Emit a fossil map that reposurgeon can use.

Simulate CVS's default ignore patterns in a generated .gitignore?

Michael Haggerty writes:
> In 2009 I added tests demonstrating some of the erroneous behavior
> of git-cvsimport.  The failing tests in t9601-t9603 [in the git
> tree] are concrete examples of the problems mentioned in the
> manpage.

Mine the cvs2svn design notes for idea:
http://cvs2svn.tigris.org/source/browse/*checkout*/cvs2svn/trunk/doc/design-notes.txt

Merge the msysgit port patches from http://repo.or.cz/w/cvsps/4msysgit.git?

Discard branch-root tags if they have a name that's a prefix of a tip tag.

Make regression-test loads that create funky and invalid tags.  Add some
more test loads derived from CVS repos found in the wild.

git-cvsimport is a wrapper around cvsps; talk with its maintainer
and simplify that code.

Branch-merge detection.  Not sure it's possible but I want to look
into it.  Are merge_ tags standardized enough to use?

Way in the future...does this belong in C?  Maybe not.  Performance
is going to be limited by network latency anyway, so Python or Ruby
might be better tools - code size would drop by about x4, I think.

Bugs on the git-cvsimport page (some others have been fixed):

    Problems related to timestamps:

	If any files were ever "cvs import"ed more than once (e.g., import
	of more than one vendor release) the HEAD contains the wrong
	content.

    Problems related to branches:

	All files from the branching point are added to a branch even if
	never added in CVS.

	This applies to files added to the source branch after a daughter
	branch was created: if previously no commit was made on the
	daughter branch they will erroneously be added to the daughter
	branch in git.

Can we fix any of these?  We should develop test loads to demonstrate.

