[Gnu-arch-users] Re: tla update edge case

[Miles Bader]

Aaron Bentley <address@hidden> writes:
>> IOW, it updates the current tree with respect to VERS `underneath' any
>> changes between VERS and the current tree.  This formulation is
>> _incredibly_ useful (when you're dealing with a fork of VERS, this is
>> almost always what you want to do), and I find it `just makes sense.'
>
> Neat.  I just tried it somewhere where I normally use star-merge, and it 
> worked like a charm.  Produced some scary-ass change reports, though. 
> Does it have any particular advantages over star-merge?

It has the normal `update advantage' that conflicts result in _your_
changes being left in the .rej files, not the changes in VERS, which I
find is often much easier to deal with (whether or not you prefer this
to `star-merge --three-way' is a matter of taste and circumstance --
--three-way is often nice, but sometimes really weird; .rej files from
update are usually very straightforward).

Also, it feels a lot less `magic' than star-merge, which is nice.

-Miles
-- 
((lambda (x) (list x x)) (lambda (x) (list x x)))