Re: [Boost-users] [Context] Mac OS X support?

Nat Linden wrote:
Anyway -- it seems likely to me that Boost.Context will support Mac OS X 10.5+ "out of the box." I intend to check that.
If it's important for Boost.Context to support Mac OS X <= 10.4, it might be fairly straightforward to implement the Boost.Context API on top of Pth. Reference [2] contains my macro-based attempt to establish equivalences between the ucontext API and the Pth API.
However, at this moment in history, one must ask how important it still is to support Mac OS X <= 10.4?
I'm completely ignorant about today's marketshare of Mac OS X 10.4, but I can tell you I'd rather like to see it supported for a while. Personally, I own an i686 machine with Mac OS X 10.6 and a PPC machine with Mac OS X 10.4. The latter is quite old and won't be able to pull anything more modern than 10.4. Therefore my only option for PPC testing is under Mac OS X 10.4. Since Apple's switch to x86 is historically close to the release of 10.5 this is probably true for more developers. Another, perhaps more general reason to support 10.4 is that, compared to both the older and the newer versions of Mac OS X it's still quite usable. I can't very well imagine anyone still using 10.3 but 10.4 doesn't seem strange. My (vague) impression is that some software suppliers still attempt to support 10.4 for this reason. Of course, both of these "arguments" could be easily overridden by some decent marketshare statistics. (Completely unrelated: hi, I'm Julian. This is my first post to the Boost users list.) (About equally related: what is the best way to reply to an existing thread? The instructions in the digest header don't seem to care much as long as you make sure that the subject field is correct. The Boost mailing lists website on the other hand suggests that one should rather not reply to digests at all and rather use per-message mailing or reply through the GMane web interface. Any opinions on this matter would be greatly appreciated.)

On Wed, Mar 9, 2011 at 8:46 PM, Julian Gonggrijp
Nat Linden wrote:
However, at this moment in history, one must ask how important it still is to support Mac OS X <= 10.4?
I'm completely ignorant about today's marketshare of Mac OS X 10.4, but I can tell you I'd rather like to see it supported for a while. Personally, I own an i686 machine with Mac OS X 10.6 and a PPC machine with Mac OS X 10.4. The latter is quite old and won't be able to pull anything more modern than 10.4. Therefore my only option for PPC testing is under Mac OS X 10.4.
Mmm, I have no hard data either (anyone?) -- but it seems to me that the importance of testing on OS X 10.4 depends not only on how many such machines are still in use, but on how many of those machines the owners are still installing new software. Their reluctance to upgrade the OS might suggest a corresponding lack of enthusiasm about buying/upgrading other applications. Not trying to incite a flame war; I'm honestly interested in the answer, willing to be convinced that OS X 10.4 is in fact an important target platform for the forthcoming Boost.Context. At the moment, though, I'm inclined to doubt it.

On 3/10/11 11:44 AM, in article AANLkTinnD32N=EqjKtrj3QmEFhOuVhPSR_zaqVHJ0PB1@mail.gmail.com, "Nat Linden" wrote:
On Wed, Mar 9, 2011 at 8:46 PM, Julian Gonggrijp
wrote: Nat Linden wrote:
However, at this moment in history, one must ask how important it still is to support Mac OS X <= 10.4?
I'm completely ignorant about today's marketshare of Mac OS X 10.4, but I can tell you I'd rather like to see it supported for a while. Personally, I own an i686 machine with Mac OS X 10.6 and a PPC machine with Mac OS X 10.4. The latter is quite old and won't be able to pull anything more modern than 10.4. Therefore my only option for PPC testing is under Mac OS X 10.4.
Mmm, I have no hard data either (anyone?) -- but it seems to me that the importance of testing on OS X 10.4 depends not only on how many such machines are still in use, but on how many of those machines the owners are still installing new software.
Their reluctance to upgrade the OS might suggest a corresponding lack of enthusiasm about buying/upgrading other applications.
Not trying to incite a flame war; I'm honestly interested in the answer, willing to be convinced that OS X 10.4 is in fact an important target platform for the forthcoming Boost.Context. At the moment, though, I'm inclined to doubt it.
I am an avid "Mac Guy" but even I would really think twice about trying to support OS X 10.4.11. For one thing those who develop will need testing machines which are not really easy to come by in working order. Then you have to make sure that all the boost-build infrastructure works on 10.4. I would say 10.5 is the farthest back you should probably go. Again, no hard numbers but those hanging onto their 10.4 machines are in places that either they can't upgrade them (institutions) or refuse to for what ever reason. Mike Jackson

On Mar 10, 2011, at 8:54 AM, Michael Jackson wrote:
On 3/10/11 11:44 AM, in article AANLkTinnD32N=EqjKtrj3QmEFhOuVhPSR_zaqVHJ0PB1@mail.gmail.com, "Nat Linden" wrote:
On Wed, Mar 9, 2011 at 8:46 PM, Julian Gonggrijp
wrote: Nat Linden wrote:
However, at this moment in history, one must ask how important it still is to support Mac OS X <= 10.4?
I'm completely ignorant about today's marketshare of Mac OS X 10.4, but I can tell you I'd rather like to see it supported for a while. Personally, I own an i686 machine with Mac OS X 10.6 and a PPC machine with Mac OS X 10.4. The latter is quite old and won't be able to pull anything more modern than 10.4. Therefore my only option for PPC testing is under Mac OS X 10.4.
Mmm, I have no hard data either (anyone?) -- but it seems to me that the importance of testing on OS X 10.4 depends not only on how many such machines are still in use, but on how many of those machines the owners are still installing new software.
Their reluctance to upgrade the OS might suggest a corresponding lack of enthusiasm about buying/upgrading other applications.
Not trying to incite a flame war; I'm honestly interested in the answer, willing to be convinced that OS X 10.4 is in fact an important target platform for the forthcoming Boost.Context. At the moment, though, I'm inclined to doubt it.
I am an avid "Mac Guy" but even I would really think twice about trying to support OS X 10.4.11. For one thing those who develop will need testing machines which are not really easy to come by in working order. Then you have to make sure that all the boost-build infrastructure works on 10.4. I would say 10.5 is the farthest back you should probably go. Again, no hard numbers but those hanging onto their 10.4 machines are in places that either they can't upgrade them (institutions) or refuse to for what ever reason.
I would like to advocate in favor of continued support for OS X 10.4. We use the boost libraries to develop tools that our customers use to develop products for end users. We develop for both Mac and Windows, and we have hundreds of thousands of end users who use Macs. Our customers tell us that continued 10.4 support is still required. As with others, I have no hard numbers, but I know what our customers tell us they need. - Rush

On 10 March 2011 11:20, Rush Manbert
I would like to advocate in favor of continued support for OS X 10.4. We use the boost libraries to develop tools that our customers use to develop products for end users. We develop for both Mac and Windows, and we have hundreds of thousands of end users who use Macs. Our customers tell us that continued 10.4 support is still required. As with others, I have no hard numbers, but I know what our customers tell us they need.
Customers *always* say this, because it doesn't cost them anything extra even though it increases your development/support costs. So if you want numbers, charge your customers extra if they want to keep using a sufficiently old system. :-) Another way is to ship the software for 10.4 as a separate product and see how many people download/buy it. Even Firefox 4 is dropping support for 10.4. Unless someone is volunteering to take on the burden of supporting it specifically, IMHO I think we we should drop it. Note: besides outdated hardware, 10.4 is the last system which supported the OS 9 emulator. -- Nevin ":-)" Liber mailto:nevin@eviloverlord.com (847) 691-1404

Nat Linden wrote:
However, at this moment in history, one must ask how important it still is to support Mac OS X <= 10.4?
I'm completely ignorant about today's marketshare of Mac OS X 10.4, but I can tell you I'd rather like to see it supported for a while. Personally, I own an i686 machine with Mac OS X 10.6 and a PPC machine with Mac OS X 10.4. The latter is quite old and won't be able to pull anything more modern than 10.4. Therefore my only option for PPC testing is under Mac OS X 10.4.
Mmm, I have no hard data either (anyone?) -- but it seems to me that the importance of testing on OS X 10.4 depends not only on how many such machines are still in use, but on how many of those machines the owners are still installing new software.
I agree completely to your line of reasoning. I'm an example of a 10.4 machine owner that still installs new software, but again, one person is a small sample size. -Julian

Just to let you know that here it is another Mac OS X 10.4 boost user.
Mauricio.
On Thu, Mar 10, 2011 at 4:42 PM, Julian Gonggrijp
Nat Linden wrote:
However, at this moment in history, one must ask how important it still is to support Mac OS X <= 10.4?
I'm completely ignorant about today's marketshare of Mac OS X 10.4, but I can tell you I'd rather like to see it supported for a while. Personally, I own an i686 machine with Mac OS X 10.6 and a PPC machine with Mac OS X 10.4. The latter is quite old and won't be able to pull anything more modern than 10.4. Therefore my only option for PPC testing is under Mac OS X 10.4.
Mmm, I have no hard data either (anyone?) -- but it seems to me that the importance of testing on OS X 10.4 depends not only on how many such machines are still in use, but on how many of those machines the owners are still installing new software.
I agree completely to your line of reasoning. I'm an example of a 10.4 machine owner that still installs new software, but again, one person is a small sample size.
-Julian _______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users
participants (6)
-
Julian Gonggrijp
-
Mauricio Gomes
-
Michael Jackson
-
Nat Linden
-
Nevin Liber
-
Rush Manbert