On Tue, 2005-06-07 at 14:42 +0800, James Devenish wrote: > In message <[EMAIL PROTECTED]> > on Tue, Jun 07, 2005 at 01:39:03PM +0800, Craig Ringer wrote: > > > [...] and putting a knife through developers' prospects. I imagine > > he's done a good job of doing exactly that, though. > > Really? > > Porting doesn't seem that hard. > > Is it expected that developers will just be able to cross-compile all > their apps and blindly hope for the best? If developers should be > testing and/or profiling their apps on the target platform, that doubles > the testing and doubles the hardware required. I haven't read the > porting document, but given past experiences (again, Solaris comes to > mind), it does seem a bit naive to assume that porting is no problem > (even if Apple's high-level interfaces "should" make it painless).
That's true. It will increase testing requirements, especially at first. Again, well organized developers with good test suites etc will suffer less. > Again, all the same problems as transitions from 68k to PPC, OS9 to OSX, > etc. The transitions are entirely plausible, but still painful, and > Apple makes the pain keep on coming. For example: Apple's been rather > "stupid" recently and been selling new Macs that only run Tiger, i.e. > not Panther. Oooh, great. I'm still smarting from the macs that won't run OS9 anymore (but then I'm stuck in a legacy nightmare here at the POST). You're quite right in that another big change now won't be popular. We've had: System 6 -> MacOS 7 m68k -> PPC Old World -> New World (doesn't affect most app devs) MacOS 9 -> Clasic & MacOS/X and now: MacOS/PPC -> MacOS/x86 Yeah... it doesn't help developers view it as a nice stable platform to develop for. Even so, I don't see this as likely to be a particularly bad blow - and the potential gains are IMO significant. -- Craig Ringer
