Echoing a post: bzr vs. git

Matthew D. Fuller fullermd at
Tue Nov 11 13:52:09 GMT 2008

> The choice is between using something that's mildly inconvenient all
> the time, and something that's incredibly convenient most of the
> time and having to use something else in the cases where it's
> useless, and to me that's not even a choice.  1 is better than 2,
> but 2 excellent is better than 1 sufficient.

Or, by analogy.  A hammer can put in both nails and screws.  A
screwdriver is only good for screws, but it's real good for them.  The
screwdriver is a lot better some of the time, but there's overhead in
having to know two tools, have two tools, and know when to use each.
If there were practically no screws, it may not be worth having and
knowing anything but a hammer.

But I contend that firstly, there are a lot of screws.  First and a
halfly, there are even more screws than nails.  And secondly, the
overhead in learning and having both a hammer and a screwdriver, and
knowing when to use each, is miniscule.  Any one of the 2.5 conditions
alone would be sufficient to justify having both around, and I think
all of them are overwhelmingly true.

Now, there are attributes of the bzr interface that make it much
harder than it has to be for you to choose to use a hammer all the
time if you want.  -r should just accept revid's without having to add
the "revid:" disambiguator.  Various commands should make it easier to
show revid's.  I absolutely agree those things should be fixed.  But
you don't have to pawn off the revnos to do it.

Matthew Fuller     (MF4839)   |  fullermd at
Systems/Network Administrator |
           On the Internet, nobody can hear you scream.

More information about the bazaar mailing list