IRC logs of #tryton for Thursday, 2009-02-19 #tryton log beginning Thu Feb 19 00:00:01 CET 2009
2009-02-19 00:00 <CIA-10> tryton: ced roundup * #796/AttributeError: 'Screen' object has no attribute 'views': [need-eg] Could you describe step by step what you do? And give which version do you use?
2009-02-19 00:17 -!- paola_( has joined #tryton
2009-02-19 01:53 -!- igor__(i=igor@ has joined #tryton
2009-02-19 05:20 -!- yangoon( has joined #tryton
2009-02-19 06:34 -!- bechamel(n=user@ has joined #tryton
2009-02-19 07:19 -!- sharkcz( has joined #tryton
2009-02-19 07:40 -!- Timitos(n=Timitos@ has joined #tryton
2009-02-19 08:06 -!- Gedd(n=ged@ has joined #tryton
2009-02-19 08:11 -!- carlos(n=carlos@ has joined #tryton
2009-02-19 08:30 -!- paola( has joined #tryton
2009-02-19 08:40 -!- bechamel(n=user@ has joined #tryton
2009-02-19 09:23 -!- bechamel(n=user@ has joined #tryton
2009-02-19 09:35 -!- simahawk( has joined #tryton
2009-02-19 09:35 -!- simahawk( has left #tryton
2009-02-19 10:00 -!- cedk(n=ced@gentoo/developer/cedk) has joined #tryton
2009-02-19 10:00 -!- cedric_b_( has joined #tryton
2009-02-19 10:19 -!- tekknokrat( has joined #tryton
2009-02-19 10:32 -!- cedric_b__( has joined #tryton
2009-02-19 11:03 -!- cedric_b_( has joined #tryton
2009-02-19 11:05 <CIA-10> tryton: Mathias Behrle <> default * 273:7bc5e5bb2bbb website/locale/de_DE/LC_MESSAGES/tryton.po: Update translation for de_DE
2009-02-19 11:18 -!- cedric_b( has joined #tryton
2009-02-19 11:20 <yangoon> bechamel: did you do the zazzle shop?
2009-02-19 11:21 <bechamel> yangoon: yes
2009-02-19 11:21 <yangoon> gr8
2009-02-19 11:21 <yangoon> but for the logo size on the shirt I am prefering
2009-02-19 11:22 <yangoon> bechamel: can you change it?
2009-02-19 11:23 <bechamel> yangoon: zazzle allow to tweak the layout iirc
2009-02-19 11:25 <yangoon> bechamel: yes, correct
2009-02-19 11:32 <bechamel> yangoon: but you are right, the shirt is nicer with a bigger logo
2009-02-19 11:33 <yangoon> bechamel: it seems to be possible to save different templates
2009-02-19 11:35 <bechamel> yangoon: yes, i choose zazzle over cafepress because with cafepress you have to pay to provide several designs for the same item (dark shirt in this case)
2009-02-19 11:55 <udono> bechamel: how long did it take to order a t-shirt?
2009-02-19 11:56 <bechamel> udono: i don't know, i didn't try yet
2009-02-19 11:57 <bechamel> udono: btw there is also a
2009-02-19 12:15 <CIA-10> tryton: * r437 /wiki/ Edited wiki page through web user interface.
2009-02-19 12:15 <CIA-10> tryton: * r438 /wiki/ Edited wiki page through web user interface.
2009-02-19 13:03 <CIA-10> tryton: C?dric Krier <> default * 1349:878535a1d4d0 trytond/trytond/osv/ Fix add_null in SQL query for issue795
2009-02-19 13:17 -!- paola_( has joined #tryton
2009-02-19 13:59 <CIA-10> tryton: C?dric Krier <> default * 1559:d867cda68b93 trytond/trytond/ir/ Use os separator in filename when calling file_open
2009-02-19 13:59 <CIA-10> tryton: C?dric Krier <> default * 1560:62295fcc7e35 trytond/trytond/report/ Clause temporary file before remove it
2009-02-19 14:12 -!- ikks(i=igor@ has joined #tryton
2009-02-19 14:42 -!- bechamel(n=user@ has joined #tryton
2009-02-19 14:47 <CIA-10> tryton: C?dric Krier <> default * 1350:b554778ad371 trytond/trytond/ir/ Use os separator in filename when calling file_open
2009-02-19 14:47 <CIA-10> tryton: C?dric Krier <> default * 1351:ae365a338610 trytond/trytond/report/ Clause temporary file before remove it
2009-02-19 15:10 -!- saxa( has joined #tryton
2009-02-19 16:00 -!- enlightx( has joined #tryton
2009-02-19 16:13 <udono> bechamel: the t-shirts seems not to work with white color on black ground:
2009-02-19 16:14 <udono> bechamel: "Please note that for designing on all styles aside from Zazzle Basic Dark T-shirts, white areas will not be printed regardless of image format."
2009-02-19 16:23 <bechamel> udono: the png containing the logo is a png with transparency
2009-02-19 16:23 <bechamel> udono: white color is considered as transparent for non-dark shirts
2009-02-19 17:59 -!- juanfer(n=juanfer@ has joined #tryton
2009-02-19 18:08 <Timitos> cedk: in account_invoice there are some lines in on_change_product() to retrieve the unit_price. wouldn´t it be better to have one function on product.template or product.product an pass all the necessary parameters to it link (quantity, product, uom, company.currency, _invoice.currency and price type (purchase, sale,...)? same for sale and purchase?
2009-02-19 18:09 <Timitos> this way it would be easier to adapt price calculation with custom modules
2009-02-19 18:27 <cedk> Timitos: I don't know because it will be any way module specific
2009-02-19 18:29 <cedk> Timitos: for the invoice, this unit_price is realy an approximative value
2009-02-19 18:30 <cedk> Timitos: and invoice must not be used to make sale or purchase
2009-02-19 18:31 <Timitos> cedk: i do not understand. why is unit_price for invoice approximative?
2009-02-19 18:31 <cedk> Timitos: because it doesn't care about any kind of price list
2009-02-19 18:32 <cedk> Timitos: it takes cost_price or list_price
2009-02-19 18:32 <Timitos> cedk: for me price computation in sale, purchase and invoice is quite similar. so if i create one function get_price on product.product or product.template i could use this function for all three modules.
2009-02-19 18:32 <Timitos> cedk: sale does the same. it takes list_price for the moment
2009-02-19 18:33 <cedk> Timitos: yes perhaps
2009-02-19 18:34 <Timitos> cedk: and i also would develop pricelist for invoice. because it is not good if there cannot be used pricelist their. like this it is in openerp and i don´t like that
2009-02-19 18:34 <Timitos> cedk: i do not see any reason why pricelist should not be used on invoice
2009-02-19 18:35 <cedk> Timitos: so ok, you could work on it and provide a patch
2009-02-19 18:35 <Timitos> cedk: ok. i will provide pathes for account_invoice, sale and purchase
2009-02-19 18:36 <cedk> Timitos: fill an issue in roundup to share the work
2009-02-19 18:36 <Timitos> cedk: ok
2009-02-19 18:42 -!- paola( has joined #tryton
2009-02-19 18:53 -!- X0d_of_N0d( has joined #tryton
2009-02-19 18:54 <X0d_of_N0d> ACTION waves @ udono
2009-02-19 19:04 <udono> X0d_of_N0d: hey man
2009-02-19 19:05 <udono> X0d_of_N0d: nice to read you!
2009-02-19 19:06 <X0d_of_N0d> udono: back @ ya man
2009-02-19 19:07 <X0d_of_N0d> man, have you ever used a Kinesis kbd? It's like I'm relearning to type....
2009-02-19 19:07 <X0d_of_N0d> so what were you wondering?
2009-02-19 19:08 <udono> X0d_of_N0d: what is a kinesis kbd? never heared/used
2009-02-19 19:09 <udono> X0d_of_N0d: Iam on planning how to realise a clean ldap auth
2009-02-19 19:10 <X0d_of_N0d> < a friend got one and didn't like it so I'm trying it
2009-02-19 19:11 <X0d_of_N0d> udono: ldap auth should be done the same way all other ldap stuff...
2009-02-19 19:11 <udono> X0d_of_N0d: huh, so you are writing on a broken kbd :-)
2009-02-19 19:12 <X0d_of_N0d> not broken... just really strange.
2009-02-19 19:12 <X0d_of_N0d> heh
2009-02-19 19:12 <udono> X0d_of_N0d: yes, is there a standard way for ldap auth?
2009-02-19 19:12 <udono> X0d_of_N0d: I think of:
2009-02-19 19:12 <udono> 1. anon bind
2009-02-19 19:12 <udono> 2. check if user exists
2009-02-19 19:13 <udono> 2.1 If user not exist: auth fail
2009-02-19 19:13 <X0d_of_N0d> udono: no...not really
2009-02-19 19:13 <udono> 2.2 is user(name) exist: auth bind with password
2009-02-19 19:14 <udono> 3. if auth bind nok: auth fails
2009-02-19 19:14 <X0d_of_N0d> the standard way to bind is to bind as user... so simple_bind_s(user,passwd)
2009-02-19 19:14 <X0d_of_N0d> if
2009-02-19 19:14 <X0d_of_N0d> if that fails then auth failes
2009-02-19 19:14 <udono> 4. if auth bind ok: auth ok
2009-02-19 19:15 <X0d_of_N0d> you never need to check for a user because auth bind does that for you
2009-02-19 19:15 -!- kleinerdrache( has joined #tryton
2009-02-19 19:15 <udono> X0d_of_N0d: ok, understand
2009-02-19 19:15 <udono> X0d_of_N0d: how to combine with ldap_base?
2009-02-19 19:16 <udono> X0d_of_N0d: there we have resources/server
2009-02-19 19:18 <udono> X0d_of_N0d: if we make directly simple_bind_s(user,passwd) we need there in server definition an auth bind, with password too?
2009-02-19 19:18 <X0d_of_N0d> no you'd bind with the passwd supplied by the user... if the ldap setting for user bind is used that is
2009-02-19 19:19 <udono> X0d_of_N0d: so ldap_auth doesn't need ldap_base?
2009-02-19 19:19 <udono> X0d_of_N0d: srry, wrong
2009-02-19 19:20 <udono> X0d_of_N0d: don't understand: "if the ldap setting for user bind is used that is"
2009-02-19 19:21 <X0d_of_N0d> the user is filter+": "+username+ldap_base
2009-02-19 19:22 <X0d_of_N0d> erm... s/:/=/
2009-02-19 19:23 <X0d_of_N0d> e.g.: uid=joe,dc=company
2009-02-19 19:25 <X0d_of_N0d> so iirc the call would be l.simple_bind_s('uid=joe,dc=company','passwd')
2009-02-19 19:26 <udono> X0d_of_N0d: ah, ok, understand
2009-02-19 19:27 <X0d_of_N0d> but this is all only if the user's server is newer than 5 years or so, or configured correctly...
2009-02-19 19:27 <udono> X0d_of_N0d: so I put on res_user a link to a ldap_resource?
2009-02-19 19:28 <udono> X0d_of_N0d: and vice versa
2009-02-19 19:28 <X0d_of_N0d> hold on, lemme look at the code
2009-02-19 19:28 -!- vengfulsquirrel( has joined #tryton
2009-02-19 19:38 <X0d_of_N0d> we need to override the orm for res_user so that queries for users search ldap
2009-02-19 19:38 <X0d_of_N0d> this isn't needed for auth, but for looking at or adding users it would be
2009-02-19 19:39 <udono> X0d_of_N0d: yes
2009-02-19 19:40 <X0d_of_N0d> which was the next thing I was going to work on before I got pulled away to do a bunch of other stuff here at work
2009-02-19 19:40 <X0d_of_N0d> ACTION grumbles
2009-02-19 19:41 <udono> X0d_of_N0d: ...but just for ldap_auth: are res_user m2o ldap_base combined?
2009-02-19 19:41 <X0d_of_N0d> basically we need to implment that in the base ldap module
2009-02-19 19:42 <udono> X0d_of_N0d: the res_user m2o ldap_base ?
2009-02-19 19:42 <udono> X0d_of_N0d: it can be inherited from ldap_auth. So you can use ldap_base without ldap_auth
2009-02-19 19:43 <X0d_of_N0d> hold up for a sec
2009-02-19 19:43 <udono> ACTION thinks that somone only likes to use ldap for an addressbook, but not for auth.
2009-02-19 19:43 <X0d_of_N0d> res_user would be seperate from ldap base, yes
2009-02-19 19:44 <X0d_of_N0d> ldap_auth would override res_user
2009-02-19 19:47 <udono> X0d_of_N0d: yes
2009-02-19 19:48 <X0d_of_N0d> man, this kb is really making it hard to move around, sorry it's taking so long
2009-02-19 19:48 <udono> X0d_of_N0d: but you have the double size ESC key?
2009-02-19 19:49 <udono> X0d_of_N0d: and keys for copy, cut and paste :-)
2009-02-19 19:49 <X0d_of_N0d> I almost never use esc, I use emacs
2009-02-19 19:50 <X0d_of_N0d> and now ctrl is under my thumb instead of my pinkey
2009-02-19 19:51 <X0d_of_N0d> this is kind of interesting
2009-02-19 19:52 <X0d_of_N0d> User already has it's own methods for create write and read
2009-02-19 19:52 <udono> X0d_of_N0d: one last question. Which should be the behavoir when ldap auth fails. Should there be a fallback to tryton standard authentication mechanism?
2009-02-19 19:52 <udono> X0d_of_N0d: or should this be customizable via a checkbox?
2009-02-19 19:52 <cedk> udono: by the way, you should ask to intuxication guys to put your ldap modules under tryton category
2009-02-19 19:53 <udono> cedk: yes, I will do so. thanks
2009-02-19 19:53 <cedk> udono: I was just searching the modules and it was a little difficult :-)
2009-02-19 19:54 <udono> cedk: I will just wait if they are tested and debugged a bit more.
2009-02-19 19:54 <udono> cedk: I give you the links:
2009-02-19 19:54 <X0d_of_N0d> udono: using nsswitch as a model we should fail back to tryton db if ldap fails, but only if ldap fails
2009-02-19 19:55 <udono> cedk: is not working for now
2009-02-19 19:56 <udono> cedk: is ok, but there is still the translation issue open for the error messages.
2009-02-19 19:56 <udono> X0d_of_N0d: ldap fails means no connection to the ldap server?
2009-02-19 19:56 <udono> X0d_of_N0d: and not authentication on ldap fails
2009-02-19 19:58 <X0d_of_N0d> no, it means that auth fails against ldap
2009-02-19 19:58 <udono> X0d_of_N0d: ok
2009-02-19 20:01 <X0d_of_N0d> udono: hey man, btw, sorry I'm not more involved in this
2009-02-19 20:05 <udono> X0d_of_N0d: no problem, I think I can handle the ldap_auth stuff. The orm part I don't know enough for now. Maybe later you find some time to specify your Ideas for this in the wiki.
2009-02-19 20:06 <X0d_of_N0d> udono: i can give you veague idea...
2009-02-19 20:07 <X0d_of_N0d> I'll try to add stuff to the wiki though
2009-02-19 20:08 <udono> X0d_of_N0d: Yeah, an idea even a veague is very welcome.
2009-02-19 20:08 <udono> X0d_of_N0d: Thanks a lot!
2009-02-19 20:08 -!- cristi_an(i=5978d3ce@gateway/web/ajax/ has joined #tryton
2009-02-19 20:09 <X0d_of_N0d> udono: take a look at res/
2009-02-19 20:09 <cristi_an> when i create a custom module all code is written on server side in the module...nothing on client side ?
2009-02-19 20:10 <X0d_of_N0d> cristi_an: afaik
2009-02-19 20:10 <X0d_of_N0d> cristi_an: yes
2009-02-19 20:10 <cristi_an> just wnated to be sure,about this
2009-02-19 20:11 <cristi_an> cause for simple moudles i have done...ithis was the case
2009-02-19 20:11 <cristi_an> and client acted like a browser
2009-02-19 20:11 <cristi_an> wonder if i want to change how a control react on a key ..
2009-02-19 20:12 <cristi_an> fopress
2009-02-19 20:13 <X0d_of_N0d> cristi_an: i belive user configuration is stored in the db, if you're talking about adding a hot key to a form or something I don't know
2009-02-19 20:15 <X0d_of_N0d> udono: so the User class inherits OSV then modifies the create, write, delete, etc.. methods of OSV
2009-02-19 20:15 -!- vengfulsquirrel( has joined #tryton
2009-02-19 20:16 <X0d_of_N0d> udono: likewise an ldap_OSV would override read, write, create, etc in order to query ldap instead of postgres
2009-02-19 20:17 <X0d_of_N0d> then ldap_user would inherit ldap_OSV, and everything else should be (more or less) transparent
2009-02-19 20:18 <X0d_of_N0d> vengfulsquirrel: how are things going in your neck of the woods man? sorry I've been so out of the loop, my boss pulled me off tryton stuff to work on our ecr system
2009-02-19 20:20 <cedk> X0d_of_N0d: but be carefull about the transaction
2009-02-19 20:20 <X0d_of_N0d> ACTION looks at cedk
2009-02-19 20:21 <cedk> if you update stuff outside the cursor transaction, in case of exception the change will not be rollbacked
2009-02-19 20:22 <cedk> X0d_of_N0d: for this I suggest to keep the write, create, etc... into the database and add also write, create, ... on the ldap server
2009-02-19 20:22 <vengfulsquirrel> X0d_of_N0d: Yeah I have been kind of distracted with extending Tryton to work with my company minus production. I barely started the base production module and it still needs a lot of work. I'm going to start working on it some more today though.
2009-02-19 20:24 <X0d_of_N0d> cedk: I'd rather handle this all in ldap, replicated data can lead to difficult to diagnose bugs and security problems
2009-02-19 20:25 <X0d_of_N0d> vengfulsquirrel: cool, it looks like I'll be stuck on this stuff for at least the next week or so.
2009-02-19 20:26 <X0d_of_N0d> cedk: if ldap can't commit a transaction it raises an error, as long as we don't do bulk transactions or try to update more than one DN at a time then ldap should take care of the rollback issue
2009-02-19 20:26 <cedk> X0d_of_N0d: yes, use only the data from ldap but I suggest to still perform the operation on database to care of the transaction and the constraint
2009-02-19 20:27 <cedk> X0d_of_N0d: any way, you need an entry in res_user table for each user because there is a foreign key on every table that point to res_user
2009-02-19 20:27 <cedk> X0d_of_N0d: write_uid and create_uid
2009-02-19 20:28 <X0d_of_N0d> cedk: hum....
2009-02-19 20:29 <cedk> X0d_of_N0d: it is not an issue if you override each function and put the data from ldap
2009-02-19 20:31 <X0d_of_N0d> are there any constraints on other tables that checks if the wirte and create uids exist in the db?
2009-02-19 20:31 <cedk> X0d_of_N0d: perhaps, any module could add it
2009-02-19 20:32 <cedk> X0d_of_N0d: bbl
2009-02-19 20:33 <X0d_of_N0d> cedk: if it's not there it's better, thanks
2009-02-19 20:45 <udono> X0d_of_N0d: cedk another way could be something like handshake communication: catch the read, write, change and direct it to ldap. After this read the results from ldap, check them if they are equal with the sended data and after this read, write, change the tryton data.
2009-02-19 20:46 <X0d_of_N0d> ldap does all this
2009-02-19 20:46 <X0d_of_N0d> ldap will tell you if the write failed and why
2009-02-19 20:46 <X0d_of_N0d> you just need to check the return value for the write
2009-02-19 20:50 <X0d_of_N0d> udono: so if you tried to add a user and all the fields were right, but the telephone number was invalid then nothing would be committed until all information was correct
2009-02-19 20:55 <cedk> X0d_of_N0d: this is not enough for the transaction
2009-02-19 20:56 <cedk> I think the best it to do the same on both side database and ldap
2009-02-19 20:56 <cedk> the transaction could fail later
2009-02-19 20:57 <cedk> but of course the ldap data will not be rollbacked
2009-02-19 20:57 -!- tekknokrat( has left #tryton
2009-02-19 20:58 <cedk> there is no good solution since we duplicate the transaction system
2009-02-19 20:58 <cedk> so you need one master, and I think it must be the data in database
2009-02-19 20:58 <cedk> as it is part of the all
2009-02-19 21:05 <X0d_of_N0d> how could the transaction fail later?
2009-02-19 21:14 <X0d_of_N0d> I don't quite understand
2009-02-19 21:35 -!- enlightx( has joined #tryton
2009-02-19 22:04 <udono> X0d_of_N0d: good night, see you.
2009-02-19 22:11 <X0d_of_N0d> udono: later man
2009-02-19 22:31 -!- ikks_(n=igor@ has joined #tryton
2009-02-19 23:11 -!- tekknokrat( has joined #tryton
2009-02-19 23:19 -!- ikks_(n=igor@ has joined #tryton
2009-02-19 23:29 -!- vengfulsquirrel( has joined #tryton

Generated by 2.17.3 by Marius Gedminas - find it at!