On Sat, Mar 16, 2013 at 2:47 PM, Joe Abley <jabley(_at_)hopcount(_dot_)ca>
wrote:
What useful history can you possibly get when each file is only ever
substantively changed by publishing another file?
Aue Te Ariki! He toki ki roto taku mahuna!
On 2013-03-16, at 14:21, James Cloos <cloos(_at_)jhcloos(_dot_)com> wrote:
"JL" == John Levine <johnl(_at_)taugh(_dot_)com> writes:
JL> In practice, rsync works great. I pull the RFCs and I-Ds every night.
The benefits git would bring are history and reduced badwidth per pull
when updating.
'reduced bandwidth'? I suppose you lose on the initial 'what changed'
conversation with git vs rsync, but really beyond that you're going to
xfer the same bits...
this seems specious as well, to me at least.
The latter, of course, is only really relevant when the daily pull is
done over a low-bandwidth link. (The impact of the cron job was very
noticeable over a DS0; it is not at all noticeable over doccis.)
-JimC
--
James Cloos <cloos(_at_)jhcloos(_dot_)com> OpenPGP: 1024D/ED7DAEA6