On 10/01/2013 01:35 PM, Mathias Gaunard wrote:

On 01/10/13 13:02, Stephen Kelly wrote:

...

Hi there,

Following from

http://thread.gmane.org/gmane.comp.lib.boost.devel/243923/focus=243924

I am posting a more focused and specific request to review a patch to bump compiler requirements to require support for partial template specialization.

So basically it removes support for Macintosh Programmer's Workshop and bump Sun C++ to 5.3?

Is that correct?

Not according to the commit message. Thanks, Steve.

On 10/01/2013 02:28 PM, Beman Dawes wrote:

On Tue, Oct 1, 2013 at 7:37 AM, Stephen Kelly wrote:

...

On 10/01/2013 01:35 PM, Mathias Gaunard wrote:

...

On 01/10/13 13:02, Stephen Kelly wrote:

...

Hi there,

Following from

http://thread.gmane.org/gmane.comp.lib.boost.devel/243923/focus=243924

...

I am posting a more focused and specific request to review a patch to bump compiler requirements to require support for partial template specialization. So basically it removes support for Macintosh Programmer's Workshop and bump Sun C++ to 5.3?

Is that correct? Not according to the commit message.

Hum... My interpretation of Mathias' question

I don't understand even the motivation for Mathias' question. It seems to be a restatement of what I wrote in the commit message of the commit I posted, with an added 'basically', and a mistake (5.3 instead of 5.4, which is what I wrote in my commit message). I don't know if he simply did not read my commit message at all. If he did, he also missed this:

Version 5.3 seems to have had some bugs with BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION, so bump above that in the abundance of caution.

Even if Mathias did get the version number correct, his 'question' wouldn't add anything, because all of the information he seems to be looking for is in the commit message. I don't understand his motivation for the 'question'. It is pre-answered by the commit message. I guess he just didn't read it and decided to reply anyway. Mathias, can you confirm?

about the patch is that, yes, the patch basically removes support for Macintosh Programmer's Workshop and bump Sun C++ to 5.3?

So I'm confused by your answer. Isn't that exactly what the patch does?

The patch bumps the SunPro requirement to 5.4, as I wrote in the commit message, and as a reading of the code shows: -#if __SUNPRO_CC < 0x400 +#if __SUNPRO_CC < 0x540 #error "Compiler not supported or configured - please reconfigure" #endif I did typo in the comment just above that. Fixed in the newly attached patch. Thanks, Steve.

On 10/01/2013 04:14 PM, Mathias Gaunard wrote:

On 01/10/13 14:49, Stephen Kelly wrote:

...

I don't understand even the motivation for Mathias' question.

It seems to be a restatement of what I wrote in the commit message of the commit I posted, with an added 'basically', and a mistake (5.3 instead of 5.4, which is what I wrote in my commit message).

I don't know if he simply did not read my commit message at all. If he did, he also missed this:

My bad, I went through the diff and looked at what it changed, but didn't see the commit message at the top.

Yes, that's unfortunate. It caused a lot of confusion and mails which are best avoided, distracting attention away from the actual point of the email.

Anyway I don't think dropping those compilers is a problem. It would be interesting to see how much of Boost can be simplified by removing BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION everywhere.

Now that there's no confusion anymore (I hope), let's see if anyone else has anything to say too. Thanks, Steve.

On 10/01/2013 06:25 PM, Robert Ramey wrote:

Stephen Kelly wrote:

...

...

Anyway I don't think dropping those compilers is a problem. It would be interesting to see how much of Boost can be simplified by removing BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION everywhere. wouldn't this break libraries which refer to BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION ?

No. Think deeper about it, and maybe grep for it. Thanks, Steve.