I've been following this thread with interest. May I throw out a
few observations?
1) "No 64-bit Carbon in Leopard" is not the same as "no 64-bit
Carbon ever".
I was wondering if someone else was going to mention that. ;-)
(Of course, Eric hinted we may learn more about this after the
HIToolbox session on Thursday.) As we all know, sometimes you
have to cut features to ship,
Carbon 64-bit support in Tabby (10.6) just isn't going to cut it
for a lot of people. Obviously it would signal the continued
support of Carbon, which many of us would see as a positive, but
people who have been planning 64-bit Carbon products, just waiting
for said support to appear, are unlikely to be willing to sit
around twiddling their thumbs hoping it actually appears in 10.6.
In addition to the obvious timeframe aspect, there's the "fool me
once, shame on you; fool me twice, shame on me" effect in which
people may not trust Apple if they renege this time on their last
promise for 64-bit support.
Well actually, I could wait for 10.6. My main motivation for going
64-bit has been to maximize the performance of scientific apps
running under Intel. I'm not out of memory yet (in fact, if
anything, my memory footprint has gone down over time, thanks to some
helpful tips from Ian Ollmann and the like), so I am thankfully not
feeling the crunch in that regard.
Now, last year at WWDC, I was faced with the decision of choosing
between two possible 64-bit migration paths. The first would have
been to break my application apart into two halves (a 32-bit GUI and
a 64-bit everything else) and find a way to share memory between
them. The second would have been to concentrate on removing
deprecated APIs and have a Carbon executable ready to go 64-bit by
the time Leopard rolled around.
I decided to go with the second option, but it has been a long road
full of obstacles. Getting rid of every last vestige of Quickdraw,
for example, has been arduous, and I am still not there yet. I
probably won't make if for Leopard either, since I have had to turn
my attention away from Mac development these past few months, and I
can certainly appreciate how much effort must have gone into bringing
64-bit to the GUI levels of OS X. If it's simply a matter of 64-bit
Carbon not making the fall deadline, I can accept that. If they have
abandoned it entirely, on the other hand, I will feel rather
betrayed, since going with Plan B will have set me back about a year.
Now, I realize there will be people who say removing deprecated APIs
is worthwhile regardless of the 64-bit issue. Please keep in mind,
however, that I am essentially a one-man show as far as Mac
development is concerned, and I can't even devote all my time to
that, so it's a question of prioritization. Had I stuck with the
initial plan, I may have had 64-bit capability by now. Removing
deprecated APIs may mean my application will look nicer with its
resolution-independent graphics and what not, but really, for a
scientific package, the former would have been much more valuable.
-Ted
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Carbon-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/carbon-dev/email@hidden
