Originally created by: *anonymous
Originally created by: dak@gnu.org
Originally owned by: dak@gnu.org
Remove convert-ly rule \shiftOff -> \undo\shiftOn
For the rationale, see
<URL:https://codereview.appspot.com/190500043/#msg1>
The presence of the rule will cause similarly surprising effects to
users as it would to the current state of documentation translations.
Even if it were properly mentioned in the diagnostic description of
the conversion rule, people would rarely look closely enough to
notice.
Backup-2015-12-01: #4242
Backup-2015-12-01: #4384
Issues: #4384
Backup-2016-12-10: #4242
Backup-2016-12-10: #4384
Originally posted by: dak@gnu.org
(No comment was entered for this change.)
Owner: dak@gnu.org
Originally posted by: dak@gnu.org
(No comment was entered for this change.)
Blocking: lilypond:3229
Originally posted by: dak@gnu.org
The corresponding Google issue of the problematic change is issue 4242.
Labels: -Type-Enhancement Type-Maintainability
Related
Issues:
#4242Originally posted by: dak@gnu.org
Aaand the rule has botched up the English manuals in the translation branch as well via
commit [rca016880c383a964f3d3542cfba64ada0e1d41cd]
Author: Trevor Daniels <t.daniels@treda.co.uk>
Date: Wed Apr 29 16:19:05 2015 +0100
Issue 3687 (part 0): run convert-ly
Run convert-ly on the doc files to be changed to bring them
up to date.
and has then been "translated" into
commit [reb64473ed20845c5872250192c680f53ef7ee251]
Author: Jean-Charles Malahieude <lilyfan@orange.fr>
Date: Fri May 8 18:20:07 2015 +0200
Doc-fr: updates LM and NR
This is going to need a lot of cleanup in addition to removing the rule.
Cc: k-ohara5...@oco.net
Related
Issues:
#3687Originally posted by: dak@gnu.org
Patchy the autobot says: passes tests. Not terribly surprising since a convert-ly rule change is nothing Patchy would cover
Labels: -Patch-new Patch-review
Originally posted by: lily...@orange.fr
I would have warned if I'd noticed anything in the logs, but nothing just after a merge of master into translation, nor after a full build from scratch before re-merging into staging.
Originally posted by: dak@gnu.org
The convert-ly rule replaced valid code by valid code, so there is little to warn about from Patchy.
The problem is just that as a rule, the replaced code is not wanted. Particularly not in the documentation, but quite likely not elsewhere either.
Originally posted by: pkx1... (code.google.com)@gmail.com
If it helps I manually tested this patch with a full make doc and it passed.
Originally posted by: dak@gnu.org
Well, there is really nothing worthwhile for automatic testing since the Patch just removes one rule action for convert-ly and that rule is very well separated from anything else. So there is really nothing at all to test here, only to discuss.
The reversal of the rule should be accompanied with a cleanup of the affected documentation files. In fact, the documentation files should be cleaned up even when keeping the rule, the only difference being that with the rule kept it would be unsafe to also rewind the version numbers.
That cleanup will be too late for the currently rolling release I think. So we still have a two-week time window to get that under wraps. The cleanup would warrant Patchy testing once it is prepared: while it is simple to do, it will be mostly manual and manual changes have the potential for accidents.
So at any rate: thanks for testing this, and sorry for not having made clearer that this effort would likely have been better spent on a different phase of cleaning up.
Originally posted by: dak@gnu.org
Clean up documentation as well
http://codereview.appspot.com/233290043
Labels: -Patch-review Patch-new
Originally posted by: pkx1... (code.google.com)@gmail.com
Patchy the autobot says: passes tests. includes a full make doc
Labels: -Patch-new Patch-review
Originally posted by: pkx1... (code.google.com)@gmail.com
Patch on countdown for May 19th
Labels: -Patch-review Patch-countdown
Originally posted by: dak@gnu.org
I'll be applying this before the end of countdown once the other issues on "Patch-push" have been pushed in order to commit issue 3229. So holler soon if this should be split into separate patches or you have other considerations.
Related
Issues:
#3229Originally posted by: dak@gnu.org
Pushed to staging as
commit [r6d2a63ff019f548037a513176dc09cead6e53408]
Author: David Kastrup <dak@gnu.org>
Date: Sun May 17 19:19:07 2015 +0200
Run scripts/auxiliar/update-with-convert-ly.sh
Also some whitespace fixes.
commit [r7e9e338479796af1ef2ef66fca98240ff6da273a]
Author: David Kastrup <dak@gnu.org>
Date: Mon May 11 16:15:54 2015 +0200
Issue 4384/2: Undo effects of \shiftOff convert-ly rule
Includes the effects of several manual operations in consequence
commit [rc1793a8d3c9943293ebcd5725dcbfa26bb2d99ad]
Author: David Kastrup <dak@gnu.org>
Date: Sun May 10 14:17:48 2015 +0200
Issue 4384: Remove convert-ly rule \shiftOff -> \undo\shiftOn
For the rationale, see
<URL:https://codereview.appspot.com/190500043/#msg1>
The presence of the rule will cause similarly surprising effects to
users as it would to the current state of documentation translations.
Even if it were properly mentioned in the diagnostic description of
the conversion rule, people would rarely look closely enough to
notice.
Labels: -Patch-countdown Fixed_2_19_21
Status: Fixed
Related
Issues:
#4384Originally posted by: dak@gnu.org
(No comment was entered for this change.)
Blocking: -lilypond:3229
Originally posted by: PhilEHol...@googlemail.com
Verified from git
Status: Verified
Diff:
Related
Issues:
#3229Related
Issues:
#3229