summaryrefslogtreecommitdiffstats
path: root/doc
diff options
context:
space:
mode:
authorChris Larson <clarson@kergoth.com>2003-09-02 17:55:01 +0000
committerChris Larson <clarson@kergoth.com>2003-09-02 17:55:01 +0000
commit3325c317ba478becb641f787b758038352d8f904 (patch)
treec6f2cc583b381c1b44961271f96f4e47ee18c9cd /doc
parent03177c5115dd4a785fe7bca5cd777696dc437a8e (diff)
downloadbitbake-3325c317ba478becb641f787b758038352d8f904.tar.gz
Update TODO
Diffstat (limited to 'doc')
-rw-r--r--doc/TODO55
1 files changed, 7 insertions, 48 deletions
diff --git a/doc/TODO b/doc/TODO
index a7450eb27..4d9cfdb45 100644
--- a/doc/TODO
+++ b/doc/TODO
@@ -6,22 +6,6 @@ TODO:
[ ] API Sanity Audit.
[ ] Comment/Docs Audit.
[.] Implement API testing using doctest.
- [x] BUG: If we only run an update_data after a load of an .oe,
- but not for a .oeclass, if multiple .oeclasses define, say, for
- example DEPENDS_prepend, one overrides the other, and we lose
- some of the deps. If we update_data after each .oeclass load,
- then DEPENDS_prepend affects DEPENDS immediately, but subsequently,
- it applies against a DEPENDS variable which doesnt yet exist,
- because the oeclass (in the case of base/INHERIT) is loaded prior
- to defining DEPENDS in the .oe. The only solution I see is to
- pick up _append/_prepend/_delete during the .oe feeder, and
- append each of them to a list of tasks to do during the
- update_data run. I.e. base sets DEPENDS_prepend to content/patcher,
- then content/patcher ends up in the 'to prepend' list, and is then
- prepended when update_data runs after the completion of the .oe
- parse.
- [x] Teach 'include' to fetch if necessary.
- [x] Move this into handle() rather than include.
[ ] Alter 'export' handling, to support exporting a var without
defining the variable.
[ ] Add support for 'anonymous' OE functions in .oe{,class} files.
@@ -33,19 +17,8 @@ TODO:
knowledge of which packages are to be built.
* Build requirements
- [x] Code will not be checked out into a temporary area
- [x] Solve by leaving SRC_URI blank, and setting
- S appropriately.
[ ] Code will not be automatically updated from CVS
[ ] Solve by making Fetch use md5.
- [x] Follow dependencies
- [x] Use oemake
- [x] Default .oe argument for oebuild
- [x] Default 'command' argument for oebuild
- [x] Allow execution with a command, without .oe
- [x] Allow execution with .oe, without command
- [x] Default .oe argument for oemake
- [x] Allow specifying a command on oemake execution
[x] Must be able to execute 'oemake' within a subpackage
of an upstream package, and yet retain access to the
toplevel TMPDIR. Perhaps specify TMPDIR through
@@ -57,34 +30,20 @@ TODO:
[ ] Allow switching to a debug build for the entirety
of the build. Alter CONFIG for created .pro files,
and add -g to appropriate flags.
- [x] Nightly builds will use the same .oe files as those
- used upstream. This is easily solved, since the use
- of OE upstream already gives us the necessary event
- handling.
- [x] Does not require that the build be done as root.
- [x] Simplicity of .oe file naming (PV/PR inside .oe)
+ [ ] Requires some form of either if block,
+ or anonymous python function to do the
+ logic without lambda madness.
[ ] Tests for versioning
+ [ ] Far, far beyond the scope of what OE is
+ designed to do.
[ ] Wrapper Makefile -> oemake
- [x] Command as <cmd> as opposed to do_<cmd>
+ [ ] Not part of OE, only requires that we can
+ override the path to Makefile.
[ ] Target device filesystem locations as metadata
items so that all the packages can install things
into proper locations. This is highly distribution
specific,
[ ] Real life test case: Opie
- [x] default PN,PV,PR,CATEGORY in oe.conf
- [x] Allow override of PN,PV,PR,CATEGORY within a .oe
- [x] Allow filename to not contain PV, PR, CATEGORY
- Needed because a .oe within an upstream source
- will likely prefer to hold the version as an item
- of metadata, rather than as the filename.
- [x] Set TMPDIR to a path off of the original
- TOPDIR, then set TOPDIR to the dir the .oe resides in,
- in _all_ cases. This wont break functionality for most,
- but will allow relative paths to behave properly, thereby
- fixing this issue.
- In this way we can sanely handle .oe files that
- use relative paths extensively, such as those
- using '.' as their 'S' variable.
[ ] oebuild
[ ] Implement and test check_md5, taking nostamp into account.