hello? *** reinhard has joined #gnuenterprise hail, chillywi1ly yo *** dimas has quit IRC *** btami has joined #gnuenterprise *** Morphous has joined #gnuenterprise *** Amorphous has quit IRC *** kilo has joined #gnuenterprise *** btami has quit IRC hi reinhard: gacvs --selftest breaks DB000: gnue.common.apps.checktype.TypeError: "row" is expected to be of but is of and has value None latest svn and rebuilt db completely *** holycow has quit IRC *** ajmitch has quit IRC *** reinhard has quit IRC *** ajmitch has joined #gnuenterprise *** reinhard has joined #gnuenterprise *** dsmith has joined #gnuenterprise *** kilo has quit IRC *** cpele has joined #gnuenterprise *** cpele has quit IRC *** jamest has joined #gnuenterprise *** dimas has joined #gnuenterprise reinhard: dropdowns working for you now? did you change something? can't remember seeing a commit mail let me svn update oh yeah there comes stuff :) hmm btw they seem to be a lot more usable than they were if it doesn't work for you then be sure and send all complaints to um um derek :) yeah, that's it, derek ok that's where they always go anyway actually there is a new config option dropdownMode auto,gtk1,windows windows is btami and co's setup gtk1 is more like the old way derek: when i drop down a dropdown it's not always the correct entry that is highlighted for example i have "foo" selected when i drop down, in the list "bar" is highlighted i think that happens when i set the value of the dropdown via trigger or when i display an existing record -- was that ok this way :) ? yeah, he'll get right on that i however may need another day or two before I can dig into it again that stuff is messy btw :) ;) jamest: so you think it will be no fun for derek to fix it? ;-) i think he'll have lots of fun *** dyfet has joined #gnuenterprise *** wayneg has quit IRC *** reinhard_ has joined #gnuenterprise *** reinhard has quit IRC *** reinhard has quit IRC *** reinhard has joined #gnuenterprise *** Vee has quit IRC *** Vee has joined #gnuenterprise *** wayneg has joined #gnuenterprise jamest: btw, could you verify the useParameters=1 date bug? to be honest, i forgot about it :( i'll try and test it tonight *** jcater has joined #gnuenterprise *** dyfet has quit IRC i have an understanding problem of conditions my understanding is that a GCondition instance is used as the top level node of a condition tree can a GCondition have more than one child? anyone know of a way I can rad a whole sql file and get it to execute just like using psql and redirecting stdin....preferably without having to parse the sql file ;) programatically with psycopg I have some drop statements but if those objects don't exist it throws an exception which I have been catching but the transaction is aborted...I tried turning autocommit on too hmmm, I suppose I can exec psql using the os module hmmmmmm doh select * from foo where bar != 'x'; does *not* return rows where bar is NULL??? reinhard: I believe so. is that logical? Sure If it's NULL, it's not even there. if it's NULL, it is not 'x' There isn't something there that is not x. so there is a basic difference between "!=" and "not =" You can't compare the thing in bar to x, because there isn;t anything there. you have to use "is not null" i think i have to use where not bar = 'x' instead of where bar != 'x' select * from foo where bar != 'x' or bar is null; * dsmith guesses doh where not bar = 'x' doesn't work, either no "=" and "null" do not mix whether or not it's prefixed with "!" i don't get the logic if "bar = 'x'" is false set theory then "not bar = 'x'" must be true boolean theory :) sql is built on set theory so i reallly have to "or bar is null " yes thanks it works with "or is null" can anybody answer the question if a GCondition can have more than one child? I think it can you guys rewriting gconditions? that scares me jcater: no we aren't we are trying to *use* them :) what johannes did is add a method ah, ok reinhard: iirc, multiple children of GCondition implies "and" to let a condition tree decide if it is true or false for a given dictionary of field/value pairs this allows you to quickly combine multiple conditions on the flyt s/flyt/fly/ this is also what i read from the code but then i don't understand combineConditions children = cond1._children[:] cond1._children = [] _and = GCand(cond1) _and._children = children if len(cond2._children) > 1: _and2 = GCand(cond1) _and2._children = cond2._children[:] else: cond1._children.append(cond2._children[0]) this all could be replaced by cond1._children.extend (cond2._children) ? also, buildPrefixFromTree doesn't handle GCondition at all well, is combineConditions used internally by GCondition when it has more than one child? * jcater doesn't know what buildPrefixFromTree is... * jcater needs to look jcater: combineConditions is used "externally" to combine the static condition of a datasource with the condition given in the query buildPrefixFromTree was done by jcater according to svn praise :) currently we use buildPrefixFromTree and buildTreeFromPrefix internally in appserver to be able to transport a condition tree over a RPC connenction i'm not sure if these functions are used outside appserver anywhere if not, we would like very much to rewrite them, as they butcher non-trivial condition trees completeley beyond all recognition reinhard: I'm almost certain it's not used externally but don't know where it's used internally we grepped through the whole gnue tree and it's only used in appserver then I say do what you need to do with it ok we're digging around in conditions quite a lot lately as we stand now in CVS head, appserver can handle calculated fields in conditions that is you can write a python method that returns a value from a calculation (say total invoice value or something like that) and you can use that function like a field even in conditions appserver breaks apart the condition sends the part that can be translated into SQL to the database and checks the rest of the conditions itself *** dneighbo has joined #gnuenterprise *** someon has joined #gnuenterprise Anyone awake? *** derek has quit IRC kinda Grins.... Kinda wondered after seeing the log... reinhard: I saw you were online and doing stuff with AppServer... I'm looking at doing a re-write of an extensible proprietary system, and like the progress that AppServer is making... *** someon has left #gnuenterprise *** dsmith has quit IRC *** reinhard has quit IRC *** dsmith has joined #gnuenterprise *** jbailey has joined #gnuenterprise *** sjc has joined #gnuenterprise *** kilo has joined #gnuenterprise *** johnf has joined #gnuenterprise *** johnf has quit IRC *** dsmith has quit IRC *** jamest has left #gnuenterprise *** kilo has quit IRC *** sjc has quit IRC *** jamest has joined #gnuenterprise *** jcater has quit IRC *** dcmwai has joined #gnuenterprise *** dcmwai has quit IRC *** dcmwai has joined #gnuenterprise *** dcmwai has quit IRC *** dcmwai has joined #gnuenterprise *** dcmwai has quit IRC *** jbailey has quit IRC *** jcater has joined #gnuenterprise *** jcater has quit IRC *** kreiszner has joined #gnuenterprise *** kreiszner has joined #gnuenterprise