Is there any interest in indentation patches?

Is there any interest in indentation patches?

[Richard Hartmann]

I have been going through a few functions and there is no clear style
when it comes to indentation. Some files use whitespaces, some use tabs,
many use a combination of the two. Some use 2 chars, some use 4. No idea
if the commit noise is wanted, but I could prepare a batch of patches to
fix this. It's work no one wants to do, anyway :p

Is that wasted effort or a worthwhile thing to do?


Richard

Re: Is there any interest in indentation patches?

[Andrey Borzenkov]

[-- Attachment #1: Type: text/plain, Size: 835 bytes --]

On Sunday 20 July 2008, Richard Hartmann wrote:
> 
> I have been going through a few functions and there is no clear style
> when it comes to indentation. Some files use whitespaces, some use tabs,
> many use a combination of the two. Some use 2 chars, some use 4. No idea
> if the commit noise is wanted, but I could prepare a batch of patches to
> fix this. It's work no one wants to do, anyway :p
> 

If you will start it, may I ask to add standard mode lines for vi
(and possibly emacs) so that selected style is ensured in the future?

> Is that wasted effort or a worthwhile thing to do?
> 

In general, I guess, yes. The main problem with functions is they
have overly long lines, specifically where _arguments is used.

May be we could switch off line folding although I always has mixed
feelings about it.

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

Re: Is there any interest in indentation patches?

[Richard Hartmann]

On Sun, Jul 20, 2008 at 07:40, Andrey Borzenkov <arvidjaar@newmail.ru> wrote:

> If you will start it, may I ask to add standard mode lines for vi
> (and possibly emacs) so that selected style is ensured in the future?

Yes, if I do it, that would be a Good Thing.


> In general, I guess, yes. The main problem with functions is they
> have overly long lines, specifically where _arguments is used.

Should be doable :)


Richard

Re: Is there any interest in indentation patches?

[Nikolai Weibull]

On Sun, Jul 20, 2008 at 10:33, Richard Hartmann
<richih.mailinglist@gmail.com> wrote:

> On Sun, Jul 20, 2008 at 07:40, Andrey Borzenkov <arvidjaar@newmail.ru> wrote:

>> If you will start it, may I ask to add standard mode lines for vi
>> (and possibly emacs) so that selected style is ensured in the future?

> Yes, if I do it, that would be a Good Thing.

Please don't.  Adding mode-lines is not the right way to maintain
consistency.  Commit hooks, as you mention elsewhere, are far
superior.

Re: Is there any interest in indentation patches?

[Richard Hartmann]

On Sun, Jul 20, 2008 at 11:49, Nikolai Weibull <now@bitwi.se> wrote:

> Please don't.  Adding mode-lines is not the right way to maintain
> consistency.  Commit hooks, as you mention elsewhere, are far
> superior.

The modeline I would have the most interest in would be

  set ft=zsh

so my highlighting works correctly :)


Richard

Re: Is there any interest in indentation patches?

[Nikolai Weibull]

On Sun, Jul 20, 2008 at 14:17, Richard Hartmann
<richih.mailinglist@gmail.com> wrote:
> On Sun, Jul 20, 2008 at 11:49, Nikolai Weibull <now@bitwi.se> wrote:

>> Please don't.  Adding mode-lines is not the right way to maintain
>> consistency.  Commit hooks, as you mention elsewhere, are far
>> superior.

> The modeline I would have the most interest in would be
>
>  set ft=zsh
>
> so my highlighting works correctly :)

That will be fixed in Vim 7.3.  Bram didn't want to add my addition to
$VIMRUNTIME/scripts.vim before 7.2.  Anyway, it will look for

  emulate zsh

and the like, which should be in more or less every Zsh script anyway.

Also, you can always add your own patterns to your own filetype.vim, for example

au BufNewFile,BufRead /usr/share/zsh/**/* setf  zsh

(or something similar).

Re: Is there any interest in indentation patches?

[Bart Schaefer]

On Jul 20,  2:21am, Richard Hartmann wrote:
}
} I have been going through a few functions

You mean shell functions as opposed to C functions?

} and there is no clear style when it comes to indentation. Some files
} use whitespaces, some use tabs, many use a combination of the two.
} Some use 2 chars, some use 4. No idea if the commit noise is wanted,
} but I could prepare a batch of patches to fix this.

My personal preference would be that this NOT be done.  It makes it
difficult to see from one file revision to the next what's a semantic
change and what was merely whitespace being fixed up.

It's one thing if the indentation is actually so bad that it impairs
readability -- in that case I'd be more inclined to fix it.  Changing
strictly for the sake of consistency does not seem to me worthwhile.
To maintain it going forward, we'd have to implement a policy of
reformatting every function that is contributed to the sources.

Re: Is there any interest in indentation patches?

[Richard Hartmann]

On Sun, Jul 20, 2008 at 10:05, Bart Schaefer <schaefer@brasslantern.com> wrote:

> You mean shell functions as opposed to C functions?

Yes.


> My personal preference would be that this NOT be done.  It makes it
> difficult to see from one file revision to the next what's a semantic
> change and what was merely whitespace being fixed up.

Yes, that is the the one huge downside of whitespace commits.
If there was a single coding style, that one huge commit would
(possibly) be OK. That is exactly why I asked the ML.


> It's one thing if the indentation is actually so bad that it impairs
> readability -- in that case I'd be more inclined to fix it.  Changing
> strictly for the sake of consistency does not seem to me worthwhile.
> To maintain it going forward, we'd have to implement a policy of
> reformatting every function that is contributed to the sources.

Once the initial effort is finished, maintaining the status quo could
be done via commit hooks and a little checker script.


Richard