X-Git-Url: https://arthur.barton.de/cgi-bin/gitweb.cgi?p=bup.git;a=blobdiff_plain;f=HACKING;h=cb576bdf7e64efa813218169b4d9adb314b4e948;hp=31f48a556336b133af5cfc46fa2ed0526ec97336;hb=HEAD;hpb=0144a07edc1a00b065bda208c36986b8201fb15a diff --git a/HACKING b/HACKING index 31f48a5..0c8c719 100644 --- a/HACKING +++ b/HACKING @@ -19,10 +19,21 @@ via email. Current Trajectory ================== -Now that we've finished the 0.31 release, we're working on 0.32, and +Now that we've finished the 0.33 release, we're working on 0.34, and although we're not certain which new features will be included, we're considering: + - Migrating hashsplitting to C. + + - Automatically splitting trees to avoid having to save large tree + objects for large directories even if only a few files have + changed or been added (e.g. maildirs). + + - Moving all of the compoents of the index to sqlite. Right now the + main index is an mmapped file, and the hard link and metadata + databases are pickled. As a result the index isn't transactional + and suffers from bugs caused by "skew" across the components. + - Better VFS performance for large repositories (i.e. fuse, ls, web...). @@ -45,7 +56,7 @@ posted to the list, or post your own. (See "ways to help" below.) More specific ways to help ========================== -Testing -- yes please. +Testing -- yes please. With respect to patches, bup development is handled via the mailing list, and all patches should be sent to the list for review (see @@ -69,6 +80,11 @@ For example: ./pytest test/int/test_git.py ./pytest test/ext/test-ftp +If you have the xdist module installed, then you can specify its `-n` +option to run the tests in parallel (e.g. `./pytest -nauto ...`), or +you can specify `-j` to make, which will be translated to xdist with +`-j` becoming `-nauto` and `-jN` becoming `-nN`. + Internal tests that test bup's code directly are located in test/int, and external tests that test bup from the outside, typically by running the executable, are located in test/ext. @@ -135,3 +151,9 @@ ideas here aren't altogether terrible: In particular, we've been paying at least some attention to the bits regarding Acked-by:, Reported-by:, Tested-by: and Reviewed-by:. + +