Re: [Gnu-arch-users] GNU Arch review - am I accurate?

[Brian May]

>>>>> "David" == David A Wheeler <address@hidden> writes:

    David> There are some things I didn't see:
    David> * Is anyone currently working on automated caching?

For what purpose?

There is currently mirrors which serves:
* backups.
* faster access.
* off line access, eg. laptop computers.

There is also revision libraries (and pristine trees) which serves:
* faster speed for some operations.

I suspect when you ask you question, you consider these to deficient,
but in what way?

    David> * Is it even slightly plausible to change the default
    David> filename/tagname conventions so arch will

edit {arch}/=tagging-method ?

    David> work more easily with common tools (e.g., vi/vim, more, csh,
    David> bash, Windows (it doesn't handle long names well))?
    David> Conventions are so arbitrary, yet the ones arch uses
    David> seem designed to cause unnecessary problems.

Not sure what problems you are referring to here?

Possibly I have misunderstood what you are referring to.

    David> * Is there any reason that "mv" and "move" couldn't be the
    David> same thing (and let mv-id or an mv flag be the id mover)?

Why do you see this as being an issue? Personally I don't care either
way.

    David> * Has anyone thought about the "signing of signing" issue
    David> (A signs A's code, B accepts it, C accepts B's, and we
    David> have a chain of signatures from all 3 showing the transition)?
    David> Centralized systems don't need this as much, but distributed
    David> systems need more if you're going to show where code came from.

I don't think you need a chain of signatures, all you would need is to
support multiple signatures. That has been debated in the past, read
the archives ;-).

    David> * Is there an intent to fix the remaining problems in the
    David> native Windows port (e.g., symlinks, newline oddities)?

I believe people are working on the Windows port, see the mailing list
archives, I haven't read them in detail.

    David> Also - has anyone tried to compare BitKeeper and Arch in
    David> detail?  For example, BitKeeper claims its 3-way merge is
    David> better than anyone's, and Monotone claims its 3-way merge
    David> is better than Arch's, but I'd love to see a more detailed
    David> comparison.  I did find some specific information on
    David> BitKeeper, for example:
    David> http://www.kerneltraffic.org/kernel-traffic/kt20030323_210.txt

See <URL:http://better-scm.berlios.de/comparison/>.

BitKeeper does come with a fancy GUI for merging, I haven't used it
enough to come to any conclusion if it helps or not. As I am in the
process of contributing changes to Arch, I suspect I wouldn't be
eligible for the free software license of BitKeeper anymore either, so
I can't easily check.
-- 
Brian May <address@hidden>