Sunday, 2011-02-27

*** tpb has joined #melange00:00 [freenode-info] channel flooding and no channel staff around to help? Please check with freenode support:
*** dhaun has quit IRC00:01
*** dr__house has quit IRC00:41
*** m4k3r has quit IRC02:33
*** dr__house has joined #melange02:36
*** dr__house has joined #melange02:36
*** dr__house has quit IRC03:11
*** dr__house has joined #melange03:11
*** dr__house has quit IRC03:36
*** tansell-laptop has joined #melange03:43
*** chocolaate-maan has joined #melange04:22
-chocolaate-maan- download this
chocolaate-maanelite botnet
*** chocolaate-maan has left #melange04:22
tpb<> (at
*** schultmc has quit IRC06:40
*** schultmc has joined #melange06:46
*** m4k3r has joined #melange06:51
*** m4k3r has quit IRC07:02
*** chocolaate-maan has joined #melange07:08
*** chocolaate-maan has left #melange07:08
*** madrazr has joined #melange07:21
*** ChanServ sets mode: +o madrazr07:21
*** madrazr has quit IRC07:41
*** madrazr has joined #melange09:05
*** ChanServ sets mode: +o madrazr09:05
*** tansell-laptop has quit IRC09:06
*** Pline has joined #melange09:45
-Pline- bots
Plinehacking software download
*** Pline has left #melange09:45
tpb<> (at
*** westi|gone has quit IRC10:08
*** westi has joined #melange10:11
*** madrazr has quit IRC11:03
*** slingshot316 has joined #melange11:04
slingshot316hey guys! i have a small issue with the development server11:06
*** madrazr has joined #melange11:53
*** ChanServ sets mode: +o madrazr11:53
slingshot316madrazr: there?12:55
madrazrslingshot316: heya!12:56
slingshot316madrazr: i need some help12:56
madrazrslingshot316: yeah?12:56
slingshot316madrazr: when i'm running the appserver, i'm gettin the following warnings and page not found eroor12:58
tpbTitle: Plain Text code - 7 lines - codepad (at
madrazrslingshot316: ah13:01
madrazrwe all have been running into this problem recently13:01
madrazrslingshot316: sudo easy_install ssl13:01
madrazrslingshot316: rather, install the Python ssl package in some way on your machine13:02
slingshot316oh i installed opensll with no luck13:02
madrazrslingshot316: install python bindings for openssl13:02
slingshot316madrazr: and what about the page not found at localhost:8080/site/show/site13:02
madrazrslingshot316: openssl installs the OpenSSL package13:03
slingshot316madrazr: will do that now13:03
madrazrslingshot316: that is fine13:03
madrazrif you are on the tip of the default branch13:03
madrazrmost of the URLs on the development version are broken at the moment13:03
madrazrwe are reworking on a lot of stuff13:03
slingshot316madrazr: oh okay okay13:03
slingshot316madrazr: so i wont be able to run the local development server at the moment?13:05
madrazrslingshot316: sure it will be13:06
madrazrslingshot316: but not all the URLs are working13:06
madrazrslingshot316: you may get 404 for many URLs13:06
slingshot316madrazr: but the main page is itself giving a 40413:06
madrazrslingshot316: I don't think so13:07
madrazrlocalhost:8080 is the main page13:07
madrazrit works for me13:07
slingshot316well, its not working for me :(13:08
slingshot316madrazr: it says request url : http://localhost:8080/site/show/site not found13:08
slingshot316madrazr: its actually redirecting me to localhost:8080/site/show/site13:09
madrazrslingshot316: the home page is http://localhost:8080/13:09
madrazrslingshot316: set your site settings13:10
slingshot316madrazr: i know m trying the same13:11
slingshot316madrazr: site settings as in? wher?13:12
madrazrslingshot316: ah wait13:12
madrazrthis is going to be tricky13:12
madrazrslingshot316: what version are you using?13:13
slingshot316madrazr: you mean python ?13:13
madrazrslingshot316: how updated are you wrt to our default branch?13:13
madrazrslingshot316: no Melange13:13
slingshot316madrazr: well, i cloned it from
tpbTitle: Revision 2b657a4220: / (at
slingshot316madrazr: i dont know which version it is13:15
madrazrslingshot316: can you give the output of hg tip?13:15
madrazr$ hg tip13:15
slingshot316one second13:15
tpbTitle: Plain Text code - 5 lines - codepad (at
*** m4k3r has joined #melange13:19
madrazrslingshot316: can you just run hg update once?13:20
slingshot316madrazr: sure13:21
slingshot316madrazr: one second13:23
slingshot316madrazr: 0 files updated, 0 files merged, 0 files removed, 0 files unresolved13:24
madrazrslingshot316: Oh13:25
madrazrslingshot316: now for your development, not to see the 404s13:25
madrazropen, app/soc/modules/soc_core/callback.py13:26
madrazrline 8413:26
madrazrdisable the if line and the return13:26
slingshot316one second13:26
slingshot316madrazr: yea i commented the two lines13:27
madrazrslingshot316: run your server now13:27
slingshot316yea one second13:27
slingshot316madrazr: whoa! it worked13:28
madrazrslingshot316: but don't commit that code13:28
slingshot316madrazr: but what was the problem ?13:28
madrazrslingshot316: it should not be removed actually, but since things are slightly broken13:28
madrazrslingshot316: there is no problem13:28
madrazrslingshot316: since we are working on new stuff13:29
madrazrwe have disabled all the old stuff13:29
slingshot316madrazr: okay okay13:29
slingshot316madrazr: okay13:29
slingshot316madrazr: one more thing13:29
madrazrslingshot316: yeah?13:30
slingshot316madrazr: i have python-imaging installed but it still says WARNING  2011-02-27 11:05:32,687] Could not initialize images API; you are likely missing the Python "PIL" module. ImportError: No module named _imaging13:30
madrazrslingshot316: I think PIL is installed for a different version of Python13:32
madrazrand even imaging library13:32
slingshot316madrazr: oh right, sorry i forgot m in the virtualenv13:32
slingshot316madrazr: my bad13:32
slingshot316madrazr: so when should i put those lines back again ?13:34
madrazrslingshot316: you should not have taken it out in the first place :D13:34
madrazrslingshot316: but yeah, just create some data13:34
madrazrand put it back13:35
madrazrslingshot316: site settings13:35
madrazrorg admins, mentors, students, student proposals13:35
madrazrstudent projects13:35
madrazris the data you need to create13:35
madrazryou can use localhost:8080/seed_db13:36
slingshot316madrazr: http://localhost:8080/seed_db this does the same right?13:36
madrazrto create some of the data13:36
slingshot316yea :P13:36
madrazrslingshot316: but seed_db doesn't seed many items13:36
madrazrlike student proposals13:36
madrazrstudent projects13:36
slingshot316madrazr: yea i'll do that manually13:36
madrazrwhich are very important13:36
slingshot316madrazr: okay and i need to change the path to the datastore from temp right?13:36
slingshot316madrazr: and you're sure i can use 2.7 for sdk right?13:38
madrazrslingshot316: I am using Python2.7 right now13:38
madrazras we are talking together13:38
madrazrI am doing some development work on Melange13:38
madrazron Python2.713:38
madrazrslingshot316: if that is sure enough :P13:38
slingshot316madrazr: more than enough13:39
slingshot316madrazr: thank you so much for all the guidance, you saved my day :)13:39
slingshot316madrazr: thank you :)13:39
madrazrslingshot316: NP :)13:39
slingshot316madrazr: hope i;m not irritating you13:39
*** dhaun has joined #melange13:51
madrazrslingshot316: definitely not13:54
madrazrdon't worry13:54
madrazrslingshot316: we always welcome our new contributors with open arms :)13:54
slingshot316madrazr: yea i noticed that, very sweet :)13:56
slingshot316madrazr: how is everyone at FOSSEE?13:57
madrazrslingshot316: we are doing good13:57
madrazrslingshot316: we have few new comers13:57
slingshot316madrazr: oh thats nice, let me know if you will guys are organizing any events like scipy13:59
madrazrslingshot316: we will organizing another SciPy in 2011 :P14:01
slingshot316madrazr: lol14:01
slingshot316madrazr: instead of running --datastore_path everytime, is there a way to change it permanently?14:03
madrazrslingshot316: not that I know of14:03
madrazrslingshot316: but there is bash14:03
madrazrslingshot316: if you can do tricks there :)14:03
madrazrslingshot316: for example, I have a bash shortcut, smel14:04
madrazrwhich does this14:04
madrazrcd /media/python/workspace/melange14:04
madrazrpython2.7 ./thirdparty/google_appengine/ --datastore_path=/media/python/dev_appserver.datastore --blobstore_path=/media/python/dev_appserver.blobstore --port=8000 --enable_sendmail --show_mail_body --allow_skipped_files build14:04
slingshot316madrazr: yea, i can do that in bash14:04
madrazrpretty nifty14:04
madrazrslingshot316: then do it :P14:04
slingshot316madrazr: will do that, thank you :)14:05
madrazrslingshot316: NP14:05
slingshot316madrazr: after filling the data and putting back those two lines, can i commit?14:17
madrazrslingshot316: sure14:17
slingshot316yea locally14:17
madrazrslingshot316: but you cannot push back to the main repo yet14:18
madrazrslingshot316: yeah sure locally14:18
slingshot316madrazr: okay :)14:18
madrazryou can even push to your own version if you are running it somewher14:18
slingshot316madrazr: no right now, m not running it anywhere :P14:18
madrazrslingshot316: Ok14:19
*** slingshot316 has quit IRC15:50
*** slingshot316 has joined #melange17:20
*** dr__house has joined #melange17:52
*** dr__house has joined #melange17:52
*** dr__house has quit IRC18:11
*** dr__house has joined #melange18:12
*** dr__house has joined #melange18:12
*** m4k3r has quit IRC18:12
*** m4k3r has joined #melange18:12
*** madrazr has quit IRC18:34
*** madrazr has joined #melange18:45
*** ChanServ sets mode: +o madrazr18:45
*** slingshot316 has quit IRC18:58
antrik14:32 <@madrazr> slingshot316: since we are working on new stuff20:19
antrik14:32 <@madrazr> we have disabled all the old stuff20:19
antrikmadrazr: it seems you still need some mentoring on good development practices... :-P20:19
madrazrantrik: can you please explain?20:19
antrikmaster should always be kept in a working state as much as possible20:20
madrazrantrik: why it is not a good development practice to disable stuff?20:20
antrikso people can work on unrelated features20:20
madrazrantrik: well20:20
antrikif you do major changes, do them on a branch, and merge once they are ready20:20
madrazrantrik: it really depends20:21
madrazrantrik: our point is, when you are doing 10s of changes a day20:21
madrazrkeep them on master20:21
madrazrantrik: that is the development line right?20:22
madrazrantrik: and then you have release branches?20:22
antrikno, only changes that don't (knowingly) break existing functionality should be applied to master directly20:22
antrikotherwise it makes life really hard for anyone working on some other features20:23
madrazrantrik: when there is no one working on some other feature?20:23
madrazrantrik: when everyone is working on the same thing?20:23
antrikyou never know that, especially in a community project20:24
antrikI have a feature I intend to work on in the near future20:24
madrazrantrik: I see that most of the projects do their development on master and once the master reaches the release state, they branch that off into a maintenance branch20:24
madrazrare they all doing it wrong then?20:24
antrik(well, most likely someone else will do the actuall work, but that's beside the point...)20:24
madrazrantrik: talk to the community first, if it is such a case20:24
madrazrantrik: why do you want to hide that from the community then?20:25
antrikmadrazr: I'm not sure about "most"20:25
antrikprojects that do it *right*, never commit patches to master that are known to break existing functionality20:25
antrikremember my rants about bisectability? :-)20:26
madrazrantrik: sure :)20:26
antrikit should always be possible to pick an arbitrary revision on master, and have a more or less working system20:26
madrazrantrik: for the rants20:26
madrazrnot for the earlier20:26
madrazrantrik: I don't quite agree here20:27
madrazrantrik: it means every point on the development line should be in a working state20:27
antrikyou can do temporary commits on branches if that helps you, but before commiting them to master, better rework them so each patch goes from one working state to another working state20:27
madrazrantrik: which means you want to rewrite history?20:28
antriksure :-)20:28
madrazrantrik: why?20:28
madrazrwhy do you encourage that?20:28
antrikbecause it doesn't matter in the end whether you first had a partial or buggy implementation of some change before getting to the final version. it's only the final changes that matter.20:29
madrazrantrik: anyways seems like we are getting sidelined20:29
madrazrantrik: well, in the end nothing matters. Whether you wrote the code or not also doesn't matter. All that matters is get me a releasable software that just works20:30
antrikwhen people look at history, they want answers to question like "what change did introduce this bug?"20:31
madrazrantrik: then why maintain history at all?20:31
madrazrantrik: people may also want to know how you arrived at what you arrived at20:31
antrikor they want to do actions like "revert that change that turned out to be problematic in the end"20:31
antriknope, that only confuses20:31
madrazrantrik: you say that20:32
antrikif it's necessary to explain why something was done in a specific manner, it should be explained in code comments and/or in the commit message of the final, cleaned up change20:32
madrazrantrik: but I am interested in knowing it20:32
antrikthe purpose of revision control is not to satisfy curiosity :-)20:33
antrikbut rather to help the development process20:33
madrazrantrik: which is what I am telling :D20:33
madrazrantrik: the purpose of revision control system is to aid the development process20:33
madrazrantrik: should be used in the way that best helps the community that is using it20:33
antriknope. having dead-ends, half-baked features etc. in the history doesn't help development; it only makes it harder to see the relevant things20:34
madrazrantrik: there may be best practices20:34
madrazrbut it is not guaranteed that, those best practices are best practices for every one out there?20:34
antrikit's only important what changes ultimately made it to the main branch(es); how they were created doesn't matter after the fact20:34
madrazrantrik: thats again your perception20:35
antriknah, that's logical if you think about it...20:35
antrikand what's more, it's proven in many large projects, including Linux and X.Org among many many others20:35
madrazrantrik: sure it works for them20:36
madrazrthats why they use it20:36
antrikwell, one of my favourite quotes: "learn from other people's mistakes. you don't have enough time to make them all yourself."20:37
antrikif you followed any project that uses good practices, you would see yourself how they are helpful. in lack of that, you'd better believe they know what they are doing...20:39
madrazrantrik: sure, mistakes involve circumstances too20:39
madrazrantrik: and circumstances may be different20:39
antrikI've had my share of "I don't think it's a good idea, even if everyone else is doing it that way" myself... in most cases, I was proven wrong :-)20:40
madrazrantrik: "most"20:41
madrazrantrik: you don't want to say "all" here?20:41
madrazrantrik: that means, there were cases, where you were right?20:41
antrikI would have to think obout it more20:46
madrazrantrik: please do and let me know20:47
antrikthere are probably cases where I just haven't been proven wrong yet ;-)20:47
madrazrantrik: anyways let us not reach a moot point20:47
madrazrI hope this discussion is for Melange's good20:48
antrikBTW, this is not an abstract comparision. branch management is actually one of the things I had my own ideas about, and was proven wrong20:48
madrazrantrik: so I just have few questions20:48
antrikwas proven wrong through both personal experience, and following other projects20:48
antrikin fact I witnessed myself how a change of process *vastly* improved the quality of X.Org development20:48
antrikBTW, you are right that we got sidetracked a bit:20:49
antrikrewriting history is indeed a somewhat controversial topic, and besides the point20:49
antrikwhether you keep the original development history or not -- you still should do development on branches. this way any meandering is clearly contained on the branches, and doesn't interfere with other development20:50
antrik(either while happening or while reading history after the fact)20:50
madrazrantrik: well correct20:53
madrazrI agree, if it is meandering, I completely agree with your point20:53
madrazrantrik: but when your whole code base is changing?20:53
madrazrantrik: please listen to this20:53
madrazrantrik: and let me know, if there is something wrong20:53
madrazrantrik: unless someone from the community who has not spoken to the devs here wants to do something knows exactly what he is doing20:54
madrazrantrik: it is very very very likely that whatever he does changes20:54
madrazrantrik: so he better do his changes against the latest development version20:54
madrazrantrik: taking this radical change off the master branch, reduces visibility for new comers20:55
madrazrantrik: take any Revision Control System tool today20:55
madrazrantrik: even the best among all, github.com20:55
madrazrby default they only show the changes happening on the master branch20:55
madrazrantrik: and for a new comer, it is very hard to know where the whole development is happening20:56
madrazrantrik: what is your take on it?20:56
madrazrantrik: Melange code base is in flux now20:56
madrazrantrik: we are almost in the process of rewriting20:56
madrazrantrik: we don't want to do another Python 3.0 like epic fail20:57
madrazrantrik: taking 5 years to change everything else that happens in parallel20:57
madrazrantrik: Oh well, in Python 3.x case, it is 13 years20:57
madrazr8 years of 3.x development time + 5 years to transform20:57
*** dr__house has quit IRC21:03
antrikactually, when people do development while the whole code base is in flux, it's *better* to de the development on the old code base, and forward port it later21:08
antrikmuch easier than working with a moving target...21:08
antrikand even more importantly, the feature can be landed before the rewrite is finished21:08
antrikthe feature I have in mind should be available for this year's GSoC, not at some indefinite point in the future21:08
antrikalso, if you are changing the whole code base at once, you are probably doing something wrong anyways... it's almost always better to do incremental changes21:09
antrikyou named one example of the problems a complete rewrite leads to yourself... there are many more21:10
madrazrantrik: again, "it is almost always"21:15
madrazrantrik: which is not always21:15
madrazrantrik: there will be cases, when you want to do that21:15
madrazrantrik: yes, we were doing things wrong previously21:16
madrazrantrik: we know that21:16
madrazrantrik: we have been knowing that from a long time now21:16
antrikhehe... I could name one example, but that would be evil ;-)21:16
madrazrantrik: this is our chance to correct ourselves and move forward21:16
madrazrantrik: please21:16
madrazrantrik: if you want may be on PM21:16
madrazrantrik: but I know what that example is, anyways :D21:16
antrikbut these are really extremely rare, like software that didn't really work at all in the original version...21:16
madrazrantrik: it is a known secret between we 2 :D21:17
madrazrantrik: you made the point my friend21:17
antrikoh, are you still following the list?21:17
madrazrantrik: yes, really21:17
antrikah, cool :-)21:17
madrazrantrik: I always want to do some coding for Hurd, but get very little time :(21:17
madrazrantrik: I follow most of the mails except the mails from the guy whose name I can neither spell nor say :P21:18
madrazrantrik: it is just too hard for me :D21:18
antrikhm... not sure what guy you mean :-)21:19
antrikso, what would be the right branch to work on for features that should land soon?21:23
madrazrantrik: well depends on what the feature is21:25
madrazrantrik: ouch wait21:25
madrazrstupid me21:25
madrazrWTB sleep I guess :D21:25
madrazrantrik: well we have resolved to the agreement that we won't land any features until we rewrite our views layer21:26
madrazrantrik: although if you have a bug fix to make, we can talk about it21:27
antrikthat's a very very bad decision21:27
madrazrantrik: I agree21:27
madrazrIn the most idealistic of the cases, that should not have happened21:27
antrikit's not like adding a new feature to the working version hurts anyone21:28
madrazrantrik: but unfortunately we don't live in ideal world. We all have limited time21:28
antrikyou can still work on your rewrite in the meantime, if you think it's necessary...21:28
antrikmind you, I'm not asking you to do the implementation. I just want to know where to work from21:29
madrazrantrik: sure it won't hurt any one, other than the guy who spends his time working on it, which gets changed later21:29
antrikthe feature is extremely important to us21:29
antrik(admittedly, I should have started on it much earlier...)21:29
madrazrantrik: I see your point21:30
antrikwell, if "the guy" wants it that way, it can hardly hurt him?... :-)21:30
madrazrantrik: sure then21:30
madrazrantrik: I can do some mentoring in that case21:30
madrazrantrik: it is just about re-registering a URL21:31
madrazras a legacy URL21:31
madrazrantrik: one extra line of code21:31
madrazrantrik: you don't need another branch for it21:31
madrazrantrik: well, if it is a bug fix, there is a bug fix branch (which I don't think is public atm)21:31
madrazrantrik: but I can push to make that branch public21:32
madrazrantrik: but I wonder if there are any bug fixes now21:32
madrazrantrik: (I can push != I can run hg push) :D21:32
madrazrantrik: I can ask the person who has the branch locally to push it :P21:32
madrazrantrik: anyways nice to know that you want to contribute to Melange21:33
madrazrantrik: we will be excited!21:33
antrikare you saying the branch containin the code actually running the live site only lives in someone's private repository?...21:33
madrazrantrik: and I personally will be blessed21:33
madrazrantrik: no way21:33
antrikhehe... it's totally selfish ;-)21:33
madrazrantrik: I never said that :D21:33
antrikhm... I must have misunderstood then :-)21:34
madrazrantrik: I can dig into the point in the tree at which we made the last release21:34
madrazrantrik: v0-8-2011022521:34
antrikI meant to write to the mailing list, but as we are already at it, I guess I can just as well explain what the feature is...21:34
madrazrantrik: thats the tag21:34
madrazrantrik: and the current live version is running that21:35
madrazrantrik: sure21:35
madrazrantrik: and if you want to talk to the entire team21:35
madrazryou can as well write to the list21:35
antrikwell, the more relevant question is where the code lives which will go live in the near future... and if there is no branch for that purpose yet, where it should be based off21:35
madrazrantrik: I told you I guess, unless something is a bug fix, the code won't go live in the near future21:36
madrazrantrik: although you can base your patch against, v0-8-20110225 tag21:36
*** m4k3r has quit IRC21:41
madrazrantrik: also if you are not convinced with what I have said, please feel free to write to the mailing list21:41
antrikheh, still talks about GHOP...21:42
tpbTitle: soc - Project Hosting on Google Code (at
madrazrantrik: whats wrong with that?21:43
madrazrantrik: did Google ever announce that it won't do GHOP?21:43
madrazrantrik: they only said they are doing a program called GCI21:43
madrazrantrik: anyways, thanks for pointing that out21:43
madrazrantrik: I was just kidding there21:44
madrazrantrik: I will ask some one who has access to edit that page, to change it21:44
madrazrantrik: unfortunately I don't have access21:44
madrazrantrik: thanks for pointing that out21:47
*** madrazr has quit IRC22:02
*** madrazr has joined #melange22:12
*** ChanServ sets mode: +o madrazr22:12
*** dhaun has quit IRC22:19
*** dr__house has joined #melange23:07
*** dr__house has joined #melange23:07

Generated by 2.13.1 by Marius Gedminas - find it at!