Linked by Eugenia Loli-Queru on Sat 30th Jul 2005 02:19 UTC, submitted by wordtech
According to an announcement posted on the Carbon developer's mailing list, Metrowerks announced at AdHoc that the forthcoming release of CodeWarrior 10 will be the last for OS X. This isn't surprising given that Apple is transitioning to Intel chips and Metrowerks has exited the Intel market, but it's still the end of an era. CodeWarrior literally saved Apple's bacon during the transition to PowerPC in the early 1990s by shipping the first working set of developer tools for the new platform. And since then CodeWarrior has been the main toolkit for commercial development on the Mac (especially pre-XCode).
Thread beginning with comment 11276
To read all comments associated with this story, please click here.
To read all comments associated with this story, please click here.
Member since:
Member since:2005-07-06
but it is very buggy, slow, and unpolished at the same time: Excellent FUD!
Care to elaborate why you think that is FUD?
I am currently porting a large application from CW from Xcode, and I fully agree with the original author's opinion: Xcode is buggy and slow and the workflow is counter-intuitive and unnecessary complicated.
Member since:2005-06-30
Recent Original Stories
Recent Comments
Headlines
Random Comments
Random Stories
Random OS Link
Member since:
I've been a user of both CodeWarrior (version 9) and Xcode, and CodeWarrior definitely has/had advantages over Xcode.
Sure, Xcode is free, but it's also bloated, slow and buggy. CodeWarrior is simpler, MUCH faster - in both compiler time (though that can be blamed on gcc), and interface snappyness - this is Apple's fault in Xcode. It also does some things _much_ better than Xcode, such as being able to disassemble the code and see relative hex addresses next to lines of code, for example - which makes going from hex addressed in crashlogs to the code that made the crash so much easier (as opposed to Apple's recommended way is to use a bunch of complicated command line tools straight from their Technical Note).
Another example is that project files from Xcode are not compatible with each other, and that Xcode versions are tied to the OS versions - that is, you can't run Xcode 2 (tiger's) on Panther, and same for Xcode 1.0 on Jaguar (where the dev tool was called Project Builder). This makes it hell for Open Source projects, which can't guarantee that all developers run the same OS. In one such project, we have like 4 versions of the project files, each for a different version of Apple's dev tools, to make sure all the developers can contribute. When a file is added to the project, every project file has to be updated with the change, its a major pain in the ass. Something like CodeWarrior 9, however, supports Mac OS X 10.1 to 10.4, making a single project file be able to be used regardless of what OS version developers use (though obviously CW is not the perfect choice for developing an OOS project, since CW is a commercial app, but that fact still doesn't make Xcode any better).
It's a shame - but it seems its CodeWarrior people were just out of luck. They figured PPC is here to stay (a reasonable assumption, especially if Freescale, their parent company is interested in making PPC better and advancing it). So, they sell their x86 compiler (which they owned and was actually pretty good!), to Symbian. And then Apple announces its switching to x86. Go figure.
With CodeWarrior, there was at least some competition to Xcode (although not much since Apple undercut CW by giving Xcode away), so that urged Apple to improve Xcode to make it more attracting to people migrating from CW. Unfortunately, now there's no longer a need for them to do this: Xcode will be the only choice for Cocoa and Carbon apps (not counting stuff like RealBasic). That's too bad, hopefully it won't stagnate even more, and actually continue to improve - there's great potential in Xcode and it does some things well, but it is very buggy, slow, and unpolished at the same time, and Apple needs to address that.