I have not looked at yet. I have to be honest that I find re-testing a bit tedious, although I do have some standard test scripts I could try. I'll look at over the weekend at the latest.
I thought I posted a note this morning about an alternate approach: rather than modifying class.c just call the plugin entry points like getobject or definestruct directly with the arguments set up in szargs and process the results. That way we don't have to keep retesting class.c and you don't have to depend on me distributing a new version (or users of your plugins downloading a special version of class.dll). Does that make sense? --- In [email protected], "entropyreduction" <alancampbelllists+ya...@...> wrote: > > > > --- In [email protected], "brucexs" <bswitzer@> wrote: > > > > > > > > > > vec.c I've got dated 03/08/2008 > > > > > Yes, that is the latest. > > Have a look at class.c in ClassAndVec.zip, > http://tech.groups.yahoo.com/group/power-pro/sfupld?path=/0_TEMP_/AlansPluginProvisional/ > > defineSomething(...) takes care of define and definestruct. Too much > factoring for you? Lots of args, price for trying to minmise changes to your > code. > > Only behavioural change is absent (null string) minVarSize treated same as 0 > (both result in OBJECT_DEFAULT_MINVARSIZE). Can change that easily back to > the way it was. > > Runs your sample set of scripts doing person/address okay. Haven't got > around to sucking the turing machine stuff out of the help file. Fancy > posting those scripts up to files area, or somewhere. he asked lazily? >
