RE: Brain Dump of Towards FIT 1.0 OpenSpace session
"Charlie Poole" <[email protected]> Fri, 22 Aug 2003 14:03:03 -0700
| Newsgroups | gmane.comp.programming.tools.fit.devel |
|---|---|
| Message-ID | <[email protected]> |
Thanks John... that clarifies it for me. I agree with your take on the tradeoffs. For now, it seems unnecssary. For the future, it would be nice if we could follow Michael's lead and keep the framework so simple that it's easy to port. If it comes down to having a central single code base, I'll argue that Java isn't appropriate. But for now, I'll shut up. :-) Charlie Poole [email protected] www.pooleconsulting.com www.charliepoole.org > -----Original Message----- > From: [email protected] > [mailto:[email protected]]On Behalf Of > [email protected] > Sent: Friday, August 22, 2003 12:09 PM > To: [email protected] > Subject: Re: [Fit-dev] Brain Dump of Towards FIT 1.0 OpenSpace session > > > Well, the basic idea is simply to separate FIT into one > Java component that does most of the work, and a > series of smaller, language specific components that > take care of the details of interfacing with the fixtures > and the SUT. > > The single code base is the FIT Java component. > It's really only feasible if the main component grows > large enough to be a pain to port, and the language > dependent components are small enough. At the > moment, I don't really see that. I may continue > thinking about it for a while, but unless what we > want for 1.0 causes the base to grow quite a bit, > I don't think there's any advantage in it. > > Clearly, Parse belongs in the main component. > TypeAdapter belongs in the language dependent > component, and Fixture straddles them. > > John Roth > > ----- Original Message ----- > From: "Charlie Poole" > <[email protected]> > To: "FIT Developers" <[email protected]> > Sent: Friday, August 22, 2003 1:28 PM > Subject: RE: [Fit-dev] Brain Dump of Towards FIT 1.0 OpenSpace session > > > > John, > > > > Can you take this a bit further? What would you put in the single > > codebase? What language would it be in. My own take is that something > > like ansi C would be needed if it were to work in all environments, and > > that the resulting feature set would therefore stop short of anything > > depending upon reflection. > > > > But I may not be understanding where you're going. > > > > Charlie > > > > > -----Original Message----- > > > From: [email protected] > > > [mailto:[email protected]]On Behalf Of > > > [email protected] > > > Sent: Friday, August 22, 2003 3:46 AM > > > To: [email protected] > > > Subject: Re: [Fit-dev] Brain Dump of Towards FIT 1.0 OpenSpace session > > > > > > > > > > > > ----- Original Message ----- > > > From: "Charlie Poole" > > > <[email protected]> > > > To: "FIT Developers" > <[email protected]> > > > Sent: Friday, August 22, 2003 12:21 AM > > > Subject: RE: [Fit-dev] Brain Dump of Towards FIT 1.0 OpenSpace session > > > > > > > > > > I think you need native implementations for /some/ applications but > > > > probably not for all or most. But having a native implementation is > > > > probably a big factor in getting many people to adopt Fit. > > > > > > > > Besides, this is open source and people like to tinker with it. > > > > > > One of the things I like about separating things is that there's one > > > code base to put most new features in. I think the native adapters > > > would be significantly more stable. They're also the place where there > > > would be the most variation. > > > > > > John Roth > > > > > > > > Charlie > > > > > > > > > -----Original Message----- > > > > > From: [email protected] > > > > > [mailto:[email protected]]On Behalf Of Jim Little > > > > > Sent: Thursday, August 21, 2003 5:05 PM > > > > > To: 'FIT Developers' > > > > > Subject: RE: [Fit-dev] Brain Dump of Towards FIT 1.0 OpenSpace > session > > > > > > > > > > > > > > > From: Al Chou > > > > > > Rather than ... having FIT re-implemented in numerous > > > > > > languages, why not have one core FIT that allows fixtures > > > > > > to be written in whatever language you choose? > > > > > > > > > > That's an interesting idea. But would it work? Consider that we > need > > > > > to be able to write fixtures that interface with the system > > > under test. > > > > > Those fixtures need to be in the programmer's native > > > language, as he may > > > > > not know Java. We also want (I presume) to make Fit inviting > > > for people > > > > > to adopt... can we use this approach and still do that? > > > > > > > > > > Jim > > > > > > > > > > > > > > _______________________________________________ > > > Fit-dev mailing list > > > [email protected] > > > http://lists.freepan.org/mailman/listinfo.cgi/fit-dev > > > > > > > > > > > _______________________________________________ > > Fit-dev mailing list > > [email protected] > > http://lists.freepan.org/mailman/listinfo.cgi/fit-dev > > > _______________________________________________ > Fit-dev mailing list > [email protected] > http://lists.freepan.org/mailman/listinfo.cgi/fit-dev >