llvm dev - Nov 2013 - [LLVMdev] [cfe-dev] RFC: A proposal to move toward using C++11 features in LLVM & Clang / bounding support for old host compilers

On 10/30/13 03:17 AM, Chandler Carruth wrote:
> On Mon, Oct 28, 2013 at 6:07 PM, "C. Bergström" 
> <cbergstrom at pathscale.com <mailto:cbergstrom at
pathscale.com>> wrote:
>
>     On 10/29/13 07:27 AM, Chandler Carruth wrote:
>
>         On Mon, Oct 28, 2013 at 5:06 PM, "C. Bergström"
>         <cbergstrom at pathscale.com <mailto:cbergstrom at
pathscale.com>
>         <mailto:cbergstrom at pathscale.com
>         <mailto:cbergstrom at pathscale.com>>> wrote:
>
>             fuzzy://How much "heads up"
>
>
>         One full release cycle, so approximately 6 months before a release
>
>     If it's 3-6 months from *today* before something hits clang svn
>     trunk that should be enough time to address any problems.
>
>
> No, it's 1 month, maybe 2 before something hits trunk, and over 6 
> months before something hits a release.
I'm objecting to 1 month for svn trunk - 2 months notice is even pushing 
it imho.

1) This is imho not the small change which it's being presented as
2) As someone else stated - there are projects tracking svn trunk and 
telling them to just stop doing that and follow the previous release is 
untenable without sufficient notice. (It just doesn't seem fair)

May I humbly propose you create a c++11-development branch 
now/later/anytime and let people start using that. In parallel to that 
let people know that pieces of the c++11 branch will potentially start 
merging Feb 1st 2014. (roughly 3 months from today). This gives people 
time to review things before they hit trunk, test, discuss and 
experiment in a way that virtual discussions simply can't flush out. 
This hopefully won't hurt your target of the release-after-next using 
more modern toolchains and is *hopefully* a win-win in your view.

On Tue, Oct 29, 2013 at 6:11 PM, "C. Bergström" <cbergstrom at
pathscale.com>wrote:
> On 10/30/13 03:17 AM, Chandler Carruth wrote:
>
>> On Mon, Oct 28, 2013 at 6:07 PM, "C. Bergström"
<cbergstrom at pathscale.com<mailto:
>> cbergstrom at pathscale.**com <cbergstrom at
pathscale.com>>> wrote:
>>
>>     On 10/29/13 07:27 AM, Chandler Carruth wrote:
>>
>>         On Mon, Oct 28, 2013 at 5:06 PM, "C. Bergström"
>>         <cbergstrom at pathscale.com <mailto:cbergstrom at
pathscale.**com<cbergstrom at pathscale.com>
>> >
>>         <mailto:cbergstrom at pathscale.**com <cbergstrom at
pathscale.com>
>>
>>         <mailto:cbergstrom at pathscale.**com <cbergstrom at
pathscale.com>>>>
>> wrote:
>>
>>             fuzzy://How much "heads up"
>>
>>
>>         One full release cycle, so approximately 6 months before a
release
>>
>>     If it's 3-6 months from *today* before something hits clang svn
>>     trunk that should be enough time to address any problems.
>>
>>
>> No, it's 1 month, maybe 2 before something hits trunk, and over 6
months
>> before something hits a release.
>>
> I'm objecting to 1 month for svn trunk - 2 months notice is even
pushing
> it imho.
>
> 1) This is imho not the small change which it's being presented as
> 2) As someone else stated - there are projects tracking svn trunk and
> telling them to just stop doing that and follow the previous release is
> untenable without sufficient notice. (It just doesn't seem fair)
>
If such projects exist and don't want to switch to C++11 mode, I think we
should let them speak up, and not delay our own plans on the hypothesis
that they exist. (Maybe they'll say that this is sufficient notice, maybe
not, but we can't really guess.)

> May I humbly propose you create a c++11-development branch
> now/later/anytime and let people start using that. In parallel to that let
> people know that pieces of the c++11 branch will potentially start merging
> Feb 1st 2014. (roughly 3 months from today). This gives people time to
> review things before they hit trunk, test, discuss and experiment in a way
> that virtual discussions simply can't flush out. This hopefully
won't hurt
> your target of the release-after-next using more modern toolchains and is
> *hopefully* a win-win in your view.

I don't see how this helps anything. We don't /want/ to have the hassle
of
some people developing on a branch and some on trunk, so we would
essentially have trunk stagnating and everyone developing on the branch.
And then we'd merge the branch back again. Net result: exactly the same as
if the people who aren't ready for c++11 stick with the 3.4 release.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20131029/103db8b2/attachment.html>

On 10/30/13 08:24 AM, Richard Smith wrote:
>
>     May I humbly propose you create a c++11-development branch
>     now/later/anytime and let people start using that. In parallel to
>     that let people know that pieces of the c++11 branch will
>     potentially start merging Feb 1st 2014. (roughly 3 months from
>     today). This gives people time to review things before they hit
>     trunk, test, discuss and experiment in a way that virtual
>     discussions simply can't flush out. This hopefully won't hurt
your
>     target of the release-after-next using more modern toolchains and
>     is *hopefully* a win-win in your view.
>
>
> I don't see how this helps anything. We don't /want/ to have the 
> hassle of some people developing on a branch and some on trunk, so we 
> would essentially have trunk stagnating and everyone developing on the 
> branch. And then we'd merge the branch back again. Net result: exactly 
> the same as if the people who aren't ready for c++11 stick with the 
> 3.4 release.
No its entirely not the same

1) Branch development is very common - I can't imagine *everyone* will 
use c++11 immediately and it causing some instant stall on trunk if a 
branch was made
2) As stated before - a branch lets a *real* discussion happen *before* 
it hits trunk. With any big feature change - isn't it best to be a 
little conservative instead of potentially break 1st and fix-it-later 
approach.

Right now we're having a virtual discussion about things which may/may 
not happen. A branch with code is a tangible which can allow *real* 
testing and feedback. I think it's being worked on independently, but I 
haven't even seen an updated style guide. I don't even know what's 
exactly being proposed at this time. "c++11" is still fuzzy...

clang/llvm has been around for 10 years?? Will waiting an extra month or 
doing some work in a branch 1st really kill anyone? Doesn't google and 
friends already use a branch model for development/test before it's merged.

On Oct 29, 2013, at 6:11 PM, C. Bergström <cbergstrom at pathscale.com>
wrote:
>> No, it's 1 month, maybe 2 before something hits trunk, and over 6
months before something hits a release.
> I'm objecting to 1 month for svn trunk - 2 months notice is even
pushing it imho.
> 
> 1) This is imho not the small change which it's being presented as
> 2) As someone else stated - there are projects tracking svn trunk and
telling them to just stop doing that and follow the previous release is
untenable without sufficient notice. (It just doesn't seem fair)
> 
> May I humbly propose you create a c++11-development branch
now/later/anytime and let people start using that. In parallel to that let
people know that pieces of the c++11 branch will potentially start merging Feb
1st 2014. (roughly 3 months from today). This gives people time to review things
before they hit trunk, test, discuss and experiment in a way that virtual
discussions simply can't flush out. This hopefully won't hurt your
target of the release-after-next using more modern toolchains and is *hopefully*
a win-win in your view.
Umm, no.  LLVM development happens on trunk, not on branches.  Please see the
developer policy if you are confused by this.

I'm not sure what you're concerned about here, do you have any concrete
problem in mind?

The practical effect of this decision is that as soon as 3.4 branches, a *lot*
of codebase churn will happen.  All of which would be pointless to do on a
branch.

-Chris

Richard Smith <richard at metafoo.co.uk> writes:
>     I'm objecting to 1 month for svn trunk - 2 months notice is even
>     pushing it imho.
>     
>     1) This is imho not the small change which it's being presented as
>     2) As someone else stated - there are projects tracking svn trunk
>     and telling them to just stop doing that and follow the previous
>     release is untenable without sufficient notice. (It just doesn't
>     seem fair)
>
> If such projects exist and don't want to switch to C++11 mode, I think
> we should let them speak up, and not delay our own plans on the
> hypothesis that they exist. (Maybe they'll say that this is sufficient
> notice, maybe not, but we can't really guess.)
I'm not objecting to C++11 at all.  I would love it!  But I do agree
that 1-2 months notice for a toolchain update on trunk is too short.
I've been hammered countless times about not working off trunk and I am
moving us in the direction of working off trunk (or at least something
close to it).  Short-notice toolchain changes on trunk are going to
hamper that effort.

It is a massive undertaking to test all of our compiler-related software
once we change the toolchain used to build it.  If there is consensus on
which toolchain to move to about 4-6 months out, that seems like
sufficient time to do testing.  This means that for the next release, a
decision on toolchain needs to be made pretty soon after cutting the
previous release.

                            -David