Difference between --date-order and reverse chronological order?

Difference between --date-order and reverse chronological order?

[Dun Peal]

Hi.

The git-log manpage defines --date-order:

"This option is similar to --topo-order in the sense that no parent
comes before all of its children, but otherwise things are still
ordered in the commit timestamp order."

But by Git's definition, for a child commit to be created, its parent
must already exist. So even in reverse chronological order, all
parents should come after all their children, no?

Thanks, D.

Re: Difference between --date-order and reverse chronological order?

[Junio C Hamano]

Dun Peal <dunpealer@gmail.com> writes:

> But by Git's definition, for a child commit to be created, its parent
> must already exist. So even in reverse chronological order, all
> parents should come after all their children, no?

I think "distributed" and "your clock may be skewed" would solve that
puzzlement ;-)

Re: Difference between --date-order and reverse chronological order?

[Johannes Sixt]

Am 27.04.2011 19:36, schrieb Junio C Hamano:
> Dun Peal <dunpealer@gmail.com> writes:
> 
>> But by Git's definition, for a child commit to be created, its parent
>> must already exist. So even in reverse chronological order, all
>> parents should come after all their children, no?
> 
> I think "distributed" and "your clock may be skewed" would solve that
> puzzlement ;-)

Are you saying that given this history:

      E----D
    /     /
   A--B--C

* we can get D-C-B-E-A or D-E-C-B-A with --topo-order

* we can get the above plus D-C-E-B-A with --date-order

* and with neither --topo-order nor --date-order we can also get
D-E-A-C-B or D-C-B-A-E if there was sufficient clock skew when the
commits were created. How would such a clock skew have looked like?

-- Hannes