I just tried a CVS-update to see if I could contribute back an adaption that allows for extra ssh options. I've got a client/partner site that I sync with that hosts ssh on a non-standard port, and I need to inject "-p 100" in a few places.
This is do-able, by adding the option to the site-aliases data BUT...
Current CVS scrambled my diff in a way that makes it really hard to compare changes. Indenting is all over the show. Specifically in sql.sync.inc. Maybe more.
Can we run coder.module or a beautifier over the current release please? Tabs are bad.
(the editor seems to have had tabs set at 8 spaces! a replace of \t with ' 'x8 helps a lot)
Comments
Comment #1
dman commentedPS, I'm not contributing this as a patch, because it would be huge and far-reaching, and hard to evaluate (even though it's just whitespace). I'd rather an active committer looked at it. It's probably just a recent oversight that's easily reversed in a working copy.
Comment #2
dman commented... may have been partially an CVS oddness. Another update (using -f to blow away my copy) reduces (but does not eliminate) the tabs.
// $Id: sync.sql.inc,v 1.8 2009/12/28 23:18:37 greg1anderson Exp $
seems to have had a bunch.
// $Id: sync.sql.inc,v 1.10 2010/01/08 20:13:27 greg1anderson Exp $
Has only three tabs
Comment #3
greg.1.anderson commentedThis is my fault. I have a %blinking& editor that screws things up under its default settings. It will bite me if I don't remember to set them.
The transformation is predictable (for me); I will fix it.
Sorry for the inconvenience.
Comment #4
dman commented:-B I know how that happens!
One day I thought that enabling the 'strip trailing whitespace' on my editor would be a good idea...
It wasn't!
Thanks for looking at it.
Comment #5
greg.1.anderson commentedFixed and committed. The patchfile was droll, and therefore omitted.
Comment #6
dman commentedWoot. I see it. Thanks heaps!
I'll open an issue to suggest my ssh options need. My changes are now back to the original 6 lines, not the 50 I saw earlier :-B
Comment #7
greg.1.anderson commentedIs what you need any different than #671542: Backend invoke shouldn't always insist on prohibiting password authentication?
Comment #8
dman commentedYeah, ... damn, probably~!!!! #686384: Allow the site-aliases configs to define extra options when invoking ssh