I'll see if I can bake a build from source myself and share it. <span></span><br><br>On Tuesday, October 13, 2015, Austin Seipp <<a href="mailto:austin@well-typed.com">austin@well-typed.com</a>> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">I wasn't aware it completely broke on El Capitan, but I think that's a<br>
pretty extreme case (where a completely usable release simply breaks<br>
because a hugely popular OS changes) that would warrant a 7.8.5<br>
release, to be honest. Presumably, this would really be the only<br>
change, so mostly OS X users would be affected by its availability.<br>
<br>
In general people like to use the last 3 GHC releases anyway, and we<br>
haven't even *released* 8.0 yet. And at that point, 7.8.x will be the<br>
last 'in the chain' - so I can assume people would _still_ want 7.8 on<br>
OS X presumably, even more than a year from today until 8.2!<br>
<br>
At that point (8.2-ish era) if Apple keeps breaking things, we may not<br>
do another release, but in this case I think the timescale for<br>
user/library support is probably worth it. Otherwise OS X users are<br>
going to have to fix it themselves, and it might be ugly (propagating<br>
random fixes around the internet isn't going to work well, and we'll<br>
just keep getting bug reports about this because of that - we've<br>
already done this dance with OS X.)<br>
<br>
I'll note this down and think about it more this week, since we'd like<br>
to get 7.10.3 out soon as well...<br>
<br>
On Tue, Oct 13, 2015 at 7:55 PM, Adam Foltzer <<a href="javascript:;" onclick="_e(event, 'cvml', 'acfoltzer@gmail.com')">acfoltzer@gmail.com</a>> wrote:<br>
> 7.8.4 is was released only ten months ago; it surprises me that we'd<br>
> consider it acceptable to have such a recent release fail to work on new<br>
> machines when the fix is so small as to not even require a patch to the<br>
> source tree. In particular, I remember hearing proponents of 7.10's changes<br>
> reassure detractors that if they choose, they can continue to use 7.8.4,<br>
> which is no longer an option.<br>
><br>
> I'd ask that we (1) reconsider the decision to stop 7.8 releases, and (2) if<br>
> we stand by the decision, document the mitigations for users who would like<br>
> to build their own working version of the compiler. I just finished a build<br>
> from a source dist with `unix-2.7.1.0` and it still triggers the El Capitan<br>
> error, so I suspect it's not as simple as just dropping the new package into<br>
> `/libraries/unix/`.<br>
><br>
> On Tue, Oct 13, 2015 at 4:19 PM, Carter Schonwald<br>
> <<a href="javascript:;" onclick="_e(event, 'cvml', 'carter.schonwald@gmail.com')">carter.schonwald@gmail.com</a>> wrote:<br>
>><br>
>> There are no plans to do a new 7.8 release at this Time. Even doing a<br>
>> point release for 7.10.3 is still speculative.<br>
>><br>
>> We should reflect the capitan issue in the installation documentation.<br>
>> And or remark that only 7.10 or newer is supported on El Capitan.<br>
>><br>
>> If a user needs a 7.8 build that works on capitan, they will need to build<br>
>> a source dist with an updated UNIX package.<br>
>><br>
>><br>
>> On Tuesday, October 13, 2015, Adam Foltzer <<a href="javascript:;" onclick="_e(event, 'cvml', 'acfoltzer@gmail.com')">acfoltzer@gmail.com</a>> wrote:<br>
>>><br>
>>> Hello,<br>
>>><br>
>>> The issue was brought up on this list [1] that the released versions of<br>
>>> GHC 7.8 do not work on El Capitan due to an older `unix` package. In a<br>
>>> subsequent GHC weekly news [2], it was announced that since El Capitan<br>
>>> wasn't yet released, it wasn't worthwhile to put more developer time into<br>
>>> 7.8.<br>
>>><br>
>>> Now that El Capitan is out, and the existing 7.8 binaries continue to not<br>
>>> work, I'd like to reopen this issue and request a point release with an<br>
>>> updated `unix` package.<br>
>>><br>
>>> Thank you,<br>
>>><br>
>>> Adam Foltzer<br>
>>> Research Engineer<br>
>>> Galois, Inc.<br>
>>><br>
>>> [1]: <a href="https://mail.haskell.org/pipermail/ghc-devs/2015-July/009403.html" target="_blank">https://mail.haskell.org/pipermail/ghc-devs/2015-July/009403.html</a><br>
>>> [2]:<br>
>>> <a href="https://ghc.haskell.org/trac/ghc/blog/weekly20150721#MacOSXElCapitansupport" target="_blank">https://ghc.haskell.org/trac/ghc/blog/weekly20150721#MacOSXElCapitansupport</a><br>
><br>
><br>
><br>
> _______________________________________________<br>
> ghc-devs mailing list<br>
> <a href="javascript:;" onclick="_e(event, 'cvml', 'ghc-devs@haskell.org')">ghc-devs@haskell.org</a><br>
> <a href="http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs" target="_blank">http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs</a><br>
><br>
<br>
<br>
<br>
--<br>
Regards,<br>
<br>
Austin Seipp, Haskell Consultant<br>
Well-Typed LLP, <a href="http://www.well-typed.com/" target="_blank">http://www.well-typed.com/</a><br>
</blockquote>