IRC logs of #tryton for Wednesday, 2014-09-10 #tryton log beginning Wed Sep 10 00:00:01 CEST 2014
-!- pokoli(~pokoli@unaffiliated/pokoli) has joined #tryton00:05
-!- gremly(~gremly@ has joined #tryton02:49
-!- alisonken1lap3( has joined #tryton03:13
-!- digitalsatori(~Thunderbi@ has joined #tryton06:02
-!- digitalsatori(~Thunderbi@ has joined #tryton06:08
-!- digitalsatori(~Thunderbi@ has joined #tryton06:24
-!- frispete_( has joined #tryton06:41
-!- yangoon1( has joined #tryton07:02
-!- pobsteta( has joined #tryton07:21
-!- nicoe( has joined #tryton07:42
-!- nicoe_(~nicoe@ has joined #tryton07:52
-!- bechamel( has joined #tryton08:00
-!- nicoe_(~nicoe@ has joined #tryton08:12
-!- grasbauer( has joined #tryton08:15
-!- marius_( has joined #tryton08:31
-!- nicoe(~nicoe@ has joined #tryton08:50
-!- jeancavallo1( has joined #tryton08:57
-!- jeancavallo( has joined #tryton08:57
-!- cedk(~ced@gentoo/developer/cedk) has joined #tryton09:08
-!- jeancavallo( has joined #tryton09:19
-!- jeancavallo1( has joined #tryton09:19
-!- rpit( has joined #tryton09:22
-!- vcardon( has joined #tryton09:27
-!- prksh(~prksh@ has joined #tryton09:31
-!- rpit( has joined #tryton09:38
-!- nicoe_(~nicoe@ has joined #tryton09:40
-!- vcardon( has joined #tryton09:55
-!- ytz( has joined #tryton10:34
-!- nicoe_(~nicoe@ has joined #tryton10:39
-!- pobsteta( has joined #tryton10:40
-!- ytz( has joined #tryton10:48
-!- grasbauer( has joined #tryton11:04
grasbauerACTION Sublab confirmed as Location for Sprints and Breakfast at Unconference/Sunday11:06
pokoligrasbauer: there are beers for breakfast??? ;)11:07
grasbauerpokoli: of course!11:08
grasbauerpokoli: there is a 3d printer11:10
pokoligrasbauer: so I can print beers? amazing :P11:11
-!- VaticanCameos(~pritishc@ has joined #tryton11:12
grasbauernicoe_: btw - did you benchmarks withtryton client and gtk311:19
nicoe_grasbauer: not yet11:20
grasbauernicoe_: because on arm the performance of gtk2 is poor - mostly addText11:20
grasbauernicoe_: did a gtkperf and combobox and addText are the bottlenecks11:22
grasbauernicoe_: need to check what is the diffrence between chromebook and banana pi - gtkperf on chtomebook with default setings: 11s - on banana 32s11:26
nicoe_And with gtk3 the perfs were better?11:27
grasbauernicoe_: didn't check yet11:27
-!- vcardon( has joined #tryton11:31
grasbauernicoe_: possibly the difference between ARM Mali-T604 and Mali-400 MP is significant - but on the presentation yesterday I decided to start Tryton via X2Go because of the slow rendering ... yesterday our client had an audit as supplier for Airbus - one point without issues: software integration - a guy told me, that he would love to use a system like tryton - "We are using SAP and it sucks" ;)11:37
nicoe_grasbauer: That's great :D11:43
cedkgrasbauer: what do you mean by addText?11:44
grasbauercedk: GtkTextView - Add text - time:  0,58 here on my chromebook - on banana pi it takes 4,XX11:45
cedkgrasbauer: there is no such method as "add text" on TextView11:47
grasbauercedk:  this check only renders text to a box and is slooooooooow - need to investigate - but the gtk rendering on banana pi is a pain11:51
cedkgrasbauer: I don't understand11:52
grasbauercedk: this is chromebook
grasbauercedk: on banana it takes 37s - so tryton looks like responding slow, but it's a rendering issue11:55
cedkgrasbauer: do you have 2d acceleration?11:56
grasbauercedk: don't know - set up X, xfce with a lightweight theme and run in this issue - I was in hurry and haven't the time to investigate in deep - but out of the box tryton was near unusable11:59
cedkgrasbauer: for me, it looks like you get a plain old vesa driver12:01
grasbauercedk: i think its fbdev12:02
cedkgrasbauer: I even think it is not GTK that is slow but pango12:04
grasbauercedk: possibly - it's doing the font rendering, right?12:04
cedkgrasbauer: yes12:06
cedkgrasbauer: also you should check glx, with glxinfo12:07
-!- prksh(~prksh@ has joined #tryton12:08
-!- nicoe_(~nicoe@ has joined #tryton12:09
grasbauercedk: I will do some tests and report the results - or better I send you a banana for free ;)12:11
marius_is it possible to reconcile payable and receivable accounts?12:31
marius_what's the proper way of doing that?12:31
cedkmarius_: under "Financial>Entries>Reconcile lines"12:32
marius_cedk, I have to select an account there12:33
marius_so after I select it, I don't see payable lines when looking at receivable, etc12:33
pokolimarius_: so you want to reconcile a payable with a receivable?12:34
marius_yeah, something like that12:34
marius_I don't know what's the proper way of doing that12:34
pokolimarius_: so to do it correctly, you have to create a move, from payable to receivable (to move the balance)12:35
pokolimarius_: and the reconcile both (payable with payable, receivable with receivable)12:35
pokolimarius_: this gives the full traceability12:35
cedkmarius_: you must create a move between the two accounts12:35
cedksuch case should really be rare12:36
cedkmaybe you can reconcile the line alone and put the write-off to the other account and than reconcile with the new line12:37
marius_I lose party on account_move when writing off12:40
marius_cedk, I have some parties which both buy from me and I sell to them ;)12:40
cedkmarius_: for the exact same amount?12:41
cedkmarius_: at the exact same date?12:42
marius_not for the same amount, not for the same date12:43
cedkmarius_: so they can not be reconciled12:44
marius_so whats the right way of using receivables as payments and the other way around?12:45
cedkmarius_: I don't understand what you want12:45
pokolimarius_: so you want to balance the payments, and only pay the diferens between what you sold and what you bought?12:47
marius_pokoli, yes12:47
cedkmarius_: look at the party and make a payment of the difference between both amount12:48
pokolimarius_: I think we have a wizard for this, but I don't remember on which module12:49
cedkpokoli: probably it could land in std12:50
pokolicedk: we have a lot of fields that don't exist on std, but I will try to find it and open an issue to add it to standard account12:51
cedkbut the difficulty will be later when receiving the bank statement because you will have to split lines12:51
cedkpokoli: you have a lot of non std field in the accounting12:51
cedkpokoli: this sounds very bad12:51
pokolicedk: for example, payment_type, bank_account, and so12:51
pokolicedk: this is the wizard
cedkpokoli: ha yes the out of mind fields12:53
cedkpokoli: all things that has nothing to do in accounting12:54
pokolicedk: won't discuss about that ;)12:55
pokolicedk: so if we remove all this "out of mind fields" we can introduce this in std?12:55
-!- prksh(~prksh@ has joined #tryton13:03
cedkpokoli: no, I mean it must create a payment13:03
pokolicedk: currently we only create move_line, so the payment can be manually created from there13:05
pokolicedk: so when importing the bank statement you have the correct line with the amount13:06
pokolicedk: and all the old lines get reconciled13:06
cedkpokoli: except that you reconcile before doing the payment13:08
cedkpokoli: and account_payment add a payment field on statement for reconciliation13:09
pokolicedk: yes, but the result of the wizard is a new move line which is not reconciled + all the reconciled ones, so the amount is the same, but grouped in one line13:09
-!- mariomop( has joined #tryton13:11
cedkpokoli: yes but the invoices linked are marked as paid13:11
cedkpokoli: I always found wrong to modify the accounting for operational facilities13:12
-!- smarro(~sebastian@ has joined #tryton13:13
pokolicedk: yes, invoiced are marked as paid13:13
cedkpokoli: so -> wrong13:14
pokolicedk: yes and not13:16
pokolicedk: as some invoices are paid, but others not13:16
pokolicedk: so the only way to do it right is to do it manually13:16
-!- marius_( has joined #tryton13:16
cedkpokoli: no the right way is the make a payment (with clearing)13:20
cedkpokoli: and such payment clearing move should reconcile once succeed13:20
pokolicedk: so basically the PayLine wizard, should be able to group payments by some criteria13:23
pokolicedk: but the payment has a link with only one move (which is right), so how to keep the link between all the compensated moves and the payment?13:24
cedkpokoli: no it should not group payments13:24
pokolicedk: so I don't understand you13:26
pokolimarius_: you requirement is to only send one line to the bank? The amount of this line will be the diference between payables and receivables, isn't it?13:27
cedkpokoli: grouping payments has no meaning13:32
cedkpokoli: you want to have only one payment not many grouped13:32
pokolicedk: Yes, one payment with the diference amount, so I don't understand how you will manage to renconcile all the invoices when the payment has succeed13:33
cedkpokoli: I don't know  but it is the goal13:34
cedkpokoli: otherwise your design is a failure if something goes wrong13:34
pokolicedk: yes from the invoicing point of view, not for the accounting point of view13:35
pokolicedk: as you will still have on move with the pending amount13:35
cedkpokoli: I want both point of view13:35
cedkpokoli: yes but it breaks all the other modules like dunning, account_invoice, sale, purchase etc.13:36
pokolicedk: understood, so I see your point13:37
marius_pokoli, yes13:39
pokolimarius_: so with the solution I explained you will have all the invoices marked as paid13:39
pokolimarius_: which is not the reality, as there is some diference that must be paid13:40
pokolimarius_: if it's right for you, just create the intermediate move and go ahead13:41
marius_pokoli, you mean between receivable account and payable account?13:43
pokolimarius_: yes, so you move the payable amounts to the receivable ones for example13:43
marius_ok, thanks13:44
-!- kstenger(~karla@ has joined #tryton13:49
-!- bechamel( has joined #tryton13:53
-!- LordVan(~LordVan@gentoo/developer/LordVan) has joined #tryton13:58
cedkpokoli: maybe if we could have a way in case of payment failure to completly delete this grouped move14:05
-!- vcardon( has joined #tryton14:09
-!- moldy(~asdf@unaffiliated/moldy) has joined #tryton14:57
-!- alisonken1home( has joined #tryton15:18
-!- smarro(~sebastian@ has joined #tryton15:19
-!- alisonken1home( has joined #tryton15:20
pokolicedk: why we can not delete a failed payment?15:20
pokolicedk: sorry, you talk about the move, so you can unreconcile it, and then create another move using the same origins (so invoices are marked as pending another time)15:25
cedkpokoli: doesn't work because you can not be sure the the account of the invoice will match the not reconciled account of your new move15:27
-!- Timitos( has joined #tryton15:27
cedkpokoli: I mean indeed it is not a problem if invoice is marked as paid during the payment process15:27
cedkpokoli: but at the end, if the payment fails, the invoice must be back to validated15:28
pokolicedk: i was thinking, that the move must be related to only one invoice, So if we fill the origin all will work as expected15:29
pokolicedk: so the user must select an invoice to relate the resulting, if some of the moves are related to an invoice15:30
pokolicedk: so in this case, all the compensated invoices will be marked as paid, and you will only have one invoice that have some part (or the total) pending to pay15:31
cedkpokoli: sounds like a dirty hack15:32
cedkpokoli: more over how do you cancel such things15:32
pokolicedk: invoices has a many2one to moves and we may change this to a one2Many field, so you can modify the lines to pay by creating new moves related to the invoice15:34
pokolicedk: so to cancel you can delete the new created moves (or create a compensation moves if not possible to delete)15:35
cedkpokoli: why not but doesn't fix the original issue15:36
-!- mfladischer( has joined #tryton15:37
pokolicedk: why you don't fix the original issue?15:37
cedkpokoli: OK in some way, but you still need something to create the move and do all the linking15:40
cedkalso I'm not sure you can have an invoice doing more than 1 move15:40
moldyi am considering giving tryton a try at my company sometime soon15:41
moldyi read that it does not really support serial numbers, is that still true?15:41
cedkpokoli: what about, creating the grouping move with reonciliation etc.15:42
cedkpokoli: using a special journal for that15:42
pokolicedk: I'm not sure about the invoice doing more than one move too15:42
cedkpokoli: so on payment fail, if the move is on this *special* journal, payment cancel it15:42
cedkby cancel, I mean delete it if not posted15:43
pokolicedk: and post when confirmed?15:44
pokolicedk: makes sense15:44
cedkpokoli: post when set on statement like for clearing15:45
pokolicedk: yes, post when you really now that this is what happened15:45
cedkpokoli: so now the wizard, should probably be design with care because you must have the ability to select which lines to group15:46
pokolicedk: what we have now, it ensures that you don't mix lines from diferent parties, and this kind of things15:47
cedkpokoli: I think about opening a view with parties that have both kind of lines15:50
pokolicedk: we have this on the lines to pay view15:50
cedkpokoli: from there, open a view per parties with the lines15:50
cedkpokoli: and using a wizard group the selected one15:51
cedkstill one issue is that payment often requires id like end_to_end_id of SEPA15:52
cedkso for now, it uses the origin of the move and for example if it is an invoice take the number15:52
cedknow I guess it must check if it is an other move then take all the origin15:53
pokolicedk: I guess we can get all the origins15:53
pokolicedk: we have this
pokolicedk: so having a domain on this field and ordering with party you have your view15:56
cedkpokoli: don't understand15:58
pokolicedk: don't worry, it's basically how to show all the moves that have another move that can be compensated16:00
cedkpokoli: don't need of that16:00
cedkpokoli: just party with at least unreconciled line in receivable and payable16:01
cedkbut I will never do a function field for that16:01
pokolicedk: so how you will know all the parties that have a unreconcile line for payable and receivable16:01
cedkpokoli: indeed it is just parties with amount_payble != 0 and amount_receivabel != 016:03
cedkindeed no need of anything for that16:03
pokolicedk: right, so you need a way to open receivable and payable lines (can be selected in the wizard) and select the ones you want to group16:06
cedkpokoli: we have already 2 relate16:08
cedkpokoli: better to convert into one relate with tab domains16:08
pokolicedk: makes sense, but with a all tab in order to be able to select both16:09
pokolicedk: I will open an issue for this as can be done easly16:09
cedkpokoli: can put a link to this chat16:14
pokolicedk: I explained a llitle bit the issue, as the chat is large and can be confusing16:15
pokolimoldy: nobody answered your question, did they?16:26
pokolimoldy: I think you need the stock_lot module in order to manage serial numbers16:27
moldypokoli: not yet.16:29
moldypokoli: thanks. i will look into that.16:29
cedkyes, a serial number == a lot with 1 product16:29
moldythis sounds a little unclean, but i will take a look at it, maybe it's just me16:30
-!- alisonken1lap3( has joined #tryton16:31
pokolimoldy: why you say unclean?16:34
moldypokoli: it's just my intuition16:37
pokolimoldy: so just give i try, and if you confirm your intiution we will be happy to hear your thoughts16:38
moldypokoli: yes, i will give it a try. thanks.16:40
moldypokoli: what is the right way to install tryton in a virtualenv, especially how to handle pygtk?16:40
pokolimoldy: install the server on a virtualenv16:41
moldypokoli: ok, i did16:41
cedkmoldy: what is your OS?16:41
pokolimoldy: and install pygtk outside of the virtualenv, and run the client from outside of the virtualenv16:41
moldypokoli: so, installing the client in a virtualenv is not supported?16:42
cedkmoldy: it is but pygtk is complicated16:43
cedkmoldy: you can also use a virtualenv with system-packages16:43
cedkbut indeed, what is your OS?16:43
cedkTryton is packages on many OS16:43
moldycedk: debian16:43
moldythanks. i'd still like to use virtualenv, if it's easily possible.16:49
moldyW: Failed to fetch  404  Not Found16:51
pokolimoldy: you should use jessie-X.X , where X.X corresponds to the series you want to use16:52
pokolimoldy: latests stable series are 3.2, so i will recommend you using jessie-3.216:53
moldyi really hate insalling such things globally to my system16:55
moldyit creates all sorts of headaches :(16:55
pokolimoldy: you won't have problems if you only install the client16:56
-!- vernichon( has joined #tryton16:56
moldyi will give it a try16:56
moldythe server is running. where is the documentaton on how to connect to it?16:57
moldythe client says that it cannot connect. i guess i am missing some configuration, like db and credentials...17:02
moldypokoli: thanks, but i am looking for the server configuration17:05
moldyi just found /etc/trytond.conf17:05
-!- bechamel( has joined #tryton17:07
moldy"can not connect to server"17:08
moldytrytond is 3.2.2, but tryton is 3.2.3 -- is that a problem?17:10
pokolimoldy: no, the series are the same17:13
pokolimoldy: are you connection from different machines?17:13
moldypokoli: ok, thanks. i can now connect to the server. next problem seems to be that the db is empty, it is lacking the schema.17:13
moldypokoli: no, all on localhost17:13
pokolimoldy: just create a new one from the client17:13
moldy"already exists". so i need to delete the existing db first?17:14
moldyunknown problem17:17
moldygeez, what a PITA17:17
cedkmoldy: could be 2 issues:17:19
moldy"permission denied to create the database". which permissions are needed here? which db engine does it try to use by default? sqlite?17:19
cedkmoldy: - ipv6 vs ipv417:19
cedkmoldy: or ssl fingerprint in known_hosts17:20
moldyok, it seems to be trying to use postgres17:20
cedkmoldy: on 3.2 postresql17:20
moldyok. maybe this should be mentioned somewhere ;)17:20
cedkmoldy: it changed in trunk17:21
moldyso i need to create a postgres role and give that role permission to create new databases?17:21
moldythere is no other way to initialize the db?17:21
moldyi am totally confused about where to enter which credentials.17:22
moldythere is "login name" and "password", and then under "manage profiles", there is again "login name" (but no password), and then when i hit "create db", there is "tryton server password", and "admin password".17:23
moldyis tryton really not able to just initialize its own database when it has an existing empty db and permissions on it?17:25
cedkmoldy: no because it will be a security thread17:27
moldythat sounds weird, every other db app i know can do that17:28
cedkmoldy: indeed, it the future database will no more be created with the client17:28
moldycedk: how is it a security threat if tryton creates its db tables?17:28
cedkmoldy: it will be if it creates a DB when someone access it is interface17:28
moldyi am not talking about creating the database. i am talking about creating the tables and filling them with initial data.17:28
moldybig difference.17:29
moldycreating the db is the sysadmin's job, creating the schema and data is the application's job.17:29
moldyi created the db, i started the server, i launched the client, then i get: rogrammingError: relation "ir_module_module" does not exist17:30
cedkmoldy: It is what Tryton does17:30
moldycedk: ok, what am i doing wrong then?17:30
cedkmoldy: trytond can not check on every connection if the database is completly in sync with the code17:31
cedkmoldy: I don't know what you do17:31
moldycedk: at the moment, i just want to know how i get this thing running.17:31
moldycedk: i created an empty postgres db called "tryton". i entered the db name, db user and db password in /etc/trytond.conf. then i started the server and the client.17:32
cedkmoldy: why?17:32
moldycedk: why what? why i created the db?17:32
cedkmoldy: where do you put the db name in configuration ?17:32
cedkmoldy: where do you get all those information?17:32
moldycedk: sorry, i didn't. only the host, user and password17:32
cedkmoldy: so you can use the wizard of the client and follow the instruction17:33
moldyok, so tryton expects that he has privileges to create databases, it's not possible to use existing ones?17:33
moldythat's a bit crazy, but ok.17:33
cedkmoldy: no, you can setup the database from command line17:34
moldycedk: ok, how do i do this?17:35
moldywell, i gave the postgres tryton user CREATE rights and used the client to create th db, that seems to have worked17:37
cedkmoldy: trytond -d <db> -i all17:37
moldycedk: ah, thanks17:37
moldyso that is meant by "init a module".17:38
moldycedk, pokoli: thanks.17:41
moldyis there PDF documentation for the current release?17:42
moldythe one on seems to be very outdated17:43
cedkmoldy: it is not, it is just because it is incomplete17:45
moldycedk: oh, ok. because it says "0.1"17:46
cedkmoldy: where?17:47
moldyfirst page of the generated PF17:47
cedk -> 3.3 in the title17:47
cedkwhich PDF?17:48
moldycedk: bottom left corner, "read the docs", "PDF"17:48
cedkmoldy: don't see by the title is: Release 3.317:49
moldyok, then it's just the PDF17:49
moldythanks for your help, i will drive home now... talk to you later.17:51
-!- Telesight( has joined #tryton18:15
cedkmoldy: is such doc would have helped:
cedkpokoli: scenario fails:
cedkpokoli: Sale is not defined, I don't see why you do a sale18:20
cedkpokoli: please run scenario when you modify them18:20
pokolicedk: ohh, this was not uploaded to the review, so I commited a wrong patch sorry18:21
cedkpokoli: please backout it and push the right one18:29
pokolicedk: done18:33
-!- Telesight( has joined #tryton19:09
-!- Telesight( has joined #tryton19:10
-!- rpit( has joined #tryton20:18
-!- pablovannini( has joined #tryton20:36
-!- gremly(~gremly@ has joined #tryton20:53
-!- smarro(~sebastian@ has joined #tryton21:25
-!- Telesight( has joined #tryton22:07
-!- uranus( has joined #tryton23:26
-!- uranush( has joined #tryton23:30

Generated by 2.11.0 by Marius Gedminas - find it at!