Re: [boost] [Boost-commit] svn:boost r68912 - in branches/release: boost libs/utility
AMDG On 2/15/2011 6:54 AM, bdawes@acm.org wrote:
Author: bemandawes Date: 2011-02-15 09:54:16 EST (Tue, 15 Feb 2011) New Revision: 68912 URL: http://svn.boost.org/trac/boost/changeset/68912
Log: Merge trunk BOOST_ASSERT_MSG additions Properties modified: branches/release/boost/static_assert.hpp (contents, props changed) branches/release/libs/utility/assert.html (contents, props changed) branches/release/libs/utility/assert_test.cpp (contents, props changed) Text files modified: branches/release/boost/static_assert.hpp | 10 ++++- branches/release/libs/utility/assert.html | 74 ++++++++++++++++++++++++++++++++++----- branches/release/libs/utility/assert_test.cpp | 44 +++++++++++++++++++++++ 3 files changed, 116 insertions(+), 12 deletions(-)
First of all, this breaks the tests because it merges the tests for BOOST_ASSERT_MSG, but not the header. Second, this is a new feature. Shouldn't it wait till after 1.46 goes out? In Christ, Steven Watanabe
On Tue, Feb 15, 2011 at 10:06 AM, Steven Watanabe <watanabesj@gmail.com> wrote:
Text files modified: branches/release/boost/static_assert.hpp | 10 ++++- branches/release/libs/utility/assert.html | 74 ++++++++++++++++++++++++++++++++++----- branches/release/libs/utility/assert_test.cpp | 44 +++++++++++++++++++++++ 3 files changed, 116 insertions(+), 12 deletions(-)
First of all, this breaks the tests because it merges the tests for BOOST_ASSERT_MSG, but not the header.
Grrr. Failed to commit that part of the change. Fixed. Thanks for the quick heads up!
Second, this is a new feature. Shouldn't it wait till after 1.46 goes out?
Traditionally, yes. But (1) it seems like a pretty harmless change, and this is a "not really part of any library" merge that is always being forgotten. There was already a prior change to assert.hpp that had never been merged to release. And (2) part of the point of more time between beta and release is to allow time for more changes to get into release. We need to formalize that as we do the schedule for 1.47.0, but I'm already moving in that direction. --Beman
On Tue, Feb 15, 2011 at 11:30 AM, Beman Dawes <bdawes@acm.org> wrote:
On Tue, Feb 15, 2011 at 10:06 AM, Steven Watanabe <watanabesj@gmail.com> wrote:
Second, this is a new feature. Shouldn't it wait till after 1.46 goes out?
Traditionally, yes.
But (1) it seems like a pretty harmless change, and this is a "not really part of any library" merge that is always being forgotten. There was already a prior change to assert.hpp that had never been merged to release.
And (2) part of the point of more time between beta and release is to allow time for more changes to get into release. We need to formalize that as we do the schedule for 1.47.0, but I'm already moving in that direction.
I agree with allowing more time for changes to get into release during the beta period, however, by "changes" I think most people think "bug fixes", not new features. Note that this is a general comment and not a comment on the particular commit in question since I am not a Boost developer. --Michael Fawcett
On Tue, Feb 15, 2011 at 2:32 PM, Michael Fawcett <michael.fawcett@gmail.com> wrote:
On Tue, Feb 15, 2011 at 11:30 AM, Beman Dawes <bdawes@acm.org> wrote:
And (2) part of the point of more time between beta and release is to allow time for more changes to get into release. We need to formalize that as we do the schedule for 1.47.0, but I'm already moving in that direction.
I agree with allowing more time for changes to get into release during the beta period, however, by "changes" I think most people think "bug fixes", not new features.
Particularly if there is quite a gap between beta and release, I think risk is a better criteria. So a minor new feature that seems very low risk might be OK, but a really tricky and complex bug fix, particularly a non-showstopper, might be less desirable. But we really need to start a different thread about what a beta should be trying to achieve, what are the problems we are currently experiencing that we be able to fix, etc.
Note that this is a general comment and not a comment on the particular commit in question since I am not a Boost developer.
Hey, discussion helps! Doesn't matter if you are developer or not. --Beman
Would it be better to have release notes for this update? Regards, Michel
participants (4)
-
Beman Dawes -
Michael Fawcett -
Michel MORIN -
Steven Watanabe