The Glasgow Haskell Users Mailing List

Jens Petersen | 1 Sep 2011 04:26

Re: ANNOUNCE: GHC version 7.2.1

On 19 August 2011 17:15, Jens Petersen <juhp <at> community.haskell.org> wrote:
>> I have done a test build of 7.2.1 for Fedora [17]:
>> http://koji.fedoraproject.org/koji/taskinfo?taskID=3267317

Since the test rpms are now gone from Koji and I don't think
I will start building 7.2 properly for Fedora until after ICFP
I have placed those rpms here for now:
http://petersen.fedorapeople.org/ghc-7.2/
if anyone is looking for them.  They should work
fine with F15 and F16 (untested but probably with F14 too).
(I forgot to archive the original src rpm for the build
but I believe the one I uploaded is essentially equivalent.)

Just note that the bootstrap test rpms don't have ABI hash metadata.
I haven't committed the packaging yet to fedora git either,
if there is interest I am happy to push it to a branch there.

Jens

headers

Bryan O'Sullivan | 1 Sep 2011 19:06

Re: ANNOUNCE: GHC version 7.2.1

On Wed, Aug 31, 2011 at 7:26 PM, Jens Petersen <juhp <at> community.haskell.org> wrote:

Since the test rpms are now gone from Koji and I don't think
I will start building 7.2 properly for Fedora until after ICFP

Just in case you didn't know, 7.4 is due to come out around the time of ICFP. If you have limited time to work on this, I don't think it's worth worrying further about 7.2.

_______________________________________________
Glasgow-haskell-users mailing list
Glasgow-haskell-users <at> haskell.org
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

headers

AntC | 27 Mar 2012 00:47

Re: ANNOUNCE: GHC version 7.2.1 - {-# LANGUAGE NoTraditionalRecordSyntax #-}

> 
Ian Lynagh <igloo <at> earth.li> writes:
> 
>    =============================================================
>     The (Interactive) Glasgow Haskell Compiler -- version 7.2.1
>    =============================================================
> 
Ticket #3356 claims that {-# LANGUAGE NoTraditionalRecordSyntax #-} was 
implemented in 7.2.1.

But GHCi v7.2.1 complains "Unsupported extension: NoTraditionalRecordSyntax".

What (if anything) actually got implemented?

Is/was the plan to be able to selectively 'prune' bits of record syntax? or 
just to completely banish anything with curly brackets {and their contents} 
from patterns, expressions and data declarations?

I ask because I'd like to raise a ticket for 
http://hackage.haskell.org/trac/ghc/wiki/Records/DeclaredOverloadedRecordFields
/NoMonoRecordFields -- that is, to enable declaring within the same module 
multiple record types with the same field name.

Of course it will also help when declaring the same field name in different 
(but imported) modules.

Avoiding generating the (monomorphic) field selector function is a modest step 
towards freeing up the namespace, without in any way pre-judging how 
the 'narrow namespace issue' might get addressed.

AntC

headers

Simon Peyton-Jones | 27 Mar 2012 09:20

RE: ANNOUNCE: GHC version 7.2.1 - {-# LANGUAGE NoTraditionalRecordSyntax #-}

ghci -XNoTraditionalRecordSyntax does not complain of unsupported extensions for me.

The flag appears to just disable record construction and update syntax, and record patterns, and record
syntax in GADT declarations.  It has probably never been used.

Simon

| -----Original Message-----
| From: glasgow-haskell-users-bounces <at> haskell.org [mailto:glasgow-haskell-
| users-bounces <at> haskell.org] On Behalf Of AntC
| Sent: 26 March 2012 23:47
| To: glasgow-haskell-users <at> haskell.org
| Subject: Re: ANNOUNCE: GHC version 7.2.1 - {-# LANGUAGE
| NoTraditionalRecordSyntax #-}
| 
| >
| Ian Lynagh <igloo <at> earth.li> writes:
| >
| >    =============================================================
| >     The (Interactive) Glasgow Haskell Compiler -- version 7.2.1
| >    =============================================================
| >
| Ticket #3356 claims that {-# LANGUAGE NoTraditionalRecordSyntax #-} was
| implemented in 7.2.1.
| 
| But GHCi v7.2.1 complains "Unsupported extension: NoTraditionalRecordSyntax".
| 
| What (if anything) actually got implemented?
| 
| Is/was the plan to be able to selectively 'prune' bits of record syntax? or
| just to completely banish anything with curly brackets {and their contents}
| from patterns, expressions and data declarations?
| 
| I ask because I'd like to raise a ticket for
| http://hackage.haskell.org/trac/ghc/wiki/Records/DeclaredOverloadedRecordFiel
| ds
| /NoMonoRecordFields -- that is, to enable declaring within the same module
| multiple record types with the same field name.
| 
| Of course it will also help when declaring the same field name in different
| (but imported) modules.
| 
| Avoiding generating the (monomorphic) field selector function is a modest
| step towards freeing up the namespace, without in any way pre-judging how the
| 'narrow namespace issue' might get addressed.
| 
| AntC
| 
| 
| 
| 
| _______________________________________________
| Glasgow-haskell-users mailing list
| Glasgow-haskell-users <at> haskell.org
| http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

headers

AntC | 28 Mar 2012 00:13

Re: ANNOUNCE: GHC version 7.2.1 - {-#LANGUAGE NoTraditionalRecordSyntax#-}

Simon Peyton-Jones <simonpj <at> microsoft.com> writes:

> 
> ghci -XNoTraditionalRecordSyntax does not complain of unsupported extensions 
for me.

OK, it's effective in v7.4.1, as Ian says.

> 
> The flag appears to just disable record construction and update syntax, and 
> record patterns, and record syntax in GADT declarations.

For the record(!) it also suppresses record syntax in ordinary data 
declarations. So it 'solves' the namespacing issue by not allowing field names 
at all.

No curly brackets, no field names nowhere! Everything has to be positional.

>  It has probably never been used.
> 

Yes, the pruning is _too_ radical, the tree withered away.

AntC

headers

Ian Lynagh | 27 Mar 2012 14:12

Re: ANNOUNCE: GHC version 7.2.1 - {-# LANGUAGE NoTraditionalRecordSyntax #-}

On Mon, Mar 26, 2012 at 10:47:14PM +0000, AntC wrote:
>
> Ticket #3356 claims that {-# LANGUAGE NoTraditionalRecordSyntax #-} was 
> implemented in 7.2.1.
> 
> But GHCi v7.2.1 complains "Unsupported extension: NoTraditionalRecordSyntax".
> 
> What (if anything) actually got implemented?

It was implemented in 7.4.1.

Thanks
Ian

headers

AntC | 27 Mar 2012 22:32

Re: ANNOUNCE: GHC version 7.2.1 - {-#LANGUAGE NoTraditionalRecordSyntax#-}

Ian Lynagh <igloo <at> earth.li> writes:

> 
> On Mon, Mar 26, 2012 at 10:47:14PM +0000, AntC wrote:
> >
> > Ticket #3356 claims that {-# LANGUAGE NoTraditionalRecordSyntax #-} was 
> > implemented in 7.2.1.
> > 
> > But GHCi v7.2.1 complains "Unsupported extension: 
NoTraditionalRecordSyntax".
> > 
> > What (if anything) actually got implemented?
>
> ghci -XNoTraditionalRecordSyntax does not complain of unsupported extensions 
> for me.
> ...
> Simon
> 
> It was implemented in 7.4.1.
> 
> Thanks
> Ian
> 
Thank you Simon and Ian (and apologies for first posting to the Haskell list).

So I'll upgrade to 7.4, and play with it.

And post a ticket for the more modest idea of suppressing the (monomorphic) 
field selector function.

AntC