Skip Menu |

This queue is for tickets about the MIME-Base32 CPAN distribution.

Report information
The Basics
Id: 113838
Status: resolved
Priority: 0/
Queue: MIME-Base32
People
Owner: Nobody in particular
Requestors: SREZIC [...] cpan.org
Cc: cwhitener [...] gmail.com
AdminCc:
Bug Information
Severity: (no value)
Broken in: 1.301
Fixed in: 1.301
Attachments
signature.asc

History Show all quoted text
  Sat Apr 16 15:27:41 2016 SREZIC [...] cpan.org - Ticket created
Subject: 1.301 breaks CPAN modules (export changes?)
With the recent 1.301 some CPAN modules depending on MIME::Base32 are broken: * https://rt.cpan.org/Ticket/Display.html?id=113831 * https://rt.cpan.org/Ticket/Display.html?id=113832 * https://github.com/Songmu/p5-Riji/issues/19 * https://rt.cpan.org/Ticket/Display.html?id=113837 * https://rt.cpan.org/Ticket/Display.html?id=113836 All these modules have in common that they seem to use a lax approach to import MIME::Base32 functions, something which is not possible anymore (it seems that importing was done by specifying the strings "RFC", "crockford", and "09AV"). Not sure what to do about it. At least there should be a note about the stricter exports in the Changes file. Maybe the old behavior could be restored, that is, by specifying "RFC" both encode_rfc3548 and decode_rfc3548 would be imported, and so on.
  Sat Apr 16 15:28:28 2016 SREZIC [...] cpan.org - Reference to ticket #113831 added
  Sat Apr 16 15:28:29 2016 SREZIC [...] cpan.org - Reference to ticket #113832 added
  Sat Apr 16 15:28:29 2016 SREZIC [...] cpan.org - Reference to ticket #113837 added
  Sat Apr 16 15:28:29 2016 SREZIC [...] cpan.org - Reference to ticket #113836 added
  Sat Apr 16 15:53:13 2016 REHSACK [...] cpan.org - Correspondence added
Me neither has any idea. The patches came from Chase Whitener <chase.whitener@infotechfl.com> - maybe talk to him first. I uploaded a dev release - I wish one had checked that earlier ...
  Sat Apr 16 15:53:13 2016 The RT System itself - Status changed from 'new' to 'open'
  Sat Apr 16 16:23:52 2016 SREZIC [...] cpan.org - Correspondence added
On 2016-04-16 15:53:13, REHSACK wrote: Show quoted text
> Me neither has any idea. The patches came from Chase Whitener > <chase.whitener@infotechfl.com> - maybe talk to him first. > > I uploaded a dev release - I wish one had checked that earlier ...
AFAIK unfortunately no smoker implements a mechanism "new devel version of a CPAN module uploaded → smoke all reverse dependencies".
  Sat Apr 16 19:38:39 2016 rehsack [...] gmail.com - Correspondence added
CC: Chase Whitener <cwhitener [...] gmail.com>
Subject: Re: [rt.cpan.org #113838] 1.301 breaks CPAN modules (export changes?)
Date: Sun, 17 Apr 2016 01:38:26 +0200
To: bug-MIME-Base32 [...] rt.cpan.org
From: Jens Rehsack <rehsack [...] gmail.com>
Show quoted text
> Am 16.04.2016 um 22:23 schrieb Slaven_Rezic via RT <bug-MIME-Base32@rt.cpan.org>: > > Queue: MIME-Base32 > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=113838 > > > On 2016-04-16 15:53:13, REHSACK wrote:
>> Me neither has any idea. The patches came from Chase Whitener >> <chase.whitener@infotechfl.com> - maybe talk to him first. >> >> I uploaded a dev release - I wish one had checked that earlier ...
> > AFAIK unfortunately no smoker implements a mechanism "new devel version of a CPAN module uploaded → smoke all reverse dependencies".
Dunno - some proposed that for LMU. The reason for dev-release is that people test. But this is not the discussion here :) Please involve Chase in the ticket and ask him to review his changes. Best -- Jens Rehsack - rehsack@gmail.com
Download signature.asc
application/pgp-signature 842b

Message body not shown because it is not plain text.

  Sat Apr 16 20:14:01 2016 cwhitener [...] gmail.com - Correspondence added
This is indeed my fault. I looked through the reverse dependencies and only saw one module using MIME::Base32 and that module was already using the long name of the encode/decode methods. I should have searched through http://grep.cpan.me/?q=MIME%3A%3ABase32 instead. What I pulled out in favor of Traditional Exporter was: sub import { my( $pkg, $arg )=@_; if( defined($arg) && $arg =~ /rfc|3548/i ) { *encode = \&encode_rfc3548; *decode = \&decode_rfc3548; } else { *encode = \&encode_09AV; *decode = \&decode_09AV; } } I'll work up a quick fix to revert that and alter the verbiage on the main POD and enter a PR on the GitHub repository. Apologies. Thanks, Chase
  Sat Apr 16 20:29:44 2016 cwhitener [...] gmail.com - Correspondence added
https://github.com/perl5-utils/MIME-Base32/pull/2 reverts that change. Again, I apologize for the oversight. -- Chase
  Sun Apr 17 04:13:53 2016 REHSACK [...] cpan.org - Correspondence added
Chase, I wouldn't shoot so quick - all Slaven reported were smoke failures. Maybe none of the modules are in use neither maintained. CGI::Session::ID::Base32 - by Dan Peder - not maintained :) CGI::Session::ID::MD5_Base32 - by Dan Peder - not maintained :) Riji seems fairly active, but no changes since a year - let's see ... Firefox::Sync::Client - by Robin Daermann - untouched since 4 years ... Crypt::VERPString - by Dorian Taylor - untouched since 11 years, I would rate it unmaintained, too Since the old import was kind-of weird, I prefer use the new way. Quick solution within a week or two which restores old behavior (eg. in case import doesn't match any exported functions) with a clear line of deprecation. Slaven recommended a note into Changes what has been changed in behavior, no behavior change :) Cheers, Jens
  Sun Apr 17 04:14:58 2016 REHSACK [...] cpan.org - Cc CAPOEIRAB added
  Sun Apr 17 04:14:58 2016 REHSACK [...] cpan.org - Fixed in 1.301 added
  Thu May 18 13:05:38 2017 REHSACK [...] cpan.org - Status changed from 'open' to 'resolved'