Wednesday, 2008-11-05

*** tpb has joined #melange00:00
*** forever is now known as _Allen_00:07
*** dmitrig01 has quit IRC00:31
*** SRabbelier has joined #melange00:56
*** ChanServ sets mode: +v SRabbelier00:56
*** webchick has quit IRC03:00
*** solydzajs has joined #melange03:09
*** ChanServ sets mode: +v solydzajs03:09
*** mithro has joined #melange04:16
*** ChanServ sets mode: +v mithro04:16
*** solydzajs has quit IRC04:23
*** mithro has quit IRC05:33
*** MatthewWilkes has joined #melange05:54
*** _Allen_ has quit IRC06:12
*** Sidnei has joined #melange06:54
*** webchick has joined #melange07:19
*** MatthewWilkes has quit IRC07:42
*** MatthewWilkes has joined #melange07:42
*** webchick has quit IRC07:54
*** webchick has joined #melange07:56
*** webchick has quit IRC08:20
*** MatthewWilkes has quit IRC08:27
*** SRabbelier has quit IRC08:39
*** penyaskito has quit IRC08:39
*** scorche|sh has quit IRC08:39
*** SRabbelier has joined #melange08:40
*** penyaskito has joined #melange08:40
*** scorche|sh has joined #melange08:40
*** irc.freenode.net sets mode: +v SRabbelier08:40
*** MatthewWilkes has joined #melange09:37
*** divz has joined #melange10:54
*** penyaskito_ has joined #melange11:02
*** webchick has joined #melange11:11
*** madrazr has joined #melange11:12
*** Sidnei has quit IRC11:16
*** Lennie has joined #melange11:17
LennieHello :)11:17
*** divz has quit IRC11:21
*** penyaskito has quit IRC11:29
*** solydzajs has joined #melange11:54
*** ChanServ sets mode: +v solydzajs11:54
solydzajshi12:03
Lenniesolydzajs: Renaming the things that form the key_names of a document result in a copy, design choice or just because delete hasn't been properly implemented yet?12:04
solydzajsLennie: the delete of documents was working just fine12:04
solydzajsLennie: so when you change document you get a copy or what's the problem ?12:05
Lennie<starts app_engine, 1 sec>12:06
Lenniek12:07
Lenniesolydzajs: When changing either the partial path or link name the original document isn't deleted12:08
solydzajsLennie: oh that's interesting. I guess it's a bug. Probably cause of Sverre's refactoring12:09
Lenniesolydzajs: I'll look into it tonight then ^^12:09
solydzajsLennie: let me check12:09
Lennieokay12:09
solydzajsyep it's a bug.12:10
solydzajsThanks12:10
LennieShould I put this on the issue tracker?12:10
solydzajsI think I know what's the problem12:10
Lenniethen you'll fix it, I'll stay away from it ^^12:11
solydzajsit's complicated I need to talk with Sverre12:11
solydzajsbasically it's a problem with keynames12:12
Lenniehehe, I've been working on keyname with sverre12:12
Lenniehe's busy refactoring home and site_settings so that it can now all be dealt with in a generic way12:12
solydzajsok so if you change partial path or link name (which both are part of key) you won't find the same document and the updateOrCreate function will create new entity12:13
solydzajswhich sucks12:13
solydzajsSRabbelier: ping12:13
LennieHe's probably on his way home now ^^12:14
solydzajsok12:14
Lennieisn't it easy to just do a getkeyfieldnames and just check if there is a difference?12:15
Lennieif so the entity will be relocated12:15
Lennieso delete the old one12:15
Lenniegetkeyfieldnames will be implemented in all logic\models after the refactoring, already have a patch ready for that12:15
LennieSo, get form information -> Read the keyfieldnames from it -> changed? then delete old one else update12:16
solydzajsI will have a look at it soon or you can if you have time. I'm trying to fix former ids.12:17
Lennieok12:18
LennieI'll try to solve it then12:18
Lenniebut first I need food :D12:18
Lenniebbl12:18
*** Lennie is now known as Lennie|Food12:18
solydzajsLennie|Food: maybe we should just make partial path and link name not editable12:20
Lennie|Foodsolydzajs: That's possible but would limit the control the user has12:21
Lennie|Foodsolydzajs: Relocating documents and such should be possible right?12:21
solydzajsLennie|Food: yep we can create new one and delete old one too.12:23
solydzajsLennie|Food: it's a question of what to do with key_name.12:23
Lennie|Foodsolydzajs: We can either do it in the code or make some sort of special relocation option on existing items12:24
solydzajsLennie|Food: we will see, ttyl :-)12:24
Lennie|Foodsolydzajs: I'll see if I can find a direct way to do it, and else we'll talk about it :)12:24
Lennie|Foodbbl :)12:25
*** MatthewWilkes has quit IRC13:22
*** tlarsen has joined #melange13:37
*** ChanServ sets mode: +o tlarsen13:37
*** Lennie|Food is now known as Lennie13:40
Lenniesolydzajs: you here?13:40
solydzajsLennie: yes13:40
Lenniesolydzajs: I've been thinking, the problem is kinda the same as with the suffix I fixed. The editPost in views\models gets the enity if it exists. So you could check the keyname via the entity if it exists13:44
Lenniesolydzajs: And act upon that13:44
SRabbelierheya14:02
Lenniehi Sverre14:03
SRabbeliersolydzajs: pong14:04
tlarsenSRabbelier: Howdy.14:04
SRabbeliertlarsen: heya :)14:04
SRabbelieris the meeting now, or in an hour?14:04
solydzajsSRabbelier: in 30 minutes14:04
SRabbelierhehe14:05
solydzajswell in 2514:05
* tlarsen just wishes he would stop getting ill.14:08
SRabbeliertlarsen: did you get ill again? :O14:09
tlarsenSRabbelier: never really got over the last time completely, I suppose.14:09
SRabbeliertlarsen: have you been not taking a break as you said you would this weekend? :P14:10
tlarsenSRabbelier: I did, but stuff tends to linger with me, since I have a weakened immune system.14:10
SRabbeliertlarsen: ah, how do you get better then?14:10
SRabbeliers/better/well/14:10
ibotSRabbelier meant: tlarsen: ah, how do you get well then?14:10
tlarsenSRabbelier: it just takes time.14:11
* SRabbelier nods14:11
SRabbelierok14:11
tlarsenSRabbelier: I didn't say I have *no* immune system. :)14:11
SRabbelierhehe, good thing, that14:11
*** madrazr has left #melange14:19
solydzajstlarsen, SRabbelier: ping14:30
SRabbeliersolydzajs: pogn14:30
solydzajsok let's wait for Todd14:31
SRabbelierLennie: ping14:31
Lennieping: 4 sec14:31
LennieIRC meeting?14:31
* durin42 promises to pay attention if he isn't in a meeting at work14:31
SRabbelierdurin42: are you?14:31
solydzajsdurin42: thx14:32
solydzajsLennie: yes IRC meeting, you might want to join14:33
Lenniesolydzajs: I'm in :)14:33
tlarsenI'm here as well.14:34
solydzajsok14:35
solydzajsso first let's start with summary of what we have been working on since Mentor Summit14:35
solydzajsSRabbelier: maybe you can be first ?:-)14:36
SRabbeliersure14:36
solydzajsBasically what are you working on right now and what's the status14:36
solydzajsand what are next steps for you14:36
* SRabbelier does a git log since :P14:36
* SRabbelier is typing14:38
SRabbelierfirst I worked on the generic view code a lot, addressing comments by Todd and Pawel, then made Document, Sponsor generic, and added a sort-of implementation for Host.14:38
solydzajsok but after Mentor Summit :-)14:39
SRabbelierthis was after the summit ;)14:39
* SRabbelier is still typing14:39
SRabbelierI moved the keyName stuff from model to the logic modules. and started working on making User generic as well.14:39
SRabbelierThis took quite some time, as I had to clean up a lot of code first, but clean code is good ;)14:40
SRabbelierThen I got interrupted by Lennard, who asked me about what he could do for Melange.14:40
SRabbelierI suggested making the keyName sutf generic too14:40
SRabbelierTo do that the Site/Home Settings stuff had to be made generic too, so I started working on that14:40
SRabbelierThat is almost done now14:41
SRabbelierWas working out the last part of that before this meeting14:41
SRabbelierAfter I finish this, I'll go on with User14:41
solydzajsso basically you have almost done generic User and Host/Site Settings ?14:41
SRabbelierwhen that's generic as well, I think we have pretty much all views that we currently use, using the generic view system.14:42
SRabbelieroh14:42
SRabbelieralso, I did a major rehaul of the site naming14:42
SRabbelierthis is thoroughly tested, but I have it sitting in my private branch atm14:42
SRabbelierI'll push it when I finish site settings and have tested that too14:42
SRabbelier(we really need unit tests)14:42
SRabbelierafter that, I can get to the 'invite a host' stuff14:43
solydzajsok you will have to merge with me regarding User , since I touched a little bit of Users view code to fix former ids and other bugs there14:43
SRabbelierI have that sitting in a branch too14:43
SRabbeliersolydzajs: if you commit to svn I will rebase my work on trunk :)14:43
SRabbelier<3 git :)14:43
solydzajsSRabbelier: ok14:43
tlarsenSRabbelier: Will the views be one view that tailors itself to the role of the accessing User?  Or, is this something for later?14:43
SRabbeliertlarsen: no, atm it's still done through the self._rights field14:44
SRabbeliertlarsen: but we have a public() view for mostly anything14:44
SRabbeliertlarsen: I'm not doing that till Pawel is done with his as_table work14:44
tlarsenSRabbelier: So, the public view will show only a subset of fields, but to everyone?14:44
SRabbeliertlarsen: hopefully I can integrate that logic with as_table, so that it's possible to have the field tailor themselves :)14:45
SRabbeliertlarsen: well, not to anyone14:45
SRabbeliertlarsen: you can say that you can only see the public view if you'e ra mentor14:45
SRabbelier**you're a mentor14:45
SRabbeliertlarsen: and that you can edit it if you're an org admin or higher14:45
tlarsenSRabbelier: Ah, OK.14:45
solydzajsSRabbelier: ok sounds good.14:46
solydzajsI can be next14:46
SRabbeliersolydzajs: go ahead14:47
solydzajsok I have almost working as_table the only missing thing are tooltips but I will do it soon, do some additional code cleaning and commit.14:48
solydzajsnext things I've been working on is fixing broken former_ids14:48
solydzajsthis is also 99% done14:48
solydzajsand I just discovered one bug14:49
solydzajswhen we compare ids and emails we need to compare all of them in lower case14:49
solydzajsotherwise somebody can change User email from: [email protected] to [email protected]14:50
SRabbeliersolydzajs: aah, heh, nice one14:50
solydzajsand it's possible right now14:50
solydzajsI have this bug noted and I will work on that14:50
Lennienice find indeed14:50
tlarsensolydzajs: That isn't going to work with escaped email addresses, though.14:50
tlarsensolydzajs: The email address RFCs allow some pretty weird stuff if it is escaped, sadly.14:50
SRabbeliertlarsen: normalize them then?14:51
solydzajsok that's something we need to figure out14:51
solydzajsNext thing, I added support for Invalid accounts14:51
tlarsenhttp://en.wikipedia.org/wiki/E-mail_address#RFC_specification14:52
tpb<http://ln-s.net/2RIQ> (at en.wikipedia.org)14:52
tlarsenWe can choose to just not support quoted email addresses.14:52
solydzajsif account is in former_ids and user tries to create profile with this account he/she gets error14:52
tlarsenI don't know if Google Accounts does or not.14:52
tlarsensolydzajs: Yes, we need that feature, for the reasons we discussed during the summit.14:52
solydzajstlarsen: it's ready14:53
solydzajstlarsen: just need to clean the code and commit14:53
tlarsensolydzajs: Have we decided yet if we are going to go ahead and make the link_name immutable?14:53
SRabbeliertlarsen: no need anymore14:53
tlarsensolydzajs: I sent an email to the dev list as a result of some of the confusion about nick_name and link_name, but no one responded.14:53
SRabbeliertlarsen: with my user refactor I fixed that14:53
solydzajsLike I talked with Sverre before, I didn't want to check if user id is in former ids everytime we call checkIf functions , so I only check when user tries to create profile14:54
solydzajstlarsen: it's on my reply todo list :-)14:54
tlarsensolydzajs: That sounds fine to me.14:54
tlarsenAs far-reaching as it would be, I think we need to call link_name and nick_name something else.14:54
tlarsenToo many people are confusing nick_name with an IRC nick, for example, when it is not.14:55
solydzajsdisplay name is used in Person model already (display name for certificates)14:56
tlarsensolydzajs: Yes, I know.14:56
tlarsensolydzajs: I didn't say I had the solution, I just wanted to start the discussion.14:56
solydzajsmaybe we can call it Name on Certificate :-)14:56
solydzajsand use display name in User14:56
tlarsensolydzajs: The point is, we need to make it clear that it is *not* an IRC nick, and that it supports UTF8 characters, etc.14:57
solydzajstrue14:57
* SRabbelier agrees14:57
tlarsensolydzajs: Maybe by calling it public_id or something, and saying "the way that your User presence is publically displayed on the site".14:58
SRabbeliertlarsen: that's a rather bad description though :P14:58
tlarsenSRabbelier: so, with the User refactor, there is no need for link_name to be immutable?14:58
SRabbeliertlarsen: nope14:58
SRabbeliertlarsen: it automagacially translates between the two14:58
SRabbeliertlarsen: as we discussed at the summit14:59
tlarsenSRabbelier: OK, please suggest a better description, then.  :)  Examples in the help text would probably work better.14:59
tlarsenSRabbelier: (in the email thread)14:59
SRabbelierI think s/your User precence/your Site-wide account/14:59
solydzajshmm14:59
SRabbelierwould help a lot14:59
tlarsenI think we should rename link_name to something like url_id, since that is where we use it.14:59
SRabbeliersure15:01
tlarsenWe have gotten so accustomed to our own little vocabulary in the app, that developed slowly, that it took some input from people new to the project to see some of the points of confusion more clearly.15:01
SRabbeliertlarsen: at first I thought that most of that stuff was somehow required15:02
durin42Yes, stepping in and testing has been tricky because of the vocabulary oddities15:02
SRabbelierit wasn't till I went in head-first and started throwing stuff around till I realised how much of the code is by choice15:02
SRabbelierdurin42: can you explain?15:02
solydzajsok guys let's keep on track, we know what's the status of Sverre's work and mine.15:02
solydzajstlarsen: did you have a chance to code ?15:03
solydzajstlarsen: I know you were not feeling well for the last couple days15:03
tlarsensolydzajs: I have only been poking around with Allen's patch.  Nothing to commit.15:03
tlarsensolydzajs: It is best that I don't code under the influence of meds.  :)15:03
solydzajstlarsen: ok, what's your plan for upcoming week ?15:04
solydzajstlarsen: hehe :-)15:04
tlarsenI want to get a Site-wide Terms of Service working, using a hard-coded Question object and a hard-coded Quiz object.15:04
* SRabbelier nods15:04
tlarsenI also have been trying to figure out how to move the Page stuff into the view classes.15:05
SRabbeliertlarsen: it'll be a lot easier when my rename patch is in15:05
solydzajstlarsen: ok sounds good.15:05
SRabbelierwith how structured most of it becomes then15:05
tlarsenSRabbelier: I think that we should not have singleton view objects, but should instead instantiate them when creating the site map.15:05
SRabbeliertlarsen: right, you said that, I agree15:05
solydzajstlarsen: I agree15:05
solydzajsok15:05
solydzajsdurin42: how are you doing with free time ?;-)15:06
tlarsenSRabbelier: That way, we do not have to pass a Page object to the view.  It will simply be member attributes of the object.15:06
tlarsenI'm not finished.15:06
tlarsen:015:06
solydzajsoh ok15:06
solydzajs:-)15:06
solydzajssorry15:06
SRabbelierpwnd :P15:06
tlarsenSRabbelier: The one thing I am having trouble resolving, though, is that the views are actually methods of a class.  Each view is not a separate object.15:06
tlarsenSRabbelier: So, I'm trying to figure out how one of the "views" objects can provide all of its site map entries.15:07
solydzajstlarsen: let me know when you finish.15:07
SRabbeliertlarsen: then use the same object for multiple views15:07
tlarsenSRabbelier: yes, but then how do you uniquely associate the URL with a particular view method?15:07
SRabbeliertlarsen: have it sign up at some sort of registry15:07
tlarsenSRabbelier: the site map is such a registry.15:07
SRabbelierurl_manager.register(self._views)15:07
tlarsenSRabbelier: it is just constructed explicitly in the code currently.15:08
SRabbeliertlarsen: yeah, but it feels kind of hacky15:08
tlarsenSRabbelier: the URL was just an example.15:08
SRabbelierso much indirection going on for something relatively simple15:08
tlarsenSRabbelier: the more difficult thing is representing the "parent" relationship between pages, so that the sidebar menu and the breadcrumbs are constructed appropriately.15:08
SRabbelierI think there should be just one object in urls.py/map.py that serves a receptient for Views to register with15:09
tlarsenSRabbelier: I shouldn't have mentioned URL.  It was a red herring.15:09
SRabbeliertlarsen: yes, I understand that, and I agree15:09
tlarsenSRabbelier: the real issue is how to express the hierarchical nature of the pages.15:09
SRabbeliertlarsen: but, I think we should not try to make the site map too complicated15:09
SRabbeliertlarsen: if we keep it simple, that can just be a string argument when the page registers15:09
tlarsenSRabbelier: I think everyone agrees that the current menu hierarchy makes sense.15:09
SRabbelierwhen we limit ourselves to one level15:10
tlarsenSRabbelier: the question becomes how to convey that.15:10
tlarsenSRabbelier: yes, you could supply a string to indicate your "parent", but do you use some sort of lazy initialization?  (that is what I have been trying to work out).15:10
SRabbeliertlarsen: sure, sounds good15:10
tlarsenSRabbelier: you need to have a placeholder for when a page says my parent is "/foo", but the site map hasn't seen "/foo" registered yet.15:11
SRabbelierjust don't try to retreive untill everybody is registered15:11
SRabbeliermy pont it15:11
SRabbelier**is15:11
SRabbelieryou don't have to register with foo15:11
tlarsenSRabbelier: that is a given, but, if you look at site/map.py carefully, you will see that everything was listed in order from root to leaves.15:11
SRabbeliereach entity is pretty self contained15:11
SRabbelierwith my current refactoring15:12
SRabbeliereverything goes under /15:12
SRabbeliersince we are removing /site15:12
SRabbelierwe might as well remove the "Site"/"Admin" sub-menu15:12
SRabbelierand just display to the user what they have access too15:12
tlarsenSRabbelier: so, you have a method in each "views" class that produces its sub-menu?15:12
SRabbeliertlarsen: in the base class ofcourse, but yeah, each view could define which views it wishes to export15:13
tlarsenThe Site menu will not go away for actual logged-in Developers.  Otherwise, how do they get to the site home page settings, etc.?15:13
SRabbeliertlarsen: I am working on that with the site refactoring :)15:13
SRabbeliertlarsen: there is no need for there to be a site sub menu15:13
tlarsenSRabbelier: It will likely be very de-populated, as the views recognize Users' permissions, but at least one menu item will always be needed, I think.15:14
SRabbeliertlarsen: which one is that?15:14
tlarsenSite Settings15:14
tlarsenHow do you propose to provide Developers access to that page without a sidebar menu choice?15:15
SRabbeliernah, a Mentor won't need to see Site Settings, do they ;)15:15
tlarsenSRabbelier: No, they won't, but Developers will.15:15
SRabbelierright15:15
SRabbelierbut devs have access to everything15:15
tlarsenSRabbelier: I am not talking about conditionally displaying sidebar menus.15:15
SRabbelierso they just see this structure:15:15
SRabbelier|- User15:16
SRabbelier|- Documents15:16
SRabbelier|- Sponsor15:16
SRabbelier|- Host15:16
SRabbelier|- Site Settings15:16
SRabbelier|- Home Settings15:16
SRabbelier|- Orgs15:16
SRabbelierbut there is no need to nest everything under |- Site Settings15:16
tlarsenWhat is under Home Settings?15:17
tlarsenWho sees it?15:17
SRabbelier|- Home Settings15:17
SRabbelier  |- List15:17
SRabbelier  |- Create a new Home Setting15:17
SRabbelierit's the dev's view that allows him to edit any of the Home Settings15:17
SRabbeliersuch as the home page for Apache15:17
solydzajstlarsen: I just checked with emails in GMail: "Sorry, only letters (a-z), numbers (0-9), and periods (.) are allowed. "15:17
SRabbelierPawel Solyga: and +15:18
SRabbelierPawel Solyga: and periods can be placed anywhere with gmail...15:18
SRabbeliersolydzajs: I think best would be to strip all non-standard characters15:18
tlarsenSRabbelier: And the periods are *not* significant.15:19
solydzajsSRabbelier: "Sorry, the last character of your username must be an ascii letter (a-z) or number (0-9)"15:19
SRabbeliersolydzajs: that means we won't accept a "[email protected]" and a "[email protected]" at the same time, but that's ok I think15:19
tlarsenSRabbelier: Yes, but it means a custom comparison function.15:19
SRabbeliertlarsen: you can register both [email protected] and [email protected] though, I think15:19
SRabbeliertlarsen: obviously15:19
tlarsenSRabbelier: ugh, that's right.15:19
tlarsenSRabbelier: it will be email service-specific.15:20
tlarsenSRabbelier: sigh.15:20
SRabbeliertlarsen: just have it be generic15:20
tlarsenSRabbelier: I guess we just need to count on Google Accounts only giving us unique emails?15:20
solydzajstlarsen: I just want to check lower upper case and that's it15:20
SRabbeliertlarsen: if anyone has both "[email protected]" and "[email protected]" they'll have to pick15:20
tlarsenI think we should just stick with exact matches and be done.15:21
tlarsensolydzajs: Why do we need to normalize to all-lowercase?15:21
solydzajstlarsen: when I login with my gmail account to deployed Melange app I see my email as [email protected]15:22
tlarsensolydzajs: Does the id.email() always produce that string for you?15:23
tlarsensolydzajs: If so, that is your ID.15:23
tlarsensolydzajs: Keep in mind that the dev_appserver.py might let you do things with its fake "Google Accounts" that the deployed App Engine will not.15:23
solydzajstlarsen: yes it does, but Developer can change email right ? So I can put [email protected] in place of [email protected] and it will change it15:23
solydzajstlarsen: I'm testing on deployed15:24
solydzajstlarsen: that's how I found this bug15:24
SRabbeliersolydzajs: nice catch though :)15:24
tlarsensolydzajs: Ugh.  That is not nice.15:24
tlarsentlarsen: The problem is, according to this, it won't hold true for all (common, not bizarre, unusual) email systems that case does not matter:15:25
SRabbeliersolydzajs: can't we just do a .lower() on the id field when it goes into the store, and then again when we get it from the user?15:25
tlarsenhttp://en.wikipedia.org/wiki/E-mail_address#Local-part_normalization15:25
tpb<http://ln-s.net/2RJ6> (at en.wikipedia.org)15:25
tlarsenSRabbelier: or when we get it from the users module also?15:25
SRabbeliertlarsen: the saving part we can, the getting from users.get_current_user needs to be encapsulated so that we can wrap it15:26
SRabbeliertrivial ofcourse15:26
SRabbelierbut has to be done nonethelss15:26
tlarsenSRabbelier: except that you will break users that sign up for a Google Account where their underlying email system is case-sensitive.  That being said, those users may *already* be broken with Google Accounts.15:27
tlarsenI can do some internal checking, I think.15:27
SRabbeliersolydzajs: so why were you in there as "[email protected]" in the first place?15:27
solydzajsSRabbelier: when I created my GMail account as my login I put Pawel.Solyga15:28
solydzajsSRabbelier: of course I can login with pawel.solyga too15:28
SRabbeliersolydzajs: aaah, heh, ok15:28
solydzajsSRabbelier: but GMail remembers me as [email protected]15:28
SRabbeliersolydzajs: interesting :)15:29
solydzajstlarsen: +1 for internal checking15:29
tlarsensolydzajs: Looking now.  Also, I am done talking.  :)15:29
solydzajsWhat I will do for now is probably just lower when comparing15:29
solydzajstlarsen: ok thx :-)15:30
solydzajsdurin42: ping15:30
solydzajsok durin42 has idle time 28 mins :-)15:31
solydzajsok, I would like to welcome and thank for so far contributions to Lennie :-)15:31
SRabbelierprolly in a meeting then15:31
solydzajsHi Lennie, you there ?15:31
LennieI'll keep them coming15:31
Lennieyeah I've been silently watching this talk :)15:31
SRabbelierLennie: you sneaky bastid ;)15:32
solydzajsLennie: ok great, what do you plan to work on ?15:32
Lenniesolydzajs: at the moment I'm trying to find bugs and solve them so I can get a general understanding how everything is working15:32
Lenniesolydzajs: When I reach the point that I feel I know how everything works I'm sure I'll switch to adding new features or something :)15:32
solydzajsLennie: sounds good to me :-) I'm doing that all the time since Sverre's refactoring :D hihi15:33
* SRabbelier mutters 'test suite'15:33
* Lennie hugs Sverre, dont let the pawel bite you ^^15:33
solydzajsoh I forgot15:33
SRabbelierLennie: hehe :D15:33
solydzajsI've been looking into more tests and code coverage tools15:33
SRabbelierLennie: he kicks my ass when I break somethign  :P15:33
* tlarsen mutters to SRabbelier 'write some...'15:33
solydzajstlarsen: what are you using at G ? something internal or open source ?15:33
tlarsensolydzajs: Google uses a custom code-coverage system.  It makes use of language-specific backends that are open source, though.15:34
solydzajsSRabbelier: we will get to the point where everything works, then there will be no more refactoring without tests. If you want to refactor something write missing tests then refactor I think this should be a rule.15:35
SRabbeliersolydzajs: heh :)15:35
solydzajstlarsen: can we know what kind of open source backend is used for python ?15:35
solydzajsSRabbelier: I'm not kidding :-)15:36
durin42oh, hi15:36
solydzajsdurin42: hi :-)15:36
durin42I'm back if you still want me around15:36
SRabbeliersolydzajs: that's ok :)15:36
durin42Found a regression in stuff at work15:36
tlarsenI believe it is just the built-in profile.15:36
solydzajsdurin42: yep :-) I was wondering how are you doing with free time ? and do you have any Melange work plans for upcoming week or so ?15:37
tlarsensolydzajs: I'm not on the testing team, so I just consume the output in their lovely web UI.  :)15:37
solydzajstlarsen: :-) I saw the UI :-) it even uses Google Chart API :-)15:37
durin42solydzajs: I can probably find some time this week, hgsubversion is calming down for the moment15:37
durin42solydzajs: the problem is I'm still drowning in the infrastructure that exists15:38
solydzajsdurin42: ok. I guess you can work on more tests if that's ok with you ?15:38
SRabbelierdurin42: part of the reason I'm doing so much reason is to make that better15:38
tlarsenProfiling in App Engine:15:38
SRabbeliers/so much reasons/so much refactoring/15:38
tlarsenhttp://code.google.com/appengine/kb/commontasks.html#profiling15:38
tpb<http://ln-s.net/2RJO> (at code.google.com)15:38
SRabbeliertlarsen: that's no test coverage though, is it?15:39
durin42solydzajs: that's fine, but I need some ideas on what should be tested vs what will be refactored15:39
tlarsenSRabbelier: it lists function calls only, I think.15:40
solydzajstlarsen: rietveld is doing profiling that way15:40
tlarsenAh hah:  http://nedbatchelder.com/code/modules/coverage.html15:40
tpb<http://ln-s.net/1xDa> (at nedbatchelder.com)15:40
solydzajstlarsen: but I'm more interested in tests code coverage15:41
solydzajstlarsen: yep found it too15:41
solydzajstlarsen: this one looks really nice15:41
tlarsensolydzajs: You get that by combining the execution coverage of running all of the tests.  :)15:41
solydzajstlarsen: and nose has support for it15:41
tlarsensolydzajs: Then we should just use the one in nose.15:41
SRabbelierPawel Solyga: sounds good15:42
solydzajstlarsen: yep nose supports the one you sent link for15:42
durin42nose doesn't include coverage.py15:43
durin42but it's easy to add15:44
durin42we can put it in third party and things should work reasonably15:44
solydzajsdurin42: yep it doesn't include but it support it right ?15:44
durin42right15:44
solydzajsok good15:44
durin42you have to add coverage.py to someplace on the search path then do --with-coverage --cover-packages=<packages>15:44
solydzajsyep exactly15:45
tlarsensolydzajs: Ugh.  Google Accounts support changing the email address used to log in. :(  It uses a unique ID internally, that is obviously secret and not revealed to App Engine (for obvious security reasons).15:45
tlarsendurin42: add it somewhere in thirdparty ?15:45
durin42tlarsen: sure, I can do that15:45
tlarsendurin42: I'm still waiting on a wiki topic (or sub-topic) on how to run the fancy tests.  :)15:46
durin42tlarsen: emailed myself a reminder, will get to this first thing tonight after revising the pycon proposal15:46
solydzajstlarsen: and I'm still working on wiki topic about pylint :-)15:46
durin42hgsubversion is finally usable enough I don't notice warts in it15:46
durin42(well, notice in a "must....fix....now..." way)15:47
tlarsensolydzajs: I think the pylint article needs to be linked from the contribution topic.15:47
tlarsensolydzajs: It needs to say "run pylint before committing".15:47
tlarsensolydzajs: In the section on sending a patch to the mailing list, it should say "run pylint before sending your patch".15:47
solydzajstlarsen: yep but I need to write one more script that runs pylint only on changed code based on svn status output or using svn python binding15:48
tlarsendurin42: The Mercurial Eclipse plug-in devs are going to shortly release a beta of the plug-in with the hgsubversion support.15:48
tlarsensolydzajs: Why?  Does it take a long time to run it on all code?15:48
durin42tlarsen: I'm amazed how everyone in the hg world is suddenly rallying around this tool.15:48
solydzajsdurin42: you rock :-) hehe15:49
durin42solydzajs: it's just a hobby project, honestly15:49
durin42keeps my brain in shape15:49
solydzajs:-)15:49
tlarsendurin42: The nice thing about them releasing a beta of the plug-in is that they have an Eclipse updater URL for their beta as well as release plug-ins.15:49
durin42cool15:49
tlarsendurin42: So, Eclipse will be able to pull it down auto-magically.15:49
solydzajstlarsen: yep it takes a little and I still need to fix couple warnings. I don't want patch contributors to see warnings about our TODOs ;-)15:50
tlarsensolydzajs: Then we should just fix them.  :)15:51
SRabbeliertlarsen: right ;)15:51
solydzajs:-)15:52
Lennie:)15:52
solydzajsdurin42: ok so you will focus on tests ?15:53
durin42solydzajs: yeah15:54
solydzajsbtw nice link: http://pycheesecake.org/wiki/PythonTestingToolsTaxonomy15:54
tpb<http://ln-s.net/2RJh> (at pycheesecake.org)15:54
durin42solydzajs: I'm kind of enjoying the idea of getting better at QA15:55
durin42solydzajs: we have QA guys here and they feel like wizards to me with what they do, so the goal is to learn how it works15:55
solydzajsdurin42: ok awesome :-) yep I'm recently reading about QA a lot more :-)15:55
solydzajsdurin42: if you have any good tips don't hesitate to let me know :D15:56
SRabbeliersolydzajs: nice comparison15:56
durin42solydzajs: I'm planning on getting help from our QA team at leapfrog15:56
durin42since we all (claim) to love FLOSS15:56
SRabbelierhehe, nice15:56
solydzajsgood :-)15:57
solydzajsok did I forgot about somebody with tasks assignment ?;-)15:58
solydzajsdurin42, tlarsen, Lennie, SRabbelier, solydzajs, I think that's all :-) at least for now15:59
Lennieokay15:59
SRabbeliersolydzajs: thanks for leading this meeting :)15:59
Lenniethanks for the meeting :)15:59
solydzajsthanks for participation :-)16:00
SRabbelierawesome16:00
SRabbeliernow... on to fix those views16:00
Lennieon to fixing double documents16:00
Lenniegot an ugly fix but I'm not committing that :P16:01
tlarsensolydzajs: It is nice to have access to the source code sometimes.  :)16:03
tlarsensolydzajs: the local portion (the part before the @) is compared in lowercase.16:03
tlarsensolydzajs: and, the RFC says that the domain part (the part after the @) has to be case-insensitive anyway.16:03
solydzajstlarsen: ok so basically we will compare in lowercase too :-)16:04
tlarsensolydzajs: Yes, but that won't catch the "Google Mail ignores dots and stuff after the plus (+)" cases.16:04
tlarsenAssuming the local portion is case-sensitive is "discouraged" by RFC 5321, so we are justified enough in doing it that way.16:05
solydzajstlarsen: ok GMail only ignores that, and Google Account can be used with other emails too right ?16:06
tlarsensolydzajs: Yes, so it would require looking at the part after the @ and seeing if it matched gmail.com or googlemail.com16:06
tlarsensolydzajs: I don't think we need to go that far.16:06
SRabbeliertlarsen: no site-specific stuff please16:06
SRabbeliertlarsen: either do it for all, or not at all16:07
tlarsensolydzajs: If someone is willfully trying to break stuff, we'll just ban them.  :)16:07
solydzajstlarsen: ok :-) so I will just compare in lowercase for now allright ?16:07
tlarsenThe Developer will need to be *very* careful when changing someone's email address, though.16:07
tlarsensolydzajs: Yes, just lowercase for now.16:07
solydzajsLet's not put that bug into Issue Tracker :-)16:08
tlarsensolydzajs: Definitely not.  What bug?  :)16:09
solydzajstlarsen: I don't know :-) what are you talking about ?;-)16:09
solydzajs:D16:09
tlarsensolydzajs: Exactly.  :)16:09
SRabbeliergood thing teh chan isn't logged, mhh16:09
SRabbelier:P16:09
Lenniesees a whole in the logging ^^16:10
Lennie*hole16:10
tlarsensolydzajs: The IRC logs have ruined our plausible deniability, though.16:10
solydzajsnobody reads IRC logs :-)16:11
solydzajs:D16:11
Lenniehmm16:12
LennieThe US government does16:13
Lenniethey see all ^^16:13
SRabbelieryup16:13
SRabbelier'terrorist, bombing, tomorrow'? :P16:13
Lennieand that's not gonna CHANGE! ^^^16:13
Lenniehaha yeah SRabbelier, too bad they didnt raid you for that SMS I send :)16:13
SRabbelier:P16:14
SRabbelierLennie: I wouldof totally had your ass if you made me miss my flight because of that ;)16:14
solydzajsLennie: you think we can expect some hackers from NSA or Homeland Security or maybe FBI to join our #melange channel soon ?;-)16:14
solydzajsSRabbelier: my flight was canceled without that ;-)16:14
Lenniesolydzajs: My guess is they are already here :)16:14
solydzajsLennie: hehe :-)16:15
SRabbeliersolydzajs: I heard! at least they paid for your hotel :)16:15
Lenniesolydzajs: What do you do in your daily life?16:16
SRabbelierLennie: he's a slacker, doesn't do anything, ever! :P16:16
solydzajsSRabbelier: yep :-) but just so you can imagine my flight was 1:38PM, and I got to the hotel at 10:00PM16:16
LennieSRabbelier: We aren't talking about you!16:16
Lennieand trust me I know :P16:17
SRabbeliersolydzajs: dude! that sucks 0.o16:17
Lenniethat's really sick :S16:17
solydzajsLennie: chasing nice chicks :-)16:18
Lenniehold on16:18
Lennieyou can code16:18
Lennieand chicks like you :P16:18
Lenniehmm16:18
Lenniesomeone see the paradox?16:18
SRabbelierLennie: it's not mutually exclusive you know ;)16:18
Lenniehehe16:18
*** Sidnei has joined #melange16:19
solydzajsLennie: well I do a lot of different stuff, in different project :-) I'm NUI Group (Natural User Interface Group) co-founder , we have bunch of projects there, I work of NUI Europe AB company working on multitouch screens and software for it, I work on Melange, I work on some iPhone/Mac Objective-C projects etc etc :-)16:19
Lenniemultitouch is cool :D16:20
Lennieyou do a lot16:20
SRabbelierLennie: you should read his shared feeds16:20
LennieI know you do SRabbelier ^^16:20
SRabbelierLennie: there's -awesome- stuff there about NUI16:20
solydzajs:-)16:20
solydzajsoh and I sleep sometimes :D16:21
SRabbelierhe just pretends16:21
solydzajshttp://www.youtube.com/watch?v=thOxW19vsTg&eurl=http://www.max3d.pl/news.php?id=40016:23
tpb<http://ln-s.net/2RKA> (at www.youtube.com)16:23
solydzajsthat's just crazy16:23
solydzajs:-)16:23
solydzajsstart trek at CNN :-)16:23
solydzajshehe16:23
*** Sidnei has quit IRC16:24
SRabbelierawesooome :D16:24
Lenniehmm16:26
Lenniedoes he really see her like that?16:26
Lennieor is it just the people who are watching :P?16:26
solydzajsyep he sees her like that16:28
SRabbeliersolydzajs: amazing16:29
Lennieno way16:32
Lennieligt can't just stop :S16:32
Lennielight16:32
Lennieits just a projection for the viewers just like those advertisements in soccer games16:33
Lennie<- doing the dishes, brb16:33
SRabbelierLennie: forcefields and/or mirrors are involved16:33
SRabbelierLennie: possibly a portable holographic transmitter for the EMH16:34
LennieB'lna reroute all auxiallary power to the main deflector disc16:34
LennieParis, set a course for the Alpha-Quadrant warp 916:34
Lennieengage!16:34
durin42solydzajs: yeah, the first time I saw that I said "oh, it's clever chroma-keying", but upon further research it's real telepresence16:35
Lennieback to dishes, stupid highlighting :P16:35
SRabbeliersolydzajs, durin42: so how do they do it? :P16:36
solydzajsdurin42: yep it is, first time I saw real telepresence was some time ago when Cisco presented it on stage.16:36
durin42SRabbelier: the theory of holographic imagery actually predates the invention of the laser16:36
durin42I don't make any claim to actually understand, I bailed on electrodynamics before I got there16:37
SRabbeliersolydzajs: oh, I saw that one!16:37
durin42but maybe I should go back armed with my better mathematics and try electrodynamics again16:37
solydzajsCNN is investing a lot in new technologies16:39
solydzajsfirst multitouch magic wall16:39
solydzajsnow this16:39
SRabbeliersolydzajs: yeah, they got money to spare16:40
solydzajslike big G :-)16:41
SRabbeliersolydzajs: not so much anymore with the financial crisis :P16:42
solydzajsdid you read Q3 Google report ?;-)16:42
SRabbelierpffff, then they can take me as an intern!16:43
Lennieback16:45
LennieSRabbelier, I hope they do16:45
Lennieyou can get me a new pen16:45
Lenniethe other is empty :P16:45
Lennieor broken16:45
SRabbelierLennie: hehe, not empty, just the paper was crappy! :P16:46
Lennienope doesnt work on my paper either :S16:46
SRabbelierLennie: pfff, try again!16:47
Lennielol16:48
Lennie^^16:48
*** tlarsen is now known as tlarsen|afk17:06
*** tlarsen|afk is now known as tlarsen17:06
tlarsenSRabbelier: how long until you check in your giant refactoring?17:07
tlarsenSRabbelier: Allen's changes will be tricky to merge, so I was just wondering.17:07
SRabbeliertlarsen: ah, I am just about done with the view refactoring, have to add a link and I'll be about done17:07
SRabbeliertlarsen: I can commit after some more testing17:07
tlarsenSRabbelier: OK, by tomorrow, then?17:08
Lenniesolydzajs, I can fix the double document problem but I dont think the solution is nice :S17:08
tlarsenSRabbelier: the Page changes are also pretty intrusive, so I don't want to work on them with a major refactoring in progress, only to have to re-do them because they won't merge.  :)17:08
solydzajstlarsen: I will commit former ids fixes tomorrow.17:08
SRabbeliertlarsen: yup, should be good17:09
solydzajsLennie: you can send the review to the list and we will comment how to improve17:09
solydzajsLennie: patch I mean, not review17:09
Lennieokay but I warn you, its uglY :P17:09
SRabbelierLennie: what you showed me earlier is not that ugly, it's just not that elegant ;)17:09
solydzajsLennie: do you know how to make it less ugly ? or not ugly at all ?17:09
SRabbelierLennie: unless you added something ugly :)17:09
LennieSRabbelier: Yeah that bit17:10
SRabbeliersolydzajs: don't worry, it's not that ugly17:10
solydzajsSRabbelier: ok :-)17:10
SRabbeliersolydzajs: Lennie has a habit of not giving himself enough credit ;)17:10
Lenniesolydzajs: I have an idea on how to make it better, but I'm not sure if I can get it to work :p17:10
LennieSRabbelier thinks I'm some sort of know-it-all :S17:11
Lenniehe's mistaken :p17:11
SRabbelierI have evidence of the contrary *subtly points at Lennie's avg grades*17:11
LennieGrades dont say that much :S17:12
Lennieteachers are just too predictable in the tests they make :P17:12
SRabbelierheh17:12
Lennieand sure I do have a head for stuff like Turing-Machines but that doesnt mean I can code ^^17:13
SRabbelierLennie: fair enough :)17:13
solydzajsok guys I will catch you tomorrow17:15
solydzajsif you need anything email me17:15
solydzajs:-)17:15
SRabbeliersolydzajs: will do ;)17:16
* SRabbelier echo "to: [email protected] subject: need help body: need 10.000 euros plx" | sendmail17:16
*** tlarsen has quit IRC17:17
solydzajsfilter: from sverre move to trash17:17
SRabbelier:(17:17
solydzajs:-)17:17
* SRabbelier sends from 'alturin'17:17
solydzajsbut for alturin you have Sverre in display name anyway17:18
solydzajs:-)17:18
SRabbelierhehe, darn17:18
solydzajsok cya guys17:18
SRabbeliercheers17:18
*** solydzajs has quit IRC17:18
*** Sidnei has joined #melange17:21
Lennieugly patch send in :P17:28
LennieI'm tired17:34
Lennienn :)17:34
SRabbelierni'night17:35
*** Lennie has quit IRC17:41
*** lh has joined #melange17:42
*** ChanServ sets mode: +o lh17:42
*** MatthewWilkes has joined #melange17:49
*** SRabbelier has quit IRC18:40
*** lh has quit IRC18:48
* Sidnei is away: I'm busy19:39
*** MatthewWilkes has quit IRC19:57
*** madrazr has joined #melange21:30
*** madrazr has quit IRC23:21
*** webchick has quit IRC23:42
*** grim has quit IRC23:42
*** webchick has joined #melange23:50
*** grim has joined #melange23:53

Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!