| Date | Commit message (Collapse) |
|---|
|
This makes it easier to notice a typo, too.
|
|
It's faster than doing it via real Rack middleware.
|
|
This needs a LOT of work.
|
|
We need to implement our own hash functions for splitting
databases across multiple files. This was totally fucking up
Rainbows!
|
|
Most (other) users only need a single file, even though
my primary use of this is for multiple files.
|
|
It can be confusing with printf-style patterns in URIs,
so just force path_pattern-users to specify "/" as their
path.
|
|
TDB supports multiple simultaneous readers and writer
*processes*, not just threads.
|
|
These allow serving pre-compressed data off disk and
on-the-fly uncompressing for the few clients that do
not accept compressed responses.
|
|
Useful as a proof-of-concept and for benchmark base.
|
|
Some other databases will be able to utilize these
variables.
|
|
It will be useful when we support other backends.
|
|
It's less typing and less likely to clash on both
the eyes and the interpreter.
|
|
Basically working, for now
|