[1.54.0][release] Provisional schedule

newer
[spirit] [lex] semantic actions to...

Daniel James

6 May 2013 6 May '13

10:53 a.m.

Hello everyone, I've put up a provisional schedule for the next release on the calendar. You can see it at: http://www.boost.org/development/ It targets a 1st July release date, and follows the usual schedule relative to that date. The deadlines for the release branch are pretty soon, particularly the deadline for new libraries which is just a week away. Also, the deadline for new features is just after C++Now. If any of this is a problem let us know now. The release after that will probably target November 1st (as normal), which will give a 4 month release cycle. If the git modularization is considered ready, this will allow a little extra time to deal with any disruption caused (possibly not enough...). How does that sound? Any thoughts? Daniel

Show replies by date

Edward Diener

6 May 6 May

12:18 p.m.

On 5/6/2013 6:53 AM, Daniel James wrote:

...

Hello everyone,

I've put up a provisional schedule for the next release on the calendar. You can see it at:

http://www.boost.org/development/

It targets a 1st July release date, and follows the usual schedule relative to that date. The deadlines for the release branch are pretty soon, particularly the deadline for new libraries which is just a week away.

If I want to get TTI into 1.54, what are the steps I must accomplish by next Monday ? Edward Diener

Daniel James

1:08 p.m.

On 6 May 2013 13:18, Edward Diener wrote:

...

On 5/6/2013 6:53 AM, Daniel James wrote:

...
Hello everyone,

I've put up a provisional schedule for the next release on the calendar. You can see it at:

http://www.boost.org/development/

It targets a 1st July release date, and follows the usual schedule relative to that date. The deadlines for the release branch are pretty soon, particularly the deadline for new libraries which is just a week away.

If I want to get TTI into 1.54, what are the steps I must accomplish by next Monday ?

The first thing to do is add it to the unit tests (by adding it to status/Jamfile.v2), and get them looking decent. I don't know if a week is enough time to do that. There's often a lot of waiting around for the test results to update. I know it's quite short notice, so I think we should be a bit lenient about this. Btw. when I try building the documentation, it seems to be missing a header. I get this error: Building the tti docs with automatic index generation disabled. Try building with --enable-index. error: Unable to find file or target named error: '/home/daniel/boost/git/libs/tti/doc/../../../boost/tti/has_member_function_with_sig.hpp' error: referred to from project at error: '.'

Edward Diener

4:20 p.m.

On 5/6/2013 9:08 AM, Daniel James wrote:

...

On 6 May 2013 13:18, Edward Diener wrote:

...
On 5/6/2013 6:53 AM, Daniel James wrote:

...
Hello everyone,

I've put up a provisional schedule for the next release on the calendar. You can see it at:

http://www.boost.org/development/

It targets a 1st July release date, and follows the usual schedule relative to that date. The deadlines for the release branch are pretty soon, particularly the deadline for new libraries which is just a week away.

If I want to get TTI into 1.54, what are the steps I must accomplish by next Monday ?

The first thing to do is add it to the unit tests (by adding it to status/Jamfile.v2), and get them looking decent.

I will work on it. If I have any Boost build/bjam questions I will ask.

...

I don't know if a week is enough time to do that. There's often a lot of waiting around for the test results to update. I know it's quite short notice, so I think we should be a bit lenient about this.

Btw. when I try building the documentation, it seems to be missing a header. I get this error:

Building the tti docs with automatic index generation disabled. Try building with --enable-index. error: Unable to find file or target named error: '/home/daniel/boost/git/libs/tti/doc/../../../boost/tti/has_member_function_with_sig.hpp' error: referred to from project at error: '.'

Thanks ! Fixed.

Edward Diener

6:15 p.m.

On 5/6/2013 9:08 AM, Daniel James wrote:

...

On 6 May 2013 13:18, Edward Diener wrote:

...
On 5/6/2013 6:53 AM, Daniel James wrote:

...
Hello everyone,

I've put up a provisional schedule for the next release on the calendar. You can see it at:

http://www.boost.org/development/

It targets a 1st July release date, and follows the usual schedule relative to that date. The deadlines for the release branch are pretty soon, particularly the deadline for new libraries which is just a week away.

If I want to get TTI into 1.54, what are the steps I must accomplish by next Monday ?

The first thing to do is add it to the unit tests (by adding it to status/Jamfile.v2), and get them looking decent.

A number of unit tests for TTI depend on the compiler supporting variadic macros. I can add all the TTI unit tests to status/Jamfile.v2 easily enough but for any compiler not supporting variadic macros that subset of the tests will fail. Is that OK, or do I have to do something different for this particular case in status/Jamfile.v2 ?

Daniel James

8:01 p.m.

On 6 May 2013 19:15, Edward Diener wrote:

...

On 5/6/2013 9:08 AM, Daniel James wrote:

...
The first thing to do is add it to the unit tests (by adding it to status/Jamfile.v2), and get them looking decent.

A number of unit tests for TTI depend on the compiler supporting variadic macros. I can add all the TTI unit tests to status/Jamfile.v2 easily enough but for any compiler not supporting variadic macros that subset of the tests will fail. Is that OK, or do I have to do something different for this particular case in status/Jamfile.v2 ?

You can specify tests that are expected to fail in status/explicit-failures-markup.xml, but you can do that after adding the tests. Which compilers will it fail for?

Edward Diener

9:04 p.m.

On 5/6/2013 4:01 PM, Daniel James wrote:

...

On 6 May 2013 19:15, Edward Diener wrote:

...
On 5/6/2013 9:08 AM, Daniel James wrote:

...
The first thing to do is add it to the unit tests (by adding it to status/Jamfile.v2), and get them looking decent.

A number of unit tests for TTI depend on the compiler supporting variadic macros. I can add all the TTI unit tests to status/Jamfile.v2 easily enough but for any compiler not supporting variadic macros that subset of the tests will fail. Is that OK, or do I have to do something different for this particular case in status/Jamfile.v2 ?

You can specify tests that are expected to fail in status/explicit-failures-markup.xml, but you can do that after adding the tests. Which compilers will it fail for?

The tests which assume support for variadic macros will fail for any compiler/version which simply does not recognize the syntax. There aren't too many around but I do currently support some pretty early versions of gcc ( prior to 4+ ) which do not support variadic macros. Of course the test matrix will probably not even test these early versions. The equivalent non-variadic versions should not have any problems. I am also still working on support for Intel and Sun C++ compilers on Linux and so have not determined yet how much of TTI will work with them.

Edward Diener

7 May 7 May

3:01 a.m.

On 05/06/2013 09:08 AM, Daniel James wrote:

...

On 6 May 2013 13:18, Edward Diener wrote:

...
On 5/6/2013 6:53 AM, Daniel James wrote:

...
Hello everyone,

I've put up a provisional schedule for the next release on the calendar. You can see it at:

http://www.boost.org/development/

It targets a 1st July release date, and follows the usual schedule relative to that date. The deadlines for the release branch are pretty soon, particularly the deadline for new libraries which is just a week away.

If I want to get TTI into 1.54, what are the steps I must accomplish by next Monday ?

The first thing to do is add it to the unit tests (by adding it to status/Jamfile.v2), and get them looking decent.

I added the tti tests to the status/Jamfile.v2. I assume the results should be showing up online when testers run the tests. I will follow those results and also continue my own testing for compilers which I have not tested yet ( Intel, Sun ). If the results of the tests are satisfactory, what are the subsequent steps I need to take to get TTI ready for the next release ? Is this described online somewhere ?

Daniel James

8 May 8 May

10:50 p.m.

On 7 May 2013 04:01, Edward Diener wrote:

...

I added the tti tests to the status/Jamfile.v2. I assume the results should be showing up online when testers run the tests. I will follow those results and also continue my own testing for compilers which I have not tested yet ( Intel, Sun ).

The test results haven't updated for a few days, I don't know why. I sent a message to the testing list to see if they can do anything.

...

If the results of the tests are satisfactory, what are the subsequent steps I need to take to get TTI ready for the next release ? Is this described online somewhere ?

You can see a list of the things to check for here: https://svn.boost.org/trac/boost/wiki/ReleasePractices/ManagerCheckList#NewL... Although, some are things that I do. I added TTI to the documentation build, I think it's okay. You can see the results at: http://boost-sandbox.sourceforge.net/libs/log/doc/html/index.html Can you check that they look okay?

Andrey Semashev

11:02 p.m.

On Wednesday 08 May 2013 23:50:30 Daniel James wrote:

...

On 7 May 2013 04:01, Edward Diener wrote:

...
I added the tti tests to the status/Jamfile.v2. I assume the results should be showing up online when testers run the tests. I will follow those results and also continue my own testing for compilers which I have not tested yet ( Intel, Sun ).

The test results haven't updated for a few days, I don't know why. I sent a message to the testing list to see if they can do anything.

I was waiting for tests to cycle to request for permission to merge Boost.Log to release. Is it reasonable to go on waiting or just perform the merge now? I ran tests locally on msvc and gcc, so I don't expect any failures.

Daniel James

11:11 p.m.

On 9 May 2013 00:02, Andrey Semashev wrote:

...

I was waiting for tests to cycle to request for permission to merge Boost.Log to release. Is it reasonable to go on waiting or just perform the merge now? I ran tests locally on msvc and gcc, so I don't expect any failures.

Please wait. We'll allow for a delay because the test results aren't running, as well as the short notice for the deadline. I also need to check that everything else is okay. Just to be clear, have you tested with mingw gcc? That seems to be the biggest problem.

Andrey Semashev

9 May 9 May

8:04 a.m.

On Thursday 09 May 2013 00:11:11 Daniel James wrote:

...

On 9 May 2013 00:02, Andrey Semashev wrote:

...
I was waiting for tests to cycle to request for permission to merge Boost.Log to release. Is it reasonable to go on waiting or just perform the merge now? I ran tests locally on msvc and gcc, so I don't expect any failures. Please wait. We'll allow for a delay because the test results aren't running, as well as the short notice for the deadline. I also need to check that everything else is okay.

Ok.

...

Just to be clear, have you tested with mingw gcc? That seems to be the biggest problem.

Yes, I tested MinGW gcc 4.7.2, the tests went fine.

Andrey Semashev

12 May 12 May

9:17 a.m.

On Thursday 09 May 2013 00:11:11 Daniel James wrote:

...

On 9 May 2013 00:02, Andrey Semashev wrote:

...
I was waiting for tests to cycle to request for permission to merge Boost.Log to release. Is it reasonable to go on waiting or just perform the merge now? I ran tests locally on msvc and gcc, so I don't expect any failures. Please wait. We'll allow for a delay because the test results aren't running, as well as the short notice for the deadline. I also need to check that everything else is okay.

Just wanted to ask about the delay you were talking about. What is the updated schedule of the release and deadlines? The current schedule states that the deadline for new libs is tomorrow.

Daniel James

10:32 a.m.

On 12 May 2013 10:17, Andrey Semashev wrote:

...

Just wanted to ask about the delay you were talking about. What is the updated schedule of the release and deadlines? The current schedule states that the deadline for new libs is tomorrow.

I'll leave the deadline as it is, but tti and log have extra time. Say, a week (but hopefully you won't need a week...). If I extend the deadline then there's the possibility of another library popping up on Wednesday, which I don't want to deal with.

Edward Diener

11:31 a.m.

On 05/12/2013 06:32 AM, Daniel James wrote:

...

On 12 May 2013 10:17, Andrey Semashev wrote:

...
Just wanted to ask about the delay you were talking about. What is the updated schedule of the release and deadlines? The current schedule states that the deadline for new libs is tomorrow.

I'll leave the deadline as it is, but tti and log have extra time. Say, a week (but hopefully you won't need a week...). If I extend the deadline then there's the possibility of another library popping up on Wednesday, which I don't want to deal with.

Another week is more than enough for tti, but it is appreciated. Even if I cannot find workarounds for Intel C++ problems or Sun C++ problems supporting gcc, clang, and vc++ completely was always my goal. Are purely documentation changes allowed after the deadline you set ?

Daniel James

11:41 a.m.

On 12 May 2013 12:31, Edward Diener wrote:

...

On 05/12/2013 06:32 AM, Daniel James wrote:

...
On 12 May 2013 10:17, Andrey Semashev wrote:

...
Just wanted to ask about the delay you were talking about. What is the updated schedule of the release and deadlines? The current schedule states that the deadline for new libs is tomorrow.

I'll leave the deadline as it is, but tti and log have extra time. Say, a week (but hopefully you won't need a week...). If I extend the deadline then there's the possibility of another library popping up on Wednesday, which I don't want to deal with.

Another week is more than enough for tti, but it is appreciated. Even if I cannot find workarounds for Intel C++ problems or Sun C++ problems supporting gcc, clang, and vc++ completely was always my goal.

Are purely documentation changes allowed after the deadline you set ?

Yes. Once in release, you just follow the same process as everything else. Documentation changes are always allowed apart from when the release branch is frozen to build a release.

Edward Diener

9 May 9 May

2:30 a.m.

On 5/8/2013 6:50 PM, Daniel James wrote:

...

On 7 May 2013 04:01, Edward Diener wrote:

...
I added the tti tests to the status/Jamfile.v2. I assume the results should be showing up online when testers run the tests. I will follow those results and also continue my own testing for compilers which I have not tested yet ( Intel, Sun ).

The test results haven't updated for a few days, I don't know why. I sent a message to the testing list to see if they can do anything.

...
If the results of the tests are satisfactory, what are the subsequent steps I need to take to get TTI ready for the next release ? Is this described online somewhere ?

You can see a list of the things to check for here:

https://svn.boost.org/trac/boost/wiki/ReleasePractices/ManagerCheckList#NewL...

Although, some are things that I do.

It does say "Release Manager's Checklist" so I will assume you will tell me what I am to do if you need me to do anything in particular.

...

I added TTI to the documentation build, I think it's okay. You can see the results at:

http://boost-sandbox.sourceforge.net/libs/log/doc/html/index.html

That's for Boost Log, not TTI. I believe Boost Log is also being added to 1.54.

Daniel James

8:02 a.m.

On 9 May 2013 03:30, Edward Diener wrote:

...

On 5/8/2013 6:50 PM, Daniel James wrote:

...
On 7 May 2013 04:01, Edward Diener wrote:

You can see a list of the things to check for here:

https://svn.boost.org/trac/boost/wiki/ReleasePractices/ManagerCheckList#NewL...

Although, some are things that I do.

It does say "Release Manager's Checklist" so I will assume you will tell me what I am to do if you need me to do anything in particular.

Okay, you can add your libraries to libs/libraries.htm, libs/libraries.txt. And libs/tti/index.html (you can just copy log's and do a search and replace). The inspect report is at (it's also linked from the testing page): http://boost.cowic.de/rc/docs-inspect-trunk.html#tti If you add license info to tti.qbk that should fix license issues. You can probably also remove copying the image and css files from 'doc/src' into your documentation, it should use them in their original location now it's included in boost. That leaves adding license and copyright info to build.txt, and renaming 'check' in some headers (I expect that's possible because they're all in detail headers so probably not part of the public interface).

...

...
I added TTI to the documentation build, I think it's okay. You can see the results at:

http://boost-sandbox.sourceforge.net/libs/log/doc/html/index.html

That's for Boost Log, not TTI. I believe Boost Log is also being added to 1.54.

http://boost-sandbox.sourceforge.net/libs/tti/doc/html/index.html

Edward Diener

12 May 12 May

7:35 a.m.

On 5/9/2013 4:02 AM, Daniel James wrote:

...

On 9 May 2013 03:30, Edward Diener wrote:

...
On 5/8/2013 6:50 PM, Daniel James wrote:

...
On 7 May 2013 04:01, Edward Diener wrote:

You can see a list of the things to check for here:

https://svn.boost.org/trac/boost/wiki/ReleasePractices/ManagerCheckList#NewL...

Although, some are things that I do.

It does say "Release Manager's Checklist" so I will assume you will tell me what I am to do if you need me to do anything in particular.

Okay, you can add your libraries to libs/libraries.htm, libs/libraries.txt. And libs/tti/index.html (you can just copy log's and do a search and replace).

Done ! I did not see a libraries.txt file anywhere.

...

The inspect report is at (it's also linked from the testing page):

http://boost.cowic.de/rc/docs-inspect-trunk.html#tti

If you add license info to tti.qbk that should fix license issues. You can probably also remove copying the image and css files from 'doc/src' into your documentation, it should use them in their original location now it's included in boost. That leaves adding license and copyright info to build.txt, and renaming 'check' in some headers (I expect that's possible because they're all in detail headers so probably not part of the public interface).

Done ! HTML gets generated correctly. I am having trouble generating my PDF file but this is probably a local problem. I still have some work on the documentation in general. I also want to work and see if I can get Intel C++ to work ( it mostly does except for introspecting class templates, where it fails the same for MPL ). The gcc, msvc, and clang compilers should pass all tests.

Daniel James

10:23 a.m.

On 12 May 2013 08:35, Edward Diener wrote:

...

I did not see a libraries.txt file anywhere.

Sorry, that should be maintainers.txt.

Edward Diener

11:45 a.m.

On 05/12/2013 06:23 AM, Daniel James wrote:

...

On 12 May 2013 08:35, Edward Diener wrote:

...
I did not see a libraries.txt file anywhere.

Sorry, that should be maintainers.txt.

Done !

Philip Bennefall

6 May 6 May

1:15 p.m.

----- Original Message ----- From: "Daniel James" To: Sent: Monday, May 06, 2013 12:53 PM Subject: [boost] [1.54.0][release] Provisional schedule Hello everyone, I've put up a provisional schedule for the next release on the calendar. You can see it at: http://www.boost.org/development/ It targets a 1st July release date, and follows the usual schedule relative to that date. The deadlines for the release branch are pretty soon, particularly the deadline for new libraries which is just a week away. Also, the deadline for new features is just after C++Now. If any of this is a problem let us know now. Hello, I was just wondering if it is possible to get a modified, Boost licensed version of detail/endian.hpp into this next release? It'd be great to get this licensing issue resolved in this next Boost version, if at all possible. Kind regards, Philip Bennefall

Daniel James

2:06 p.m.

On 6 May 2013 14:15, Philip Bennefall wrote:

...

I was just wondering if it is possible to get a modified, Boost licensed version of detail/endian.hpp into this next release? It'd be great to get this licensing issue resolved in this next Boost version, if at all possible.

I don't know. I haven't been following that. Did anyone take responsibility for it?

Rene Rivera

2:11 p.m.

On Mon, May 6, 2013 at 9:06 AM, Daniel James wrote:

...

On 6 May 2013 14:15, Philip Bennefall wrote:

...
I was just wondering if it is possible to get a modified, Boost licensed version of detail/endian.hpp into this next release? It'd be great to get this licensing issue resolved in this next Boost version, if at all possible.

I don't know. I haven't been following that. Did anyone take responsibility for it?

It's highly unlikely. There is al alternative implementation in Boost Predef. But that isn't even fully accepted as a library. I could attempt to do a standalone port from Predef into detail/endian.hpp but I doubt I could both get it done and tested to a point that it would be safe to merge into release. -- -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - grafikrobot/yahoo

Philip Bennefall

2:26 p.m.

----- Original Message ----- From: "Rene Rivera" To: Sent: Monday, May 06, 2013 4:11 PM Subject: Re: [boost] [1.54.0][release] Provisional schedule On Mon, May 6, 2013 at 9:06 AM, Daniel James wrote:

...

On 6 May 2013 14:15, Philip Bennefall wrote:

...
I was just wondering if it is possible to get a modified, Boost licensed version of detail/endian.hpp into this next release? It'd be great to get this licensing issue resolved in this next Boost version, if at all possible.

I don't know. I haven't been following that. Did anyone take responsibility for it?

It's highly unlikely. There is al alternative implementation in Boost Predef. But that isn't even fully accepted as a library. I could attempt to do a standalone port from Predef into detail/endian.hpp but I doubt I could both get it done and tested to a point that it would be safe to merge into release. Hi, I understand. I just wanted to make sure because I want to use Boost in a commercial product in the near future and I am not sure what the company will think about the ambiguity. I hope that it can be resolved - if not for the July release, then maybe for the one in November? Perhaps a ticket could be made for this? Kind regards, Philip Bennefall

Rene Rivera

2:59 p.m.

On Mon, May 6, 2013 at 9:26 AM, Philip Bennefall wrote:

...

----- Original Message ----- From: "Rene Rivera" To: Sent: Monday, May 06, 2013 4:11 PM Subject: Re: [boost] [1.54.0][release] Provisional schedule

On Mon, May 6, 2013 at 9:06 AM, Daniel James wrote:

On 6 May 2013 14:15, Philip Bennefall wrote:

...
...
I was just wondering if it is possible to get a modified, Boost licensed version of detail/endian.hpp into this next release? It'd be great to > get this licensing issue resolved in this next Boost version, if at all possible.

I don't know. I haven't been following that. Did anyone take responsibility for it?

It's highly unlikely. There is al alternative implementation in Boost Predef. But that isn't even fully accepted as a library. I could attempt to do a standalone port from Predef into detail/endian.hpp but I doubt I could both get it done and tested to a point that it would be safe to merge into release.

Hi,

I understand. I just wanted to make sure because I want to use Boost in a commercial product in the near future and I am not sure what the company will think about the ambiguity. I hope that it can be resolved - if not for the July release, then maybe for the one in November? Perhaps a ticket could be made for this?

Sure, creating a ticket would help at least in tagging it to a release milestone. But it will be resolved for the November release one way or another. -- -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - grafikrobot/yahoo

Philip Bennefall

3:35 p.m.

----- Original Message ----- From: "Rene Rivera" To: Sent: Monday, May 06, 2013 4:59 PM Subject: Re: [boost] [1.54.0][release] Provisional schedule On Mon, May 6, 2013 at 9:26 AM, Philip Bennefall wrote:

...

----- Original Message ----- From: "Rene Rivera" To: Sent: Monday, May 06, 2013 4:11 PM Subject: Re: [boost] [1.54.0][release] Provisional schedule

On Mon, May 6, 2013 at 9:06 AM, Daniel James wrote:

On 6 May 2013 14:15, Philip Bennefall wrote:

...
...
I was just wondering if it is possible to get a modified, Boost licensed version of detail/endian.hpp into this next release? It'd be great to > get this licensing issue resolved in this next Boost version, if at all possible.

I don't know. I haven't been following that. Did anyone take responsibility for it?

It's highly unlikely. There is al alternative implementation in Boost Predef. But that isn't even fully accepted as a library. I could attempt to do a standalone port from Predef into detail/endian.hpp but I doubt I could both get it done and tested to a point that it would be safe to merge into release.

Hi,

I understand. I just wanted to make sure because I want to use Boost in a commercial product in the near future and I am not sure what the company will think about the ambiguity. I hope that it can be resolved - if not for the July release, then maybe for the one in November? Perhaps a ticket could be made for this?

Sure, creating a ticket would help at least in tagging it to a release milestone. But it will be resolved for the November release one way or another. Hi again, Would you mind creating that ticket, since you're the one with the potential solution? Thank you. Kind regards, Philip Bennefall

John Maddock

4:14 p.m.

...

...
I was just wondering if it is possible to get a modified, Boost licensed version of detail/endian.hpp into this next release? It'd be great to get this licensing issue resolved in this next Boost version, if at all possible.

I don't know. I haven't been following that. Did anyone take responsibility for it?

Rene has a "clean" version, but it's tied up with the Predef lib I think. John.

Vicente J. Botet Escriba

1:52 p.m.

Le 06/05/13 12:53, Daniel James a écrit :

...

Hello everyone,

I've put up a provisional schedule for the next release on the calendar. You can see it at:

http://www.boost.org/development/

It targets a 1st July release date, and follows the usual schedule relative to that date. The deadlines for the release branch are pretty soon, particularly the deadline for new libraries which is just a week away. Also, the deadline for new features is just after C++Now. If any of this is a problem let us know now.

The release after that will probably target November 1st (as normal), which will give a 4 month release cycle. If the git modularization is considered ready, this will allow a little extra time to deal with any disruption caused (possibly not enough...).

How does that sound? Any thoughts?

This means that we have skip one release. 1.54 should be released in May, 1.55 in August and 1.56 in November, isn't it? It is Ok for me. Vicente

Daniel James

2:07 p.m.

On 6 May 2013 14:52, Vicente J. Botet Escriba wrote:

...

This means that we have skip one release. 1.54 should be released in May, 1.55 in August and 1.56 in November, isn't it?

It is Ok for me.

Yes, unless anything changes the plan is 1.54 in July, 1.55 in November.

Rob Stewart

7 May 7 May

9:29 a.m.

On May 6, 2013, at 10:07 AM, Daniel James wrote:

...

On 6 May 2013 14:52, Vicente J. Botet Escriba wrote:

...
This means that we have skip one release. 1.54 should be released in May, 1.55 in August and 1.56 in November, isn't it?

It is Ok for me.

Yes, unless anything changes the plan is 1.54 in July, 1.55 in November.

With the C++ Now and git distractions, I think it would be wiser to just skip the May release and resume the normal schedule with an August release. One month's difference in delivering a release shouldn't be a big deal for anyone, but it will avoid a lot of last minute rushing and frustration in the preparations for it. ___ Rob (Sent from my portable computation engine)

Vicente J. Botet Escriba

10:27 a.m.

Le 07/05/13 11:29, Rob Stewart a écrit :

...

On May 6, 2013, at 10:07 AM, Daniel James wrote:

...
On 6 May 2013 14:52, Vicente J. Botet Escriba wrote:

...
This means that we have skip one release. 1.54 should be released in May, 1.55 in August and 1.56 in November, isn't it?

It is Ok for me. Yes, unless anything changes the plan is 1.54 in July, 1.55 in November. With the C++ Now and git distractions, I think it would be wiser to just skip the May release and resume the normal schedule with an August release. One month's difference in delivering a release shouldn't be a big deal for anyone, but it will avoid a lot of last minute rushing and frustration in the preparations for it.

What are the worst dates for release? December-January, C++Now - May, July-August, ... Maybe we should have 3 releases a year: I would suggest begin of February/June/October or begin of Mars/July/November. Having 4 moth would let enough time to add new libraries or major changes and have a long period for the beta release so that the release contains less bugs. Best, Vicente

Daniel James

10:38 a.m.

On 7 May 2013 11:27, Vicente J. Botet Escriba wrote:

...

What are the worst dates for release? December-January, C++Now - May, July-August, ...

I think the schedule was actually picked to avoid certain disruptions. I don't know how it works out for C++Now participants.

...

Maybe we should have 3 releases a year: I would suggest begin of February/June/October or begin of Mars/July/November. Having 4 moth would let enough time to add new libraries or major changes and have a long period for the beta release so that the release contains less bugs.

FWIW I've suggested a 4 month cycle before, but I think others consider it too infrequent.

Daniel James

10:37 a.m.

On 7 May 2013 10:29, Rob Stewart wrote:

...

On May 6, 2013, at 10:07 AM, Daniel James wrote:

...
On 6 May 2013 14:52, Vicente J. Botet Escriba wrote:

...
This means that we have skip one release. 1.54 should be released in May, 1.55 in August and 1.56 in November, isn't it?

It is Ok for me.

Yes, unless anything changes the plan is 1.54 in July, 1.55 in November.

With the C++ Now and git distractions, I think it would be wiser to just skip the May release and resume the normal schedule with an August release. One month's difference in delivering a release shouldn't be a big deal for anyone, but it will avoid a lot of last minute rushing and frustration in the preparations for it.

We already have skipped the May release. We can't really avoid a last minute rush, there's always one regardless of when the release is set. There has already been 3 months preparation time for this release, another 2 months is hardly rushing. If the git transition is to happen in the next release cycle, that will require extra time.

Eric Niebler

4:17 p.m.

On 13-05-07 03:37 AM, Daniel James wrote:

...

We already have skipped the May release. We can't really avoid a last minute rush, there's always one regardless of when the release is set.

Tru dat.

...

There has already been 3 months preparation time for this release, another 2 months is hardly rushing. If the git transition is to happen in the next release cycle, that will require extra time.

+1 -- Eric Niebler Boost.org

4216

Age (days ago)

4222

Last active (days ago)

List overview

Download

34 comments

9 participants

participants (9)

Andrey Semashev
Daniel James
Edward Diener
Eric Niebler
John Maddock
Philip Bennefall
Rene Rivera
Rob Stewart
Vicente J. Botet Escriba