From: Max M. <max...@ma...> - 2003-02-13 17:15:37
|
Hi David, An even simpler and better solution than the one I suggested. = Nice! Regards, Max =09 On Wednesday, February 12, 2003, at 09:29 AM, David Teran wrote: > Hi all, > > i changed ERXGenericRecord and now it uses reflection to call=20 > Validity. Just set > > er.extensions.ERXGenericRecord.useValidity=3Dtrue > > in any Properties file that gets loaded or do this manually:=20 > System.setProperty("er.extensions.ERXGenericRecord.useValidity",=20 > "true"); > > and ERXGenericRecord will use validity. Make sure that validity=20 > classes are in the classpath by adding the framework, this is -not-=20 > necessary for compiling ERExtensions, its just important at runtime so=20= > that reflection finds the classes. > > regards david > > > On Wednesday, Feb 12, 2003, at 12:03 Europe/Berlin, David Teran wrote: > >> Hi, >> >> ok, i will implement something based on reflection and lazy=20 >> initialization to eliminate the need for linking the one framework to=20= >> the other one, should not be too slow because WO itself uses already=20= >> so much reflection. >> >> regards David >> >> On Wednesday, Feb 12, 2003, at 08:22 Europe/Berlin, Anjo Krank wrote: >> >>> Hi David, >>> >>> as I now have 4 apps in production, I'd like a solution that doesn't=20= >>> force me to change anything on them - and I'm not really comfortable=20= >>> in changing the superclass of ERXGenericRecord...why not make a=20 >>> ERValidity.framework that would bridge the two frameworks? >>> >>> >>> Cheers, Anjo >>> >>> PS: Can you send me your changes? >>> >>> Am Dienstag, 11.02.03 um 18:53 Uhr schrieb David Teran: >>> >>>> Hi, >>>> >>>> for me its also OK if ERExtensions would depend on Validity but i=20= >>>> think that ERExtensions is more or less the 'root' framework=20 >>>> besides ERJars. But even if we would use=20 >>>> 'editingContextShouldValidateChanges' we would always have a=20 >>>> dependency on Validity. So if Validity should not depend on=20 >>>> ERExtensions then it means either that ERExtensions will depend on=20= >>>> Validity or ERExtensions will not use Validity at all. Of course=20 >>>> this discussion would be obsolete with obj-c but does anyone have a=20= >>>> solution here? Validity simplyfies the validation of attributes a=20= >>>> lot and its so easy to use that even customers can 'generate' the=20= >>>> validation rules. >>>> >>>> regards David >>>> >>>> On Tuesday, Feb 11, 2003, at 17:21 Europe/Berlin, Michael Sacket=20 >>>> wrote: >>>> >>>>> Hey Guys, >>>>> >>>>> Jeremy and I have been discussing this issue and while we think we=20= >>>>> would prefer that Validity did not rely on ERExtensions, we aren't=20= >>>>> completely against it. Whichever works best for the most is fine=20= >>>>> with us. One of the things that has always bugged me a bit about=20= >>>>> Validity is that we even had to make GSVGenericRecord and=20 >>>>> GSVCustomObject. We always figured there ought to be a better way=20= >>>>> to hook in to the validateFor methods without having to extend=20 >>>>> EOGenericRecord One thought we had was to implement a delegate=20 >>>>> for EOEditingContext which implemented the=20 >>>>> 'editingContextShouldValidateChanges' method, and do the=20 >>>>> validation here. Any thoughts on this? >>>>> >>>>> Regards, >>>>> Michael >>>>> >>>>> >>>>> On Tuesday, February 11, 2003, at 08:05 AM, David Teran wrote: >>>>> >>>>>>> >>>>>>> Hello, >>>>>>> >>>>>>>> Anyway, I think it will be really great if Validity and >>>>>>>> ERXGenericRecord work together. What do you guys think? >>>>>>> >>>>>>> Let me ask which would depend on which: Validity on ERExtensions=20= >>>>>>> or >>>>>>> inverse, or mutual? >>>>>>> >>>>>> Hm if you ask me like this i would suggest that ERExtensions=20 >>>>>> depends on Validity because Validity is much 'smaller'. But in=20 >>>>>> real for me it does not matter if the hierarchy is >>>>>> >>>>>> com.webobjects.eocontrol.EOGenericRecord >>>>>> --- com.gammastream.validity.GSVGenericRecord >>>>>> ------ er.extensions.ERXGenericRecord >>>>>> >>>>>> or >>>>>> >>>>>> com.webobjects.eocontrol.EOGenericRecord >>>>>> --- er.extensions.ERXGenericRecord >>>>>> ------ com.gammastream.validity.GSVGenericRecord >>>>>> >>>>>> or just by copying the related methods directly into=20 >>>>>> ERXGenericRecord, but i would try to avoid copy and paste. >>>>>> >>>>>> com.webobjects.eocontrol.EOGenericRecord >>>>>> --- er.extensions.ERXGenericRecord >>>>>> >>>>>> >>>>>> >>>>>> I think one major issue would be if ERXGenericRecord would=20 >>>>>> implement validateForSave, ..Insert, ... directly because=20 >>>>>> Validity just throws an NSValidation.ValidationException with a=20= >>>>>> customized userInfo (which is deprecated in WO5.2...) without=20 >>>>>> checking the super implementation. >>>>>> >>>>>> Regards David >>>>>> >>>>>> >>>>>> >>>>>> ------------------------------------------------------- >>>>>> This SF.NET email is sponsored by: >>>>>> SourceForge Enterprise Edition + IBM + LinuxWorld =3D Something 2=20= >>>>>> See! >>>>>> http://www.vasoftware.com >>>>>> _______________________________________________ >>>>>> Wonder-disc mailing list >>>>>> Won...@li... >>>>>> https://lists.sourceforge.net/lists/listinfo/wonder-disc >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------- >>>>> This SF.NET email is sponsored by: >>>>> SourceForge Enterprise Edition + IBM + LinuxWorld =3D Something 2=20= >>>>> See! >>>>> http://www.vasoftware.com >>>>> _______________________________________________ >>>>> Wonder-disc mailing list >>>>> Won...@li... >>>>> https://lists.sourceforge.net/lists/listinfo/wonder-disc >>>>> >>>>> >>>> ----------------------------- >>>> cluster9 GmbH >>>> David Ter=E1n >>>> >>>> J=FCdenstra=DFe 13 Mail: dav...@cl... >>>> 37073 G=F6ttingen Phone: +49 (0) 551 48 83 077 >>>> Germany Fax: +49 (0) 551 48 83 = 079 >>>> >>>> >>>> cluster9 David Ter=E1n GmbH managing director: David Ter=E1n >>>> principal office: G=F6ttingen court: HRB >>>> >>>> >>>> >>>> >>>> >>>> ------------------------------------------------------- >>>> This SF.NET email is sponsored by: >>>> SourceForge Enterprise Edition + IBM + LinuxWorld=20 >>>> http://www.vasoftware.com >>>> _______________________________________________ >>>> Wonder-disc mailing list >>>> Won...@li... >>>> https://lists.sourceforge.net/lists/listinfo/wonder-disc >>>> >>>> >>>> >>> >>> >>> >>> ------------------------------------------------------- >>> This SF.NET email is sponsored by: >>> SourceForge Enterprise Edition + IBM + LinuxWorld=20 >>> http://www.vasoftware.com >>> _______________________________________________ >>> Wonder-disc mailing list >>> Won...@li... >>> https://lists.sourceforge.net/lists/listinfo/wonder-disc >>> >>> >> ----------------------------- >> cluster9 GmbH >> David Ter=E1n >> >> J=FCdenstra=DFe 13 Mail: dav...@cl... >> 37073 G=F6ttingen Phone: +49 (0) 551 48 83 077 >> Germany Fax: +49 (0) 551 48 83 = 079 >> >> >> cluster9 David Ter=E1n GmbH managing director: David Ter=E1n >> principal office: G=F6ttingen court: HRB >> >> >> >> >> >> ------------------------------------------------------- >> This SF.NET email is sponsored by: >> SourceForge Enterprise Edition + IBM + LinuxWorld=20 >> http://www.vasoftware.com >> _______________________________________________ >> Wonder-disc mailing list >> Won...@li... >> https://lists.sourceforge.net/lists/listinfo/wonder-disc >> >> > ----------------------------- > cluster9 GmbH > David Ter=E1n > > J=FCdenstra=DFe 13 Mail: dav...@cl... > 37073 G=F6ttingen Phone: +49 (0) 551 48 83 077 > Germany Fax: +49 (0) 551 48 83 = 079 > > > cluster9 David Ter=E1n GmbH managing director: David Ter=E1n > principal office: G=F6ttingen court: HRB > > > > > > ------------------------------------------------------- > This SF.NET email is sponsored by: > SourceForge Enterprise Edition + IBM + LinuxWorld Something 2 See! > http://www.vasoftware.com > _______________________________________________ > Wonder-disc mailing list > Won...@li... > https://lists.sourceforge.net/lists/listinfo/wonder-disc |