On 31 August 2012 18:55, Stéphane Ducasse <[email protected]> wrote: >>> >> >> Sure. And maybe I would have, if I didn't first have to get a Pharo >> 2.0 image and then install it and then figure out the various new APIs >> and so on and so on. So instead I wimped out and asked a clear >> question in a place where the experts on the relevant APIs would have >> a chance to see it :P > > :) > >>>> and the relevant code (I think) is >>>> >>>> transcriptStream := ((Smalltalk at: #FileSystem) workingDirectory >>>> / transcriptFileName) readStream >>>> >>> so, why you don't just wrap this thing with exception handler? >>> >>> [ transcriptStream := ((Smalltalk at: #FileSystem) >>> workingDirectory / transcriptFileName) readStream >>> >>> ] on: Error do: [ aha... somethign wrong with that file ] >>> >>> this will prevent image from bailing out to OS. :) >> >> Well, no. Catching Error is a terrible idea. I'll just touch the file >> if it doesn't already exist. >> >> Thanks for taking the trouble to answer the question, Igor. If it's >> any consolation, this means Dale & I can get our CI stuff working for >> Pharo 2.0 that must quicker. > > Pay attention that 2.0 is bleeding edge for real. > We will take some action this week to improve the situation.
I understand :) In this particular case it's probably just unfamiliarity on my part. I know, at least, that my Control library works in both Pharo 1.4 and 2.0, and while ContextPart's left alone I shouldn't need to touch code much to keep it working. It's just the CI part that's currently broken, again, probably because I haven't bothered reading the documentation properly yet :) frank > Stef > >
