Front page | perl.perl5.porters |
Postings from March 2023
Re: perlbug UX
Thread Previous
|
Thread Next
From:
Zefram via perl5-porters
Date:
March 2, 2023 22:41
Subject:
Re: perlbug UX
Message ID:
ZAEmIaN+4Wq+KZ+M@fysh.org
Craig A. Berry wrote:
>https://www.rfc-editor.org/rfc/rfc7763.html
Useful reference, thanks.
>BUT, it looks like charset is required to go with it -- do we know
>what that is? Once upon a time we were trying to figure out how to
>determine the character set of the report
Guessing the charset of user input may well not do for the present use
case. It would have made sense in the context of sending email que email,
where the concern is only to correctly describe the body. But now we're
specifically supporting sticking the output of perlbug into the GitHub
bug tracker without further editing. There are two routes by which
this can happen: for GitHub users perlbug advises submitting the output
(which it saved to a file) to GitHub. For non GitHub users perlbug sends
an email message to p5p, and we're supporting a recipient submitting it
to GitHub. The choice of charset needs to support getting the report,
by either route, into GitHub with whatever charset GitHub is expecting.
There's a need here for some experimentation with multiple means by which
reports get conveyed into GitHub. People have spoken of pasting; is this
an operation on an octet sequence or on characters? Do people also use
other means of inserting file content into GitHub (some "insert this
file" operation in the editor)? In the recipient-of-email case, does
the operation differ according to MUA? I can easily imagine that some
routes of conveyance would require the report to originate with whatever
charset GitHub wants while others would achieve implicit conversion.
Or, for less work, you could just specify that the charset is ASCII,
making it robust to all manner of later handling. The representation
of non-ASCII characters in reports can be left to markup.
>It's not a specification, but basic instructions for writing Markdown
>for GitHub are here:
I'll read that as a starting point, and it may well also be useful for
the perlbug template to point at that. But some more definite kind of
spec is required. We don't need the complete story on fancy markup,
but the instructions do need to be complete on how to to represent
text to avoid it being mistaken for markup, including the edge cases.
Think especially about what happens if a bug report is concerned with
Perl code that processes Markdown text.
-zefram
Thread Previous
|
Thread Next