[Fwd: A new user experience]

[John A Meinel]

> The following message was posted to the mercurial list, regarding an
> unpleasant initial experience with mercurial. I figured it wouldn't hurt
> to post it here to be assured that bzr doesn't currently have any
> "features" like this, and to help all the coders remember to avoid
> anything like this in the future.
>
> I would experiment with this in bzr myself, but alas, I can't run bzr
> any more. Apparently it now requires breezy, and I'm timidly sticking
> with hoary for a few more weeks because I can't afford to risk any
> downtime on this machine right now.
>
> Kevin
>

Interesting. To my understanding, bzr does indeed have this same
performance. "bzr revert" reverts the entire tree back to it's current
state. And the current state of an uncommitted tree is completely empty.

On the other hand, I believe we ignore .o files by default.

I do believe there is work being done (if not completed) such that doing a
revert on a working tree will at least create backup files, or possibly
not delete files that have newly been added.

There was certainly a discussion about it within the last month (probably
week).

Thanks for a little reminder, though.

John
=:->

>
> -------- Original Message --------
> Subject: A new user experience
> Date: Tue, 18 Oct 2005 23:15:12 +0100
> From: Dan Lewis <actiondan at gmail.com>
> To: mercurial at selenic.com
>
> Hello all,
>
> I'm new to mercurial, so I just thought I'd share a new user experience
> with you. I tried an initial commit of a project I started today, using
> version 0.7. Following the advice of the QuickStart wiki page, I typed
> the following commands:
>
>     hg init
>     hg addremove
>     hg status
>
> There were a number of unwanted files shown as added (object files
> etc...), due to not setting up the .hgignore file properly. Guessing
> from the list of commands how to get back to the state before the
> addremove, I bravely typed:
>
>     hg revert
>
> ...and promptly wiped out the day's work. Don't get me wrong - I'm not
> aiming to kick anybody but myself for this mistake (which I did, several
> times), but I thought you may find the story useful. Should revert be
> this destructive? Should there have been a warning? Should I have read
> the effing manual (ok... don't answer that one).
>
> I'm going to persevere, just with a little more caution.
>
> Regards,
> Dan.
>
> _______________________________________________
> Mercurial mailing list
> Mercurial at selenic.com
> http://selenic.com/mailman/listinfo/mercurial
>
>
>