"hg up -C" is also evil - ;-)

[waldemar]

On 11/08/2010 09:17 AM, Robert Figura wrote:
> Kevin Bullock <kbullock+mercurial at ringworld.org> wrote:
>   
>> On Nov 8, 2010, at 10:37 AM, Haszlakiewicz, Eric wrote:
>>     
>>> The problem is that hg up does *not* invoke the usual merge machinery if you go from one branch to another. 
>>>       
>> That was precisely my point, which is why it might be sensible to write *.orig on an update crossing branches. The required `-C` flag is a clue that it's doing something potentially destructive to your working dir.
>>     
> Improving the workdir managment seems to be a very good idea to me. I
> just don't like backups on a file basis. Because when the same
> situation happens twice you'll have to find out which backups belong
> where. After all a DVCS is all about changesets, as in "Collections
> of files".
>
>   
>> On the other hand, having a more reliable/easy-to-use built-in shelve mechanism is high on my hg wishlist.
>>     
Maybe we could extend mq's a bit by defining a context to them in some
sensible manner.  We could then create and maintain multiple mq stacks
depending on the context.  This is based on observation that in spite of
the existence of two different implementations of shelve functionality,
many still opt to use mq for the purpose even if not quite fitting.
> I second that. Mq's strip has the nice habit of generating bundle
> files, which may be a bit too much since only a single changeset needs
> to be stored, but it does support specifying the base revision it
> belongs to.
>
> BTW, is there a hook to be called before an uncommitted workdir will be
> modified? That might trivialize the effort, at least for this this
> threads perspective =:)
>
> Kind Regards,
>   Robert Figura
>
>