IRC logs of #tryton for Friday, 2009-02-13 #tryton log beginning Fri Feb 13 00:00:01 CET 2009
2009-02-13 01:51 -!- ikks(i=igor@ has joined #tryton
2009-02-13 02:06 -!- vengfulsquirrel( has joined #tryton
2009-02-13 05:01 -!- ikks(i=igor@ has joined #tryton
2009-02-13 05:19 -!- yangoon( has joined #tryton
2009-02-13 07:13 -!- nicoe( has joined #tryton
2009-02-13 07:21 -!- Timitos(n=Timitos@ has joined #tryton
2009-02-13 07:44 -!- sharkcz( has joined #tryton
2009-02-13 07:52 -!- paola( has joined #tryton
2009-02-13 08:54 -!- bechamel(n=user@ has joined #tryton
2009-02-13 09:00 -!- carlos(n=carlos@ has joined #tryton
2009-02-13 09:28 -!- nicoe( has joined #tryton
2009-02-13 09:28 -!- juanfer(n=juanfer@ has joined #tryton
2009-02-13 09:39 -!- Gedd( has joined #tryton
2009-02-13 10:12 -!- cedk(n=ced@gentoo/developer/cedk) has joined #tryton
2009-02-13 10:17 -!- paola( has joined #tryton
2009-02-13 10:42 -!- tekknokrat( has joined #tryton
2009-02-13 12:11 -!- bechamel(n=user@ has joined #tryton
2009-02-13 12:15 -!- ikks(i=igor@ has joined #tryton
2009-02-13 12:20 <CIA-10> tryton: matb roundup * #778/Refactorization: logging of pooler in database update: [new] Since refactorization of the pooler the name of the database which is updated is no more logged (for switch -u all). Would be nice to have ...
2009-02-13 13:31 <CIA-10> tryton: matb roundup * #776/genshi/template/ UndefinedError(key, owner=owner): Doesn't work here. Definitely all calls to attributes/functions of company are not recognized. Deleting the call to company.header reults in U ...
2009-02-13 13:32 -!- ikks_(n=igor@ has joined #tryton
2009-02-13 13:37 <CIA-10> tryton: C?dric Krier <> default * 1544:2798019662d1 trytond/trytond/ Add logging info when init pool for issue778
2009-02-13 13:37 <CIA-10> tryton: C?dric Krier <> default * 1545:56545a0f35cb trytond/trytond/backend/postgresql/ Add logging info when connect to database
2009-02-13 13:37 <CIA-10> tryton: ced roundup * #778/Refactorization: logging of pooler in database update: [chatting] Fix with changeset 2798019662d1
2009-02-13 14:07 <cedk> yangoon: did you send your module?
2009-02-13 14:08 -!- simahawk( has joined #tryton
2009-02-13 14:12 -!- paola( has joined #tryton
2009-02-13 14:13 -!- enlightx( has joined #tryton
2009-02-13 14:26 <cedk> yangoon: I receive it
2009-02-13 14:28 <cedk> yangoon: by the way, did you run with at least the changest 1525 ?
2009-02-13 14:34 <cedk> yangoon: ok, with your module I had the issue
2009-02-13 14:39 <CIA-10> tryton: C?dric Krier <> default * 1546:e8e6a0de00ba trytond/trytond/modules/ Look for all report names before creating missing reports for issue776
2009-02-13 14:39 <CIA-10> tryton: ced roundup * #776/genshi/template/ UndefinedError(key, owner=owner): [resolved] Fix with changeset e8e6a0de00ba
2009-02-13 15:11 -!- paola( has joined #tryton
2009-02-13 15:50 -!- juanfer(n=juanfer@ has joined #tryton
2009-02-13 16:27 <bechamel> hi everybody, i made a sort of graph of the tryton architecture:, comments are welcomed
2009-02-13 16:57 <carlos> bechamel: it looks nice
2009-02-13 16:58 <carlos> bechamel: however, shouldn't you note the DB backend split that is being done for 1.2?
2009-02-13 16:58 <carlos> I know there is no support for other db backends, but the architecture allows (or is going to allow) it
2009-02-13 16:58 <bechamel> carlos: when other backend will be supported I will update it :)
2009-02-13 16:59 <carlos> ok
2009-02-13 16:59 <carlos> bechamel: I have a question about the architecture now that I see your graph
2009-02-13 17:00 <carlos> does it mean that all report logic is on the client side?
2009-02-13 17:00 <carlos> for instance, If I create an xml-rpc client, will I need to add the logic to generate the Invoice as PDF ?
2009-02-13 17:01 <bechamel> carlos: there is a report square on the kernel
2009-02-13 17:02 <carlos> hmm, right, I missed it :-P
2009-02-13 17:02 <carlos> then, what's in the kernel and what's in the client?
2009-02-13 17:02 <carlos> if it's easy to explain it, if it's complex, don't worry too much
2009-02-13 17:02 <bechamel> carlos: all the creation of the document is made server side, the client only has to ask for it and call the correct program to show it
2009-02-13 17:03 <carlos> ok, that makes much more sense, I was worried I misunderstood the way it worked
2009-02-13 17:03 <carlos> bechamel: thanks for the explanation
2009-02-13 17:05 <bechamel> carlos: the graph is made for non-technician people, so it's more eye-candy than rigorous
2009-02-13 17:05 <bechamel> and graph are not rigorous by definition :)
2009-02-13 17:05 <carlos> indeed :-P
2009-02-13 17:56 -!- bechamel`(n=user@ has joined #tryton
2009-02-13 18:24 -!- enlightx_( has joined #tryton
2009-02-13 18:25 -!- enlightx( has joined #tryton
2009-02-13 18:29 -!- simahawk( has joined #tryton
2009-02-13 21:00 -!- vengfulsquirrel( has joined #tryton
2009-02-13 21:23 <udono> bechamel: the graph looking very nice!
2009-02-13 21:24 <udono> bechamel: missing smtp connection for outgoing EMail
2009-02-13 21:24 <udono> and caldav for next version
2009-02-13 21:24 <bechamel> udono: yes
2009-02-13 21:24 <bechamel> udono: I didn't thought about smtp
2009-02-13 21:25 <bechamel> udono: and the graph will follow the new feature
2009-02-13 21:26 <udono> bechamel: will the source be available? So we can translate print a nice illustration for the ceBIT
2009-02-13 21:28 <bechamel> udono: yes of course
2009-02-13 21:28 <bechamel> udono: i send you a mail
2009-02-13 21:30 -!- johbo( has joined #tryton
2009-02-13 21:32 <udono> bechamel: you forget odf integration with openofficeorg
2009-02-13 21:32 <bechamel> udono: it's the report stuff
2009-02-13 21:32 <udono> bechamel: and the arrow lines could be more big
2009-02-13 21:34 <bechamel> udono: you can put bigger arrow if you want, i used inkscape to create it
2009-02-13 21:37 <udono> bechamel: yes, thanks a lot!
2009-02-13 21:38 <udono> bechamel: did you take a look on the advertisment stuff?
2009-02-13 21:38 <udono> bechamel: yangoon sended it to you I think?!
2009-02-13 21:39 <bechamel> udono: yes I received it
2009-02-13 21:40 <bechamel> udono: the idea is to add a news with this content ?
2009-02-13 21:40 <udono> bechamel: yes, Its direct advertisment, so I would prefer it to news instead of press coverage.
2009-02-13 21:42 <bechamel> udono: ok
2009-02-13 22:08 <vengfulsquirrel> cedk: Hey so I don't know if you looked at my planned module breakup on the wiki but I've only started on the base production module's model.
2009-02-13 22:09 <vengfulsquirrel> The states/fields are not complete but maybe if you wanted to just give the relations a rough look over: ,
2009-02-13 22:09 <vengfulsquirrel> I still need to talk to X0d about the revisions because I'm not sure how necessary they are.
2009-02-13 22:15 <cedk> vengfulsquirrel: I don't see the need of type all over
2009-02-13 22:16 <cedk> vengfulsquirrel: why is there a principal_product in production?
2009-02-13 22:16 <cedk> vengfulsquirrel: for me input and output lines for the production are stock.move
2009-02-13 22:17 <vengfulsquirrel> The type was all over because I thought we might need it for extending it, I havne't started on the phantoms and stuff though.
2009-02-13 22:17 <carlos> cedk: is there any way to use xml-rpc from the client in Tryton 1.0?
2009-02-13 22:17 <vengfulsquirrel> There is a principal product in production because I assumed a production would be created to produce a specific product.
2009-02-13 22:17 <cedk> carlos: no
2009-02-13 22:18 <cedk> vengfulsquirrel: I think it is not needed
2009-02-13 22:18 <carlos> cedk: and is there any xmlrpc client example available?
2009-02-13 22:18 <carlos> I need a trivial one
2009-02-13 22:18 <carlos> nothing complex
2009-02-13 22:18 <cedk> vengfulsquirrel: I don't see the needs of consumed_moves
2009-02-13 22:19 <cedk> carlos: bechamel have got some I think
2009-02-13 22:19 <vengfulsquirrel> The input and output lines have to be more than stock moves because there needs to be reports of what was actually selected for production and what actually came out of the production. The products could change or be added.
2009-02-13 22:19 <cedk> vengfulsquirrel: ok, I understand the input but not the output
2009-02-13 22:20 <cedk> vengfulsquirrel: and I will call input with other name
2009-02-13 22:20 <vengfulsquirrel> It is meant to parallel the bom though but is more condensed that is why the names are similar.
2009-02-13 22:20 <cedk> vengfulsquirrel: like bom selected ...
2009-02-13 22:21 <carlos> cedk: ok, thanks
2009-02-13 22:21 <bechamel> carlos:, don't forget to allow xmlrpc on the server config
2009-02-13 22:21 <cedk> vengfulsquirrel: or bom split
2009-02-13 22:21 <bechamel> bbl
2009-02-13 22:21 <vengfulsquirrel> Also the output is necessary if for example extra "products" are produced if something goes wrong, like waste or partially completed products.
2009-02-13 22:21 <cedk> vengfulsquirrel: but this is the moves that makes that
2009-02-13 22:23 <vengfulsquirrel> The moves need to know what product they are moving though.
2009-02-13 22:23 <cedk> vengfulsquirrel: I think that the bom split is more complex
2009-02-13 22:23 <cedk> vengfulsquirrel: this will be computed on time
2009-02-13 22:23 <vengfulsquirrel> What is a bom split ?
2009-02-13 22:25 <cedk> vengfulsquirrel: it is all the boms splitted into lines
2009-02-13 22:27 <vengfulsquirrel> You mean exploding the bom ?
2009-02-13 22:28 <cedk> vengfulsquirrel: yes
2009-02-13 22:28 <vengfulsquirrel> cedk: The moves to/from consumed/produced were meant so as to not flood lost/found.
2009-02-13 22:29 <cedk> vengfulsquirrel: I don't understand, you just need to leave it in production location
2009-02-13 22:29 <vengfulsquirrel> You mean the Production Input Zone ?
2009-02-13 22:29 <vengfulsquirrel> cedk:
2009-02-13 22:30 <cedk> vengfulsquirrel: I think there is too much location in production
2009-02-13 22:30 <cedk> vengfulsquirrel: one location is enough
2009-02-13 22:32 <vengfulsquirrel> So you would just create draft moves to lower the quantitiy forecasts until the production started ?
2009-02-13 22:32 <cedk> vengfulsquirrel: yes
2009-02-13 22:33 <carlos> bechamel: thanks!
2009-02-13 22:33 <cedk> vengfulsquirrel: and even assigned
2009-02-13 22:34 <cedk> vengfulsquirrel: for me the production will be first a selection of some boms
2009-02-13 22:34 <cedk> vengfulsquirrel: one normal and some phatoms
2009-02-13 22:34 <vengfulsquirrel>
2009-02-13 22:36 <cedk> vengfulsquirrel: and so input and out will be real move that are computed from the selection of boms
2009-02-13 22:37 <cedk> vengfulsquirrel: it is no needed for me to have two steps between the selection of boms and the creation of moves
2009-02-13 22:39 <cedk> vengfulsquirrel: I don't understand unfinished state
2009-02-13 22:39 <cedk> vengfulsquirrel: can I have the source of the diagram production-order-states
2009-02-13 22:40 <cedk> vengfulsquirrel: I will change it like I see the production
2009-02-13 22:40 <vengfulsquirrel> cedk: Yeah the Unfinished state is meant to collect the repucussions of canceling a production that has been started.
2009-02-13 22:40 <vengfulsquirrel> So like some stuff must be thrown away and some stuff must be considered a Work In Progress
2009-02-13 22:40 <vengfulsquirrel> etc.
2009-02-13 22:41 <cedk> vengfulsquirrel: still don't understand
2009-02-13 22:42 <cedk> vengfulsquirrel: you want to set to done some moves ?
2009-02-13 22:44 <vengfulsquirrel> I guess that state was meant to help account for everything that was used in the production that was canceled.
2009-02-13 22:44 <vengfulsquirrel> Hold on I'm trying to upload the diagrams.
2009-02-13 22:45 <cedk> vengfulsquirrel: ok, but I don't see how this will be made in the GUI
2009-02-13 22:46 <cedk> vengfulsquirrel: it can be something other than a state, but a wizard that ask what must be move to lost and found
2009-02-13 22:46 <cedk> vengfulsquirrel: but I think this is some thing for real complex production
2009-02-13 22:49 <vengfulsquirrel>
2009-02-13 22:49 <vengfulsquirrel>
2009-02-13 22:51 <vengfulsquirrel> cedk: My plan was that you would pull the stock necessary for a production AND THEN "start" the production. In the same fashion that you would pull stock from storage to the input zone and then ship it to the customer.
2009-02-13 22:52 <vengfulsquirrel> Sorry I meant *storage->output->customer but similarly the produced->output->storage would also parallel supplier->input->storage .
2009-02-13 22:53 <vengfulsquirrel> I think it would be more convenient for the system to guess where stock should be moved to in storage after production based on what is already there. But usually the person that does that probably doesn't have anything to do with production they are just putting stuff away.
2009-02-13 22:54 <vengfulsquirrel> That is why there are steps: Select boms, Pull materials, Start production, Record production results, Put materials away.
2009-02-13 22:55 <cedk> vengfulsquirrel: I'm not sure that it must be handle in the production order
2009-02-13 22:55 <cedk> vengfulsquirrel: it must be configured with the location tree
2009-02-13 22:56 <vengfulsquirrel> What do you mean ? It would work similarly to assignation.
2009-02-13 22:56 <vengfulsquirrel> Defaulting to moving to the root: Storage Zone.
2009-02-13 22:57 <vengfulsquirrel> I think inventory management is the most difficult problem for my business and that's what I am so concerned with machine assistance when accounting for inventory.
2009-02-13 22:58 <vengfulsquirrel> So maybe I should think of a way to extend a more minimal production order to fulfill my inventory management needs.
2009-02-13 23:02 <cedk>
2009-02-13 23:03 <cedk> vengfulsquirrel: I mean if you need to move product in a pre-production storage
2009-02-13 23:04 <cedk> vengfulsquirrel: you must create this location outside the storage
2009-02-13 23:04 <cedk> vengfulsquirrel: and the scheduler will generate the internal packing that must be done
2009-02-13 23:04 <cedk> vengfulsquirrel: it is better to decompose stuff and reuse the existing mecanism
2009-02-13 23:06 <cedk> vengfulsquirrel: because I think that it is not at the production time that you move the product from the storage to the pre-production location?
2009-02-13 23:07 <vengfulsquirrel> yeah but a production order could be created weeks/months in advanced and then maybe the materials would be pulled 'right before' or maybe a week before
2009-02-13 23:08 <vengfulsquirrel> cedk: Hey there is a problem with doing all that at the draft stage.
2009-02-13 23:09 <vengfulsquirrel> Boms needs to be selected AND THEN substitutes must be selected
2009-02-13 23:11 <cedk> vengfulsquirrel: the scheduler will create the order as soon as it sees that the stock quantity of a product is lesser than 0
2009-02-13 23:11 <cedk> vengfulsquirrel: so it is not done at draft stage but long time before
2009-02-13 23:12 <cedk> vengfulsquirrel: it depends of the range of date that the scheduler evaluates
2009-02-13 23:12 <vengfulsquirrel> Yeah okay just as long as that can be turned off because that probably will only work for make to order businesses.
2009-02-13 23:13 <cedk> vengfulsquirrel: even for other way to work
2009-02-13 23:14 <cedk> vengfulsquirrel: if it is not to order, the production order must be created by a planning so we fail into the same senario
2009-02-13 23:14 <vengfulsquirrel> Yeah okay that's True
2009-02-13 23:16 <vengfulsquirrel> Okay something about your diagram: if a production stops you can't just cancel the original moves some of the materials could have been destroyed or partially altered
2009-02-13 23:16 <CIA-10> tryton: vengfulsquirrel * r436 /wiki/ Fixed links.
2009-02-13 23:17 <cedk> vengfulsquirrel: I think that must be made with a wizard
2009-02-13 23:18 <cedk> vengfulsquirrel: so you put a wizard on the button cancel
2009-02-13 23:18 <cedk> vengfulsquirrel: but this can be made with optionnal module
2009-02-13 23:18 <cedk> vengfulsquirrel: to keep the based module as simple as possible
2009-02-13 23:18 <cedk> I go to lunch
2009-02-13 23:18 <cedk> bbl
2009-02-13 23:19 <vengfulsquirrel> cedk: okay wizard might be fine, ttyl
2009-02-13 23:43 <cedk> back
2009-02-13 23:46 <vengfulsquirrel> hey
2009-02-13 23:46 <cedk> vengfulsquirrel: one last things, I think you must allow to select a phatom bom in the bom lines
2009-02-13 23:47 <vengfulsquirrel> cedk: That is meant for a seperate module though in my plan.
2009-02-13 23:47 <cedk> vengfulsquirrel: I don't think that will be possible
2009-02-13 23:47 <vengfulsquirrel> I don't see how these moves are going to work when products will need to be substituted either.
2009-02-13 23:48 <vengfulsquirrel> I think we might need to make the production order states more extensive in order to make it modular enough to add in more modules later.
2009-02-13 23:48 <cedk> vengfulsquirrel: because this will define all the boms in a production order
2009-02-13 23:48 <vengfulsquirrel> Yeah sorry what do you mean by phantom bom ?
2009-02-13 23:48 <vengfulsquirrel> I think we got this confused before.
2009-02-13 23:49 <cedk> vengfulsquirrel: it is a bom where the result is not a real product
2009-02-13 23:49 <vengfulsquirrel> Yes okay
2009-02-13 23:49 <vengfulsquirrel> So when a production bom has an input line that references a phantom product that phantom product's bom will be merged into the production bom.
2009-02-13 23:50 <cedk> vengfulsquirrel: yes
2009-02-13 23:50 <vengfulsquirrel> Yes okay that is determined at bom "explosion" time.
2009-02-13 23:50 <vengfulsquirrel> Substitutes must be determed after the bom has been exploded.
2009-02-13 23:50 <cedk> vengfulsquirrel: and if there is more than one bom for a phantom product, we must allow to select one
2009-02-13 23:51 <cedk> vengfulsquirrel: at the draft state when we change the boms selection of a production order, we need to "explode" it
2009-02-13 23:51 <vengfulsquirrel> Yeah okay so maybe that will be like Complete Lines is in the inventory.
2009-02-13 23:52 <cedk> vengfulsquirrel: yes complete or change
2009-02-13 23:52 <vengfulsquirrel> Hmmm
2009-02-13 23:52 <vengfulsquirrel> That is very confusing
2009-02-13 23:52 <cedk> vengfulsquirrel: the best will be on the fly but the simplest will be to put a compute button
2009-02-13 23:53 <vengfulsquirrel> selecting a different phantom could include more phantoms
2009-02-13 23:53 <cedk> vengfulsquirrel: yes
2009-02-13 23:53 <vengfulsquirrel> So it can be any number of steps
2009-02-13 23:53 <cedk> vengfulsquirrel: the selection stuff will be a little bit complex
2009-02-13 23:53 <vengfulsquirrel> based on user input
2009-02-13 23:54 <vengfulsquirrel> the user will just have to keep making changes and clicking Explode BOM until they are satisfied
2009-02-13 23:54 <cedk> vengfulsquirrel: I think it must be a set of line with (default bom, selected bom) ordered by the level in the bom tree
2009-02-13 23:56 <vengfulsquirrel> yeah okay I'll have to think about that more but the model will have to be much more complex
2009-02-13 23:57 <vengfulsquirrel> and I'm not sure we need that complexity in the base module
2009-02-13 23:57 <vengfulsquirrel> but i'm not sure we can not put it in either
2009-02-13 23:58 <cedk> vengfulsquirrel: I think we need in the base the selection of bom
2009-02-13 23:58 <bechamel> carlos: I just remember that i made this
2009-02-13 23:58 <vengfulsquirrel> yeah but sorry i am also thinking about substitutes
2009-02-13 23:58 <cedk> vengfulsquirrel: and phatom bom is very useful even in simple production
2009-02-13 23:59 <carlos> bechamel: don't worry, I just discovered that OpenBravoPOS is not using xmlrpc but soap
2009-02-13 23:59 <vengfulsquirrel> if you substitute a phantom product with another phantom product than a new bom must be selected
2009-02-13 23:59 <carlos> bechamel: so I need to do my own server to connect it with Tryton
2009-02-13 23:59 <carlos> bechamel: that's why my xmlrpc testing was not working
2009-02-13 23:59 <bechamel> carlos: ok

Generated by 2.17.3 by Marius Gedminas - find it at!