] <alex_joni> BenC: was that so scary?
[19:00] <BenC> alex_joni: It's all documented in source in debian/binary-custom.d/ in the git tree
[19:00] <BenC> alex_joni: the other thing is we need reassurances that it will be maintained, especially post-release
[19:01] <BenC> we generally like to see a team behind custom flavors and someone who will respond immediately when we see a build failure cause by patch conflicts
[19:01] <alex_joni> I see
[19:02] <alex_joni> well, I'll talk to abogani about it
[19:03] <alex_joni> I'm also concearned I don't have the time/energy for long-term support
[19:07] <BenC> May be best to keep it in a PPA then
[19:08] <alex_joni> ppa?
[19:41] <amitk> alex_joni: ppa = personal package archives, feature of launchpad to allow users to upload source deb package and have it spit out Ubuntu binaries. It also provides a way to host your own packages
[19:45] <alex_joni> amitk: I have a repo set up
[19:45] <alex_joni> with proper package signing, etc
[20:02] <amitk> alex_joni: in that case you probably don't care about ppa
[20:05] <pwnguin> unless you're popular and want canonical to foot the hosting bill ;)
[22:04] <alex_joni> cjwatson: any idea what this means? http://pastebin.ca/1002222
[22:06] <cjwatson> alex_joni: no, although the /dev/sda I/O error suggests hardware trouble
[22:07] <smb> alex_joni: this means the block layer returned one sector bad. 
[22:07] <alex_joni> so it's faulty HW ?
[22:09] <smb> alex_joni: somewhere yes. but depending on hw setup could be somewhere one the way down to the disk. is that a real scsi disk or sata?
[22:10] <alex_joni> sata
[22:10] <alex_joni> smb: so it could be a problematic driver of the chipset or something like that?
[22:12] <smb> alex_joni: the message itself comes from the generic block layer. normally the drivers would or could issue additional info. it could even be a bad sector on the disk. or bad cable (if the error is not always the same sector)
[22:12] <alex_joni> smb: ok, I'll poke it some more
[22:12] <smb> alex_joni: is there anything before that message?
[22:13] <alex_joni> smb: it's a livecd I built, run by another user.. I'll have him run with nosplash, and without quiet
[22:13] <smb> alex_joni: ah, ok
[22:13] <alex_joni> thanks for the pointers
[22:15] <smb> alex_joni: np :)