Requesting suggestions for a good sample "prepare-commit-msg" hook

Requesting suggestions for a good sample "prepare-commit-msg" hook

[Kaartic Sivaraam]

Hello all,

There's an attempt in progress to replace the part of the sample
"prepare-commit-msg" hook that comments the "Conflicts" part of a merge
commit message. This is being done as the "Conflicts" portion is
commented by default for quite some time (since 2014) and thus the
script in the hook is useless.

It's worth noting that it was the only part of the sample hook that
*wasn't* commented by default i.e., it was the only portion of the
sample hook that could be used just by enabling it by renaming the
hook. So, just removing that portion of the solution wouldn't be a
solution as it wouldn't be nice to have a whole sample hook to be
commented for various reasons. So, we're searching for scripts that
could both be  helpful to users and could serve as a good sample to
prove what could be done using the hooks.

In case anyone has any suggestions about what could prove to be an
alternative please send them in!

-- 
Kaartic

Re: Requesting suggestions for a good sample "prepare-commit-msg" hook

[Junio C Hamano]

Kaartic Sivaraam <kaarticsivaraam91196@gmail.com> writes:

> So, we're searching for scripts that
> could both be  helpful to users and could serve as a good sample to
> prove what could be done using the hooks.

I am not so sure that we are searching for them, to be honest (who
are we in this context anyway?)  

This sounds like a solution looking for a problem to me.  The
simplest option may be to remove the sample hook---perhaps it
outlived its usefulness.

Re: Requesting suggestions for a good sample "prepare-commit-msg" hook

[Kaartic Sivaraam]

On Wed, 2017-07-05 at 10:00 -0700, Junio C Hamano wrote:
> Kaartic Sivaraam <kaarticsivaraam91196@gmail.com> writes:
> 
> > So, we're searching for scripts that
> > could both be  helpful to users and could serve as a good sample to
> > prove what could be done using the hooks.
> 
> I am not so sure that we are searching for them, to be honest (who
> are we in this context anyway?)
I think I misinterpreted your sentence in one of the other mails (found
below),

    That sounds like a sample that is there not because it would be
    useful, but because we couldn't think of any useful example.


> This sounds like a solution looking for a problem to me.  The
> simplest option may be to remove the sample hook---perhaps it
> outlived its usefulness.
Removing it is going to be done for sure. I thought it would be even
better to replace it for the reasons I specified previously. Anyways,
in case the last path I sent[1] seems not to be useful and not other
interesting alternatives turn up in this thread then the "simplest
option" has to be chosen.


[1]: http://public-inbox.org/git/20170705165114.20662-1-kaarticsivaraam91196@gmail.com/raw

-- 
Kaartic

Re: Requesting suggestions for a good sample "prepare-commit-msg" hook

[Kaartic Sivaraam]

> On Wed, 2017-07-05 at 10:00 -0700, Junio C Hamano wrote:
> > I am not so sure that we are searching for them, to be honest (who
> > are we in this context anyway?)
> 
Imagining HTML could be used in plain-text,

<strike> I think I misinterpreted your sentence in one of the other
mails (found below), </strike> Sorry for that sloppy sentence. More
correctly, the "we" in that context is the same as the "we" in the
context of the text quoted below,

>     That sounds like a sample that is there not because it would be
>     useful, but because we couldn't think of any useful example.
> 

Link to the post that has the quoted text,
http://public-inbox.org/git/%3Cxmqqy3s7nbkm.fsf@gitster.mtv.corp.google.com%3E/

Re: Requesting suggestions for a good sample "prepare-commit-msg" hook

[Junio C Hamano]

Kaartic Sivaraam <kaarticsivaraam91196@gmail.com> writes:

>> On Wed, 2017-07-05 at 10:00 -0700, Junio C Hamano wrote:
>> > I am not so sure that we are searching for them, to be honest (who
>> > are we in this context anyway?)
>> 
> Imagining HTML could be used in plain-text,
>
> <strike> I think I misinterpreted your sentence in one of the other
> mails (found below), </strike> Sorry for that sloppy sentence. More
> correctly, the "we" in that context is the same as the "we" in the
> context of the text quoted below,
>
>>     That sounds like a sample that is there not because it would be
>>     useful, but because we couldn't think of any useful example.
>> 
>
> Link to the post that has the quoted text,
> http://public-inbox.org/git/%3Cxmqqy3s7nbkm.fsf@gitster.mtv.corp.google.com%3E/

I honestly do not see your point.  Yes, I said that the change
indicates that there is no useful example found (so far).  That does
not necessarily mean that we must find a useful example so that we
can keep this sample script, which now became useless, alive.  

I am questioning the assumption that it helps users to have a sample
for prepare-commit-msg shipped with our source, and I suspect that
it may no longer be true.  If the sole purpose of finding a useful
example is to keep the sample script alive, when the sample script
is no longer a useful thing to ship, then it does sound like "a
solution looking for a problem", no?

Re: Requesting suggestions for a good sample "prepare-commit-msg" hook

[Kaartic Sivaraam]

On Wed, 2017-07-05 at 12:19 -0700, Junio C Hamano wrote:
> I honestly do not see your point.  Yes, I said that the change
> indicates that there is no useful example found (so far).  That does
> not necessarily mean that we must find a useful example so that we
> can keep this sample script, which now became useless, alive.  
> 
> I am questioning the assumption that it helps users to have a sample
> for prepare-commit-msg shipped with our source, and I suspect that
> it may no longer be true.  If the sole purpose of finding a useful
> example is to keep the sample script alive, when the sample script
> is no longer a useful thing to ship, then it does sound like "a
> solution looking for a problem", no?
> 
Ah, now I see your point. When you said "it's no longer useful", I
thought you were mentioning only the part of the script that commented
the sign-off. Now I get that your point of not shipping the sample
script itself. Now that clears the confusion and sorry for the
confusion I may have caused.

-- 
Kaartic