You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Max Bowsher <ma...@ukf.net> on 2003/06/14 21:50:41 UTC
Suggestion: Short option for --no-ignore (RESEND)
I got no response to this.
To clarify, I'm not asking someone to actually do the change, I'm just
trying to find out whether I should bother sending a patch, and if so, which
option letter people would prefer.
Max.
Max Bowsher wrote:
> I find myself typing --no-ignore far more frequently than other long
> options.
>
> Would it be acceptable to add a short option for this?
>
> I guess -i would be the most likely candidate.
>
>
> Max.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Max Bowsher <ma...@ukf.net>.
cmpilato@collab.net wrote:
> "Max Bowsher" <ma...@ukf.net> writes:
>
>> To clarify, I'm not asking someone to actually do the change, I'm just
>> trying to find out whether I should bother sending a patch, and if so,
which
>> option letter people would prefer.
>
> To send a patch without first suffering the obligatory weeks-long
> bikeshed discussion on what precious letter of the alphabet to
> sacrifice for your short option needs would be ... well, it'd be just
> darned unfair. :-)
I knew the discussion was inevitable! So, continuing...
> Seriously, IMO, -i doesn't map well to a *no* ignores option for me.
Personally I think of it as "ignore ignores" :-)
Anyway, -n would probably be the next most logical letter, but it could
quite easily be in great demand for just about every --no-foo or --non-foo
option.
However, I could be happy with any letter which gets this change accepted.
> Do you do this operation frequently enough to justify just setting
> your .subversion/config global_ignores config option to the empty
> string?
This wouldn't work for me - my primary use of --no-ignore is to override
existing svn:ignore properties, in order to dump a list of files to delete
to return a tree to an as-checked-out condition.
>> Max Bowsher wrote:
>>> I find myself typing --no-ignore far more frequently than other long
>>> options.
>>>
>>> Would it be acceptable to add a short option for this?
>>>
>>> I guess -i would be the most likely candidate.
Max.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by cm...@collab.net.
"Max Bowsher" <ma...@ukf.net> writes:
> I got no response to this.
>
> To clarify, I'm not asking someone to actually do the change, I'm just
> trying to find out whether I should bother sending a patch, and if so, which
> option letter people would prefer.
To send a patch without first suffering the obligatory weeks-long
bikeshed discussion on what precious letter of the alphabet to
sacrifice for your short option needs would be ... well, it'd be just
darned unfair. :-)
Seriously, IMO, -i doesn't map well to a *no* ignores option for me.
Do you do this operation frequently enough to justify just setting
your .subversion/config global_ignores config option to the empty
string? If so, you could save yourself the hassle of a bikeshed
discussion, a patch composition, and the long option typing. :-)
> Max Bowsher wrote:
> > I find myself typing --no-ignore far more frequently than other long
> > options.
> >
> > Would it be acceptable to add a short option for this?
> >
> > I guess -i would be the most likely candidate.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Michael Wood <mw...@its.uct.ac.za>.
On Wed, Jun 18, 2003 at 08:13:24AM +0100, Max Bowsher wrote:
> Robert Pluim wrote:
[snip]
> > In that case it might be better to build in a different directory
> > than your source[1], in which case your cleanup tool becomes 'rm'.
> >
> > Robert
> >
> > Footnotes:
> > [1] If your build env supports it of course ;-)
>
> And there, of course, is the issue.
>
> You cannot, for example, have autoconf/automake/libtool put their
> files into a seperate tree. Generated .c files (e.g. bison) usually
> get put in the source tree as well.
[snip]
I sometimes do:
$ cp -lR blah blah-build (-l creates hard links)
$ cd blah-build
$ ./autogen.sh
[...]
$ cd ..
$ rm -rf blah-build (-f needed if this is a Subversion wc)
--
Michael Wood <mw...@its.uct.ac.za>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Max Bowsher <ma...@ukf.net>.
Robert Pluim wrote:
> Max Bowsher writes:
> > Justin Erenkrantz wrote:
> > > --On Monday, June 16, 2003 5:34 PM +0100 Max Bowsher <ma...@ukf.net>
wrote:
> > >
> > >> I find that I use --no-ignore about 10 times more often than any
other
> > long
> > >> option for any program (rough, but conservative estimate).
> > >
> > > What exactly are you doing that requires such usage
f --no-ignore? --
> > justin
> >
> > Checking for and/or purging generated files, to restore a tree to an
> > as-checked-out state.
>
> In that case it might be better to build in a different directory than
> your source[1], in which case your cleanup tool becomes 'rm'.
>
> Robert
>
> Footnotes:
> [1] If your build env supports it of course ;-)
And there, of course, is the issue.
You cannot, for example, have autoconf/automake/libtool put their files into
a seperate tree. Generated .c files (e.g. bison) usually get put in the
source tree as well.
Fortunately subversion provides a way to clean up - it's just rather long to
type!
Max.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Robert Pluim <rp...@bigfoot.com>.
Max Bowsher writes:
> Justin Erenkrantz wrote:
> > --On Monday, June 16, 2003 5:34 PM +0100 Max Bowsher <ma...@ukf.net> wrote:
> >
> >> I find that I use --no-ignore about 10 times more often than any other
> long
> >> option for any program (rough, but conservative estimate).
> >
> > What exactly are you doing that requires such usage of --no-ignore? --
> justin
>
> Checking for and/or purging generated files, to restore a tree to an
> as-checked-out state.
In that case it might be better to build in a different directory than
your source[1], in which case your cleanup tool becomes 'rm'.
Robert
Footnotes:
[1] If your build env supports it of course ;-)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Max Bowsher <ma...@ukf.net>.
Greg Stein wrote:
> On Wed, Jun 18, 2003 at 07:49:00AM +0100, Max Bowsher wrote:
>> Justin Erenkrantz wrote:
>>> --On Monday, June 16, 2003 5:34 PM +0100 Max Bowsher <ma...@ukf.net>
wrote:
>>>
>>>> I find that I use --no-ignore about 10 times more often than any other
long
>>>> option for any program (rough, but conservative estimate).
>>>
>>> What exactly are you doing that requires such usage of --no-ignore? --
justin
>>
>> Checking for and/or purging generated files, to restore a tree to an
>> as-checked-out state.
>>
>> It's not a *very* common action, but --no-ignore is the only long option
of
>> any program that I use on a regular basis.
>
> It does seem that your use case is more amenable to scripting. One of our
> design points was to make the output parseable. Seems like you'd want to
> just write a little script kind of like:
>
> #!/bin/sh
> svn status --no-ignore | grep ^I | sed 's/. *//' | xargs rm
>
> Need to deal with spaces in file names and the empty list. But it ought to
> be pretty easily scripted into a quick little "wipe the directory for me"
> script.
You are correct of course. My alternative to sed is "cut -c 8-".
The short option does have an application in viewing the cleanness of a
directory, but this is a less common task, so I am satisfied with a scripted
solution.
Max.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Greg Stein <gs...@lyra.org>.
On Wed, Jun 18, 2003 at 07:49:00AM +0100, Max Bowsher wrote:
> Justin Erenkrantz wrote:
> > --On Monday, June 16, 2003 5:34 PM +0100 Max Bowsher <ma...@ukf.net> wrote:
> >
> >> I find that I use --no-ignore about 10 times more often than any other
> long
> >> option for any program (rough, but conservative estimate).
> >
> > What exactly are you doing that requires such usage of --no-ignore? --
> justin
>
> Checking for and/or purging generated files, to restore a tree to an
> as-checked-out state.
>
> It's not a *very* common action, but --no-ignore is the only long option of
> any program that I use on a regular basis.
It does seem that your use case is more amenable to scripting. One of our
design points was to make the output parseable. Seems like you'd want to
just write a little script kind of like:
#!/bin/sh
svn status --no-ignore | grep ^I | sed 's/. *//' | xargs rm
Need to deal with spaces in file names and the empty list. But it ought to
be pretty easily scripted into a quick little "wipe the directory for me"
script.
Cheers,
-g
--
Greg Stein, http://www.lyra.org/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Max Bowsher <ma...@ukf.net>.
Justin Erenkrantz wrote:
> --On Monday, June 16, 2003 5:34 PM +0100 Max Bowsher <ma...@ukf.net> wrote:
>
>> I find that I use --no-ignore about 10 times more often than any other
long
>> option for any program (rough, but conservative estimate).
>
> What exactly are you doing that requires such usage of --no-ignore? --
justin
Checking for and/or purging generated files, to restore a tree to an
as-checked-out state.
It's not a *very* common action, but --no-ignore is the only long option of
any program that I use on a regular basis.
Max.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Justin Erenkrantz <ju...@erenkrantz.com>.
--On Monday, June 16, 2003 5:34 PM +0100 Max Bowsher <ma...@ukf.net> wrote:
> I find that I use --no-ignore about 10 times more often than any other long
> option for any program (rough, but conservative estimate).
What exactly are you doing that requires such usage of --no-ignore? -- justin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Max Bowsher <ma...@ukf.net>.
Justin Erenkrantz wrote:
> --On Saturday, June 14, 2003 10:50 PM +0100 Max Bowsher <ma...@ukf.net>
wrote:
>
>> I got no response to this.
>>
>> To clarify, I'm not asking someone to actually do the change, I'm just
>> trying to find out whether I should bother sending a patch, and if so,
which
>> option letter people would prefer.
>
> Ick on -i because that has the connotation of 'interactive' from
cp/mv/rm-land.
>
> I don't see the need for a short option here, FWIW. Short options should
only
> be reserved for something that is so distressingly common that there is an
> overwhelming groundswell of support for it, or a strong precedent. --
justin
I find that I use --no-ignore about 10 times more often than any other long
option for any program (rough, but conservative estimate).
Evidently, I'm somewhat in the minority here, given the lack of responses.
I guess I'll just shell-script around the issue.
Max.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Suggestion: Short option for --no-ignore (RESEND)
Posted by Justin Erenkrantz <ju...@erenkrantz.com>.
--On Saturday, June 14, 2003 10:50 PM +0100 Max Bowsher <ma...@ukf.net> wrote:
> I got no response to this.
>
> To clarify, I'm not asking someone to actually do the change, I'm just
> trying to find out whether I should bother sending a patch, and if so, which
> option letter people would prefer.
Ick on -i because that has the connotation of 'interactive' from cp/mv/rm-land.
I don't see the need for a short option here, FWIW. Short options should only
be reserved for something that is so distressingly common that there is an
overwhelming groundswell of support for it, or a strong precedent. -- justin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org