Re: [Monotone-devel] Re: cvs import

[Nathaniel Smith]

On Thu, Sep 14, 2006 at 11:50:03AM -0400, Shawn Pearce wrote:
> I don't know how the Monotone guys feel about it but I think Git
> is happy with the data in any order, just so long as the dependency
> chains aren't fed out of order.  Which I think nearly all changeset
> based SCMs would have an issue with.  So we should be just fine
> with the current chronological order produced by cvs2svn.

What I'm mostly concerned about wrt speed is spooling in the file
data, where you have to use some clever algorithms for efficiency.  In
particular, last I heard cvs2svn was still using co(1) to fetch file
texts, which is O(n^2) -- perhaps this has been fixed by now.
Monotone's cvs_importer uses a piece table to reconstruct file texts
efficiently, and shove them straight in to the database with good
locality.  Also, the SVN dump format gives you the file contents
associated each change at the same time it gives you that change; I
don't understand how that could be generated or consumed efficiently,
as compared to slurping in all file texts at once at the beginning.

The actual order the metadata arrives in wouldn't be that big a deal;
we actually have that same LRU cache built-in out of the box :-).

-- Nathaniel

-- 
Damn the Solar System.  Bad light; planets too distant; pestered with
comets; feeble contrivance; could make a better one myself.
  -- Lord Jeffrey