IRC logs of #tryton for Monday, 2014-12-01

chat.freenode.net #tryton log beginning Mon Dec 1 00:00:02 CET 2014
-!- digitalsatori(~Thunderbi@114.92.210.36) has joined #tryton00:11
-!- udono(~udono@ip-37-201-245-76.hsi13.unitymediagroup.de) has joined #tryton00:25
-!- digitalsatori(~Thunderbi@114.92.210.36) has joined #tryton00:39
-!- smarro_(~sebastian@190.105.93.196) has joined #tryton00:49
-!- Ayub_(60e71da5@gateway/web/freenode/ip.96.231.29.165) has joined #tryton01:24
Ayub_Pilou: I did not get a chance to thank you for your help the otherday01:25
Ayub_thanks01:25
Piloudid it work ?01:33
svidelahi there, anyone available with knowledge of the client implementation?01:44
-!- udono(~udono@ip-37-201-245-76.hsi13.unitymediagroup.de) has joined #tryton02:36
-!- digitalsatori(~Thunderbi@180.158.135.39) has joined #tryton02:59
-!- smarro_(~sebastian@190.105.93.196) has joined #tryton04:26
-!- digitalsatori(~Thunderbi@180.158.135.39) has joined #tryton04:45
-!- svidela(~sancho@189.221.150.129.cable.dyn.cableonline.com.mx) has joined #tryton05:01
-!- digitalsatori(~Thunderbi@180.158.135.39) has joined #tryton05:08
-!- svidela(~sancho@189.221.150.129.cable.dyn.cableonline.com.mx) has left #tryton05:38
-!- yangoon(~mathiasb@p549F3171.dip0.t-ipconnect.de) has joined #tryton06:01
-!- svidela(~sancho@189.221.150.129.cable.dyn.cableonline.com.mx) has joined #tryton06:08
-!- svidela(~sancho@189.221.150.129.cable.dyn.cableonline.com.mx) has joined #tryton06:12
-!- svidela(~sancho@189.221.150.129.cable.dyn.cableonline.com.mx) has joined #tryton06:16
-!- svidela(~sancho@189.221.150.129.cable.dyn.cableonline.com.mx) has left #tryton06:16
-!- svidela(~sancho@189.221.150.129.cable.dyn.cableonline.com.mx) has joined #tryton06:16
-!- horli(~gregor@chello084113024052.5.12.vie.surfer.at) has joined #tryton06:19
-!- frispete_(~frispete@p54A93A22.dip0.t-ipconnect.de) has joined #tryton06:40
-!- marius__(~marius@v100.nfq.lt) has joined #tryton07:17
-!- pobsteta(~Thunderbi@4cb54-3-88-160-87-54.fbx.proxad.net) has joined #tryton07:47
-!- Timitos(~kpreisler@host-88-217-184-172.customer.m-online.net) has joined #tryton08:21
-!- digitalsatori(~Thunderbi@180.158.135.39) has joined #tryton08:27
-!- marius__(~marius@v100.nfq.lt) has joined #tryton08:33
-!- horli(~gregor@chello084113024052.5.12.vie.surfer.at) has joined #tryton08:43
-!- ytz(~ytz@opfw01.options-it.com) has joined #tryton09:08
-!- LordVan(~LordVan@gentoo/developer/LordVan) has joined #tryton09:21
-!- nicoe(~nicoe@2a02:a03f:301b:7c00:ee55:f9ff:fe7b:f7ac) has joined #tryton09:26
-!- vcardon(~vcardon@bureau-sdsl.tranquil.it) has joined #tryton09:30
-!- nicoe(~nicoe@2a02:a03f:301b:7c00:ee55:f9ff:fe7b:f7ac) has joined #tryton09:34
-!- bechamel(~Adium@62-197-96-189.teledisnet.be) has joined #tryton09:37
marius__where is multi_server in current trytond?10:28
marius__is it deprecated?10:28
-!- digitalsatori(~Thunderbi@180.158.135.39) has joined #tryton10:29
-!- jeancavallo(~giovanni@LPuteaux-656-1-230-52.w80-12.abo.wanadoo.fr) has joined #tryton10:35
marius__https://bugs.tryton.org/issue4373?@ok_message=msg%2019036%20created%0Aissue%204373%20messages%20edited%20ok&@template=item10:55
marius__any ideas anyone10:55
pokolimarius__:  have you tryied with the SELECT .. FOR UPDATE?11:02
marius__pokoli, but products_by_location happens after the first one finishes11:03
marius__in both situations11:03
marius__[Mon Dec 01 09:47:27 2014] INFO:assign_try:139656558614272 -- STOCK STOP ASSIGN11:04
marius__ is at the end of assign_try method11:04
pokolimarius__: https://bugs.tryton.org/msg1899711:05
marius__I don't understand this message11:05
marius__can you explain?11:05
marius__repeatble  vs seriazable11:06
pokolimarius__: http://www.postgresql.org/docs/current/static/transaction-iso.html11:09
marius__I only see SET TRANSACTION READ ONLY11:11
marius__and LOCK IN EXCLUSIVE MODE11:12
marius__nothing about SERIALIZABLE / REPEATABLE in the log11:13
marius__pokoli, where do you suggest putting FOR UPDATE? into products_by_location query?11:14
pokolimarius__: about transactions http://hg.tryton.org/trytond/file/b8e46cc95a39/trytond/backend/postgresql/database.py#l7711:16
pokolimarius__: which was introduced on http://hg.tryton.org/trytond/changeset/0a4097cbe66211:17
-!- horli(~gregor@chello084113024052.5.12.vie.surfer.at) has joined #tryton11:18
pokolimarius__: for update you have to change two selects11:19
pokolimarius__: http://hg.tryton.org/modules/stock/file/b9947becd5e5/move.py#l104111:19
pokolimarius__: and http://hg.tryton.org/modules/stock/file/b9947becd5e5/move.py#l105111:20
marius__what about just http://hg.tryton.org/modules/stock/file/b9947becd5e5/move.py#l107911:21
marius__or it won't work for union?11:21
marius__FOR UPDATE is not allowed in GROUP BY11:22
pokolimarius__: on last line will affect also the period table, which is not updated11:28
-!- corro(~robin@77-58-75-66.dclient.hispeed.ch) has joined #tryton11:28
pokolimarius__: I imagin you can remove the group_by of the partial queries (on which you have the FOR UPDATE) and it should work  as they are grouped at the end from same keys11:33
marius__pokoli, what about doing something like11:34
marius__commit, lock, start transaction ... commit11:34
marius__instead of start transaction, lock .. commit11:35
marius__because as I understand it, it uses snapshot from transaction start?11:35
pokolimarius__: you will commit any data written previously of assign_try, which is not correct11:36
pokolimarius__: yes, it's reading the data from the transaction start11:37
pokolimarius__: AFAIU using serialized transaction will solve the issue11:37
marius__pokoli, what about starting it just after lock11:42
pokolimarius__: with which objective?11:44
marius__to "take" a new snapshot for transaction11:44
pokolimarius__: you will read the data at the begin of the transaction becaouse of repetable read11:52
marius__but removing group by clause seems like overkill11:52
-!- mariomop(~quassel@host30.186-109-104.telecom.net.ar) has joined #tryton12:01
-!- jeancavallo(~giovanni@LPuteaux-656-1-230-52.w80-12.abo.wanadoo.fr) has joined #tryton12:02
pokolimarius__: why overkill? they will be grouped at the end, so no need to regroup them12:06
marius__ok i'll check it if it solves the problem12:09
-!- kstenger(~karla@200.124.209.158) has joined #tryton12:11
marius__pokoli, it's not so easy because of SUM, etc fields12:15
marius__doesn't work without group by ;)12:15
marius__+ FOR UPDATE doesn't work together inside UNION12:16
marius__i mean I get SQL syntax error12:16
marius__NotSupportedError: FOR UPDATE is not allowed with UNION/INTERSECT/EXCEPT12:17
pokolimarius__: yes, I understand you get the SQL error12:17
pokolimarius__: can you update the issue with what you have done and the errors?12:17
marius__pokoli, any ideas what's next?12:21
pokolimarius__: I imagine that the only solution is to use SERIALIZABLE TRANSACTION level12:23
marius__how would I do that>?12:23
marius__can I change it somethow inside assign_try or something?12:24
pokolimarius__: you have to change it for all the transactions :)12:24
pokolimarius__: basically its the reverse patch of http://hg.tryton.org/trytond/changeset/0a4097cbe66212:25
marius__can I do it like that but just for assign_try?12:25
marius__like Transaction().cursor.conn.set_isolation_level(..whatever..)12:26
pokolimarius__: set transcation isolation leel must be called before any query :)12:30
marius__i was unable to repeat the issue with serializable transaction level12:34
marius__pokoli, so are you suggesting reverting http://hg.tryton.org/trytond/changeset/0a4097cbe66212:36
marius__performance issues? etc?12:36
marius__sideeffects?12:36
-!- smarro(~sebastian@190.105.93.196) has joined #tryton12:38
marius__i'm still thinking about moving lock before transaction start, so that new snapshot is "taken" after getting lock12:43
pokolimarius__: so I think the issue will be the same as far as you use serializable12:47
pokolimarius__: I don't know why repeatable read was introduced, so i can not comment about it12:48
pokolimarius__: let's wait for ced's comments :)12:48
yangoon pokoli AFAIR repeatable read performed much better as strict serialization12:52
pokoliyangoon: that's what i imagine too, but I think consistency is more important that better performance12:54
-!- horli(~gregor@chello084113024052.5.12.vie.surfer.at) has joined #tryton13:14
-!- smarro(~sebastian@190.105.93.196) has joined #tryton13:27
marius__pokoli, so i'm submitting review for revert13:27
pokolimarius__: thanks for your efforts :)13:33
-!- marius__(~marius@v100.nfq.lt) has joined #tryton13:36
-!- meanmicio(~meanmicio@fsf/member/meanmicio) has joined #tryton13:39
-!- marius__(~marius@v100.nfq.lt) has joined #tryton14:20
marius__pokoli, do I have to bother about review bot http://codereview.tryton.org/9931002 ?14:22
udonomarius__: hi, do not care the review bot in this case.14:24
-!- cedk(~ced@gentoo/developer/cedk) has joined #tryton14:25
-!- smarro(~sebastian@190.105.93.196) has joined #tryton14:46
-!- marius__(~marius@v100.nfq.lt) has joined #tryton14:47
-!- cedk(~ced@gentoo/developer/cedk) has joined #tryton15:04
-!- digitalsatori(~Thunderbi@114.92.210.36) has joined #tryton15:19
-!- smarro(~sebastian@190.105.93.196) has joined #tryton16:08
-!- marius__(~marius@112.185.140.82.static.lrtc.lt) has joined #tryton16:51
-!- TheCowboy(~TheCowboy@wsip-98-191-208-111.dc.dc.cox.net) has joined #tryton16:59
-!- newzen(~newzen@201.242.159.40) has joined #tryton17:10
buxyHow I can I change the default date format to be YYYY-MM-DD for all customers marked as English speaking?17:36
buxyI didn't find any translatable string representing the default date format.17:36
buxy(I care about how the date are printed on reports, not about how they are displayed in the client FWIW)17:37
buxyBah I found it.17:38
buxyI just add to go in the detailed view of the language.17:38
-!- hiaselhans(~Thunderbi@chello212186043057.408.14.vie.surfer.at) has joined #tryton17:55
-!- pobsteta(~Thunderbi@4cb54-3-88-160-87-54.fbx.proxad.net) has joined #tryton18:42
-!- ytz(~ytz@opfw01.options-it.com) has joined #tryton18:48
-!- sunny_dealmeida(~quassel@210.89.41.97) has joined #tryton19:02
-!- Telesight(~anthony@4daedff9.ftth.telfortglasvezel.nl) has joined #tryton19:06
-!- udono(~udono@ip-37-201-245-76.hsi13.unitymediagroup.de) has joined #tryton19:13
kstengerdo you know of any example of a tab beeing visible/invisible according to some conditions in the same form view?19:21
bechamelkstenger: iirc, there are some extra tabs on the invoice depending of the invoice type19:27
kstengerI just found one example :) thanks bechamel19:27
kstengertryton keeps impressing me with it's simplicity19:28
bechamelkstenger: great :)19:28
-!- cedk(~ced@gentoo/developer/cedk) has joined #tryton20:00
-!- vcardon(~vcardon@bureau-sdsl.tranquil.it) has left #tryton20:17
-!- xrg_(~panos@130.43.55.131.dsl.dyn.forthnet.gr) has joined #tryton20:21
-!- hiaselhans(~Thunderbi@chello212186043057.408.14.vie.surfer.at) has joined #tryton20:39
-!- TheCowboy(~TheCowboy@ip68-98-183-236.dc.dc.cox.net) has joined #tryton21:31
-!- pilou`(~pilou@181.204.19.93.rev.sfr.net) has joined #tryton22:27
-!- sharkcz`(~sharkcz@2a01:8c00:ffb3:160:250:43ff:fe3c:3b5d) has joined #tryton22:28
-!- yangoon1(~mathiasb@p549F3171.dip0.t-ipconnect.de) has joined #tryton22:28
-!- Pilou(~pilou@pdpc/supporter/active/pilou) has joined #tryton22:59

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!