From: Maxime Devos <maximedevos@telenet.be>
To: Maxim Cournoyer <maxim.cournoyer@gmail.com>
Cc: guile-devel@gnu.org
Subject: Re: [PATCH v4 4/4] load: Display modules depth in output when using %load-verbosely.
Date: Tue, 3 Oct 2023 21:03:21 +0200 [thread overview]
Message-ID: <f8f71dc7-3270-5282-32b7-91a28800b43a@telenet.be> (raw)
In-Reply-To: <87jzs53suy.fsf@gmail.com>
[-- Attachment #1.1.1: Type: text/plain, Size: 1970 bytes --]
Op 02-10-2023 om 18:13 schreef Maxim Cournoyer:
>> Something I didn't notice previously:
>>
>> Op 25-09-2023 om 16:29 schreef Maxim Cournoyer:
>>> + if (scm_is_string (args)) {
>>> + /* C code written for 3.9 and earlier expects this function to
>>> + take a single argument (the file name). */
>>> + filename = args;
>>> + depth = scm_from_int(0);
>>> + }
>> IIUC, args is always a list and never a string. However, it might be
>> a list of one element (being a string) or two elements.
> Surprisingly perhaps, this works, I guess because a string is an array.
> See the 2009 commit 31ab99de563027fe2bceb60bbd712407fcaf868e which
> exploits the same feature.
While apparently it works, that does not look like a feature to me but
rather an accident, though I don't know what hypothetical feature you
are referring to. Also, sure strings are arrays, but you don't seem to
be using that anywhere, so I'm not following.
> %load-hook is carefully crafted as to accept both one or two arguments
> (for backward compatibility). I didn't pay attention to %load-announce
> and I'm surprised it's a public API, as its sole function seems to be
> the default %load-hook value.
The default value might be, but it's supposed to be overwritable, and
according to the old API it's a one-argument procedure, so if there is
an old application/library setting it to something accepting only a
single argument, there is your API break.
To prevent future API breaks, I propose documenting turning %load-hook
into a keyword argument procedure with #:allow-other-keys, as something
like:
(lambda* (file-name #:key (depth the-default) #:allow-other-keys)
...)
and in the documentation mention that more keywords may be added in the
future (and hence #:allow-other-keys).
I think it's quite plausible that there will be more such arguments in
the future!
Best regards,
Maxime Devos.
[-- Attachment #1.1.2: OpenPGP public key --]
[-- Type: application/pgp-keys, Size: 929 bytes --]
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 236 bytes --]
next prev parent reply other threads:[~2023-10-03 19:03 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-25 14:28 [PATCH v4 0/4] Add module depth information to %load-verbosely output Maxim Cournoyer
2023-09-25 14:28 ` [PATCH v4 1/4] (ice-9 boot-9): Fix typo Maxim Cournoyer
2023-09-25 14:28 ` [PATCH v4 2/4] .dir-locals: Set c-basic-offset to 2 for c-mode Maxim Cournoyer
2023-09-25 14:29 ` [PATCH v4 3/4] guix.scm: Update guile package native inputs Maxim Cournoyer
2023-09-25 14:29 ` [PATCH v4 4/4] load: Display modules depth in output when using %load-verbosely Maxim Cournoyer
2023-09-28 14:23 ` Maxime Devos
2023-10-02 16:13 ` Maxim Cournoyer
2023-10-03 19:03 ` Maxime Devos [this message]
2023-10-04 1:42 ` Maxim Cournoyer
2023-10-10 21:29 ` Maxime Devos
2023-10-11 2:36 ` Maxim Cournoyer
2023-10-11 12:37 ` Maxime Devos
2023-10-22 4:14 ` Maxim Cournoyer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: https://www.gnu.org/software/guile/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=f8f71dc7-3270-5282-32b7-91a28800b43a@telenet.be \
--to=maximedevos@telenet.be \
--cc=guile-devel@gnu.org \
--cc=maxim.cournoyer@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).