aboutsummaryrefslogtreecommitdiff
path: root/doc
diff options
context:
space:
mode:
authorTim Harder <radhermit@gmail.com>2019-10-05 10:23:14 -0600
committerTim Harder <radhermit@gmail.com>2019-10-05 10:23:53 -0600
commitaf786bb03d052e3a9a631d3d6d0275fc45cab723 (patch)
treef2dd111419d639d4ee0b5558e1485b06a27bcd8b /doc
parentpkgcheck scan: split network options into its own group (diff)
downloadpkgcheck-af786bb03d052e3a9a631d3d6d0275fc45cab723.tar.gz
pkgcheck-af786bb03d052e3a9a631d3d6d0275fc45cab723.tar.bz2
pkgcheck-af786bb03d052e3a9a631d3d6d0275fc45cab723.zip
doc: drop todo entries that are irrelevant or have been finished
[skip ci]
Diffstat (limited to 'doc')
-rw-r--r--doc/dev/todo.rst8
1 files changed, 0 insertions, 8 deletions
diff --git a/doc/dev/todo.rst b/doc/dev/todo.rst
index 36cd4bf0..af35bb89 100644
--- a/doc/dev/todo.rst
+++ b/doc/dev/todo.rst
@@ -1,17 +1,9 @@
- back out package.provided hack in visibility, fix start prototype
-- tests (dar)
-- silence profile deprecation warnings; pkgcore mod, but noting it here since
- it's bit more relevant to pkgcheck
- rework filter-env so it dumps a struct representing the parsing instead of
doing it inline; add python bindings, use that for source flow analysis
- look into trying to identify unused functions in an ebuild; eclasses, and
bash support for f() { echo "monkeys"; };x=f;${!x}; makes this potentially
likely to be a major false positive source.
-- subshelled dies
- some form of exemption syntax, for spots where the ebuild is doing something
normally bad, but valid in this scenario
- some form of -jN test (complex/hard)
-- parallelize the bugger (potentially hard). can probably thread it splitting
- it such that check x handles checks using globals x (cache mainly), other
- checks being bits that don't; gain may not be much since the major time
- consumers are visibility and unported_mod_x