[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Bug: svn revert annoys the user by behaving differently from all other commands

From: Markus Bertheau ☭ <twanger_at_bluetwanger.de>
Date: 2005-03-22 12:37:39 CET

В Втр, 22/03/2005 в 11:45 +0100, Nathan Maman пишет:

> The person said to forget the change, the computer never decided
> anything. But I understand the need to protect as much as possible from
> the human mistakes.

I was arguing against 'svn forget' in general. I see little point in it.

> If the user does not read the documentation, how did he hear about 'svn
> revert' in the first place? ;-)

He reads svn help and then svn help revert: "Restore pristine working
copy file (undo most local edits)." He does not read that the usage
information for svn revert includes a --recursive argument. He assumes
(and rightly so) that the command behaves like all other svn commands he
used before. He issues svn revert and gets the usage message again. He
is annoyed. Next time he already knows about svn revert but don't
remembers the fact that it's not recursive by default. He is again
annoyed plus he is interrupted in his work flow.

> Jokes apart, if that is want you want, you would probably be better off
> disabling the 'svn revert' command completely at your site: if a user
> wants to get back to the old version, he has to use the operating system
> rm/del and then 'svn update' the working copy... that is assuming the
> user knows what he is doing when he does a rm/del.
>
> So what you need now is an 'undo rm/del' command that comes from your
> operating system.

svn revert is fine in itself, it just shouldn't annoy the user. I don't
see the point of that paragraph.

> >Plus, at 1 o'clock I might think that I really want to forget them
> >forever, and two hours later still notice that I was mistaken. Errare
> >humanum est. The computer should account for that.
> >
> >
> Suppose during these two hours you have made 5 reverts. According to
> you: should we keep all of them? what would be the correct course of
> action?

Given the fact that at present time, all these local changes are lost,
it would be an improvement already to just have the last local mod after
a revert available.

> > If your ambitions are overdrawn, you can't succeed. I don't want to
> > sacrifice the concept itself for an overdrawn interpretation of it.

> I agree completely. But that does not answer the questions:
>
> 1. how many reverts should we keep?

More than zero is an improvement. I suggest 1. More than one revert in a
row are sufficiently seldom.

> 2. for how long?

Until the next revert.

> 3. should we keep them for any file type and any file size?

I don't see why not. In regard to this svn revert should behave
similarly to what other commands do with extraordinary things.

> PS: if you think this discussion does not belong to the list, just say
> so and we will keep it for ourselves.

I think it's fine here.

Markus

-- 
Markus Bertheau ☭ <twanger@bluetwanger.de>
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Tue Mar 22 12:40:14 2005

This is an archived mail posted to the Subversion Users mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.