> Yes.. I'm a bit uncomfortable with some of the thing we want to do, > because, frankly, it really looks like Squeak, and using Squeak would > probably be easier and more sane. And if we want an operating system, > it's possible to run squeak as an os (well, some did ;-) and without > even saying it's an os, just running a livecd with a barebone linux > system that automatically launch squeak wouldn't be very complex :-) > > On the other hand, we've got this great framework (openstep), and > really, really great development tools like Gorm, and they are > available now -- that's definitely an edge.
Not only does GNUstep/Objective-C provide us with a great development environment, the fact that OS X uses a similar setup provides us with a huge amount of high-quality open source software to draw from. We're not going to get that with Smalltalk/Squeak. Using the same environment as a major software vendor also means that we have a larger pool of potential developers to draw from. And we need all the developers we can get right now. While using Smalltalk/Squeak might appear to have more of what we want, I think it would be a short term solution. We know for a fact that OS X/Cocoa is not going away any time soon, and will only keep getting better. We also already have an investment in Objective-C, and I think it will continue to provide us with a flexible development environment for quite some time: GNUstep will be better for us in the long run, even though it will mean more up front work. >> Also an advantage of this "portable solution" is, that it puts almost no >> barrier on trying the Etoile, where full-featured OS is large barier. With >> portable solution someone can start by tinkering around in his free time without >> requiring to give up his work. Nice way for attracting people. > > Frankly, at the moment we should not divert our efforts with that. I > think the best thing would be to focus *one* environment (preferably a > livecd, but could be debian, or whatever..), and provide tgz for the > courageous that want to use étoilé outside that "official" > environment. I think both of you guys are right, but I'm going to have to side with Nicolas on this one. We don't have the resources to build 3 or 4 virtual machines and do extensive testing on them. We need one, perfected solution. All open source environments I've tried are a far cry from the simplicity and usability of OS X -- based on that, I feel very strongly that we should pick one architecture and one platform and build for that. That doesn't say that we wouldn't expand out as we got more resources, but I would rather we do one really integrated project where we can focus more of our energy on features, instead of trying to make Etoile run everywhere and spreading ourselves thin just trying to gain compatibility. Also, I think a LiveCD is a great idea, because it does offer some of that portability (not as easily portable as Stefan proposes, but it's still less of a commitment then making someone do a full install just to try something out). J.
