Type traits extension & Bitfield reviews
I've just seen these libraries hanging out in the review queue for quite some time now. Considering their potential usefulness adn my personnal interest, I consider stepping up as a review manager for these. What's the formal sequence to follow to apply for such a task ?
Hi Joel, ----- Original Message ----- From: "Joel Falcou" <joel.falcou@lri.fr> To: <boost@lists.boost.org> Sent: Friday, November 05, 2010 5:41 PM Subject: [boost] Type traits extension & Bitfield reviews
I've just seen these libraries hanging out in the review queue for quite some time now. Considering their potential usefulness adn my personnal interest, I consider stepping up as a review manager for these.
happy to hear that there is someone interested by managing the review of Bitfields. Brian Bartman implemented a tuple-like interface on top of mine in its GSoC project "bit_masks" (see the sandbox SOC\2010\bit_masks), and I think that it will be better to review both library at once and even better to join the libraries in only one. I have had not enough time to review the state of last version, but I was quite satisfyed by the work at the end of the GSoC project. Maybe you can take a look. There is yet one issue, Bitfields are endian-safe only if the storage is endian safe, so we need Boost.Endian which provides an endian-safe storage. Would you be interested in been the review manager for Boost.Endian?
What's the formal sequence to follow to apply for such a task ?
I think you should contatc the review wizard John Phillips and Ronald Garcia. BTW, I don't reach to see which are the extensions of TypeTraitsExt. Where in the documentation can we find them? Best regards, Vicente
On 06/11/10 01:46, vicente.botet wrote:
I think you should contatc the review wizard John Phillips and Ronald Garcia.
BTW, I don't reach to see which are the extensions of TypeTraitsExt. Where in the documentation can we find them? IIRC there is a link in the review queue page. It's basically soem
Maybe we can sort them up as TypeTraits > Endian > Bitfield ? traits that checks if types supports ==, !=,<, > and so on.
On 06/11/10 14:08, Frédéric Bron wrote:
Yep, I may restart working on that! Thanks for pushing.
We may want to synchronize at some point. I have been developing a macro to geenrating such traits check for arbitrary member or free function. Maybe it could be agglomerated. Anyway, I'm really want them to be in ;) hence my itnerest into reviewing that instead of whining about lack of review manager :p
On 11/6/2010 9:28 AM, joel falcou wrote:
On 06/11/10 14:08, Frédéric Bron wrote:
Yep, I may restart working on that! Thanks for pushing.
We may want to synchronize at some point. I have been developing a macro to geenrating such traits check for arbitrary member or free function. Maybe it could be agglomerated. Anyway, I'm really want them to be in ;) hence my itnerest into reviewing that instead of whining about lack of review manager :p
I have also been working on a type traits introspection library but it is not documented yet. I did get your files from you that are the "introspection.zip" in the Boost vault, which I have studied but I prefer the technique in the CTL which is a bit more flexible barring a VC++ ( and gcc prior to 4.0 ) bug on which I am working on a workaround. I also have looked at Frederick Bron's code in 'is_comparable.hpp'. It seems that should be fairly easily extended to other operators, although I may not understand all the issues involved. My type traits introspection library is only interested in introspecting a type at compile time and is not interested in free functions. Nor does it touch on the operators work done by Frederick Bron, so I feel that should go forward on its own. Nonetheless I want to synchronize my efforts with others who are working in this area, probably offline. Eddie Diener
On 12/11/10 16:09, Edward Diener wrote:
I have also been working on a type traits introspection library but it is not documented yet. I did get your files from you that are the "introspection.zip" in the Boost vault, which I have studied but I prefer the technique in the CTL which is a bit more flexible barring a VC++ ( and gcc prior to 4.0 ) bug on which I am working on a workaround.
Yup, introspecton was a ad hoc fit for a need I found out so it's clearly not the proper way to do that.
I also have looked at Frederick Bron's code in 'is_comparable.hpp'. It seems that should be fairly easily extended to other operators, although I may not understand all the issues involved. Well,t hat was what mly macro set was trying to do
My type traits introspection library is only interested in introspecting a type at compile time and is not interested in free functions. Nor does it touch on the operators work done by Frederick Bron, so I feel that should go forward on its own. Nonetheless I want to synchronize my efforts with others who are working in this area, probably offline. As you wish, the free function detection is a compelte different matter (as for the namespace inclusion check that I had at some point) but it could be glad indeed to synchronize and try to gather up related techniques like these.
On 11/12/2010 10:17 AM, joel falcou wrote:
On 12/11/10 16:09, Edward Diener wrote:
I have also been working on a type traits introspection library but it is not documented yet. I did get your files from you that are the "introspection.zip" in the Boost vault, which I have studied but I prefer the technique in the CTL which is a bit more flexible barring a VC++ ( and gcc prior to 4.0 ) bug on which I am working on a workaround.
Yup, introspecton was a ad hoc fit for a need I found out so it's clearly not the proper way to do that.
I also have looked at Frederick Bron's code in 'is_comparable.hpp'. It seems that should be fairly easily extended to other operators, although I may not understand all the issues involved. Well,t hat was what mly macro set was trying to do
My type traits introspection library is only interested in introspecting a type at compile time and is not interested in free functions. Nor does it touch on the operators work done by Frederick Bron, so I feel that should go forward on its own. Nonetheless I want to synchronize my efforts with others who are working in this area, probably offline. As you wish, the free function detection is a compelte different matter (as for the namespace inclusion check that I had at some point) but it could be glad indeed to synchronize and try to gather up related techniques like these.
I will try to get my tti library documented soon and the one remaining bug workaround done. Then I will post a notice on this NG and you can see what I have done.
happy to hear that there is someone interested by managing the review of Bitfields.
Brian Bartman implemented a tuple-like interface on top of mine in its GSoC project "bit_masks" (see the sandbox SOC\2010\bit_masks), and I think that it will be better to review both library at once and even better to join the libraries in only one.
I have had not enough time to review the state of last version, but I was quite satisfyed by the work at the end of the GSoC project. Maybe you can take a look.
There is yet one issue, Bitfields are endian-safe only if the storage is endian safe, so we need Boost.Endian which provides an endian-safe storage. Would you be interested in been the review manager for Boost.Endian?
Here is a link to my SOC project http://svn.boost.org/svn/boost/sandbox/SOC/2010/bit_masks/. The current status of the project is unfinished however I plan on completing it and submitting it for formal review some time near the end of December or early January. (School and work are preventing me from having enough time to finish it currently) -- thanks, Brian Bartman
Brian Bartman wrote:
Here is a link to my SOC project http://svn.boost.org/svn/boost/sandbox/SOC/2010/bit_masks/. The current status of the project is unfinished however I plan on completing it and submitting it for formal review some time near the end of December or early January. (School and work are preventing me from having enough time to finish it currently)
How does one download from the sandbox? Is the only way to checkout the sandbox branch and update from it? What is the most convenient way to do this? Robert Ramey
On Tue, Nov 9, 2010 at 11:31 AM, Robert Ramey <ramey@rrsd.com> wrote:
Brian Bartman wrote:
Here is a link to my SOC project http://svn.boost.org/svn/boost/sandbox/SOC/2010/bit_masks/. The current status of the project is unfinished however I plan on completing it and submitting it for formal review some time near the end of December or early January. (School and work are preventing me from having enough time to finish it currently)
How does one download from the sandbox? Is the only way to checkout the sandbox branch and update from it? What is the most convenient way to do this?
I don't believe this is the right place for this question it may be better suited to the users list. There are instructions here regarding how to check out from the sandbox. https://svn.boost.org/trac/boost/
Robert Ramey
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
-- thanks, Brian Bartman
participants (7)
-
Brian Bartman -
Edward Diener -
Frédéric Bron -
joel falcou -
Joel Falcou
-
Robert Ramey -
vicente.botet