site_archiver@lists.apple.com Delivered-To: darwin-dev@lists.apple.com Ernest Prabhakar writes:
Hello everyone,
Michael wrote:
But I think it could be so much better.
For the record, I (and many others at Apple) agree that it _could_ be "so much better." Unfortunately, the solutions aren't always as simple as they might appear from the outside.
There are probably many lawyers involved, which tends to complicate things greatly. Even with all the limitations, publishing the majority of the xnu source, and the source for many of the IOKit families is very helpful to my work as a driver developer. Many times, the source is much more helpful that the documentation, and can go a long way towards making up for poor or missing documentation. My main wish is that Apple would open their development process a little more, and make the Darwin source available for development versions of MacOSX. Darwin source snapshots (and they wouldn't even have to be buildable) roughly corresponding to the various tiger seeds would have helped me a great deal. Also, the ability to discuss some of the problems I was encountering on the darwin lists would have probably saved Apple employee time, and definitely would have saved my time. Just to name one concrete example, when testing a piece of 64-bit code in a tiger seed, I discovered a problem in a 64-bit area of libSystem which was fixed by the release. (This actually happened 3 times for 2 different issues: Bug #3944194, #3973045 and #3994249.) To verify that it wasn't my code, I downloaded some example code and verified it crashed the same way, yet ran in 32-bit mode. I then opened a bugreporter ticket. Between debugging my specific code, finding the & building the example code, and filing the radar request, I wasted roughly 8-10 hours for 3944194, and longer for 3973045. An Apple tech support person had to look at each bug, and I was still left with the nagging feeling that it may be my own problem because developers generally get no feedback from bug reports. Had the Darwin 8.0 codebase been open and discussable on these lists, I'm sure I either would have seen a note that "foo doesn't work in 64-bit mode", or I could have gotten a quick answer to a "I'm seeing foo failing in 64-bit mode" query on darwin-dev@, or I'd have been able to look at the source to verify the bug was in the Darwin code and would have been able to make a more intelligent bug report. In any case, Apple and I would have both saved time and money. Cheers, Drew _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.appl... This email sent to site_archiver@lists.apple.com
participants (1)
-
Andrew Gallatin