--- Log opened Thu Sep 02 00:00:17 2010 00:26 -!- jmil [~jmil@cpe-98-154-246-74.socal.res.rr.com] has joined #hplusroadmap 00:28 < jmil> kanzure: who do you work with at UT? 00:50 -!- Phreedom [~quassel@109.254.6.63] has quit [Ping timeout: 260 seconds] 02:11 -!- QuantumG [~qg@rtfm.insomnia.org] has quit [Remote host closed the connection] 02:15 -!- shepazu [~schepers@2001:660:330f:38:21e:52ff:fe81:88fa] has quit [Quit: shepazu] 02:27 -!- elmom [~elmom@hoasnet-fe29dd00-137.dhcp.inet.fi] has quit [Remote host closed the connection] 02:41 -!- wolfspraul [~wolfsprau@lucia.q-ag.de] has quit [Quit: Lost terminal] 02:43 -!- wolfspraul [~wolfsprau@lucia.q-ag.de] has joined #hplusroadmap 02:59 -!- superkuh [~hukrepus@unaffiliated/superkuh] has quit [Ping timeout: 265 seconds] 03:00 -!- superkuh [hukrepus@unaffiliated/superkuh] has joined #hplusroadmap 03:02 -!- shepazu [~schepers@2001:660:330f:38:21e:52ff:fe81:88fa] has joined #hplusroadmap 03:07 -!- shepazu [~schepers@2001:660:330f:38:21e:52ff:fe81:88fa] has quit [Quit: shepazu] 03:14 -!- elmom [~elmom@hoasnet-fe29dd00-137.dhcp.inet.fi] has joined #hplusroadmap 03:20 -!- augur [~augur@208-58-6-161.c3-0.slvr-ubr1.lnh-slvr.md.cable.rcn.com] has quit [Remote host closed the connection] 03:34 -!- QuantumG [~qg@rtfm.insomnia.org] has joined #hplusroadmap 03:56 -!- jennifer2 [~jennifer@c-67-180-253-94.hsd1.ca.comcast.net] has quit [Ping timeout: 264 seconds] 04:02 -!- mheld [~mheld@pool-173-76-224-45.bstnma.fios.verizon.net] has joined #hplusroadmap 04:29 < QuantumG> Visions of the Future: The Biotech Revolution is on tv 04:29 < QuantumG> random chick identified as a "transhumanist" 04:30 < QuantumG> I already forgot her name.. something like "cerwan" 04:31 < QuantumG> I hate Michio Kaku. 04:43 -!- katsmeow-afk [~someone@75-120-14-250.dyn.centurytel.net] has quit [Ping timeout: 265 seconds] 04:44 -!- any19811340 [~someone@75-120-14-250.dyn.centurytel.net] has joined #hplusroadmap 04:49 -!- shepazu [~schepers@mna75-10-82-243-125-40.fbx.proxad.net] has joined #hplusroadmap 05:29 -!- shepazu [~schepers@mna75-10-82-243-125-40.fbx.proxad.net] has quit [Quit: shepazu] 06:04 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 06:13 -!- mheld [~mheld@pool-173-76-224-45.bstnma.fios.verizon.net] has quit [Quit: mheld] 07:00 -!- splicer [~splicer@92.39.2.22] has quit [] 07:02 < kanzure> jmil: matt campbell 07:02 < jmil> oh wow cool http://www.me.utexas.edu/~campbell/index.htm 07:03 < jmil> are you grad student? 07:03 < jmil> i'm in chris chen's lab at penn 07:03 < jmil> for 2 years as postdoc 07:04 < jmil> kanzure: how long u been there? 07:05 < kanzure> nah, i was an undergrad when i was working with matt 07:05 < kanzure> but i was living the life of a grad student :) 07:06 < kanzure> jrayhawk: mah user sez "can't delete the hooks directory in reprap.git ..." hmm 07:06 < kanzure> jrayhawk: did you want me to get rid of hooks or hooks_OLD 07:07 < kanzure> oops 07:07 < kanzure> jmil: did you want me to get rid of hooks or hooks_OLD 07:07 -!- mheld [~mheld@c-75-69-89-109.hsd1.ma.comcast.net] has joined #hplusroadmap 07:10 -!- mheld_ [~mheld@c-75-69-89-109.hsd1.ma.comcast.net] has joined #hplusroadmap 07:12 -!- mheld [~mheld@c-75-69-89-109.hsd1.ma.comcast.net] has quit [Ping timeout: 240 seconds] 07:12 -!- mheld_ is now known as mheld 07:12 < jmil> kanzure: hooks_OLD 07:12 < jmil> i was able to rename it, just not delete it 07:14 < kanzure> the deed has been done 07:17 < jmil> kanzure: thx. but when i do ls -la i get "s" for some of the file permissions. what is "s" ? is that keeping us from listing the file directory? 07:17 < kanzure> why are all of the author email adresses broken in the commit messages 07:17 < kanzure> it's sticky bit 07:18 < kanzure> it means that when you create a file in the subdirectory that all of the permissions of the folder apply to the file 07:18 < jmil> kanzure: what do you mean? not broken for me 07:18 < jmil> ah, the sticky bit 07:18 < jmil> but now we cannot browse: https://secure.diyhpl.us/repos/reprap/ 07:18 < kanzure> erikdebruijn 07:18 < kanzure> i know that this is not erik's email address 07:19 < jmil> how do you browse the git repo when it is named reprap.git and you don't have an active tree checked out? 07:19 < kanzure> jmil: you mean https://secure.diyhpl.us/repos/reprap/ikiwiki.cgi but it doesn't matter for this project 07:19 < kanzure> i think the cgit link would be more relevant 07:19 < kanzure> https://secure.diyhpl.us/cgit/reprap/ 07:24 < jmil> oh i didn't know where the right link was 07:24 < kanzure> sorry i forget to tell people these impotant things 07:24 < jmil> to browse the git repo 07:24 < kanzure> *important 07:24 < jmil> but how can i run "git status" from the command line? 07:24 < jmil> it always tells me it's not a repo 07:25 < kanzure> i gues you should clone it into your home directory somewhere 07:25 < kanzure> /srv/git/reprap.git/ is technically not a repo (like a git repo that you work in and edit files in) 07:26 < kanzure> also, brb (just woke up) 07:28 < jmil> kanzure: me too. there's something wrong with the files that were uploaded, i'm getting different browsing in my home computer 07:30 < jmil> i'm going to recreate the entire repo on your server from scratch, see if that fixes it 07:31 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Ping timeout: 245 seconds] 07:31 < kanzure> do you need me to install git-svn? 07:31 < jmil> oh ya! please 07:32 < jmil> maybe it's the --use-log-author option 07:32 < jmil> http://www.kernel.org/pub/software/scm/git/docs/git-svn.html 07:32 < jmil> the reprap SVN server is very slow so the command takes a while 07:32 < jmil> a long while 07:33 < jmil> kanzure: lemme know when u have it installed 07:34 < kanzure> Failed to fetch http://debian.cs.pdx.edu/pool/main/g/git/git-svn_1.7.1-1_all.deb 404 Not Found 07:34 < kanzure> wtf 07:35 < jmil> kanzure: does SVN not have the email address contained in a commit? maybe the data are not there to pull into git? 07:36 < jmil> e.g. on sourceforge SVN i don't see any email addresses for the authors: 07:36 < jmil> http://reprap.svn.sourceforge.net/viewvc/reprap/ 07:41 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 08:16 -!- shepazu [~schepers@eduroam-0-104.enst.fr] has joined #hplusroadmap 08:16 < jmil> kanzure: u there? if you can get git svn installed i can update 08:18 < kanzure> one sec 08:18 < jmil> oh, it looks like i just need to do git svn rebase 08:18 < jmil> that will fix it 08:18 < jmil> all my local repos on this HD have no email addresses either 08:18 < jmil> I think it's not in SVN repos 08:18 < jmil> so don't worry about it actually 08:19 < jmil> :) 08:19 < kanzure> basically git-svn doesn't exist in this debian repository 08:20 < kanzure> i'm trying to figure out why 08:21 -!- Phreedom [~quassel@109.254.6.63] has joined #hplusroadmap 08:21 < kanzure> ok http://debian.cs.pdx.edu/pool/main/g/git/git-svn_1.7.1-1.1_all.deb i'll take it 08:21 < kanzure> jmil: it should be installed now 08:25 < jmil> kanzure: i can't find it. i just ssh'ed in again, it's not in my path though 08:29 < jmil> oh, found in /usr/lib/git-core/ 08:32 -!- wolfspraul [~wolfsprau@lucia.q-ag.de] has quit [Quit: leaving] 08:36 -!- shepazu [~schepers@eduroam-0-104.enst.fr] has quit [Quit: shepazu] 08:45 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Quit: Leaving] 09:00 < kanzure> huh it really should be in your path 09:00 < kanzure> maybe it was "git svn" not "git-svn" 09:21 -!- mheld [~mheld@c-75-69-89-109.hsd1.ma.comcast.net] has left #hplusroadmap [] 09:26 -!- superkuh [hukrepus@unaffiliated/superkuh] has quit [Ping timeout: 265 seconds] 09:26 -!- Alystair [Alystair@bas1-toronto10-1279397903.dsl.bell.ca] has joined #hplusroadmap 09:26 -!- augur [~augur@129-2-175-79.wireless.umd.edu] has joined #hplusroadmap 09:39 -!- dustbin [~chatzilla@71.145.146.189] has joined #hplusroadmap 09:41 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 10:08 -!- patrik [~patrik@h126n1c1o261.bredband.skanova.com] has joined #hplusroadmap 10:08 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Quit: Leaving] 10:09 -!- patrik is now known as Guest68002 10:09 -!- Guest68002 is now known as splicer 10:15 -!- superkuh [~hukrepus@unaffiliated/superkuh] has joined #hplusroadmap 10:37 < jmil> kanzure: u there? i try to run git clone http://diyhpl.us/srv/git/reprap.git 10:37 < jmil> but no go 10:41 -!- jennifer2 [~jennifer@c-67-180-253-94.hsd1.ca.comcast.net] has joined #hplusroadmap 10:41 < jmil> kanzure: have you pushed/pulled successfully via git with this server before? 10:57 -!- panax [panax@goldstandard.eng.usf.edu] has joined #hplusroadmap 11:00 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 11:04 < kanzure> jmil: check the email again 11:04 < kanzure> git clone git+ssh://.. 11:08 < jmil> oh i thought the git+ssh meant you can use git or ssh with that :// URL... lol 11:08 < jmil> lemme try again 11:09 < jmil> that is working, thanks 11:09 < jmil> i had to do git_ssh://username@.... 11:09 < jmil> git+ssh://... rather 11:10 < jmil> how can we enable unrestricted checkout? i am finishing testing then i will post back to the reprap-dev listserv 11:10 < jmil> kanzure: ping in case you went afk 11:13 < jmil> kanzure: how can we enable just the git:// URL instead of git+ssh:// ? i'm new to git-based serving 11:17 < epitron> you need to run a git daemon for git:// 11:17 < epitron> git can also be served over http 11:18 < epitron> also, i think git+ssh is a legacy thing... it should just be ssh://, or what i use, user@host:path/to.git 11:19 < epitron> http://www.kernel.org/pub/software/scm/git/docs/git-daemon.html 11:21 < jmil> epitron: thank you. kanzure is the server admin, so i think he has a plan already 11:21 < jmil> we are playing with moving reprap from SVN to git. would be WAY better 11:22 < epitron> i'd just use a git daemon. it's fastest. :) 11:22 < epitron> yep! 11:22 < epitron> svn is the new cvs 11:22 < epitron> you could host it on github btw... that has a lot of nice features. :) 11:22 < epitron> the forking and "pull requests" are nice 11:22 < epitron> plus free public exposure 11:25 < jmil> github is sub-optimal because it only allows 10 collaborators and 2 GB storage. the repo is 2GB already, and we need way more people to commit to the main trunk 11:26 < jmil> so i'm trying to host it on kanzure server, then we can push a read-only copy up onto github so people can fork as desired 11:26 < kanzure> jmil: "unrestricted checkout" should alerady work 11:27 < jmil> URL? 11:27 < kanzure> jmil: git:// should be working.. 11:27 < kanzure> hm hold on 11:27 < jmil> i tried with git:// and didn't get it 11:30 -!- jennifer2 [~jennifer@c-67-180-253-94.hsd1.ca.comcast.net] has quit [Ping timeout: 276 seconds] 11:36 < kanzure> jmil: still looking into it 11:37 < kanzure> jmil: try git clone http://diyhpl.us/reprap.git 11:40 < kanzure> jrayhawk: it says that i have to run git update-server-info; shouldn't a daemon somewhere be taking care of this eventually? 11:46 < epitron> jmil: maybe you should prune your repository :) 11:46 < epitron> with git, you have to download the entire commit history 11:46 < jmil> prune repo for what purpose? 11:46 < epitron> remove some ancient history 11:46 < epitron> to reduce the repo size 11:46 < jmil> naw, we need for posterity :P 11:47 < epitron> do you really? :) 11:47 < epitron> DO YOU REALLY???!? 11:47 < jmil> well, it's academic project. so i would say yes 11:47 < epitron> well, just ask yourself 11:47 < epitron> what would linus do 11:47 < kanzure> kill babies? 11:47 < epitron> i think he makes a new git repo for each version 11:48 < jmil> linus does not build physical objects... :D 11:48 < jmil> I think he is h-minus 11:48 < epitron> yeah but he's a smart mofo 11:48 < epitron> and he made git for himself 11:48 < jmil> ya wrote git in 1 week 11:48 < epitron> so if you don't act a little like him, there's going to be an impedence mismatch 11:48 < jmil> Git itself has complete history 11:48 < jmil> checkout git repo of git, and read first commit. it's hilarious 11:48 < epitron> hahah 11:48 < epitron> nice 11:49 < jmil> first commit is something like "well, shit this may actually work" 11:49 < epitron> anyhow, 2gb is a big fuckin repo 11:49 < epitron> that's unwieldy 11:49 < epitron> it's easy to break it up by versions 11:50 < epitron> http://git.videolan.org/?p=vlc.git;a=forks 11:50 < epitron> that's how VLC does it 11:50 < kanzure> why do we need 2 GB to represent a few hundred components and parts in an assembly with maybe less than 5000 revisions? 11:50 < kanzure> :) 11:51 < epitron> no idea! 11:53 < kanzure> jmil: is anonymous clone working now? 11:54 < epitron> http://git.kernel.org/?p=git/git.git;a=log;pg=230 11:54 < epitron> there she is 11:54 < kanzure> what am i thinking 11:55 < kanzure> okay, so basically 2 GB is pretty normal for a project that doesn't know what it's doing in terms of revision control 11:55 < kanzure> i mean, people still use .doc format for revision control sometimes 11:55 < epitron> hahah so true 11:55 < kanzure> so on vehicleforge, the whole military vehicle stint 11:55 < kanzure> there's no way in hell that the projects are going to be less than 10 GB a pop 11:55 < kanzure> i'm thinking multiple terabytes per war machine 11:55 < kanzure> (especially when you add up all the simulation results that people want to keep around for laughs and giggles) 11:56 < epitron> git isn't designed for huge binary repos 11:56 < epitron> that's the thing 11:56 < epitron> it's designed for little textfiles 11:56 < kanzure> yeah :( 11:56 < epitron> there's probably something for those 11:56 < kanzure> not to my knowledge 11:56 < kanzure> unless you want to go proprietary 11:56 < kanzure> which, i don't 11:56 < epitron> rsync with snapshots? :) 11:56 < kanzure> you are crazy 11:56 < epitron> snapshots are hardlinks 11:56 < epitron> they're quite efficient 11:57 < epitron> share those inodes 11:57 < epitron> i guess rsync snapshots don't support modifying one byte in a big file 11:57 < kanzure> joe is probably going to suggest revision control on top of RAID3 or something crazy like that 11:57 < epitron> it supports synching that one byte efficiently, but probably not storing it 11:57 < kanzure> how about this: no binary files to describe your warmachine 11:57 < kanzure> i like that policy. 11:57 < kanzure> yeah, i think i'll adopt it 11:58 < epitron> that sounds good! 11:58 < kanzure> hehe 11:58 < epitron> also who's making warmachines here? 11:58 < kanzure> epitron: DARPA thing for SKDB 11:58 < epitron> you're getting funded by darpa? 11:58 < kanzure> hopefully 11:58 < epitron> yeee :) 11:59 < epitron> btw, this is an interesting alternative SCM: http://www.fossil-scm.org/index.html/doc/tip/www/index.wiki 11:59 < kanzure> they want to support "complex, realistic cyber-electro-mechanical war fighting machines" 11:59 < epitron> zed shaw uses it for his projects... (eg: mongrel2) 11:59 < kanzure> "fossil scm" is an awful name 11:59 < epitron> it's an analogy for old commits 12:00 < epitron> what's neat about it is that it's kinda a hybrid between svn and git... optional centralization for efficiency 12:00 < epitron> but still distributed 12:00 < epitron> and it has a built-in wiki/bug tracker 12:00 < epitron> this is a fossil webserver: http://mongrel2.org/home 12:00 < kanzure> epitron: i like bugs everywhere 12:00 < epitron> all the pages are fossil wiki pages 12:00 < kanzure> it keeps your bug state in your dcvs 12:00 < kanzure> under a folder called .be/ 12:00 < epitron> yeah, fossil does that too 12:00 < epitron> and the wiki 12:01 < kanzure> reprap.git on diyhpl.us doubles as an ikiwiki 12:01 < epitron> okay 12:01 < kanzure> http://diyhpl.us/reprap/ 12:01 < kanzure> actaully wait 12:01 < kanzure> jrayhawk: why is that content still there? jmil changed the repo i thought 12:02 -!- mheld [~mheld@c-75-69-89-109.hsd1.ma.comcast.net] has joined #hplusroadmap 12:04 -!- mheld [~mheld@c-75-69-89-109.hsd1.ma.comcast.net] has left #hplusroadmap [] 12:13 -!- jennifer2 [~jennifer@c-67-180-253-94.hsd1.ca.comcast.net] has joined #hplusroadmap 12:28 < jmil> kanzure: i just modified the reprap.git folder, nothing else 12:28 < kanzure> manually? 12:28 < jmil> i don't know what ikiwiki is, nor what you mean by the entire repo is also a wiki 12:28 < jmil> figured you would tell me eventuallly 12:28 < kanzure> ikiwiki is a front-end to the wiki 12:28 < kanzure> ugh i screwed that up 12:28 < kanzure> ikiwiki is a front-end to the git repository 12:29 < jmil> right, like this: 12:31 < kanzure> ? 12:31 < jmil> ugh can't find the link 12:31 < jmil> but i thought you had it working with cgit 12:31 < jmil> something or other 12:31 < jmil> i was browsing the repo on your site 12:31 < kanzure> right, cgit is another front-end 12:31 < jmil> going afk for a bit... 12:40 < epitron> ikiwiki converts some datafiles (that can be stored in a git repo) into html pages 12:40 < epitron> apparently you can edit these datafiles somehow 12:40 < epitron> :D 12:43 -!- gnusha_ [~gnusha@131.252.130.248] has joined #hplusroadmap 12:43 -!- Topic for #hplusroadmap: http://www.youtube.com/watch?v=-n39RK4inzg http://designfiles.org/~bryan/hplus-summit-2009/ http://heybryan.org/ 12:43 -!- Topic set by kanzure [] [Mon Jan 11 11:14:52 2010] 12:43 [Users #hplusroadmap] 12:43 [ [X-Pro]Endos] [ davemenninger] [ fenn ] [ jmswatcher ] [ nchaimov ] [ QuantumG ] 12:43 [ Alystair ] [ dbolser ] [ ferrouswheel] [ joshcryer ] [ niftyzero ] [ saurik ] 12:43 [ any19811340 ] [ drazak ] [ ghchinoy ] [ jrayhawk ] [ nimak ] [ splicer ] 12:43 [ augur ] [ dustbin ] [ gnusha ] [ kanzure ] [ nsh ] [ superkuh ] 12:43 [ bkero ] [ egeste ] [ gnusha_ ] [ kardan ] [ Overand ] [ ToyKeeper ] 12:43 [ CIA-36 ] [ elmom ] [ Ian_Daniher ] [ kive ] [ panax ] [ uniqanomaly__] 12:43 [ clemux ] [ ENKI-][ ] [ JayDugger ] [ kristianpaul] [ patrickmclaren] [ Utopiah ] 12:43 [ clever ] [ epitron ] [ jennifer2 ] [ mage2 ] [ Phreedom ] 12:43 [ Daeken ] [ eridu ] [ jmil ] [ masked ] [ phryk ] 12:43 -!- Irssi: #hplusroadmap: Total of 52 nicks [0 ops, 0 halfops, 0 voices, 52 normal] 12:43 -!- Channel #hplusroadmap created Fri Jan 29 23:48:24 2010 12:43 -!- saurik_ [~saurik@carrier.saurik.com] has joined #hplusroadmap 12:43 -!- jrayhawk_ [~jrayhawk@orchitis.omgwallhack.org] has joined #hplusroadmap 12:43 -!- Irssi: Join to #hplusroadmap was synced in 7 secs 12:47 -!- saurik [~saurik@carrier.saurik.com] has quit [Remote host closed the connection] 12:47 -!- jrayhawk [~jrayhawk@orchitis.omgwallhack.org] has quit [Remote host closed the connection] 12:47 -!- gnusha [~gnusha@131.252.130.248] has quit [Remote host closed the connection] 12:57 < jmil> kanzure: found it: https://secure.diyhpl.us/cgit/reprap/ 12:57 < jmil> that one works 12:59 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Remote host closed the connection] 12:59 < jmil> kanzure: this command works now too: 12:59 < jmil> git clone http://diyhpl.us/reprap.git 13:01 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 13:06 < kanzure> hi jrayhawk_ 13:06 < kanzure> jrayhawk_: there were notes and questions i left for you http://gnusha.org/logs/2010-09-02.log 13:08 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Ping timeout: 245 seconds] 13:14 -!- genehacker [~chatzilla@wireless-128-62-175-233.public.utexas.edu] has joined #hplusroadmap 13:14 -!- amaruk [~freeze@p54B29108.dip0.t-ipconnect.de] has joined #hplusroadmap 13:15 < kanzure> oh shit, solidworks 2011 13:15 < kanzure> http://solidworkslaunch.com/ 13:16 -!- augur [~augur@129-2-175-79.wireless.umd.edu] has quit [Ping timeout: 264 seconds] 13:17 -!- amaruk [~freeze@p54B29108.dip0.t-ipconnect.de] has left #hplusroadmap [] 13:19 < jrayhawk_> oh hewwo 13:20 -!- jrayhawk_ is now known as jrayhawk 13:23 < jmil> kanzure: is jrayhawk server admin for gnusha? 13:23 < jrayhawk> Yeah. 13:23 < jmil> niiiice. hi jrayhawk good to meet u 13:24 < jrayhawk> Insofar as I occasionally break stuff claiming to make improvements. 13:25 < jrayhawk> So do you want Ikiwiki up and running or not? 13:25 < jmil> i dunno kanzure has plans for reprap git hosting, i think ikiwiki is part of that, but i dunno how it works with a git repo 13:25 < jmil> kanzure: this is excellent written piece: http://gnusha.org/ 13:25 < jmil> well done sir 13:26 < jrayhawk> Normally there's a post-update hook that checks the log for updated files and renders and/or publishes those to a directory tree. 13:27 < jrayhawk> Optionally, there is also a web interface for editing the contents of the tree and commiting them back to the git repository. 13:28 -!- genehacker [~chatzilla@wireless-128-62-175-233.public.utexas.edu] has quit [Remote host closed the connection] 13:29 < jrayhawk> There are some compelling features within Ikiwiki for managing and aggregating data stored within markdown files for whatever purposes you would like; it's often used for blogging and bug tracking. 13:29 < kanzure> jrayhawk: maybe the reason why the wiki front hasn't been changed is because jmil changed the hooks 13:29 < kanzure> makes sense to me. 13:29 < jrayhawk> Yeah. I can fix that if you want. 13:29 * kanzure nods 13:29 < kanzure> please? 13:30 < jmil> oh don't we need my hooks directory instead of what was there? kanzure you said i think that i should replace the reprap.git directory 13:30 < jmil> so that's what i did 13:30 < kanzure> do you have any special hooks of your own? 13:30 < jmil> what do hooks do? ;) 13:30 < jmil> i actually don't know 13:30 < jmil> i just copied my git svn directory up there 13:31 < kanzure> when you push changes to a git repository, there are "post commit hooks" that are "hooked" into the process 13:31 < jmil> correction: i was able to rename the hooks directory to hooks_OLD. actually tho kanzure deleted it for me, so blame him :) 13:31 < jrayhawk> I'm working on it. 13:31 < kanzure> these can do just about anything- i used to have some that would tweet on twitter and complain in here (in this channel) 13:31 < jmil> kanzure: ah. so the are like cgi scripts that jump to work 13:31 < jmil> cool 13:31 < kanzure> there's also pre-commit hooks you can throw in 13:31 < jrayhawk> The most common hook is to send commit messages to a mailing list. 13:32 < jmil> niiice 13:32 < jmil> kanzure: is the idea to give every reprap developer a login account to this git server? 13:32 < jrayhawk> The scariest hooks are the ones that reject commits on the basis of static analysis or coding style restrictions 13:32 < kanzure> jmil: i'd be happy to 13:32 < kanzure> "your code sucks" would be an awesome hook 13:32 < jmil> kanzure: wow that would be cool 13:32 < jmil> lol good hook too 13:33 < jmil> it would be cool if reprap dev's had login/commit access 13:33 < kanzure> that can be done 13:33 < jmil> ya i think that tiny shell would be fine for just commit access 13:33 < jrayhawk> I assume they only need commit access? 13:33 < jrayhawk> Yeah, you can do that 13:33 < kanzure> pinyshell 13:33 < kanzure> jmil: i guess i should have mentioned that to you 13:33 < jmil> ya that's it. i couldn't remember what you said 13:33 < kanzure> i know you logged in and were wondering wtf 13:33 < jmil> and you got my vi skillz pulled out of the back of my memory. thanks! 13:33 < jmil> lol ya 13:33 < jrayhawk> jmil: ssh into gnusha and run 'newuser' 13:34 < kanzure> jmil can do that? i thought sudoers only 13:34 < jrayhawk> for whomever you want to create an account. 13:34 < jrayhawk> Oh, aybe. Hmm. 13:34 < jmil> oh that creates new pinyshell user account? 13:34 < kanzure> yeah it's sudoers only 13:35 < jrayhawk> Ikiwiki should be working again, FWIW 13:35 < jrayhawk> I'll go make that command workable for you. 13:35 < kanzure> jrayhawk: wait, shouldn't it to be sudoers only? 13:35 < kanzure> i'm not really sure i care either way 13:35 < jrayhawk> I'll add an exception for jsmiller. 13:36 * kanzure shrugs 13:36 < kanzure> okay 13:38 < jrayhawk> Okay, that should work for the jsmiller user, now. 13:38 < jrayhawk> jmil: You can assign an arbitrary password and users can log in with and change with 'passwd' 13:39 < jrayhawk> also, if you want to play around with ikiwiki without polluting reprap.git with commits, you can 'newrepo whatever-you-want-it-to-be-named' 13:39 < kanzure> when it asks for an email address for the user 13:39 < kanzure> it would be nice if you use their actual email address :P 13:39 < kanzure> so that we can keep track of it 13:40 < jrayhawk> oh yeah, 'addaccess reprap whatever-username-you-created' will also be necessary 13:41 < kanzure> hm i think i forgot that part for jmil 13:41 < jrayhawk> He's the owner, it looks like. 13:41 < kanzure> well that's weird 13:41 < jmil> ok, we'll have to ask adrian how he wants to handle things 13:41 < jrayhawk> oh, no, i guess not 13:41 < jrayhawk> I'll make him the owner. 13:41 < jmil> wait, owner of what? 13:41 < jmil> teh reprap.git dir? 13:41 < jrayhawk> Yeah. 13:41 < jmil> ah 13:42 < jmil> then i could rename it 13:42 < kanzure> why would you rename it 13:42 < jrayhawk> Looks like that directory isn't shared. I should fix that, too. 13:42 < kanzure> i think what happened was that jmil just overwrote the entire directory 13:42 < jmil> my server at home went on the fritz, but it's the one i was using to pull makerbot SVN down into git, then push back up onto github every 5 min 13:42 < kanzure> with his scp move 13:42 < uniqanomaly__> "How God propelled Stephen Hawking into the bestsellers lists 13:42 < uniqanomaly__> Tim Radford suspects divine intervention was partly responsible for the miraculous success of A Brief History of Time" 13:42 < uniqanomaly__> omfg 13:43 < jmil> uniqanomaly__: i heard he was in a black hole for a while. that did it. 13:43 < jmil> kanzure: i was trying to avoid overwriting the directory, so i was trying to rename it and push a new one up, but i didn't have write access to the parent dir 13:44 < jmil> i don't need to rename it now tho, you are right 13:44 < jmil> damage already done 13:44 < kanzure> O_o 13:44 < jmil> sorry 13:44 < kanzure> are you one of those users that do crazy things? :) 13:44 < kanzure> like connect to irc via cronjobs 13:45 < jrayhawk> I'm going to break that git repo for about fifteen minutes to sort some problems out. 13:45 < jmil> i am a cronjob. you have just failed the turing test. do not pass go. do not collect $200. 13:45 -!- augur [~augur@129-2-175-79.wireless.umd.edu] has joined #hplusroadmap 13:45 < jmil> jrayhawk: you can delete the whole thing, i have the real one in my home dir on gnusha now 13:46 < jmil> i was trying to learn how to push it --mirror into that git dir 13:46 < jmil> kanzure: no i won't do crazy things, for the most part ;) 13:49 < kanzure> i only act like an admin 13:49 < kanzure> oops, scroll error 13:53 < jmil> your debian-fu is awesome, this is my first time on a debian box. will have to nix ubuntu for debian on our lab computer 13:53 < kanzure> ubuntu is based off of debian 13:54 < kanzure> jmil: so, the gnusha.org text is a little old by now 13:54 < kanzure> it was written earlier this year as a project update 13:54 < kanzure> there's a few additional parts of the story that are worth telling 13:55 < kanzure> i have some funding for these and other crazy projects 13:55 < kanzure> so i was going to invest $50 in anybody that would write me a grant proposal 13:55 < kanzure> and then if they complete and carry it out, i'd pay them $100 next time, and keep on doubling it 13:55 < kanzure> until i run out of money or until they fail (or something) 13:55 < kanzure> i can afford to waste $50 (it's like a bad dinner or something) but not $50,000 on a failed super reprap or whatever 13:59 < jmil> wow that's dangerous 13:59 < jmil> 13:59 < kanzure> heh :) 13:59 < kanzure> there's a reason i haven't announced it yet though 13:59 < kanzure> i've been working on some software infrastructure issues.. 13:59 < dustbin> I like the idea 13:59 < kanzure> jrayhawk has been working on piny (which actually saves me a huge pain in the ass) 14:00 < jmil> you just did announce it lol 14:00 < kanzure> and in the meantime, i've been working on a usable open source CAD kernel 14:00 < jmil> first post! 14:00 < kanzure> jmil: well, i mean, "money serving time is open"-sorta-announce 14:00 < jmil> CAD kernel? 14:00 < kanzure> yes 14:00 < jmil> what does that do? 14:00 < jmil> like linux kernel? 14:00 < kanzure> it's a ridiculously technical piece of software that you shouldn't have to worry about 14:00 < kanzure> basically, it's a way to design mechanical components 14:00 < jmil> now i have to know 14:01 < jmil> like openscad? 14:01 < kanzure> have you ever used solidworks, autocad, pro/e, or catia? 14:01 < jmil> and mscad? 14:01 < jmil> kanzure: i'm all blender, baby 14:01 < kanzure> sorta like openscad, except openscad only exports STL files (which are not CAD) 14:01 < kanzure> blender is not CAD either 14:01 < kanzure> in fact, i wrote a whole wiki page on this 14:01 < jmil> you can export .DXF from blender 14:01 < kanzure> http://diyhpl.us/diyhpluswiki/cadfaq/ 14:01 < kanzure> .DXF is rarely CAD 14:01 < jmil> lucy, you have some 'splainin to do 14:01 < kanzure> for various reasons, trust me ;) 14:02 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 14:02 < jmil> omg did you just write that whole web page in 17 seconds? 14:02 < kanzure> i would love to use openscad but the fact that it only does STL output makes it more or less useless for my purposes 14:02 < jrayhawk> this does not look like a git-svn repository 14:02 < jmil> so you need to join openscad dev team 14:02 < jmil> jrayhawk: howso? 14:03 < jmil> jrayhawk: check the one in my home dir 14:03 < kanzure> jmil: openscad is just cgal+opencsg+otherstuff.. opencsg isn't what i want, and CGAL certainly isn't either 14:03 < jmil> ok i will need to do some more reading 14:03 < dustbin> kanzure: I would be interested in investing some programming time working on a CAD project 14:03 < kanzure> jmil: there are two open source CAD kernels out there though 14:03 < kanzure> dustbin: one sec 14:03 < kanzure> jmil: http://brlcad.org/ 14:03 < kanzure> jmil: http://opencascade.org/ 14:04 -!- jennifer2 [~jennifer@c-67-180-253-94.hsd1.ca.comcast.net] has quit [Ping timeout: 260 seconds] 14:04 < kanzure> and i guess http://pythonocc.org/ for python wrappers for opencascade 14:04 < kanzure> for various reasons neither kernel is quite what i need 14:04 < kanzure> brlcad doesn't export STEP, for instance 14:04 < jmil> "user Bryan Bishop wrote a script in Perl to create geometry via MGED. Bryan shows in http://brlcad.svn.sourceforge.net/viewvc/brlcad/brlcad/trunk/src/proc-db/spiral.pl how to procedurally generate a spiraling pattern of hollowed-out boxes." 14:04 < jmil> niiiice 14:04 < kanzure> and opencascade is such a bloated mess and retarded that it's not worth spending any more of my time with 14:04 < kanzure> jmil: oh shit, that's right! 14:04 < kanzure> i forgot about that 14:05 < kanzure> it was for the automatic design of a spiral microfluidic algae/water separator 14:05 < kanzure> dustbin: so, uh, the CAD kernel i've been writing is all in python 14:05 < kanzure> the main goal of the kernel is STEP export, nothing else matters to me at the moment 14:05 < dustbin> ok 14:06 < kanzure> in the future i'll implement an actual SDAI interface for STEP standards compliance, but right now it's just raw and whatever-the-hell-makes-it-work 14:06 < jmil> http://www.step.org/ ? 14:06 < jmil> lol 14:06 < kanzure> jmil: http://en.wikipedia.org/wiki/STEP-File 14:06 < kanzure> dustbin: the idea is to have a python interpreter open and just type out commands 14:06 < kanzure> sphere.fuse(box) 14:06 < dustbin> ok 14:06 < kanzure> and such.. 14:06 < kanzure> but nobody is going to use this if they can't see what they are making 14:07 < kanzure> so for the past week i've been dabbling with an opengl/glut or maybe wxpython viewer/visualizer 14:07 < jrayhawk> hum dee dum 14:07 < kanzure> dustbin: do you have any experience with python or OpenGL or wxWidgets/wxPython? 14:07 < dustbin> I have opengl experience not much python experience though 14:07 < kanzure> cool! 14:08 < kanzure> so uh, i actually have a question then 14:08 < dustbin> heh k 14:08 < kanzure> i've been making the visualizer and i need to figure out the user interaction model 14:08 < kanzure> i have no idea wtf i want 14:08 < kanzure> like, should i implement rotation with arrow keys? or is that panning only 14:09 < dustbin> I can only offer opinions on that stuff 14:09 < kanzure> please :) 14:09 < kanzure> jrayhawk: was that hum dee dum you being annoyed 14:09 -!- augur [~augur@129-2-175-79.wireless.umd.edu] has quit [Ping timeout: 240 seconds] 14:09 < dustbin> I don't have any research knowledge to back it up atm, but IMO you should use meta keys and mouse or arrow keys 14:09 < jrayhawk> no i am waiting for a push and checkout to finish because this repo is the size of the moon 14:10 < dustbin> so shift plus hold left mouse and drag causes one movement type 14:10 < dustbin> alt plus left click and drag does something else 14:10 < dustbin> the same thing should happen with shift plus arrow keys 14:11 < kanzure> jmil: http://en.wikipedia.org/wiki/Solid_modeling 14:11 < jrayhawk> hmm. I should add spaces to the list of allowed ikiwiki characters, apparently. 14:11 < jmil> jrayhawk: it's all the binaries. it's because they are design files so we don't want to trash them. even the old ones should be saved for posterity 14:11 < jmil> probably 14:11 < jmil> not really my decision to make tho, prolly Adrian would agree tho as he said his plan would be to migrate all commits as well 14:12 < kanzure> jrayhawk: i was meaning to ask you about this 14:12 < kanzure> the DARPA project wants 200 MB *minimum* expected repository size 14:12 < kanzure> and seeing how reprap.git is 1 GB (although someone said yesterday 2 GB) 14:12 < kanzure> i suspect that some are multiple terabytes for any modern war machine 14:12 < kanzure> what do i dooooooo 14:14 < jmil> kanzure: what darpa project? 14:14 < kanzure> jmil: DARPA announced that they wanted to fund a "sourceforge for vehicles and other cyber-electromechanical systems" (aka SKDB) 14:14 < kanzure> so i have a team going after it.. 14:14 < jmil> kanzure: ya reprap.git is 1 GB, i thought 2 GB because i was also measuring the current working directory 14:15 < jmil> kanzure: how can i help? 14:15 < kanzure> do you know a prime contractor? 14:15 < kanzure> i have a few leads at the moment but i don't know which (or if any) will work out 14:16 < jmil> we can write a shell script if you just need some prime numbers 14:16 < jmil> you mean to ironworks things? 14:16 < jmil> for robotz? 14:16 < kanzure> ironworks? 14:16 < jmil> i just assume you mean big metal robotz 14:17 < kanzure> in the defense contracting industry, a "prime contractor" is the company that formally submits the offer/proposal/grant-application to DARPA or the DoD 14:17 < kanzure> little small time guys usually don't submit it directly 14:17 < jmil> oh. university then? 14:17 < kanzure> that's one of my options but no, not necessarily 14:17 < kanzure> an example of this would be if i knew some guys at boeing' 14:17 < jmil> oh some of the reprap guyz can help with that maybe 14:17 < jmil> ask in #reprap 14:18 < kanzure> ask for what? 14:18 < jmil> contacts at boeing 14:18 < kanzure> okay 14:19 < kanzure> jmil: btw, i should mention that management of the gada prize is being transferred over to me 14:20 < kanzure> just so you know.. 14:20 < jmil> oh shit lol 14:20 < jmil> good to know your conflicts of interest 14:21 < kanzure> what's the conflict? 14:21 < kanzure> i was just trying to explain my reaction in #reprap 14:21 < jmil> oh lol 14:22 < jrayhawk> git://diyhpl.us/reprap 14:22 < kanzure> woo it's working 14:22 < kanzure> Receiving objects: 8% (3898/44969), 2.82 MiB | 352 KiB/s 14:22 < kanzure> not a bad speed 14:23 < jrayhawk> https://secure.diyhpl.us/cgit/reprap/ 14:23 < jrayhawk> http://diyhpl.us/reprap/ 14:23 < jmil> jrayhawk: the git:// link now updates object pull, thanks!!!!!!!!!!! 14:23 < jmil> i'm getting 1.4 MiB/s 14:23 < jmil> excellent 14:23 < kanzure> !!11111oneoneone 14:24 < jmil> jrayhawk: oh shit, you already made a commit to it, we have to reverse that 14:24 < jrayhawk> whyfor 14:24 < kanzure> what's wrong with a commit? 14:24 < jmil> we need master to be a pristine pull and rebase from the SVN because we are not ready to finalize the move yet 14:24 < jrayhawk> oh, troublesome 14:24 < kanzure> huh? 14:24 < kanzure> so then what was the point 14:24 < kanzure> er, i mean.. 14:24 < kanzure> didn't people already fork the repos on github anyway 14:25 < jrayhawk> Horm. 14:26 < jrayhawk> I'll go see if I can fix that. 14:32 < jmil> kanzure: no it's never been on github 14:32 < kanzure> my bad 14:32 < kanzure> makerbot has been, i think? 14:32 < jmil> github only allows 10 collaborators (read: commit access) and 2 GB max 14:32 < jmil> they have been talking about transitioning to git for a year or two 14:33 < jmil> now they are finally going to do it, but i don't want it to go badly, it has to be right the first time 14:33 < kanzure> yeah i saw adrian just raw committing svn files? 14:33 < kanzure> i mean, without commit history 14:33 < jmil> so we need complete SVN history, easy commit access 14:34 < jmil> ya kanzure check that thread... i corrected him that we can get total SVN history and he said 14:34 < jmil> I'm not proposing to do it this way as the final solution - I'm just 14:34 < jmil> experimenting at the moment. 14:34 < jmil> When it's all sorted, I'll re-do it with the svn history intact. 14:34 < jmil> Indeed, what I may well do is to re-organise the svn repository to the 14:34 < jmil> structure of the git repository as the last thing done in svn, then 14:34 < jmil> move the lot across. 14:34 -!- jennifer2 [~jennifer@c-67-180-253-94.hsd1.ca.comcast.net] has joined #hplusroadmap 14:36 < jmil> so my planned infrastructure is one server (yours?) where the main git repo lives, all official devs have commit access there 14:36 < kanzure> jmil: i was wondering, can i have a shell account on seas.upenn.edu (or w/e) for grabbing papers? 14:36 < jmil> kanzure: i can't do that because you said that... 14:36 < jmil> i can pull papers for you and send them to you if you send me the links to the papers 14:36 < jmil> because it's all for academic purposes... 14:36 < kanzure> "all official devs have commit access there" sometimes it's just one dev with commit access, but either way is ok 14:36 < kanzure> jmil: i usually grab about 100 papers/day 14:37 < kanzure> i recommend not offering your manual services :D 14:37 < jmil> kanzure: i'm trying to keep it the same as what I think reprap SVN is used to, and it has multiple committers 14:37 < kanzure> jmil: sounds good to me 14:37 < jmil> then i want a cron job to push changes onto github 14:37 < jmil> so there is a read-only there for the vast majority of other uses to fork and pull 14:37 < kanzure> how about a commit hook :) 14:38 < kanzure> others can fork, pull and clone from the gnusha/diyhplus domain btw 14:38 < jmil> kanzure: oh ya, a hook! but prolly a cron job just in case, right? 14:38 < kanzure> if you do a hook then whenever something gets committed, github gets updated too 14:38 < jmil> kanzure: but it's not as good as including all the people that are very familiar with github 14:38 < jrayhawk> Only if github is unreliable. 14:38 < jmil> oh github is reliable. i have the makerbot repo working this way, with a cron job 14:38 < jmil> the only difference is people would have commit access 14:39 < QuantumG> kanzure: you should track down Ben Brockert .. he'd be local to you now (working at Armadillo Aerospace) and I just saw him asking about laser cutting. 14:39 < kanzure> Ben Brockert 14:39 < kanzure> got it 14:40 < QuantumG> you could invite him to some techshop event or something.. if you've found any of those that are worthwhile 14:40 < kanzure> QuantumG: you mean in the austin area? 14:40 < kanzure> also do you prefer your biodome or gmail address 14:40 < QuantumG> gmail 14:40 < QuantumG> I dunno how far he is away from you.. but it's the same state. 14:41 < QuantumG> (biggest state in the US, but anyway..) 14:41 < dustbin> about 3 to 4 hours 14:42 < kanzure> wait where is he? 14:42 < dustbin> armadillo indicates dallas 14:42 < kanzure> hah, yeah that's not happening 14:42 < kanzure> QuantumG: but thanks :) 14:43 < jrayhawk> Okay, that git repo should be in better shape. 14:43 < QuantumG> just an idea. know anyone in Texas that does laser cutting/drilling? 14:43 < jmil> jrayhawk: now how do i not break it by committing to it? 14:43 < kanzure> QuantumG: tons.. i just wrote out this huge email to him and discarded it 14:43 < jrayhawk> use git tools rather than manual filesystem operations :) 14:43 < jmil> jrayhawk: ok 14:44 < jmil> jrayhawk and kanzure this is what i follow for makerbot and reprap svn repo: 14:44 < jmil> http://www.fnokd.com/2008/08/20/mirroring-svn-repository-to-github/ 14:44 < jrayhawk> So, git push ssh://diyhpl.us/srv/git/reprap master or what have you 14:44 < jmil> perfect 14:44 < kanzure> jrayhawk: you should probably read that fnokd.com link and complain about how wrong it is or something 14:44 < jmil> then basically i envision every reprap dev has his/her own branch. Then only one committer is responsible for merging to master 14:44 < jmil> but all branches are pushed onto github 14:44 < kanzure> oh wait it looks relatively right 14:45 < jmil> yay!!!!!!!!!!!!!!1111!!1!1one 14:45 < jrayhawk> There's no real branch security for git, so I'm not sure how well your plan will work out in practice. 14:45 < jrayhawk> I guess it depends on the people involved. 14:46 < kanzure> won't someone say "hey that shouldn't be in master"? 14:46 < jrayhawk> The git model is more designed for people to each maintain their own repositories. 14:47 < jmil> jrayhawk: you gotta see linus' philosophy on it 14:47 < jmil> he has it *exactly* right 14:47 < jmil> that's what we will do here 14:47 < jmil> kanzure: you too 14:48 < kanzure> hm? 14:48 < kanzure> oh, you mean i should see linus talk about git 14:48 < jmil> we're taking linus' philosophy 14:48 < jmil> ya 14:48 < kanzure> yeah, i remember his talks on git 14:50 < jmil> kanzure: so when people say "that shouldn't be in master" then it's up to the maintainer to make a decision. since anyone can be a node and all git repos are equal, it's up to the community to prune or support the "main" master trunk 14:50 < QuantumG> Rockwell, Texas 14:51 < QuantumG> guessing that's a suburb of Dallas 14:51 < jmil> then if something breaks, the maintainer doesn't have to fix it. the maintainer instead says to the committer... this is broken, please fix it, rebase, and send back to me 14:51 < jmil> it's less of a problem with reprap where it's mostly design files tho 14:51 < jmil> so i think (maybe naively) taht it has a better chance of working for us 14:51 < jmil> it's worked very well for MakerBot/G3Firmware, for which I have commit access 14:52 < kanzure> heh i can imagine some flame wars now 14:52 < kanzure> about whether or not to commit before or after physical test 14:52 < kanzure> never commit untested code, right? right? 14:52 < jmil> lol 14:52 < jmil> basically, here: 14:52 < jmil> http://github.com/makerbot/G3Firmware/network 14:52 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Ping timeout: 245 seconds] 14:53 < jmil> a bunch of us can commit onto github, but phooky is really the one with his head wrapped around the code best 14:53 < jmil> so he prunes or rejects our changes as he sees fit. it's up to us to convince him 14:53 < jmil> or do better and replace his value 14:53 < jmil> but it'd be impossible to do better and replace him, so instead socially he ends up holding the keys 14:53 < kanzure> neato: 14:53 < kanzure> Title: Growth and replication of red rain cells at 121 oC and their red fluorescence 14:53 < jmil> link? 14:53 < kanzure> pdf: http://arxiv.org/pdf/1008.4960v1 14:54 < kanzure> http://arxiv.org/abs/1008.4960 14:54 < jmil> what's a rain cell?? 14:54 < jmil> oh, a thermophile 14:55 < kanzure> commentary: http://science.slashdot.org/story/10/09/02/1958231/DNA-Less-Red-Rain-Cells-Reproduce-At-121-C 14:55 < kanzure> red rain is something like acid rain i think? 14:55 < kanzure> so yeah, extremophile 14:56 < superkuh> Or someone mistaking complex lipid structures for living cells. 14:56 < jmil> kanzure: lol they autoclave, then claim the reaction is simply from temp 14:56 < jmil> but autoclave is very high pressure too 14:56 < jmil> they completely ignore 14:56 < jmil> 14:57 * jmil smacks head 14:57 < jmil> ouch 14:58 < jmil> 121 C for 2 hours at atmospheric pressure == no liquid growth medium left 14:58 < jmil> lol 14:58 < superkuh> Unless you are refluxing it... 14:59 < jmil> counting cells is also not the *best* way to do this, you need to do optical absorbance measurements 14:59 < jmil> superkuh: good point 14:59 < kanzure> "My research field is hyperthermophilic Archaea that grows at 90C or more and I know the existence of microbes that can grow at even higher temperatures, so this part of the claim is feasible." i want some 15:00 < jmil> why do you want some? they might be toxic 15:00 < jmil> why didn't they isolate the DNA???? 15:00 < kanzure> jmil: read the comments.. wikipedia delivers, as usual 15:01 < jmil> you mean slashdot? 15:01 < kanzure> yes 15:01 < kanzure> but the comments were talking about the wikipedia article 15:02 < jmil> link? 15:04 < kanzure> http://science.slashdot.org/comments.pl?sid=1775196&cid=33456718 15:08 < jmil> i don't see wikipedia link 15:08 < jmil> but this quote is good: To go from "our test found no DNA" to "there is no DNA" to "they must be extraterrestrial" to "they look like the dust clouds in Monocerous" is a series of leaps that go wayyy ahead of the available evidence, in my view. 15:09 < kanzure> read the child comments 15:09 < jmil> i think prions should have voting rights in the US. 15:09 < jmil> just saying... 15:10 < jmil> i had so much trouble getting them to sign an NDA tho. 15:11 < jmil> kanzure: he's also not saying they could possibly grow in the rain or the clouds, as they need LB medium 15:12 -!- shepazu [~schepers@81.253.56.19] has joined #hplusroadmap 15:12 < jmil> I like the TEM of the emulsion though. it's easy to make lipid bilayers and double membranes in a test tube. any size you want. 15:14 < QuantumG> prions are overqualified to vote in the US 15:16 -!- shepazu [~schepers@81.253.56.19] has quit [Ping timeout: 240 seconds] 15:16 -!- panax [panax@goldstandard.eng.usf.edu] has quit [] 15:23 < jmil> QuantumG: i resemble that comment 15:24 < QuantumG> heh 16:07 -!- ybit [~ybit@unaffiliated/ybit] has joined #hplusroadmap 16:10 < [X-Pro]Endos> anyone here know a little about setting up pbs for the management of scientific computing clusters? 16:15 -!- jennifer2 [~jennifer@c-67-180-253-94.hsd1.ca.comcast.net] has quit [Ping timeout: 240 seconds] 16:25 < jmil> kanzure: don't you have pub access through UT? 16:32 < kanzure> i can always use backups 16:33 -!- mheld [~mheld@pool-173-76-224-45.bstnma.fios.verizon.net] has joined #hplusroadmap 17:02 -!- mheld [~mheld@pool-173-76-224-45.bstnma.fios.verizon.net] has left #hplusroadmap [] 17:04 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 17:08 -!- klafka [~klafka@cpe-66-66-5-254.rochester.res.rr.com] has joined #hplusroadmap 17:11 -!- wolfspraul [~wolfsprau@lucia.q-ag.de] has joined #hplusroadmap 17:30 < kanzure> jmil: would you like to email the reprap-dev list with the deets or should i? 17:32 < kanzure> let's flood this guy's inbox with suggestions for ridiculous courses 17:32 < kanzure> http://www.polymathlectures.org/mod/forum/discuss.php?d=20 17:32 < kanzure> "If you could learn any subject at this very moment without applying it toward a degree, which would it be?" 17:33 < jmil> kanzure: let's wait... i'd like to have a plan with Adrian first 17:33 < jmil> for migration and user management, commits, etc 17:34 < jmil> the problem is that neither of us know how people usually function with the SVN repo 17:34 < kanzure> presumably "poorly" 17:34 < jmil> prolly we'll keep master as the main tracking SVN pristine, then have a second branch people can commit to as they migrate to git 17:35 < jmil> then at some point adrian will pull the plug on SVN, and we'll tell people to commit to master isntaead 17:35 < jmil> instead 17:35 < jrayhawk> Is there some room I should monitor and provide information on infrastructure limitations and opportunities? 17:36 < jmil> we can try in #reprap but i don't think any of the committers are in there (even though 100 peeps) 17:36 < jmil> i asked earlier 17:37 < jrayhawk> BTW you are free to rape a shitload of bandwidth on that system. 17:37 < kanzure> fantastic 17:37 < jrayhawk> That colocation facility also hosts, e.g. freedesktop.org 17:38 < jmil> i asked anyone have commit access to SVN reprap repo? We're trying to setup the git version and wanted to get more info on the workflow for the SVN repo 17:38 < jmil> [8:37PM] jmil: so we can assist with the migration 17:38 < jmil> [8:37PM] jmil: it will be grand 17:39 < jmil> response: crickets 17:39 < kanzure> jmil: irc is a variable lag communication medium 17:39 < jmil> i'll just ask Adrian 17:39 < jmil> kanzure: #reprap is usually quite populous but it's mostly on Euro time 17:39 < jmil> so now's not the best time 17:43 < jmil> kanzure and jrayhawk i'm trying one more git-svn flag to try to get the author email address back... might work 17:43 -!- XPSkeith [8da459b8@gateway/web/freenode/ip.141.164.89.184] has joined #hplusroadmap 17:44 < kanzure> ah so it was there? 17:44 < kanzure> XPSkeith: hi 17:45 < XPSkeith> hi 17:46 < jrayhawk> From previous conversions I've seen done, you need to manually parse commits in order to preserve everything you'd want to, and it takes *forever* because SVN is painfully slow. 17:46 < jrayhawk> I can probably dig up some expertise on the subject if you want. 17:49 -!- any19811340 is now known as katsmeow 17:53 < jrayhawk> And you'd need to make a pretty sophisticated system if you wanted to move stuff back and forth. 17:55 -!- kristianpaul [~kristianp@190.7.138.180] has quit [Remote host closed the connection] 17:57 -!- kristianpaul [~kristianp@190.7.138.180] has joined #hplusroadmap 17:58 -!- XPSkeith [8da459b8@gateway/web/freenode/ip.141.164.89.184] has quit [Quit: Page closed] 18:07 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Remote host closed the connection] 18:08 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 18:13 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Ping timeout: 245 seconds] 18:14 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 18:14 -!- kristianpaul [~kristianp@190.7.138.180] has quit [Quit: leaving] 18:16 -!- kristianpaul [~kristianp@190.7.138.180] has joined #hplusroadmap 18:17 -!- klafka [~klafka@cpe-66-66-5-254.rochester.res.rr.com] has quit [Quit: This computer has gone to sleep] 18:20 -!- klafka [~klafka@cpe-66-66-5-254.rochester.res.rr.com] has joined #hplusroadmap 18:24 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Ping timeout: 245 seconds] 18:24 < jmil> jrayhawk: what do you mean we'd want to manually parse commits? 18:26 < jmil> jrayhawk can you make me a new .git folder to test a new git repo push? 18:26 < kanzure> jmil: just run "newrepo" 18:26 < jmil> do i have to cd into the git folder? I don't have write perms there 18:27 < kanzure> you could make the git folder in your home directory if it's non-critical that it's on /srv/git/ 18:27 < jmil> i want to test it in /srv/git with the cgit system in my browser 18:28 < kanzure> what repo name would you like 18:28 < jmil> reprap_test would be great, thanks! 18:29 < kanzure> done 18:29 < jmil> kanzure: i found the command to change adrian-bowyer@cb376a5e-1013-0410-a455-b6b1f9ac8223 to adrian-bowyer 18:29 < jmil> etc. 18:29 < kanzure> er, what about his actual email address? 18:29 < jmil> it's not in the SVN commits as far as i can tell 18:29 < kanzure> oh it's sourceforge.. the commits are probably just sourceforge usernames :( 18:29 < jmil> ya, that's what i figure 18:30 < kanzure> the commit *authors 18:30 < jmil> do you think the @ string means anything? 18:30 < kanzure> a funky hash no doubt.. i don't know what the meaning is, however 18:30 < jmil> they're all the exact same it seems like 18:30 < kanzure> for each author or for all authors? 18:30 < jmil> all authors 18:30 < kanzure> huh. uh.. 18:31 < jmil> how do i wipe the git repo in reprap.git and re-push up a fresh copy? 18:31 < kanzure> jrayhawk: did you do some work on reprap.git in particular? 18:31 < jmil> i tried git push --mirror but it didn't work, says everything is up to date 18:31 -!- kristianpaul [~kristianp@190.7.138.180] has quit [Remote host closed the connection] 18:32 < jmil> kanzure: ya jrayhawk said to keep from messing it up i need to use git instead of file copy/rm 18:32 < kanzure> so wait, you want to re-upload the repository? may i ask why 18:32 < jmil> see above... 18:33 < jmil> to re-upload a repo without the @ string 18:33 < jmil> but i don't want to undo jrayhawk 's efforts 18:33 < kanzure> there's a way to use git tree filters to edit all of the commit messages/authors 18:33 < kanzure> so if joe did something worth keeping, i can throw some filter-foo at it 18:34 < kanzure> and if not, i'll scratch the repository? 18:34 -!- kristianpaul [~kristianp@190.7.138.180] has joined #hplusroadmap 18:34 < kanzure> hi kristianpaul 18:34 < kanzure> jmil: how about you upload the fresh one to reprap_test for now? 18:34 < jmil> joe got the folder shared, the cgit working, the git daemon running correctly... 18:35 < jmil> ok, i'll try that 18:35 < kanzure> or do you need me to reset reprap_test first 18:35 < jmil> reset it how? 18:35 < kanzure> back to when i first created it - i haven't checked if you have done anything with it in the past few min 18:35 < jmil> i bet gnusha has fatter pipe than me here, so i'll prolly create a new one in my home dir on gnusha instead 18:36 < jmil> oh no didn't do anything yet 18:36 < kanzure> okay. 18:36 < kanzure> jrayhawk: reprap_test.git has a commit message that says the following: 18:36 < kanzure> commit f5d0b4f6fb0bd84a421046eb2c08b9d8df94405e 18:36 < kanzure> Author: root 18:36 < kanzure> Date: Mon Jul 19 00:08:14 2010 -0700 18:36 < kanzure> Initial content. 18:36 < kanzure> i'm guessing this is the ikiwiki initial content commit 18:39 < kanzure> uh i guess i was thinking the date was weird but nevermind, it's probably just a copied repo 18:39 < kanzure> which makes for interesting mergers in the future? 18:39 < jmil> lol 18:40 < jmil> i will wipe the dir? 18:40 < kanzure> which one? 18:40 < jmil> reprap_test.git 18:40 < kanzure> have at it 18:40 < jmil> i started a new svn fetch in a screen instance. did control-a-d to detach 18:41 < kanzure> right 18:41 < jmil> now when i type fg it doesn't get me back in! 18:41 < kanzure> screen -x 18:41 < jmil> it says no such job 18:41 < jmil> that got it thanks!!! 18:42 < jmil> oh i don't have perms to wipe that reprap_test.git 18:42 < kanzure> yeah.. one sec, i'll do it 18:42 < kanzure> wait, how wiped do you want it? 18:42 < jmil> but i'll need to be able to push git repo into there too... 18:42 < jmil> rm -rf 18:42 < jmil> unless it has setup stuff i can push on top of? 18:42 < kanzure> why does it have to be in /srv/ then? 18:43 < kanzure> oh right, cgit 18:43 < jmil> will cgit just automagically do everything for any .git it finds in /srv? 18:43 < kanzure> yes 18:43 < jrayhawk> Yeah, sorry, part of the pain in the ass of git and ikiwiki is not having initial content, so newrepo arbitrarily creates some. 18:43 < jmil> niiiice 18:43 < jrayhawk> I got rid of this on the main reprap.git 18:43 < jmil> what is reprap-backup.git? 18:43 < kanzure> probably joe being safe 18:43 < jmil> oh good :) 18:43 < jmil> thanks joe 18:44 < kanzure> so wait, you don't have push privledges on reprap_test.git? 18:44 < jmil> is joe jrayhawk ? 18:44 < jrayhawk> The old one you had in place with the wonky permissions. 18:44 < kanzure> jrayhawk: yes 18:44 < kanzure> jrayhawk: i created reprap_test.git with newrepo and addaccess; shouldn't this cover it? 18:44 < kanzure> i mean, the jsmiller user should be able to push to it right? 18:45 < jrayhawk> He might need to relogin since Unix permissions aren't dynamic. 18:45 < jmil> i think i need to git clone --mirror 18:45 < jrayhawk> Depends on what he's doing. 18:45 < kanzure> jrayhawk: i think he's running git svn and wants to push stuff into reprap_test.git 18:45 < jmil> yes that's it kanzure 18:45 < kanzure> i guess if you're running git svn it's going to create its own repository 18:46 < jrayhawk> Yeah, he's doing that locally on gnusha; he'll need to relogin. 18:46 < jrayhawk> Sorry, unix sucks. 18:46 -!- genehacker [~chatzilla@cpe-66-68-110-194.austin.res.rr.com] has joined #hplusroadmap 18:46 < kanzure> ah right because i added him to the user group 18:46 < jmil> ya i'm making repo in home 18:46 < kanzure> jrayhawk: in this case, reprap_test.git is already created, so his repo in ~ will have to be copied over to /srv/git/? i'm not sure on the flow in this case 18:47 < jmil> i was able to do it for reprap.git because i had write perms 18:47 < kanzure> if you relogin you'll have write permission 18:47 < jmil> so i rm -rf the contents of the folder, then git clone --mirror 18:47 < kanzure> aha you rm -rf'd.. 18:47 < jrayhawk> aaaaaaa 18:47 < jmil> ok thx 18:47 < kanzure> um 18:47 < jmil> um what? 18:48 < jrayhawk> oh, different folder 18:48 < kanzure> jrayhawk: right 18:48 -!- augur [~augur@208-58-6-161.c3-0.slvr-ubr1.lnh-slvr.md.cable.rcn.com] has joined #hplusroadmap 18:48 < kanzure> but still, rm -rf'ing stuff shouldn't be necessary 18:48 < jrayhawk> It's necessary for clones. 18:48 < jmil> ya i think so too 18:48 < jrayhawk> To get rid of clones, specifically. 18:48 < kanzure> i guess piny doesn't really let you pick a repository already in existence 18:48 < jrayhawk> Yeah, I should correct that. 18:48 < kanzure> ok. so rm -rf goodness for now 18:48 < jmil> better to be safe and clean since we want this to be source for lots of people 18:48 < kanzure> "goodness" 18:49 < jmil> if it were for just me i wouldn't care about dangling clones 18:49 < jmil> i've never met an rm -rf i didn't like 18:49 * jmil just wants to make sure he never gets admin privileges 18:50 < jrayhawk> I'm going to do some work to make this a little less screwy for you. 18:51 < jmil> ok, thx. maybe i can learn more about the hooks you did for reprap.git so we can move them to reprap_test.git, then potentially delete reprap.git and rename _test to reprap.git 18:51 < jmil> then we'd be golden, i think 18:51 < jmil> then we could delete the _backup.git 18:51 < jmil> and then make a new _backup.git 18:51 < jmil> and then we'd be ready to rock 18:52 < jrayhawk> why aren't you just pushing to reprap.git 18:52 < jmil> because i need to overwrite it completely, see above 18:53 < kanzure> he's redoing the svn conversion to fix the author strings 18:53 < kanzure> i suggested we just use a git tree filter 18:53 < jmil> all authors have an @HASH string 18:53 * kanzure shrugs 18:53 < jrayhawk> I think you can git push -f 18:53 < jmil> i don't know what is git tree filter, so i'm doing the hard way 18:53 < jmil> git push --mirror -f ? 18:53 < jrayhawk> What's the 'mirror' for? 18:53 < kanzure> jmil: filters are nasty and evil and only crazy people use them :P 18:53 < jmil> lol 18:54 < jmil> --mirror is supposed to overwrite the contents completely, and keep remote refs etc. 18:54 < jmil> but it wasn't working 18:54 < kanzure> "Lets you rewrite git revision history by rewriting the branches mentioned in the , applying custom filters on each revision. Those filters can modify each tree (e.g. removing a file or running a perl rewrite on all files) or information about each commit." 18:54 < jmil> because i tried that already 18:55 < jmil> i'm also making a text file (damn you vi!!!!!!!) so i can remember what i did 18:55 < jmil> and then i will write down what you did 18:55 < jmil> hopefully 18:55 < jmil> if i understand... 18:55 < jrayhawk> Why would a bare repo need a remote? 18:55 < jrayhawk> And you have no tags. 18:55 < jrayhawk> And only one branch. 18:57 < jmil> i'm considering the working tree in my home directory on gnusha the main tree where i'm slurping down the git svn 18:57 < jrayhawk> Yeah. 18:57 < jmil> i don't want to do this directly in the /srv/git folder 18:57 < jrayhawk> Yeah, that makes sense. 18:57 < QuantumG> heh, kanzure: Bryan, I'll contact you off list in the next couple of weeks. (Your website is timing out, btw.) .. there you go, answering the question asked has benefits 18:58 < jmil> but the stuff in /srv/git will eventually need to also know that there's an SVN remote 18:58 < jmil> so i was trying to mirror my home git repo into /srv/git 18:58 < kanzure> i'm trying to overcome my anxiety of redoing all of heybryan.org's url paths.. it's like, what, 5 years of posting links on the web? that will now all be broken if i finally fix my apache config? 18:59 < jmil> kanzure: you need perl 18:59 < kanzure> no, you never need perl 18:59 < jmil> lol 18:59 < kanzure> i really don't want to write redirects for each and every incoming link (a few hundred thousand) 19:00 < kanzure> oh well. i'll get over it. 19:00 < jmil> why you need to redo the url paths? just keep the old ones and make new ones fit the new schema 19:00 < jmil> then you have some historic caché 19:00 < kanzure> because legacy support is already the name of the game 19:00 < jmil> and character 19:00 < kanzure> there's no point in adding on new legacies 19:01 < jmil> "back in my day... the URLs used to read like this..." 19:01 < jrayhawk> You can use rewriterules to translate the old to new. 19:01 < jrayhawk> You can also hit yourself in the head with a hammer. 19:01 < jrayhawk> It feels basically the same! 19:01 < jmil> lmao 19:01 < kanzure> yeah that was what i was trying to avoid with "i really don't want to write redirects for each and every incoming link (a few hundred thousand)" 19:01 < kanzure> :( 19:01 < jrayhawk> Well, it would be classes of links. 19:01 * kanzure goes off to hit his head with a hammer 19:02 -!- drazak_ [~ahdfadkfa@drazak.net] has joined #hplusroadmap 19:02 -!- drazak_ [~ahdfadkfa@drazak.net] has quit [Client Quit] 19:02 < jrayhawk> I assume most of your stuff is static anyway, so most of the namespaces translation could be pretty well automated. 19:02 < kanzure> yes 19:02 < jmil> you could do the CNN way, where every 404 redirects to your homepage. that won't piss off anyone. 19:02 < kanzure> uh won't it? 19:02 < kanzure> maybe i'll 404 redirect to the internet archive :) 19:02 < jmil> you miss my sarcasm 19:02 < kanzure> CNN and sarcasm don't mix well :( 19:03 < kanzure> hehe 19:03 < jrayhawk> You can brute force it with md5sums and directory paths for long enough to get search engines redirected permanently, at least. 19:03 < kanzure> what? 19:04 < jmil> what is old schema, and what is new schema? 19:04 < kanzure> there was no old schema (it was just legacy support) 19:04 < kanzure> and there is no new schema (yet) 19:04 < jmil> ?? what?? 19:04 < kanzure> there was no schema, it was a pile of crap accumulated over a decade 19:04 < jmil> i just mean how do old urls look 19:04 < jmil> i can't browse your site to find out b/c it is down, naturally 19:04 < kanzure> naturally :) 19:05 < jrayhawk> Get a list of md5sums and old paths, get a list of md5sums and new paths, create 301s programatically based on those lists. 19:05 < jrayhawk> After a few weeks, get rid of the redirect rules. 19:05 < jrayhawk> (I assume they will be a little slow) 19:05 < kanzure> okay maybe i'll spend some time thinking up new paths 19:05 < kanzure> jmil: here's a 2008 version http://web.archive.org/web/20080729061837/http://heybryan.org/ 19:07 < jmil> so you want new URL style for your 11,000 bookmarks? 19:08 < jmil> the bookmark links don't work in the way back machine, so i can't see what happens when you click one of those number-ordered links 19:08 < kanzure> well, if it's the bookmarks then i'd just want to write a better parser for the bookmark format that i was using 19:08 < jmil> so what is it then? 19:09 < kanzure> it's the crap spewed on / 19:09 < kanzure> and some other stuff, like /~bbishop/docs/* and /books/* and /~bryan/* 19:09 < kanzure> oh and /docs/* 19:09 < kanzure> basically it looked something like this :P http://eugen.leitl.org/ 19:10 < jrayhawk> I'm working on your push problem, jmil, just so you know. 19:10 < jmil> oh thanks jrayhawk ! 19:12 < jrayhawk> Okay, it looks like refs/remotes/ are not preserved during a clone operation, so there isn't any point trying to push remotes to a bare repository. 19:12 < jrayhawk> oh wait, I meant to mirror, didn't I 19:12 < jrayhawk> hum dee dum ignore me i am dumb 19:14 < jrayhawk> Nope, even on git clone --mirror 19:15 < jmil> so i had copied the .git/svn/.metadata file 19:15 < jmil> that's all we need 19:15 < jmil> i think 19:15 < jmil> going afk for a while... 19:19 < jrayhawk> jrayhawk@gnusha:~$ git clone ~jsmiller/reprap_test/ 19:19 < jrayhawk> Initialized empty Git repository in /home/jrayhawk/reprap_test/.git/ 19:19 < jrayhawk> warning: remote HEAD refers to nonexistent ref, unable to checkout. 19:21 < jrayhawk> perhaps you should stop manually screwing around inside of git repos 19:21 < gnusha_> it hurts 19:21 < gnusha_> :( 19:24 < jrayhawk> jmil is going to keep us on our toes 19:24 < kanzure> wait until you meet the reprap developers 19:24 < jrayhawk> this is why god invented pinysell 19:24 < jrayhawk> pinyshell 19:25 < jrayhawk> jmil is one of those people who are never 'lost', merely 'exploring', isn't he 19:25 < jrayhawk> i call those people 'scientists' 19:25 < jrayhawk> and they are sadistic bastards 19:32 < jrayhawk> Anyway, when you have what you want, just git push --force /srv/git/reprap.git master 19:34 < jrayhawk> Dangling objects will be garbage collected whenver cron.daily does its thing. 19:34 < kanzure> i'm still suspicious about crond 19:35 < kanzure> 20:50 < fenn> can't you run a cron job? 19:35 < kanzure> 20:50 < kanzure> probably, but cron and i have never got along 19:35 < kanzure> 20:51 < pmetzger> ??? 19:35 < kanzure> 20:51 < pmetzger> what's hard about cron? 19:35 < kanzure> 20:51 < pmetzger> put script in file. put line in crontab saying "run script every day at hour X" 19:35 < kanzure> 20:52 < kanzure> and then it doesn't happen 19:35 < kanzure> 20:55 < fenn> that's been my experience too :\ 19:35 < kanzure> ^ from http://gnusha.org/logs/2010-07-01.log 19:35 < kanzure> 20:55 < fenn> maybe something's just broken on davinci 19:36 < jrayhawk> I've had some problems with cron and ubuntu, but I think that's just because upstart is a giant pile of garbage and doesn't take well to being containerized. 19:36 < jrayhawk> Because for some reason upstart wants to handle all hotplug stuff. 19:36 < kanzure> is upstart some cron-specific stuff on ubuntu? 19:37 < jrayhawk> Upstart is Ubuntu's nutty init replacement. 19:37 < kanzure> or is that rc0-3? 19:37 < kanzure> i see 19:37 < kanzure> yeah that doesn't sound safe 19:37 < jrayhawk> Other distributions have tragicly started using it. 19:37 < jrayhawk> Some people are working on something more like what OS X has that looks a lot less insane. 19:38 < jrayhawk> http://0pointer.de/blog/projects/systemd.html 19:38 < jrayhawk> They're on the fence about whether or not to replace cron, I think. 19:51 < jrayhawk> kanzure: it's possible your cron woes are caused by system downtime, in which case anacron might be more your speed...? 19:52 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined #hplusroadmap 19:53 < kanzure> nah, davinci had 100+ day uptimes and the cronjobs i wanted were much less than 100d lengths 19:53 < kanzure> intervals 20:06 -!- mheld [~mheld@pool-173-76-224-45.bstnma.fios.verizon.net] has joined #hplusroadmap 20:06 -!- mheld [~mheld@pool-173-76-224-45.bstnma.fios.verizon.net] has quit [Client Quit] 20:12 -!- wolfspraul [~wolfsprau@lucia.q-ag.de] has quit [Quit: leaving] 20:14 -!- Aly2 [Alystair@bas1-toronto10-1279397903.dsl.bell.ca] has joined #hplusroadmap 20:14 -!- Alystair [Alystair@bas1-toronto10-1279397903.dsl.bell.ca] has quit [Disconnected by services] 20:14 -!- Aly2 is now known as Alystair 20:29 -!- Netsplit *.net <-> *.split quits: niftyzero, dustbin, nchaimov 20:37 < jrayhawk> Other cron gotchas are the need to make things in /etc/cron.x executable, and that cron does not source /etc/profile 20:37 < jrayhawk> Which means, for instance, that, even when running as root, the various sbins are not in the PATH 20:37 -!- Netsplit over, joins: niftyzero 20:39 -!- dustbin [~chatzilla@71.145.146.189] has joined #hplusroadmap 20:39 -!- nchaimov [nchaimov@c-71-237-208-209.hsd1.or.comcast.net] has joined #hplusroadmap 20:39 < jrayhawk> This is all mitigated somewhat by the fact that cron emails root under all but perfect execution, so any problems should make themselves real obvious if you've properly configured mail. 20:44 -!- Noahj [~noa@24.38.179.66] has joined #hplusroadmap 20:45 < jrayhawk> it can also mail other people if you set the MAILTO variable 21:10 -!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Ping timeout: 245 seconds] 21:19 -!- augur [~augur@208-58-6-161.c3-0.slvr-ubr1.lnh-slvr.md.cable.rcn.com] has quit [Ping timeout: 265 seconds] 21:19 -!- augur [~augur@208-58-6-161.c3-0.slvr-ubr1.lnh-slvr.md.cable.rcn.com] has joined #hplusroadmap 21:31 -!- splicer [~patrik@h126n1c1o261.bredband.skanova.com] has quit [Remote host closed the connection] 21:39 -!- jennifer2 [~jennifer@70-36-134-77.dsl.dynamic.sonic.net] has joined #hplusroadmap 21:52 < kanzure> long and relatively boring thread about sourceforge vs. github 21:52 < kanzure> http://groups.google.com/group/openmanufacturing/browse_thread/thread/a901eef6424702c8 21:58 -!- Noahj [~noa@24.38.179.66] has quit [Ping timeout: 240 seconds] 22:03 -!- dustbin [~chatzilla@71.145.146.189] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722145641]] 22:04 < jmil> kanzure jrayhawk i'm back... trying to nuke the contents of reprap_test.git 22:04 < jmil> but don't have permission (yes i logged out, and logged back in) 22:07 < kanzure> jmil: both joe and i are suggesting that you don't use file utilities.. just use git 22:07 < jmil> ok 22:07 < kanzure> 19:32 < jrayhawk> Anyway, when you have what you want, just git push --force /srv/git/reprap.git master 22:07 < kanzure> or, in this case, reprap_test.git 22:09 < jmil> problem: 22:09 < jmil> git clone --mirror ~/reprap_test 22:09 < jmil> fatal: destination path 'reprap_test.git' already exists and is not an empty directory. 22:09 < kanzure> um 22:09 < kanzure> that's not what we said to do 22:10 < jmil> i want to push a bare .git directory 22:10 < jmil> will that work or will it try to checkout that .git into /srv/git/ 22:10 < jmil> ? 22:10 < kanzure> git push --force /srv/git/reprap_test.git master 22:11 < jmil> git push --force /srv/git/reprap_test.git master 22:11 < jmil> fatal: '/srv/git/reprap_test.git' does not appear to be a git repository 22:11 < jmil> fatal: The remote end hung up unexpectedly 22:11 < kanzure> orly 22:11 < kanzure> hm 22:12 < kanzure> wow wtf happened to this repository 22:12 < jmil> well i didn't do it to reprap_test... 22:12 < kanzure> it's been oblitterated 22:12 < jmil> i don't have perms :) 22:12 -!- Noahj [~noa@24.38.179.66] has joined #hplusroadmap 22:12 < jrayhawk> Okay, so, explain to me one more time why you're not using git push --force /srv/git/reprap.git master 22:12 < kanzure> "su gnusha" has a funny ring to it 22:14 < kanzure> jmil: ok try pushing again to reprap_test 22:14 < jrayhawk> WHY DOES THE TEST REPOSITORY EXIST AT ALL IF YOU JUST INTEND TO OVERWRITE THE MAIN ONE ANYWAY?! 22:15 < kanzure> jrayhawk: is there some reasonable way for us to keep track of what we're doing separately on a server? maybe a log file somewhere "don't touch /blah/blah/blah because it's a bomb" 22:15 < jmil> jrayhawk: because you did a lot of work on the hooks directory and i didn't want to overwrite your efforts 22:15 < kanzure> jrayhawk: because reprap.git is the only good copy right now 22:15 < jrayhawk> YOU WILL NOT DO THAT 22:15 < kanzure> jrayhawk: i wouldn't be so sure.. he's been using rm -rf everywhere 22:16 < jmil> ok, well i'm new to hooks (obviously) and when i asked kanzure to delete the renamed hooks_OLD directory it broke cgit display and git:// checkout 22:16 < jmil> or something 22:16 < jrayhawk> Just use git commands and be aware that things like --force can delete history and we'll all be happy. 22:17 < jmil> ok 22:17 < jrayhawk> There is no way to modify hooks through git commands. 22:17 < jmil> kanzure nuked it for me because i was uploading to the folder via SFTP because the git:// command wasn't working 22:18 < jrayhawk> writes would need to happen through ssh:// (or git+ssh://, which is the same thing) 22:19 < jrayhawk> The git:// interface is read only since the only means of identifying or authorizing people over it is about as complicated as just using SSH. 22:19 < jmil> problem: 22:19 < jmil> git push --force /srv/git/reprap_test.git master 22:19 < jmil> Counting objects: 45033, done. 22:19 < jmil> Delta compression using up to 4 threads. 22:19 < jmil> Compressing objects: 100% (15629/15629), done. 22:19 < jmil> fatal: Unable to create temporary file: Permission denied 22:19 < jmil> error: pack-objects died of signal 13 22:19 < jmil> error: pack-objects died with strange error 22:19 < jmil> error: failed to push some refs to '/srv/git/reprap_test.git' 22:19 < jrayhawk> < jrayhawk> Okay, so, explain to me one more time why you're not using git push --force /srv/git/reprap.git master 22:20 < kanzure> jmil: i redid the reprap_test repo and you'd have to relogin to fix that 22:20 < kanzure> but again, you should probably just push to reprap.git 22:21 < jmil> i'm really sorry guys, please go easy on me 22:21 < kanzure> :) 22:21 < jmil> i'm pushing to reprap.git now 22:21 < kanzure> is it working? 22:21 < jmil> so far 22:21 < jrayhawk> I am just an angry person in general, don't mind me :) 22:21 < jmil> jrayhawk: really i was trying to save you work 22:21 < kanzure> joe is an angry sysadmin 22:21 < jmil> because kanzure nuked hooks_OLD for me and you then had to go recreate it and fix stuff 22:22 < jmil> so when u weren't here i asked kanzure to make me a new dir so as not to disturb you more 22:22 < jmil> problem: 22:22 < jmil> remote: error: denying non-fast-forward refs/heads/master (you should pull first) 22:22 < jmil> To /srv/git/reprap.git 22:22 < jmil> ! [remote rejected] master -> master (non-fast-forward) 22:22 < jmil> error: failed to push some refs to '/srv/git/reprap.git' 22:22 < jrayhawk> Hmm. 22:22 < jrayhawk> Don't pull. 22:22 < jmil> this is why i was trying to do a --mirror instead of a --force 22:23 < kanzure> are the commit hashes different? 22:23 < jmil> different than what? 22:23 < jrayhawk> Yeah, but the --force is supposed to ignore that. 22:23 < jmil> --force 22:23 < jmil> Usually, the command refuses to update a remote ref that is not an ancestor of the local ref used to overwrite it. This flag disables the check. This can cause the remote repository to lose commits; use it with care. 22:23 < jmil> http://www.kernel.org/pub/software/scm/git/docs/git-push.html 22:23 < jrayhawk> And just push the new ref, consequences be damned 22:24 < jmil> i think git clone --mirror is best 22:24 < jrayhawk> Oh, I bet I know. Just a second. 22:26 < jmil> if i just nuke the contents of the dir then it will be fine 22:26 < jmil> kanzure: since we are taking away the @ hash author string, the sha1 of each commit will necessarily be different 22:27 < jmil> jrayhawk: i will just nuke the contents and try again then? or will that mean you will have to recreate the hooks again 22:27 < jmil> ? 22:28 < jrayhawk> I'm going to work out how to do this the "right" way 22:28 < jmil> look at --mirror 22:28 < jmil> git clone --mirror ~/reprap_test/ 22:28 < jmil> fatal: destination path 'reprap_test.git' already exists and is not an empty directory. 22:29 < jmil> so just nuke the dir 22:29 < jmil> jrayhawk and kanzure the commits are completely and totally different. There are ZERO common ancestors 22:30 -!- katsmeow is now known as katsmeow-afk 22:30 < jrayhawk> Yeah, that's what the --force is for. 22:30 < jrayhawk> All the old objects will become disconnected and cleaned up on the next garbage collect. 22:35 < jmil> is the hooks dir all you need to keep? 22:35 < kanzure> heh 22:37 < jrayhawk> Ah, I see, it's a configuration variable. 22:39 < jrayhawk> git init --bare --shared puts receive.denyNonFastForwards = true in place 22:39 < jrayhawk> Sorry about that :( 22:39 < kanzure> hm? who should do that? 22:39 < jmil> whoa 22:39 < kanzure> and on which repository 22:39 < kanzure> er, i mean, in which .git folder 22:40 -!- ferrouswheel [~joel@121-98-81-17.bitstream.orcon.net.nz] has quit [Ping timeout: 260 seconds] 22:40 < jrayhawk> It's on the bare repo in /srv/git 22:40 < jrayhawk> I'm pushing the force update now. 22:41 < jrayhawk> It's really spectacularly irritating that it rejects the ref update *after* all the objects are uploaded. 22:42 < jmil> kanzure: doesn't matter to me which one, either one should allow us to test the cgit functionality? 22:42 < jrayhawk> Okay! That's in betterish shape. 22:42 < jmil> should i try again? 22:42 < jmil> to the reprap.git or reprap_test.git? 22:42 < kanzure> jmil: jrayhawk just did it 22:42 < kanzure> so, no 22:43 < jmil> taht is beautiful, thank you jrayhawk and kanzure 22:43 < kanzure> https://secure.diyhpl.us/cgit/reprap/ 22:44 < jrayhawk> Are you planning on completely overwriting the development history again, or can I put that config safeguard back in place? 22:44 < kanzure> jrayhawk: i like how cgit gives a hexdump -A output on binary files :) 22:44 < kanzure> https://secure.diyhpl.us/cgit/reprap/tree/trunk/reprap/mechanics/alloy-extruder/AoIs/clamp.aoi 22:45 < jmil> jrayhawk: you can put the safeguard back in place. THANKS! 22:45 < jmil> pending adrian's approval 22:45 < kanzure> jmil: huh? shouldn't it be disabled until the final svn conversion is done 22:46 < jrayhawk> I will wait for that, then. 22:46 < jmil> kanzure: naw, future pushes will be from the one in my home dir which will have ancestors in the current reprap.git 22:46 < jmil> so in my home dir I will continue to git svn fetch and git svn rebase 22:47 < jmil> this ends up being a straight fast-forward because we are never going to commit to master directly 22:47 < jmil> just yet 22:47 < kanzure> ok. i didn't know about git-svn-fetch 22:47 < jmil> b/c the one in my home dir has the correct author flag to suppress the @ hash author tag on future fetches 22:48 < jmil> nice fat pipes, btw 22:48 < jmil> i'm getting 1 mb/sec pulling down to local hd 22:48 < jmil> niiiiice 22:48 < jmil> git clone git://diyhpl.us/reprap 22:49 < kanzure> our davinci server was doing 10 MB/sec before it died and went to compuheaven 22:49 < kanzure> (it was on me.utexas.edu) 22:50 -!- jennifer2 [~jennifer@70-36-134-77.dsl.dynamic.sonic.net] has quit [Ping timeout: 265 seconds] 22:52 < jrayhawk> I've seen 100MB across the NERO network, but I guess that's probably cheating. 22:52 < jrayhawk> err 100Mb 22:52 < kanzure> no man's porn should ever come that fast 22:52 < kanzure> what am i sayingBBBBBBBB? 22:53 < jrayhawk> I am sorry for internet yelling :( 22:56 -!- jennifer2 [~jennifer@70-36-134-77.dsl.dynamic.sonic.net] has joined #hplusroadmap 22:59 < jrayhawk> no wait, it was 100MB i am just dumb 23:00 < jrayhawk> kanzure: so might '< jrayhawk> Which means, for instance, that, even when running as root, the various sbins are not in the PATH' explain your troubles with and subsequent distrust of cron? 23:00 < jrayhawk> because i think cron should be your friend 23:01 < kanzure> what about normal bin? 23:02 < jrayhawk> Normal bin works fine. 23:02 < kanzure> hrm 23:02 < kanzure> this requires some experimenting 23:02 < jrayhawk> I usually use full paths out of paranoid habit, though. 23:02 < kanzure> i should become more paranoid 23:03 < jrayhawk> Keeping in mind that things that get executed out of bin will inherit the non-sbin PATH 23:04 < jrayhawk> #hplusroadmap: for all your sysadmin and version control needs 23:05 < jrayhawk> '/window 1 23:05 < jrayhawk> whoops 23:07 < jrayhawk> all discussions of transhumanism should be subsequently redirected to #git 23:12 < jrayhawk> And, of course, you should set up proper mail handling. 23:15 < jmil> jrayhawk kanzure can we plan how to execute git commit to that reprap repo for 20 developers? 23:15 < jmil> should they all share a git@gnusha login or each user gets own account, or what? 23:15 < jmil> then they have to use git+ssh for commit? 23:15 < jrayhawk> Everyone gets their own account. You are free to add them with the 'newuser' command. 23:16 < jrayhawk> You will subsequently need to 'addaccess reprap that-user-you-just-created' 23:16 < jrayhawk> They will have a severely limited shell that will essentially only allow them access to git and the 'passwd' command. 23:16 < jmil> ok i will try a little more testing and then send instructions to adrian 23:17 < jmil> how do i setup a password for them? is one auto generated? 23:17 < jmil> and for testing i will do some commits, but then how do i reverse them? with jsmiller account I can git reset? 23:18 < jmil> can all accounts that i create execute git-reset? 23:18 < jrayhawk> You'll specify a password. The password doesn't really matter since they can change it later. 23:18 < jmil> does gnusha have command line password generator so it is pseudo secure? 23:19 < jrayhawk> I just installed apg and pwgen, use whichever you like. 23:20 < jmil> can jsmiller account execute git reset? and can accounts i create execute git-reset? 23:20 < jrayhawk> You can run that locally and push whatever refs you like. 23:20 < jrayhawk> Well, other than that denyNonFastForwards part 23:21 < jrayhawk> If I were to disable that, you could push whatever you liked! 23:21 < jmil> so i commit in ~, then push to reprap.git. then i git-reset in ~, then push --mirror? 23:23 < jrayhawk> I suppose if you wanted? Most people manually push their tags and branches so as not to clutter things. 23:23 < jmil> no i'm asking about git-reset since i will be doing test commits right now 23:23 -!- jennifer2 [~jennifer@70-36-134-77.dsl.dynamic.sonic.net] has quit [Ping timeout: 245 seconds] 23:24 < jmil> what would keep other users from doing git push --mirror and overwriting things? 23:25 < jrayhawk> Well, they're not allowed to delete history, what with the receive.denyNonFastForward 23:25 < jmil> but that means i also could not delete the test commits i'm about to do 23:26 < jrayhawk> Probably not, no. 23:27 < jrayhawk> You can revert them, though. 23:27 < jrayhawk> Or branch. 23:27 < jrayhawk> Ikiwiki won't pay attention to the new branch, but it'll be accessible through cgit. 23:28 < jrayhawk> https://secure.diyhpl.us/cgit/reprap/refs/ 23:29 < jmil> what do you mean revert them. is that not the git-reset command? 23:29 < jrayhawk> If you'd like, I can remove the receive.denyNonFastForwards restriction for the time being. 23:29 < jrayhawk> A revert is a commit that undoes a previous commit. 23:29 < jrayhawk> Both commits are in the commit log. 23:29 < jmil> ya, maybe let's do that, removing the receive.denyNonFastForwards restriction 23:29 < jrayhawk> s/are in/wind up/ 23:30 < jrayhawk> Okay. 23:30 < jrayhawk> In that case you can --force whatever pushes you like. 23:30 < jrayhawk> That's all set. 23:30 < jmil> excellent, thx 23:30 < jmil> ok i'm gonna try user creation, committing, reverting 23:31 < jmil> oh... how do we setup ssh-keys for committing? 23:31 < jmil> easy or hard? 23:31 < jrayhawk> I... haven't really provisioned for that and I really should. 23:31 < jrayhawk> For the time being you can send them to me and I'll set them up manually. 23:31 < jmil> send them to me? the public key? 23:32 < jmil> can you email me your email address using the upenn.edu address you have? 23:32 -!- uniqanomaly__ [~ua@dynamic-78-8-215-170.ssp.dialog.net.pl] has quit [Ping timeout: 245 seconds] 23:32 -!- uniqanomaly__ [~ua@dynamic-78-8-94-81.ssp.dialog.net.pl] has joined #hplusroadmap 23:33 < jrayhawk> You, at least, can manage your own SSH keys unless you don't know how, anything you can't deal with can be sent to me at jrayhawk+reprap@omgwallhack.org or over IRC 23:33 < jrayhawk> or XMPP jrayhawk@omgwallhack.org 23:34 -!- augur [~augur@208-58-6-161.c3-0.slvr-ubr1.lnh-slvr.md.cable.rcn.com] has quit [Remote host closed the connection] 23:34 < jrayhawk> Or you can wait for me to provide something within pinyshell to deal with that. 23:34 -!- wolfspraul [~wolfsprau@lucia.q-ag.de] has joined #hplusroadmap --- Log closed Fri Sep 03 00:00:17 2010