Still no dice on HHVM even with installing php5-dev. Added PHP 5.6 to see how
we fare. Will drop the allowed failure when 5.6 is gold or when the test suite
is passing. Whichever comes first.
Seems the format being returned doesn't start with ip... updated RegEx to check
for a 45 character alphanumeric instead of ip + 43 character alpha numeric
Probably need to rename the minify flag since it now also represents compiling the CSS. Better yet, probably better to eliminate it entirely and make it an assumption that you should only be compiling when on the "local" environment or when the minified file is completely missing.
I destroyed my old AYAH repo in favor of a fork of the official stuff that I updated. Had to add some vcs entry to get composer to pull from my fork and not the main package.
Characters were being split up and causing invalid sequences when using
`substr()`. Went ahead and updated to use `mb_strcut()` and forcing the
character encoding to UTF-8. I think the plan down the road will be to set the
internal encoding to UTF-8 but I am not currently sure how that could effect
the rest of the system (perhaps it won't). Closes#39
Unsure why but Gemnasium kept showing that it was refreshing but never seemed
to load the dependencies even though the file was read in correctly. Will check
in a week, thinking perhaps they don't refresh nearly as frequently as I would
have liked.
The vendor directory is now just composer stuff. Planning to rearrange some
files soon but wanted to commit these changes. AYAH is now installed via
composer from a package I am maintaining and I dropped the test helpers as it
is a dev-only requirement (and that's assuming you even want to test).
Had to rework the Model class a bit, there's some weirdness happening and I'm
unsure if it's part of the rewrite or always been busted. Won't really know for
sure until I start porting sites over to it I suppose.
Not being used and the logic is pretty damn hacky. I don't believe in JOINs so
I'm unsure if this support will be re-added in the future or if there will
simply be a backed in opinion that JOINs are the devil.
Never gets used, ended up being somewhat MySQL specific as PostgreSQL favors
letting the server handle it instead of hinting at it. Write better queries I
suppose?
