hg-git: indirect cloning screws up the graph?

[Uwe Brauer]

>>> "Sietse" == Sietse Brouwer <sbbrouwer at gmail.com> writes:

Hi Sistse,

   > Hi Uwe,
   > Both your clones know the same history, and are functionally
   > identical. The history is just displayed differently.

   > The order in which `hg log` displays commits is determined by the
   > local rev number: that incrementing number. If you have a bunch of
   > commits 0...N, and you pull in a new buch,  that new bunch will get
   > numbers N+1..., and will be displayed topmost in your `hg log` graph.

Ah I understand that explains it, thanks!
   > auctex-hg-direct pulled directly from Savannah, so it got all commits
   > in the order git presented them, and the `master` commit ended up
   > tipmost.


   > As for your other remark:

   >> hg heads | grep changeset | wc -l
   >> reveals 48 (!) Heads.

   > Nearly all those heads are tags, you can see that like so:  hg log
   > --rev 'head()' --template '{gitnode|short} -- {bookmarks} -- {tags}\n'

Ah, thanks, but now I am confused a tag creates a new head? That is odd,
I have a different, a  pure hg repo, which has 5 heads: 4 branches 1 bookmark
but it has 14 tags.

But indeed the one which is cloned shows me a lot of tags. So the git-hg
plugin converts tags to heads? That is odd. Anyhow thanks for that
command I will added to me aliases.

Uwe