[ghc-steering-committee] Please review #673: Amendment to -jsem proposal #540

[Adam Gundry]

It seems Zubin updated the proposal to take account of Sebastian's 
concerns, but then discussion has tailed off. Are we now in a position 
to accept this proposal, or are there any remaining concerns?

Cheers,

Adam



On 24/09/2024 07:44, Sebastian Graf wrote:
> I think changes to the communication protocol *are* user-facing, in that 
> any change to the protocol could implies a breaking change between all 
> cabal releases in the future and all GHC releases in the past.
> It appears that is the case for the planned release of 
> `semaphore-compat-2`, but I'm not 100% sure.
> 
> In that light, I think it's good to discuss in a proposal to ensure we 
> do not sign off lightly on such breaking changes, in particular for 
> future versions of `semaphore-compat`.
> 
> I will express support once compatibility and breaking changes are 
> properly addressed.
> 
> Sebastian
> 
> ------ Originalnachricht ------
> Von "Arnaud Spiwack" <arnaud.spiwack at tweag.io 
> <mailto:arnaud.spiwack at tweag.io>>
> An "Simon Peyton Jones" <simon.peytonjones at gmail.com 
> <mailto:simon.peytonjones at gmail.com>>
> Cc ghc-steering-committee at haskell.org 
> <mailto:ghc-steering-committee at haskell.org>
> Datum 24.09.2024 04:17:17
> Betreff Re: [ghc-steering-committee] Please review #673: Amendment to 
> -jsem proposal #540
> 
>> Quite frankly, this is barely any user-facing change at all. 
>> Technically the semaphore could previously be shared by non-Haskell 
>> processes in ways the updated proposal doesn't allow. But it's not 
>> really something people do. So it all sounds reasonable.
>>
>> On Mon, 23 Sept 2024 at 19:10, Simon Peyton Jones 
>> <simon.peytonjones at gmail.com <mailto:simon.peytonjones at gmail.com>> wrote:
>>
>>     I'm happy to support too, but I would like to see presentational
>>     changes, so the final proposal makes sense when read in 5 yrs time.
>>
>>     Simon
>>
>>     On Sat, 21 Sept 2024 at 20:45, Malte Ott <malte.ott at maralorn.de
>>     <mailto:malte.ott at maralorn.de>> wrote:
>>
>>         I agree.
>>
>>         On 2024-09-21 09:16, Eric Seidel wrote:
>>         > Hi all,
>>         >
>>         > This seems like a sensible update to an accepted proposal.
>>         >
>>         > I recommend we accept the amendment.
>>         >
>>         > Eric
>>         >
>>         > On Thu, Sep 19, 2024, at 19:00, Erik de Castro Lopo wrote:
>>         > > Hi all,
>>         > >
>>         > > This is minor updates to correct an existing approved
>>         proposal.
>>         > >
>>         > > I approve of these changes.
>>         > >
>>         > > Erik
>>         > >
>>         > > Adam Gundry wrote:
>>         > >
>>         > >> Dear Committee,
>>         > >>
>>         > >> Zubin Duggal proposes to amend proposal #540, which
>>         introduced the -jsem
>>         > >> parallelism control mechanism, so that it can avoid
>>         incompatibilities
>>         > >> arising from different system C library implementations:
>>         > >>
>>         > >> https://github.com/ghc-proposals/ghc-proposals/pull/673
>>         <https://github.com/ghc-proposals/ghc-proposals/pull/673>
>>         > >>
>>         > >>
>>         https://github.com/wz1000/ghc-proposals/blob/new-jsem/proposals/0540-jsem.rst <https://github.com/wz1000/ghc-proposals/blob/new-jsem/proposals/0540-jsem.rst>
>>         > >>
>>         > >> I'd like to nominate Eric Seidel as the shepherd, since
>>         he was the
>>         > >> shepherd of the original proposal.
>>         > >>
>>         > >> Please guide us to a conclusion as outlined in
>>         > >>
>>         https://github.com/ghc-proposals/ghc-proposals#committee-process <https://github.com/ghc-proposals/ghc-proposals#committee-process>
>>         > >>
>>         > >> Cheers,
>>         > >>
>>         > >> Adam

	
-- 
Adam Gundry, Haskell Consultant
Well-Typed LLP, https://www.well-typed.com/

Registered in England & Wales, OC335890
27 Old Gloucester Street, London WC1N 3AX, England