bug-coreutils
[Top][All Lists]
Advanced

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

Re: Bug#353911: md5sum --check checks only if _all_ are bad


From: Jim Meyering
Subject: Re: Bug#353911: md5sum --check checks only if _all_ are bad
Date: Mon, 17 Nov 2008 11:25:09 +0100

Patrick Schoenfeld <address@hidden> wrote:
...
>> > I was a bit sceptic about this copyright assignment and wondered why the
>> > HACKING file does not include any link to [1]. Just as a suggestion
>>
>> What's [1] ?
>
> sorry, accidentally removed the link:
>
> http://www.gnu.org/licenses/why-assign.html

Good idea.
I've just added that link to HACKING.

  http://git.savannah.gnu.org/gitweb/?p=coreutils.git;a=commitdiff;h=8b09875b5

Normally I'd also add your name to the THANKS list,
but since you're preparing a patch, it'll soon appear
in a commit where you're the Author.

Which reminds me.
I intend to automatically generate the THANKS file from
the git logs, so if someone is listed (proper name and email from
most recent commit) as a commit Author:, then there will be no need
to manually add their name to the THANKS file.  Of course, there will
be a list of other name/email pairs, so the switch won't remove any
names.

>> > I do not understand that. Which other programs do you mean?
>> > Utilities in other operating systems, e.g. BSDs md5sum or what do you
>> > mean? What rationale is behind this?
>>
>> Exactly.  Do any of the NetBSD, FreeBSD, OpenBSD, Solaris, HPUX,
>> AIX, etc. programs (sometimes called "md5") have an option to do this?
>> If there *is* another program with this functionality, then you'd have a
>> good argument for adding a short-named option: to be compatible with it.
>> If not, then no new short-named options, on principle: there will be
>> less risk of conflict with other vendors or evolving/future standards.
>
> Makes sense. But this isn't a strong requirement is it? I don't even
> have access to all this operating systems ;)

With google, there's no need for access to the actual systems.
All I am looking for is assurance that you've performed a
duly-diligent search ;-)

> Hmpf. For now I'm fighting with the unreasonable dependencies of the
> coreutils git version.. automake 1.10a.. from the git repository.. =(

You can just change it to 1.10.1 for now.
[thought it might be easier just to build automake from scratch,
because, if you change it...]
You'll have to be careful not to submit that change as part of your patch.

Also, in your commit log message, be sure to attribute Dan and
to include the debian bug URL.

Thanks for doing all this,

Jim




reply via email to

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