Skip Menu |

This queue is for tickets about the strictures CPAN distribution.

Report information
The Basics
Id: 107393
Status: resolved
Priority: 0/
Queue: strictures

People
Owner: Nobody in particular
Requestors: ribasushi [...] leporine.io
Cc:
AdminCc:

Bug Information
Severity: Important
Broken in: (no value)
Fixed in: (no value)



Subject: XS-basd recommends should be omitted when ! $have_compiler
Recent CPAN versions eagerly try to install all recommends, and while they properly backtrack on failure, the output is rather confusing. I suggest removing the recommends from MYMETA when the compiler isn't there. Cheers
Since we already add the XS modules as required prereqs if a compiler is detected, this would essentially make having the recommends at all pointless. It would also involve hacking around some EUMM limitations - recommends can't normally changed dynamically, since META_ADD and META_MERGE are ignored on user machines.
Subject: Re: [rt.cpan.org #107393] XS-basd recommends should be omitted when ! $have_compiler
Date: Sat, 07 Nov 2015 16:51:03 +0100
To: bug-strictures [...] rt.cpan.org
From: Peter Rabbitson <ribasushi [...] cpan.org>
On 11/07/2015 02:43 PM, Graham Knop via RT wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=107393 > > > Since we already add the XS modules as required prereqs if a compiler is detected, this would essentially make having the recommends at all pointless. > > It would also involve hacking around some EUMM limitations - recommends can't normally changed dynamically, since META_ADD and META_MERGE are ignored on user machines.
Mmm... that's all reasonable, but it is not clear from the above what is the proposed course of action: - Leave is as (wontfix) - Remove the recommends altogether - Something else ?
It sounds like WONTFIX to me. 'recommends' sounds like the right place for XS deps anyway - the spec describes them as (paraphrased) "you should install these whenever possible, but it's not the end of the world if your architecture prevents you from doing so".
I'm not certain what course of action I want to go with here. recommends is the appropriate place for these prereqs. It's unfortunate that CPAN.pm's recommends handling basically breaks one of their main use cases. I don't have any ideas other than removing the recommends or leaving things as is.
Subject: XS-based recommends should be omitted when ! $have_compiler
On Sat Nov 07 23:49:30 2015, haarg wrote: Show quoted text
> I don't have any ideas other than removing the > recommends or leaving things as is.
I just ran into a strong case for removing the recommends for good: .../bin/cpan exits with 1 even if *only recommends* failed to install. I do not intend to file further bugs against CPAN.pm about this - please do so if you think it is worth the trouble. Attached is a log of a run on a fresh 5.22.2_rc1
Subject: argh_optrequire
Download argh_optrequire
application/octet-stream 16.6k

Message body not shown because it is not plain text.

On 2016-04-11 06:38:58, RIBASUSHI wrote: Show quoted text
> On Sat Nov 07 23:49:30 2015, haarg wrote:
> > I don't have any ideas other than removing the > > recommends or leaving things as is.
> > I just ran into a strong case for removing the recommends for good: > .../bin/cpan exits with 1 even if *only recommends* failed to install.
Yes, CPAN.pm simply upgrades all 'recommends' prereqs to requires, and then treats them all identically (erroring out if something couldn't be satisfied). This has already been filed as a bug but more nudges would always be helpful, as tuits are in short supply to fix this.
Subject: Re: [rt.cpan.org #107393] XS-based recommends should be omitted when ! $have_compiler
Date: Mon, 11 Apr 2016 23:20:30 +0200
To: bug-strictures [...] rt.cpan.org
From: Peter Rabbitson <ribasushi [...] cpan.org>
On 04/11/2016 09:16 PM, Karen Etheridge via RT wrote: Show quoted text
> Yes, CPAN.pm simply upgrades all 'recommends' prereqs to requires, and > then treats them all identically
The above is incorrect. In any case the issue has been tracked to https://github.com/andk/cpanpm/issues/97, xdg++
Fixed in 2.000003