unless they are fixing regressions in the beta.

The last time I did a pull to add in the changes to std.parallelism, it pulled in a bunch of other stuff that broke peoples' code in the beta. We will never get the release done by doing things this way.

The current beta breaks are:

http://d.puremagic.com/issues/show_bug.cgi?id=6606

and S?nke Ludwig's problem with DLLs that probably resulted from a change in druntime.

Please, no more merging other than fixing these two issues.

On 6 September 2011 08:27, Walter Bright <walter at digitalmars.com> wrote:
> unless they are fixing regressions in the beta.
>
> The last time I did a pull to add in the changes to std.parallelism, it pulled in a bunch of other stuff that broke peoples' code in the beta. We will never get the release done by doing things this way.
>
> The current beta breaks are:
>
> http://d.puremagic.com/issues/show_bug.cgi?id=6606
>
> and S?nke Ludwig's problem with DLLs that probably resulted from a change in druntime.
>
> Please, no more merging other than fixing these two issues.

This is another pull request which shouldn't have been folded in. But
it's probably harmless, don't know if we need to revert it or not.
https://github.com/D-Programming-Language/druntime/commit/42495b8c76f1018132620a6128d40f170c782050

On Tuesday, September 06, 2011 09:18:46 Don Clugston wrote:
> On 6 September 2011 08:27, Walter Bright <walter at digitalmars.com> wrote:
> > unless they are fixing regressions in the beta.
> > 
> > The last time I did a pull to add in the changes to std.parallelism, it
> > pulled in a bunch of other stuff that broke peoples' code in the beta.
> > We
> > will never get the release done by doing things this way.
> > 
> > The current beta breaks are:
> > 
> > http://d.puremagic.com/issues/show_bug.cgi?id=6606
> > 
> > and S?nke Ludwig's problem with DLLs that probably resulted from a change in druntime.
> > 
> > Please, no more merging other than fixing these two issues.
> 
> This is another pull request which shouldn't have been folded in. But
> it's probably harmless, don't know if we need to revert it or not.
> https://github.com/D-Programming-Language/druntime/commit/42495b8c76f1018132
> 620a6128d40f170c782050 _______________________________________________

That commit seems utterly pointless. The function doesn't have any code in it. True, it won't break anything to leave it in, but it certainly seems to me that we should revert it.

- Jonathan M Davis

If you look at the pull request description, the intention is to create a dummy symbol to avoid a compiler error.? Not sure the exact details, but the commit message isn't enough to understand why it's there.

-Steve




>________________________________
>From: Jonathan M Davis <jmdavisProg at gmx.com>
>To: Discuss the phobos library for D <phobos at puremagic.com>
>Sent: Tuesday, September 6, 2011 3:23 AM
>Subject: Re: [phobos] Please stop merging pull requests
>
>On Tuesday, September 06, 2011 09:18:46 Don Clugston wrote:
>> On 6 September 2011 08:27, Walter Bright <walter at digitalmars.com> wrote:
>> > unless they are fixing regressions in the beta.
>> > 
>> > The last time I did a pull to add in the changes to std.parallelism, it
>> > pulled in a bunch of other stuff that broke peoples' code in the beta.
>> > We
>> > will never get the release done by doing things this way.
>> > 
>> > The current beta breaks are:
>> > 
>> > http://d.puremagic.com/issues/show_bug.cgi?id=6606
>> > 
>> > and S?nke Ludwig's problem with DLLs that probably resulted from a change in druntime.
>> > 
>> > Please, no more merging other than fixing these two issues.
>> 
>> This is another pull request which shouldn't have been folded in. But
>> it's probably harmless, don't know if we need to revert it or not.
>> https://github.com/D-Programming-Language/druntime/commit/42495b8c76f1018132
>> 620a6128d40f170c782050 _______________________________________________
>
>That commit seems utterly pointless. The function doesn't have any code in it. True, it won't break anything to leave it in, but it certainly seems to me that we should revert it.
>
>- Jonathan M Davis
>_______________________________________________
>phobos mailing list
>phobos at puremagic.com
>http://lists.puremagic.com/mailman/listinfo/phobos
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puremagic.com/pipermail/phobos/attachments/20110906/af9ea84b/attachment.html>

On 6 September 2011 11:46, Steve Schveighoffer <schveiguy at yahoo.com> wrote:
> If you look at the pull request description, the intention is to create a
> dummy symbol to avoid a compiler error.? Not sure the exact details, but the
> commit message isn't enough to understand why it's there.
> -Steve

It depends on a DMD pull request, which hasn't been pulled in (it's that pull request which makes the linker error). It doesn't make any sense on its own.

On 9/6/11 8:27 AM, Walter Bright wrote:
> Please, no more merging other than fixing these two issues.

Still, it would be great if this commit could be merged in: https://github.com/D-Programming-Language/phobos/pull/244

It is a no-risk change, since it just fixes a single word in a doc comment, but leaving the byte order documented the wrong way could confuse the heck out of people.

David

Hi,

Why not branch when releases are made? E.g. create a branch called 2.055 that is frozen except for super important backports. That way, development can continue on master without disturbing the release.

Regards,
Alex

On Tue, Sep 6, 2011 at 8:27 AM, Walter Bright <walter at digitalmars.com> wrote:
> unless they are fixing regressions in the beta.
>
> The last time I did a pull to add in the changes to std.parallelism, it pulled in a bunch of other stuff that broke peoples' code in the beta. We will never get the release done by doing things this way.
>
> The current beta breaks are:
>
> http://d.puremagic.com/issues/show_bug.cgi?id=6606
>
> and S?nke Ludwig's problem with DLLs that probably resulted from a change in druntime.
>
> Please, no more merging other than fixing these two issues.
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos
>