I'm new to drupal, drush and git, not had much sql exposure for 20 years but experienced in unix systems and all the usual software technologies. Please tell me nicely if I posted in the wrong place where I should have posted it.
I'm setting a drupal site up so I can post offline to a virtual machine on my laptop, track my development in git (fabulous tool, well done Linus). This I have working fine on several vm instances and my remote server. However, as I'd like the database backed in a git repo I was playing with ways to do that in combination with drush sql-sync and drush sql-dump when I noticed that the database that drush sql-sync produces (located with %dump in a drush alias) has differences to the one produced by drush sql-dump --result-file=..). That is, if I sql-sync from another machine and I then sql-dump the database into a file its not the same as the one that was dumped during the sync.
Why ?
Further to this I'm trying to figure out a scheme where I know there is a copy of the current database with a known name inside the git scope but its a little tricky if the commands to dump produce different results to the ones produced by sql-sync. I'm beginning to feel these sql database results are non-deterministic :-). Maybe they are ;-) (note on each machine there is only *one* database for drupal)
andy
Cheers
andy