Version information for Boost library
Does Boost have any policy regarding the way version information should be attached to a library ?
At Fri, 15 Oct 2010 20:31:19 -0400, Edward Diener wrote:
Does Boost have any policy regarding the way version information should be attached to a library ?
I don't even think I really understand the question. Could you be more specific? -- Dave Abrahams BoostPro Computing http://www.boostpro.com
On 10/17/2010 5:36 AM, David Abrahams wrote:
At Fri, 15 Oct 2010 20:31:19 -0400, Edward Diener wrote:
Does Boost have any policy regarding the way version information should be attached to a library ?
I don't even think I really understand the question. Could you be more specific?
For a given Boost library, or a library in the sandbox, what is the usual way of indicating to the end-user what is the version of that library ? I am thinking of nn.nn version number schemes, although nn.nn.nn.nn are more common in Windows and Linux. I understand that for a library which is part of the Boost distribution the version is essentially the distribution of Boost in which that library is included. But if Boost does pursue a means by which individual libraries can be updated and distributed separately I think that each library needs a way to tell the end-user what "version" it is. For libraries in the sandbox, is there any usual way to indicate to the end-user what "version" of the library is there ? Obviously I can just choose my own way, which I have done, but I am wondering if there is some more standard Boost way which has been discussed.
At Sun, 17 Oct 2010 15:57:16 -0400, Edward Diener wrote:
On 10/17/2010 5:36 AM, David Abrahams wrote:
At Fri, 15 Oct 2010 20:31:19 -0400, Edward Diener wrote:
Does Boost have any policy regarding the way version information should be attached to a library ?
I don't even think I really understand the question. Could you be more specific?
For a given Boost library, or a library in the sandbox, what is the usual way of indicating to the end-user what is the version of that library ? I am thinking of nn.nn version number schemes, although nn.nn.nn.nn are more common in Windows and Linux.
I understand that for a library which is part of the Boost distribution the version is essentially the distribution of Boost in which that library is included. But if Boost does pursue a means by which individual libraries can be updated and distributed separately I think that each library needs a way to tell the end-user what "version" it is.
For libraries in the sandbox, is there any usual way to indicate to the end-user what "version" of the library is there ?
Obviously I can just choose my own way, which I have done, but I am wondering if there is some more standard Boost way which has been discussed.
No, unfortunately not, but I believe we're going to have to standardize on one soon, as boost becomes modularized. -- Dave Abrahams BoostPro Computing http://www.boostpro.com
Edward Diener wrote:
On 10/17/2010 5:36 AM, David Abrahams wrote:
At Fri, 15 Oct 2010 20:31:19 -0400, Edward Diener wrote:
Does Boost have any policy regarding the way version information should be attached to a library ?
I don't even think I really understand the question. Could you be more specific?
For a given Boost library, or a library in the sandbox, what is the usual way of indicating to the end-user what is the version of that library ? I am thinking of nn.nn version number schemes, although nn.nn.nn.nn are more common in Windows and Linux.
I understand that for a library which is part of the Boost distribution the version is essentially the distribution of Boost in which that library is included. But if Boost does pursue a means by which individual libraries can be updated and distributed separately I think that each library needs a way to tell the end-user what "version" it is. For libraries in the sandbox, is there any usual way to indicate to the end-user what "version" of the library is there ?
Obviously I can just choose my own way, which I have done, but I am wondering if there is some more standard Boost way which has been discussed.
I had something to say about this back in May http://article.gmane.org/gmane.comp.lib.boost.devel/204044/match=ramey It didn't seem to generate much interest at the time. Robert Ramey
At Sun, 17 Oct 2010 21:38:14 -0800, Robert Ramey wrote:
I had something to say about this back in May
http://article.gmane.org/gmane.comp.lib.boost.devel/204044/match=ramey
It didn't seem to generate much interest at the time.
The 2nd half of that message is unreadable, though. Got another link? -- Dave Abrahams BoostPro Computing http://www.boostpro.com
On 18/10/10 14:52, David Abrahams wrote:
At Sun, 17 Oct 2010 21:38:14 -0800, Robert Ramey wrote:
I had something to say about this back in May
http://article.gmane.org/gmane.comp.lib.boost.devel/204044/match=ramey
It didn't seem to generate much interest at the time.
The 2nd half of that message is unreadable, though. Got another link?
It looks like it's broken everywhere: <http://boost.2283326.n4.nabble.com/Interlibrary-version-cchecking-td2668287.html> Best regards, -- Mateusz Loskot, http://mateusz.loskot.net Charter Member of OSGeo, http://osgeo.org Member of ACCU, http://accu.org
participants (4)
-
David Abrahams -
Edward Diener -
Mateusz Loskot -
Robert Ramey