llvm dev - Feb 2015 - [LLVMdev] [cfe-dev] [3.6 Release] RC3 has been tagged

On Wed, Feb 18, 2015 at 2:38 PM, Renato Golin <renato.golin at linaro.org>
wrote:
> On 18 February 2015 at 18:48, Jack Howarth
> <howarth.mailing.lists at gmail.com> wrote:
>> Define users. If compiler developers, your argument holds. For other
>> users who just want a stable compiler, the argument does not.
>
> Ok, let's turn this conversation upside down. What is *your* use for
> LLVM, and why is the release 3.6.0 so important for *you*?
>
Renato,
      In the fink project (http://finkproject.org), we have maintained
llvm packaging for each new release for awhile now...

http://fink.cvs.sourceforge.net/viewvc/fink/dists/10.7/stable/main/finkinfo/languages/llvm35.info?revision=1.3

Since the completion of the merge of clang-omp support always seems to
be one release away, starting with the 3.4.2 release I added a merge
of the clang-omp changes from upstream to our packaging so that we
have fully functional libiomp5 support. Assuming that the clang-omp
developers can find time to rebase their upstream tree on the new 3.6
release, I intend to do the same for the fink llvm36 packaging. So
yes, a stable compiler does matter to some folks.
          Jack
> cheers,
> --renato

On 18 February 2015 at 19:52, Jack Howarth
<howarth.mailing.lists at gmail.com> wrote:
> Assuming that the clang-omp
> developers can find time to rebase their upstream tree on the new 3.6
> release, I intend to do the same for the fink llvm36 packaging. So
> yes, a stable compiler does matter to some folks.
So it seems that you're one of the very few people that doesn't use
ToT. Almost everyone else uses it and the progress of LLVM kind of
assume you do.

The past releases didn't mean much until we started doing the
dot-releases, and even those didn't mean much in the first iterations.
You can't blame people for not caring that much for something that so
few people actually use it.

I think that buildbots make us rely a lot more on inter-release master
branches than we should. Maybe once more OS distributions start
relying on LLVM we'll have to match a more professional release cycle,
but for now, we have very little reason to.

I'm perfectly happy to open a bug for each one of that Phoronix run's
regressions and fix throughout the next six months...

cheers,
--renato

On Wed, Feb 18, 2015 at 3:01 PM, Renato Golin <renato.golin at linaro.org>
wrote:
> On 18 February 2015 at 19:52, Jack Howarth
> <howarth.mailing.lists at gmail.com> wrote:
>> Assuming that the clang-omp
>> developers can find time to rebase their upstream tree on the new 3.6
>> release, I intend to do the same for the fink llvm36 packaging. So
>> yes, a stable compiler does matter to some folks.
>
> So it seems that you're one of the very few people that doesn't use
> ToT. Almost everyone else uses it and the progress of LLVM kind of
> assume you do.
>
I believe MacPorts uses the major llvm releases for their toolchain
these days as the older supported systems don't have access to the
latest Xcode (due to SDK deprecation).
> The past releases didn't mean much until we started doing the
> dot-releases, and even those didn't mean much in the first iterations.
> You can't blame people for not caring that much for something that so
> few people actually use it.
Well, I assumed that llvm releases were supposed to be more than just
glorified semi-annual snapshots.
>
> I think that buildbots make us rely a lot more on inter-release master
> branches than we should. Maybe once more OS distributions start
> relying on LLVM we'll have to match a more professional release cycle,
> but for now, we have very little reason to.
>
> I'm perfectly happy to open a bug for each one of that Phoronix
run's
> regressions and fix throughout the next six months...
>
> cheers,
> --renato

On 02/18/2015 03:01 PM, Renato Golin wrote:
> So it seems that you're one of the very few people that doesn't use
> ToT. Almost everyone else uses it and the progress of LLVM kind of
> assume you do.
My company also does not use ToT.  Being able to associate a product 
with a well-known release is *very* important to us.  It enables 
communication with our customers.  It allows us to determine 
compatibility between our Clang derived product and other Clang derived 
products.  It allows us to easily discuss groups of feature sets.

I personally find it rather crazy that developers are expected to use 
ToT (and then determine stability on their own).  Why have an RC series 
at all if that is the case?

Tom.