emacs-erc
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Document ERC 5.5 regressions in new 29.1-specific version


From: J.P.
Subject: Re: Document ERC 5.5 regressions in new 29.1-specific version
Date: Mon, 24 Apr 2023 07:31:59 -0700
User-agent: Gnus/5.13 (Gnus v5.13)

Eli Zaretskii <eliz@gnu.org> writes:

>> From: "J.P." <jp@neverwas.me>
>> Cc: emacs-erc@gnu.org, bandali@gnu.org
>> Date: Sun, 23 Apr 2023 22:02:39 -0700
>> 
>> I've been mulling over whether ERC might benefit from a 29.1-specific
>> version (among other workflow-related ruminations in recent months [1]).
>> Can you see any downsides to adding something like the attached to the
>> release branch? It mainly involves emitting warnings for known
>> regressions and also adds comments and documentation to that effect.
>
> No objections, but I do have one question, see below.
>
>> Also, as might be obvious, the idea behind the proposed "5.5.0.29.1"
>> versioning scheme is to retain major/minor/patch semantics for the
>> leading components and thus allow for eventual superseding by a
>> "5.5.0.29.2" or a "5.5.1". Alternatively, we could go with something
>> like a "5.5.1-29.1" (which still compares < 5.5.1) to subtly signify
>> that this version doesn't coincide with an official ELPA release.
>> However, one possible downside might be folks thinking this is somehow
>> markedly superior to the 5.5 ELPA release, when in fact they're
>> functionally equivalent.
>
> It's up to you, I don't see why any of these two would be a problem.

I guess I'll see if Amin has any preference.

>> --- a/lisp/erc/erc-networks.el
>> +++ b/lisp/erc/erc-networks.el
>> @@ -1494,8 +1494,14 @@ erc-networks-on-MOTD-end
>>                                        (memq (erc--target-symbol erc--target)
>>                                              erc-networks--bouncer-targets)))
>>                                 proc)
>> -      (let ((m (concat "Unexpected state detected. Please report via "
>> -                       (substitute-command-keys "\\[erc-bug]") ".")))
>> +      (require 'info nil t)
>> +      (let ((m (concat "Unexpected state detected. If you've just issued an"
>> +                       " /MOTD, please know that the command is bugged in 
>> ERC"
>> +                       " 5.5 (Emacs 29) but will be fixed in the next 
>> release."
>> +                       " Otherwise, please report this occurrence via"
>> +                       (substitute-command-keys " \\[erc-bug].")
>> +                       " For additional help options, see"
>> +                       " Info:\"(erc) Getting Help and Reporting Bugs\".")))
>>          (erc-display-error-notice parsed m))))
>
> Why do we have to load info.el here?  Don't hyperlinks to Info manuals
> in doc strings work without info.el loaded?

Unfortunately, when it comes to ERC's error notices, that's only true on
master [1]. (With ERC 5.5/Emacs 29, we still get a "void-function
Info-goto-node".) Either way, linking to the manual here isn't critical,
so I'll just drop that part (and the `require' as well). Thanks for
looking at this.

[1] https://git.savannah.gnu.org/cgit/emacs.git/commit/?id=1f1cd467



reply via email to

[Prev in Thread] Current Thread [Next in Thread]