git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l
@ 2008-05-26 11:18 Mark Lawrence
  2008-05-26 12:19 ` Andreas Ericsson
  0 siblings, 1 reply; 8+ messages in thread
From: Mark Lawrence @ 2008-05-26 11:18 UTC (permalink / raw)
  To: git

Hi,

I don't know if this is a bug or just my understanding, but I'm having
trouble matching the behaviour of "cg-switch -l" with the documentation.

According to the docs, -l means that:
                                                ... local changes will
  be "kept" with your previous branch, you will get a pristine tree of
  the new branch and when you switch back to the original branch you
  will also get back the local changes.

However if I run "cg-switch -l -c testing" on a dirty (ie with local
changes) repository I do not get a pristine tree. In fact, it appears 
the local changes are carried over.

Initial state:

    mlawren@lifebook:/tmp/x$ cg-status
    Heads:
        >master     6a8a774cd1fb84f48f906a9ce3322a5ac94deb00
       R origin     cdc9a7857b93b00f515b010c9aeadda4e649b92a

    ? Changes
    ? META.yml
    ? Makefile
    ? blib/
    ? inc/
    ? pm_to_blib
    ? t/benchmark.pl
    ? t/cache.t
    ? t/hyphens.t
    ? t/synopsis.t
    ? x
    M MANIFEST
    M Makefile.PL
    M debian/changelog
    M debian/control
    M lib/XML/API.pm
    A lib/XML/API/Cache.pm
    M lib/XML/API/RSS.pm
    M lib/XML/API/XHTML.pm
    M t/02-langs.t
    M t/03-XML-API.t
    A t/04-parse.t

Create a new local branch and check status:

    mlawren@lifebook:/tmp/x$ cg-switch -l -c testing
    Creating new branch testing: 6a8a774cd1fb84f48f906a9ce3322a5ac94deb00
    Switching to branch testing...

    mlawren@lifebook:/tmp/x$ cg-status
    Heads:
        master     6a8a774cd1fb84f48f906a9ce3322a5ac94deb00
      R origin     cdc9a7857b93b00f515b010c9aeadda4e649b92a
        >testing    6a8a774cd1fb84f48f906a9ce3322a5ac94deb00

    ? Changes
    ? META.yml
    ? Makefile
    ? blib/
    ? inc/
    ? pm_to_blib
    ? t/benchmark.pl
    ? t/cache.t
    ? t/hyphens.t
    ? t/synopsis.t
    ? x
    M MANIFEST
    M Makefile.PL
    M debian/changelog
    M debian/control
    M lib/XML/API.pm
    A lib/XML/API/Cache.pm
    M lib/XML/API/RSS.pm
    M lib/XML/API/XHTML.pm
    M t/02-langs.t
    M t/03-XML-API.t
    A t/04-parse.t

Why do I still have all these modified and additional files? I fully
expected to still see the '?' file not under git/cogito control. However
the uncommitted changes are still all present.

I am also confused by the existence of the '-p' option, which by my
reading implies pretty much the same condition as if you *don't* specify
-l.

Regards,
Mark.
-- 
Mark Lawrence

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l
  2008-05-26 11:18 Mark Lawrence
@ 2008-05-26 12:19 ` Andreas Ericsson
  0 siblings, 0 replies; 8+ messages in thread
From: Andreas Ericsson @ 2008-05-26 12:19 UTC (permalink / raw)
  To: Mark Lawrence; +Cc: git

Mark Lawrence wrote:
> Hi,
> 
> I don't know if this is a bug or just my understanding, but I'm having
> trouble matching the behaviour of "cg-switch -l" with the documentation.
> 

Cogito became officially unmaintained a very long time ago (late 2006?)
and will almost certainly not work properly with any half-recent version
of git. Don't use it.

-- 
Andreas Ericsson                   andreas.ericsson@op5.se
OP5 AB                             www.op5.se
Tel: +46 8-230225                  Fax: +46 8-230231

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l
@ 2008-05-27  8:04 Mark Lawrence
  2008-05-27  8:11 ` Wincent Colaiuta
                   ` (2 more replies)
  0 siblings, 3 replies; 8+ messages in thread
From: Mark Lawrence @ 2008-05-27  8:04 UTC (permalink / raw)
  To: git

> On Mon May 26, 2008 at 02:19:04PM +0200, Andreas Ericsson wrote:
> > Mark Lawrence wrote:
> > >Hi,
> > >
> > >I don't know if this is a bug or just my understanding, but I'm having
> > >trouble matching the behaviour of "cg-switch -l" with the documentation.
> > >
> > 
> > Cogito became officially unmaintained a very long time ago (late 2006?)
> > and will almost certainly not work properly with any half-recent version
> > of git. Don't use it.
> 
> Oh. That's only <sarcasm>slightly annoying</sarcasm> to learn now, given
> that I've spent the past two years learning and converting my projects
> over to cogito :/ Could somebody *please* update the cogito web page at
> http://git.or.cz/cogito/ and put a big (maybe even blinking) notice
> informing people of this?
> 
> Bit of a shame that it's died though, because I appreciated the simpler
> interface (and some of the output I find nicer).
> 
> On a more practical note, am I likely to suffer any issues using pure
> git on repos created with cogito?

And a related question: is there any kind of cogito replacement that
provides a simplified interface? As far as versioning tools go, git is
quite low-level...

Thanks,
Mark.
-- 
Mark Lawrence

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l
  2008-05-27  8:04 [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l Mark Lawrence
@ 2008-05-27  8:11 ` Wincent Colaiuta
  2008-05-27  8:33 ` Ittay Dror
  2008-05-27  8:48 ` Jakub Narebski
  2 siblings, 0 replies; 8+ messages in thread
From: Wincent Colaiuta @ 2008-05-27  8:11 UTC (permalink / raw)
  To: Mark Lawrence; +Cc: git

El 27/5/2008, a las 10:04, Mark Lawrence escribió:
>> On Mon May 26, 2008 at 02:19:04PM +0200, Andreas Ericsson wrote:
>>> Mark Lawrence wrote:
>>>> Hi,
>>>>
>>>> I don't know if this is a bug or just my understanding, but I'm  
>>>> having
>>>> trouble matching the behaviour of "cg-switch -l" with the  
>>>> documentation.
>>>>
>>>
>>> Cogito became officially unmaintained a very long time ago (late  
>>> 2006?)
>>> and will almost certainly not work properly with any half-recent  
>>> version
>>> of git. Don't use it.
>>
>> Oh. That's only <sarcasm>slightly annoying</sarcasm> to learn now,  
>> given
>> that I've spent the past two years learning and converting my  
>> projects
>> over to cogito :/ Could somebody *please* update the cogito web  
>> page at
>> http://git.or.cz/cogito/ and put a big (maybe even blinking) notice
>> informing people of this?
>>
>> Bit of a shame that it's died though, because I appreciated the  
>> simpler
>> interface (and some of the output I find nicer).
>>
>> On a more practical note, am I likely to suffer any issues using pure
>> git on repos created with cogito?
>
> And a related question: is there any kind of cogito replacement that
> provides a simplified interface? As far as versioning tools go, git is
> quite low-level...

The reason why cogito fell out of maintenance is that git itself has  
evolved from the hard-core, close-to-the-metal tool that it was in the  
early days to be usable by mere mortals. What specific features/ 
workflows of cogito are you missing? It's likely that modern git  
already has an equivalent that is just as easy to use.

Wincent

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l
  2008-05-27  8:04 [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l Mark Lawrence
  2008-05-27  8:11 ` Wincent Colaiuta
@ 2008-05-27  8:33 ` Ittay Dror
  2008-05-27  8:48 ` Jakub Narebski
  2 siblings, 0 replies; 8+ messages in thread
From: Ittay Dror @ 2008-05-27  8:33 UTC (permalink / raw)
  To: Mark Lawrence; +Cc: git



Mark Lawrence wrote:
> And a related question: is there any kind of cogito replacement that
> provides a simplified interface? As far as versioning tools go, git is
> quite low-level...
>   
maybe easy-git is for you. http://www.gnome.org/~newren/eg.

> Thanks,
> Mark.
>   

-- 
Ittay Dror <ittayd@tikalk.com>
Tikal <http://www.tikalk.com>
Tikal Project <http://tikal.sourceforge.net>

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l
  2008-05-27  8:04 [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l Mark Lawrence
  2008-05-27  8:11 ` Wincent Colaiuta
  2008-05-27  8:33 ` Ittay Dror
@ 2008-05-27  8:48 ` Jakub Narebski
  2008-05-29 17:28   ` Petr Baudis
  2 siblings, 1 reply; 8+ messages in thread
From: Jakub Narebski @ 2008-05-27  8:48 UTC (permalink / raw)
  To: git; +Cc: git, Andreas Ericsson

Mark Lawrence <nomad@null.net> writes:

>> On Mon May 26, 2008 at 02:19:04PM +0200, Andreas Ericsson wrote:
>>> Mark Lawrence wrote:
>>>>
>>>>I don't know if this is a bug or just my understanding, but I'm having
>>>>trouble matching the behaviour of "cg-switch -l" with the documentation.
>>> 
>>> Cogito became officially unmaintained a very long time ago (late 2006?)
>>> and will almost certainly not work properly with any half-recent version
>>> of git. Don't use it.
>> 
>> Oh. That's only <sarcasm>slightly annoying</sarcasm> to learn now, given
>> that I've spent the past two years learning and converting my projects
>> over to cogito :/ Could somebody *please* update the cogito web page at
>> http://git.or.cz/cogito/ and put a big (maybe even blinking) notice
>> informing people of this?

Git Homepage (http://git.or.cz/) states:

  Cogito

    Cogito was a popular version control system on top of Git, aiming
    at seamless user interface and ease of use. It provided much
    better user interface in the past but later improvements of Git
    made it mostly obsolete and it is not maintained anymore. Please
    migrate to Git itself.

I guess that it should be mentioned stronly also on Cogito homepage,
although the date of latest release ([2006-10-19]) is a bit of hint...

BTW. maintaner (Petr Baudis <pasky@suse.cz>) welcomes patches; the
repository is at http://repo.or.cz/r/git-homepage.git
 
>> Bit of a shame that it's died though, because I appreciated the simpler
>> interface (and some of the output I find nicer).

BTW. I think one of the reasons Cogito died was that it tried to hide
index, and used single branch per repository paradigm when dealing
with remote repositories while git acquired powerful tool (git-remote)
to deal with multi-branches remote repositories... and
author/maintainer didn't have time to modify Cogito to include those
features.

I'd say that Cogito was one of forces behind making Git easier to use
(for example git-filter-branch is the equivalent of cg-admin-rewrite-hist,
a single thing that was possible in Cogito but was not possible in Git,
created when Cogito was announced to be no longer maintained/abandoned).

>> On a more practical note, am I likely to suffer any issues using pure
>> git on repos created with cogito?

I don't thinks so.
 
> And a related question: is there any kind of cogito replacement that
> provides a simplified interface? As far as versioning tools go, git is
> quite low-level...

There are two projects aiming at easier git experience: EasyGit (which
is mainly documentation improvement) and Pyrite (which will include
GUI), but IIRC both are in early stages of development.


P.S. As Cogito is OSS project, you can always [try to] revive it,
I think even without changing its name...

-- 
Jakub Narebski
Poland
ShadeHawk on #git

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l
  2008-05-27  8:48 ` Jakub Narebski
@ 2008-05-29 17:28   ` Petr Baudis
  2008-05-29 22:17     ` fmt, was " Johannes Schindelin
  0 siblings, 1 reply; 8+ messages in thread
From: Petr Baudis @ 2008-05-29 17:28 UTC (permalink / raw)
  To: Jakub Narebski; +Cc: git, Andreas Ericsson

On Tue, May 27, 2008 at 10:48:29AM +0200, Jakub Narebski wrote:
> Mark Lawrence <nomad@null.net> writes:
> 
> >> On Mon May 26, 2008 at 02:19:04PM +0200, Andreas Ericsson wrote:
> >>> Mark Lawrence wrote:
> >>>>
> >>>>I don't know if this is a bug or just my understanding, but I'm having
> >>>>trouble matching the behaviour of "cg-switch -l" with the documentation.
> >>> 
> >>> Cogito became officially unmaintained a very long time ago (late 2006?)
> >>> and will almost certainly not work properly with any half-recent version
> >>> of git. Don't use it.
> >> 
> >> Oh. That's only <sarcasm>slightly annoying</sarcasm> to learn now, given
> >> that I've spent the past two years learning and converting my projects
> >> over to cogito :/ Could somebody *please* update the cogito web page at
> >> http://git.or.cz/cogito/ and put a big (maybe even blinking) notice
> >> informing people of this?
> 
> Git Homepage (http://git.or.cz/) states:
> 
>   Cogito
> 
>     Cogito was a popular version control system on top of Git, aiming
>     at seamless user interface and ease of use. It provided much
>     better user interface in the past but later improvements of Git
>     made it mostly obsolete and it is not maintained anymore. Please
>     migrate to Git itself.
> 
> I guess that it should be mentioned stronly also on Cogito homepage,
> although the date of latest release ([2006-10-19]) is a bit of hint...

Good point. I have added a hint.

> >> Bit of a shame that it's died though, because I appreciated the simpler
> >> interface (and some of the output I find nicer).
> 
> BTW. I think one of the reasons Cogito died was that it tried to hide
> index, and used single branch per repository paradigm when dealing
> with remote repositories while git acquired powerful tool (git-remote)
> to deal with multi-branches remote repositories... and
> author/maintainer didn't have time to modify Cogito to include those
> features.

I think hiding index was not a mistake and certainly not a reason of
death. It died mainly because I lost motivation to extend it as Git
rapidly improved its user interface, and indeed also because of the work
required to implement remotes tracking (I did have some work in progress
that I'm not sure if I ever published; if anyone is interested, I can
push it somewhere).

> I'd say that Cogito was one of forces behind making Git easier to use
> (for example git-filter-branch is the equivalent of cg-admin-rewrite-hist,
> a single thing that was possible in Cogito but was not possible in Git,
> created when Cogito was announced to be no longer maintained/abandoned).

(Not single. I still use cg-export in some of my workflows to export
tree of some commit quickly. And I still did not have time to add
support for |fmt'ing git commit's -m arguments. But these are the
only things that I can think of straight away.)

> >> On a more practical note, am I likely to suffer any issues using pure
> >> git on repos created with cogito?
> 
> I don't thinks so.

Not strictly, but Cogito uses very old-style remote repositories
tracking model. I think Git still technically supports it, but you might
get confused when seeing hints and documentation expecting the modern
setup. You do not need to do anything about the central repositories but
I would advise you to simply re-clone the leaf repositories.

> P.S. As Cogito is OSS project, you can always [try to] revive it,
> I think even without changing its name...

I hoped somebody would when I announced end-of-life from my side, but
frankly I don't think it's worthwhile at this point. Cogito was Git's
counterpart for the whole of Git's early life and it shows in the code
very much - it goes at great lengths to achieve things that you would do
trivially with the newer Git interfaces. It's probably better idea to
hack away at things like EasyGit or even better improve documentation
and usage of Git itself.

-- 
				Petr "Pasky" Baudis
Whatever you can do, or dream you can, begin it.
Boldness has genius, power, and magic in it.	-- J. W. von Goethe

^ permalink raw reply	[flat|nested] 8+ messages in thread

* fmt, was Re: [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l
  2008-05-29 17:28   ` Petr Baudis
@ 2008-05-29 22:17     ` Johannes Schindelin
  0 siblings, 0 replies; 8+ messages in thread
From: Johannes Schindelin @ 2008-05-29 22:17 UTC (permalink / raw)
  To: Petr Baudis; +Cc: Jakub Narebski, git, Andreas Ericsson

Hi,

On Thu, 29 May 2008, Petr Baudis wrote:

> I still did not have time to add support for |fmt'ing git commit's -m 
> arguments.

Note that I have a git-fmt in my personal fork that might have the 
building blocks to provide a --fmt option to git-commit one day.

Ciao,
Dscho

^ permalink raw reply	[flat|nested] 8+ messages in thread

end of thread, other threads:[~2008-05-29 22:19 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-05-27  8:04 [cogito-0.18.2] Documentation/Code Mismatch: cg-switch -l Mark Lawrence
2008-05-27  8:11 ` Wincent Colaiuta
2008-05-27  8:33 ` Ittay Dror
2008-05-27  8:48 ` Jakub Narebski
2008-05-29 17:28   ` Petr Baudis
2008-05-29 22:17     ` fmt, was " Johannes Schindelin
  -- strict thread matches above, loose matches on Subject: below --
2008-05-26 11:18 Mark Lawrence
2008-05-26 12:19 ` Andreas Ericsson

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).