IRC logs of #tryton for Tuesday, 2009-09-22 #tryton log beginning Tue Sep 22 00:00:02 CEST 2009
CIA-2vengfulsquirrel roundup * #1186/KeyError when completing inventory with deactivated products: [new] If a product was ever associated with a move and then you deactivate and then you try to complete an inventory you get this error. With the ...01:50
-!- woakas(n=woakas@ has joined #tryton02:32
-!- ikks_(n=ikks@ has joined #tryton03:40
-!- yangoon( has joined #tryton05:19
CIA-2vengfulsquirrel roundup * #1187/Missing sequence_rename for shipment migration.: [new] I'm not sure if it was intentional but the sequences are not renamed when the packing->shipment tables are. I made a patch but since I'm no ...07:08
-!- MarkusB( has joined #tryton07:54
-!- cedk(n=ced@gentoo/developer/cedk) has joined #tryton08:00
CIA-2ced roundup * #1187/Missing sequence_rename for shipment migration.: [invalid] The sequence are renamed with table_rename08:05
-!- sharkcz( has joined #tryton08:09
-!- carlos( has joined #tryton08:27
-!- Timitos(n=timitos@ has joined #tryton08:29
vengfulsquirrelcedk: Hmm something weird seems to be happening, do I have to do anything for sequence renaming to take effect?08:31
cedkvengfulsquirrel: it is done in the function rename_table08:33
vengfulsquirrelYeah I see now that the functions are identical, for some reason only the table is being renamed and the sequence is just recreated from scratch.08:35
vengfulsquirrelI can't get the demo to reproduce the problem though so I guess maybe its something I've done that prevents it from happening.08:35
-!- paepke( has joined #tryton08:36
-!- dukai(n=quassel@ has joined #tryton08:38
cedkvengfulsquirrel: maybe if the sequence is not found then it is not renamed08:42
-!- essich( has joined #tryton09:10
-!- bechamel( has joined #tryton09:22
vengfulsquirrelcedk: Is the migration code only run via init=module_name ?09:33
cedkvengfulsquirrel: also on update09:35
cedkbechamel: the CAST doesn't seem to work with sqlite09:46
cedkit returns a float09:46
bechamelcedk: and it should be a decimal if I understand correctly09:49
cedkbechamel: yes but sql doesn't run the adapter for it09:49
cedkit seems that it is the sum that change the type09:53
bechamelcedk: vengfulsquirrel: a fix for
cedkbechamel: put the codereview in roundup09:56
CIA-2bch roundup * #1186/KeyError when completing inventory with deactivated products:09:58
CIA-2[chatting] Fix available here:09:58
cedkI read some doc and SQLite use float to make sum and CAST doesn't seem to make it call the adapter10:09
cedkI can add a test after the query to convert the float into Decimal10:09
cedkbut we will have float operation where we need Decimal10:10
cedkso this means that SQLite use never be use for critical accounting operation10:14
paepkecedk: thats bad for SQLite... Cause the current common use case for tryton is ERP...10:15
bechamelcedk: if someone is ready to do all the accounting by himself the odds are high that there are several users (and btw is quicker to install postgresql than to do all the account setup)10:18
-!- yangoon1( has joined #tryton10:19
vengfulsquirrelbechamel: Thanks but I won't be able to try until tomorrow.10:22
cedkpaepke, bechamel: until now we always say that SQLite is beta for demo and testing10:23
paepkecedk: i understand. for me postgres setup is the better solution. in any way.10:24
-!- tekknokrat( has joined #tryton10:24
paepkecedk: regarding to neso its an issue?10:25
paepkeis there something similar to mysql-embedded from the postgres guys?10:25
CIA-2ced roundup * #1184/TypeError: unsupported operand type(s) for /: 'float' and 'Decimal':10:26
CIA-2[testing] Patches to convert into Decimal after a sum for SQLite:10:26
cedkpaepke: I don't know10:28
bechamelcedk: "SQLite use never be use for critical accounting": otoh I know other ERP that does all the accounting in float :)10:38
-!- paepke( has joined #tryton10:38
-!- tekknokrat( has left #tryton11:00
-!- gour(n=Gour@ has joined #tryton11:11
gourhiya aa_11:20
gouraa_: what are you doing here? ;)11:20
yangoon1bechamel: cedk I am really not pleased to have code in modules for SQLite issues. Even if the current proposals for patches are surely more accurate than to do SQLite accounting in float, I would prefer the latter. We ever said that SQlite is for demo and testing purposes, because it cannot handle some requirements, that other more capable DBMS are able to do.11:23
cedkyangoon1: what is your solution?11:25
yangoon1I think we should clarify explicetely whereever there is a place to do it, that the use of SQLite doesn't provide accurate results in some situations.11:25
cedkyangoon1: simple sum of numerics return float11:26
yangoon1cedk: the test should be done in SQLite backend11:27
yangoon1cedk: wouldn't it be possible to look for calls for SUM and convert adequately?11:28
cedkyangoon1: no except if you write an SQL parser11:30
yangoon1cedk: If I see correctly there alrady are replacement actions in the sql in execute of Cursor11:34
cedkyangoon1: a replacement will not fix11:35
yangoon1cedk: yes, but it is 'kind' of sql parser11:35
cedkyangoon1: it is far from a parser11:39
udonoyangoon1: are you talking about
yangoon1udono: yes11:41
yangoon1udono: and
yangoon1for me the question is: couldn't we parse in line 257 of for SUM and convert the result by default to decimal?11:46
udonoyangoon1: but what if the sum needs to be integer?11:48
udonoyangoon1: and a working parser for SQL would be an ambitious project.11:50
yangoon1anyone knows if sqlalchemy handles it?11:51
cedkyangoon1: I check it and it doesn't seem to handle it11:52
udonocedk: yangoon1: I found 0.5.0rc1: Added func.min(), func.max(), func.sum() as "generic functions",       which basically allows for their return type to be determined       automatically.  Helps with dates on SQLite, decimal types,        others. [ticket:1160]11:55
yangoon1udono: just looking into the same...11:55
cedkudono: url ?11:56
-!- bechamel( has joined #tryton12:00
cedkudono: it doesn't solve the issue12:02
cedkudono, yangoon1 : one solution would have to fix sqlite to handle properly CAST12:10
yangoon1cedk: ack, it would be THE solution12:11
-!- ikks(n=ikks@ has joined #tryton13:03
yangoon1cedk: udono
CIA-2matb roundup * #1184/TypeError: unsupported operand type(s) for /: 'float' and 'Decimal': JFTR: Issue posted at
-!- _TiN_(i=ahenze@ has joined #tryton14:33
-!- tekknokrat( has joined #tryton15:18
-!- tekknokrat( has left #tryton15:18
-!- woakas(n=woakas@ has joined #tryton15:23
CIA-2C?dric Krier <> default * 29:46cc67c3c4c0 calendar/ Add _log_exception16:03
CIA-2C?dric Krier <> default * 30:744e7ba86dbe calendar/ Use BrowseRecord in _fbtype for easier extend16:03
CIA-2C?dric Krier <> default * 31:179de3a2af73 calendar/de_DE.csv: merge16:03
CIA-2C?dric Krier <> default * 13:c329e54651bc calendar_todo/ Remove print statement16:03
CIA-2C?dric Krier <> default * 14:b1d0133c9c42 calendar_todo/ ( de_DE.csv): merge16:03
CIA-2C?dric Krier <> default * 370:c6f5c78e64e9 account_invoice/ Fix SUM with SQLite for issue118416:09
CIA-2ced roundup * #1184/TypeError: unsupported operand type(s) for /: 'float' and 'Decimal': [resolved] Fix with changeset c6f5c78e64e916:09
CIA-2C?dric Krier <> default * 506:658f3b5c8c90 account/ ( Fix SUM with SQLite for issue118416:09
-!- jerojasro(n=jerojasr@ has joined #tryton16:26
-!- jerojasro(n=jerojasr@ has joined #tryton16:27
CIA-2ced roundup * #1185/Localized year format in timesheet reports:16:44
CIA-2[testing] Patch to review:16:44
CIA-2C?dric Krier <> default * 2020:1019bd97831c trytond/trytond/tools/ Use cast into int for range to prevent warning16:46
CIA-2C?dric Krier <> default * 2021:fa8788a27292 trytond/trytond/tools/ merge16:46
CIA-2ced roundup * #1180/KeyError: 'webdav.collection': I don't see the 401 in the dump16:51
CIA-2matb roundup * #1180/KeyError: 'webdav.collection': Second dump should contain whole sync trial.17:43
CIA-2ced roundup * #1180/KeyError: 'webdav.collection': Ok, I see it seems that it requests PROPFIND on root which is linked to any database and so it must not. So it is either a misconfiguration or an ...17:48
-!- yangoon( has joined #tryton18:50
CIA-2matb roundup * #1180/KeyError: 'webdav.collection':18:51
CIA-2Just checked again configuration on the iphone, tried with terminating slash and without for URI: http://servername:8080/test_dav/Calendars/dav1/18:51
CIA-2ced roundup * #1180/KeyError: 'webdav.collection': On 22/09/09 18:51 +0200, Mathias Behrle wrote: > > Mathias Behrle <> added the comment: > > Just checked again configuratio ...19:01
-!- paepke( has joined #tryton19:43
-!- vengfulsquirrel(n=plinux@ has joined #tryton19:50
-!- carlos( has joined #tryton20:13
vengfulsquirrelIf a product must be replaced what is the best method of operations?  Create a return shipment from that sale with onyl the returned items on it and then create a custmer shipment with the new items on it ?21:00
cedkvengfulsquirrel: why not. But it depends of the internal rules of your company21:10
vengfulsquirrelcedk: Just checking to see if it was way off, yeah I'm making the internal rules right now I guess.  Its pretty much do what you want now, but with a computerized element rules are mandatory to keep reality in sync with the computer.21:15
cedkvengfulsquirrel: yes but you may perhaps want to make more tracks by creating a new Model for return with comments, etc. that will generate the shipment and other stuffs21:16
vengfulsquirreloh right, yeah but I'm just trying to get a working minimal system right now21:23
vengfulsquirrelI have both nurseries trying to use the system now and all kinds of edge cases are cropping up.21:23
CIA-2vengfulsquirrel roundup * #1186/KeyError when completing inventory with deactivated products:22:19
CIA-2The patch prevents an exception but then results in this error:22:19
CIA-2The value of the field "Product" on "Stock Inventory Line" is not valid accordi ...22:19
CIA-2bch roundup * #1186/KeyError when completing inventory with deactivated products: Yes inactive product shouldn't appear in the inventory, but the patch is still useful if the product is set inactive while it appears in in a draf ...22:24
-!- MarkusB( has joined #tryton22:31
CIA-2vengfulsquirrel roundup * #1186/KeyError when completing inventory with deactivated products:23:10
CIA-2Yes but complete inventory button still will not work.23:10
CIA-2Or at least its not working for me unless you don't specify specific products to invento ...23:10

Generated by 2.11.0 by Marius Gedminas - find it at!