*** johannesV has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
*** kilo has joined #gnuenterprise
<kilo> good morning
*** kilo has quit IRC
*** kilo has joined #gnuenterprise
<johannesV> good morning
<kilo> hi johannesV
*** johannesV has quit IRC
*** johannesV has joined #gnuenterprise
<kilo> reinhard: how much are the datatypes 'date' and 'time' useable wrt appserver? do you still recommend not using them?
<johannesV> kilo, you can use them, of course ..
<johannesV> but you should keep in mind that backends handle date-types different
<johannesV> (until we fix that in common)
<johannesV> or find another solution to the problem
<kilo> it just hurts for me to type datetime eg date of birth...
<johannesV> ?
<johannesV> i'd say a birthdate is of type "date"
<johannesV> or do you track the time too for that information ?
<johannesV> :)
<kilo> it was my original question, that should i stick to using datetime as a safe datatype or not.
<johannesV> ah, now i understand ... sorry
<kilo> however for a GNUe-Horoscope birth time could be also important 8-)))
<johannesV> i'd say use date (since it is the *right* type)
<kilo> and exact geo coordinates...
<johannesV> :)(
<kilo> after fixing dropdowns, fixing such a little thingy should be a 10-min work for you two 8-)))))
<johannesV> i had a strict order *not* to start anything *new* today ... :)
<kilo> lol
<kilo> i'm fiddling with changes that new dropdown behaviour causes...
<kilo> could yuou help out plz?
<kilo> in a block that contains invoice lines, i have
<kilo>       <field name="fldInvProduct" field="INV_product"
<kilo>              fk_description="ITEM_shortName" fk_key="gnue_id"
<kilo>              fk_source="dts_ITEM_Item" maxLength="32"/>
<kilo>       <field name="fldInvProductPrice" field="INV_product"
<kilo>              fk_description="ITEM_price" fk_key="gnue_id"
<kilo>              fk_source="dts_ITEM_Item" readonly="Y" />
<kilo> first field is dropdown, second is a label for displaying the price of the product
<kilo> this works OK, it displays the price. but fldInvProductPrice contains the product itself (of type ITEM_Item), not the price.
<kilo> but I would need the price for later calculations. how t achieve that the second field contain the price?
<johannesV> have you tried fk_description="ITEM_price.foobar_price" ?
<kilo> I dont get it
<johannesV> how to reach the price ?
<kilo> it is in INV_product.ITEM_price, it is a num,ber
<johannesV> just a sec, i'll have a look at the gcd's
<kilo> ok, thx
<johannesV> in which gcd do i find it ?
<johannesV> item.gcd ?
<kilo> yes
<johannesV> ITEM_item.ITEM_price
<johannesV> is this the value you'd like to see ?
<kilo> yes
<johannesV> from which class ?
<johannesV> INV_product ?
<kilo> INV_Item has a product, of type ITEM_Item.
<johannesV> INV_Item.INV_product.ITEM_price
<johannesV> this is the fully qualified name of the value (as seen from INV_Item)
<kilo> all this to fk_description?
<johannesV> fk_description="INV_product.ITEM_price"
<kilo> fk_source is ITEM_Item, and ITEM_Item has no INV_product
* johannesV phone
<kilo> i'm going to commit recent state
<johannesV> so i'm a bit confused now ...
<johannesV> shortName is in the same class as the price you'd like to see...
<johannesV> why can't you use "ITEM_price" as fk_description ?
<johannesV> this should actually work if shortName is displayed correctly
<johannesV> btw. you won't need that readonly="Y", just set the entry to style="label"
<kilo> readonly is only double-cheking, a good old habit of C programmers...
*** ajmitch_ has quit IRC
<kilo> ITEM_price works, i can display it and it correctly shows the product's price.
<kilo> but the field itself contain an instance of the product, not the price.
<kilo> and i would like that it contained the price, not the gnue_id of the product
* johannesV phone ...
* kilo sharpens his knife, gets in the car, drives to Austria and cuts off that annoying phone line 8-))
<johannesV> back
<kilo> committed
<kilo> if you svn up, you'll see that product price is displayed when selecting the product from the dropdown
<johannesV> but it seems to work fine ...
<johannesV> as soon as i select another item, the price changes ...
<kilo> uncomment line 150
<johannesV> in the gfd ?
<kilo> yes
<kilo> after entering the qty, it should display that line's total
<johannesV> TypeError: int argument required
<kilo> yes. it is because a gnue_id is in fldInvProductPrice, not the price...
<johannesV> why don't you do this in a OnChange () trigger ?
<johannesV> (on appserver-side=
<johannesV> s/=/)
<kilo> hung to quantity in the gcd?
<johannesV> are create a calculated property at least
<kilo> hmmm
<johannesV> kilo: to get the 'real' price you'd have to add another datasource which you can query
<johannesV> say dtsItem_item which you query with dtsItem_item.simpleQuery ({'gnue_id': blkINV_Item.fldInvProductPrice.get ()})
<johannesV> and then you have the price
<johannesV> this is not a bug, though
<kilo> no, not a bug
<johannesV> the 'real' value of the field defined by fldInvProductPrice in fact is a gnue_id
<johannesV> but it's displayed by it's price
<kilo> and is there no way that it contained the price?
<johannesV> i'd say you do the calcs in a calculated property and use the on-change trigger (forms) to fire it
<johannesV> no, the block blkINV_Item does not provide a container for that value
<kilo> and what about your previous idea, OnChange in the gcd?
<johannesV> but maybe we could change GFField to provide the 'lookup' value for fk-fields ...
<johannesV> OnChange in the gcd would mean you've to store the result in there too ...
<kilo> ah. sure...
<johannesV> but as it looks like you're  storeing the result in blkFree....
<johannesV> but you could combine it a bit ...
<johannesV> calc-property in gcd (doing the dirty work) and in the forms-trigger just access that prop
<johannesV> INV_Item.fldMyFunnyResult (for example)
<kilo> blkFree is only because of limitatons imho, so if it could be solved any other way more elegant, i will go that way
<kilo> @roulette
<bigbrother> *BANG* Hey, who put a blank in here?!
<johannesV> where this field is connected to the calc. prop
* bigbrother reloads and spins the chambers.
<kilo> damn
<kilo> ok, thx for the help
<johannesV> np
<kilo> @dice
<bigbrother> kilo: dice <dice>d<sides>
<kilo> @dice 6d6
<bigbrother> kilo: 4, 4, 1, 3, 6, and 5
<kilo> johannesV: whoaaa, I win, Evil-Appserver loose 8-))) Thx!!!
*** johannesV_ has joined #gnuenterprise
*** johannesV has quit IRC
<kilo> @fortune
<kilo> ah
<kilo> no fortune 8-))
*** dcmwai has quit IRC
<reinhard> bbl
<reinhard> have to fix a suse server this afternoon
<reinhard> which means
<reinhard> format disk, install debian
<reinhard> :-)
<reinhard> l8r all
*** reinhard has quit IRC
<kilo> johannesV_: it is not a new problem, so you are allowed to deal with it 8-)))
<kilo> latest commit in gnue-contrib
<kilo> if you enable the trigger in line 64 in INV_Head.gfd, it causes infinite loop upon commiting
<kilo> why???
<johannesV_> there's no trigger at that line
<johannesV_> line 64 is <logic>
<johannesV_> ?
<johannesV_> ah
<johannesV_> didn't svn up
<kilo> @roulette
<bigbrother> kilo: *click*
<kilo> can you reproduce it?
<kilo> @weather LHUD
<bigbrother> kilo: The current temperature in Szeged, Hungary is 51°F (1:00 PM CET on November 19, 2004). Conditions: Partly Cloudy. Humidity: 67%. Dew Point: 43°F. Pressure: 29.44 in (997 hPa). Visibility: 19.0 miles (30.0 kilometers). UV: 0 out of 12
*** inv0lved has joined #gnuenterprise
*** kilo has quit IRC
*** inv0lved has quit IRC
*** johannesV__ has joined #gnuenterprise
*** johannesV_ has quit IRC
*** johannesV__ has quit IRC
*** lekma has joined #gnuenterprise
<lekma> hello everyone
*** wendall911 has joined #gnuenterprise
*** kilo has joined #gnuenterprise
<kilo> hi
<kilo> hi lekma
<lekma> hi kilo
*** reinhard has joined #gnuenterprise
<lekma> hi reinhard
<lekma> reinhard: if i were to reload a class while running server would replaceReferences (...) and then complete () on a class definition be enough??
* lekma wonders if his sentence makes sense
<reinhard> lekma: I'm sorry I can't tell
<reinhard> johannesV has done most of the class repository stuff and I have lost track at some point
<reinhard> what you say seems to make sense to me so you might be right
<reinhard> but to be sure you should ask johannesV
<lekma> ok , i'll ask johannes later
<lekma> in fact we already have to new methods to reload classes and modules that seem to be working
<lekma> but i wanted to be sure before submitting to you or johannes
<reinhard> lekma: how do these methods work?
<reinhard> let me rephrase
<reinhard> how are these methods activated?
<lekma> they are functions of geasSession that are available in geasSessionManager and then via xmlrpc (appserver.grpc)
<reinhard> ah
<lekma> reloadClass (classname)
<lekma> is there something wrong in this sheme?
<reinhard> we were toying with the idea to let appserver react on a signal
<reinhard> or with a pseudo business object (comparable to the /proc filesystem in linux)
<reinhard> my major concern is
<lekma> humm lot of fun in perspective
<reinhard> there will be several administrative things
<reinhard> not only reloading gcd's
<reinhard> but also stuff like killing sessions, requesting info about sessions, probably much more
<reinhard> and I think we don't want to clutter up the RPC interface
<lekma> hmm
<lekma> phone, bbl
<reinhard> in any case I agree with you that something like reloading gcd's in running appserver is needed
<reinhard> but we saw that as part of a "bigger" project named "administrative interface"
<reinhard> well my wife demands for me
<reinhard> so i'll bbl myself
<reinhard> :)
*** johannesV has joined #gnuenterprise
<lekma> i have to run, i'll try to catch up with you or johannes tomorrow
<lekma> night all
*** lekma has quit IRC
*** holycow has joined #gnuenterprise
*** wendall911 has left #gnuenterprise
*** sjc has joined #gnuenterprise
*** cilkay has quit IRC
*** bluesbaron_ has joined #gnuenterprise
*** bluesbaron_ has left #gnuenterprise
*** jcater has joined #gnuenterprise
*** cilkay has joined #gnuenterprise
*** bluesbaron_ has joined #gnuenterprise
*** bluesbaron_ has left #gnuenterprise
*** sjc has quit IRC
*** reinhard has quit IRC
*** kilo has quit IRC
*** holycow has quit IRC
*** sjc has joined #gnuenterprise
*** sjc has quit IRC
*** sjc has joined #gnuenterprise
*** shmooz has quit IRC
*** shmooz has joined #gnuenterprise
*** sjc has quit IRC
*** jbailey has quit IRC
*** cilkay has left #gnuenterprise
*** holycow has joined #gnuenterprise