Latest revision as of 17:16, 3 November 2010

TSC Decisions

The TSC supports the idea of regular patchwork days and proposes these should be on a regular timeslot around a Friday/Weekend/Monday on a monthly basis. Stefan Schmidt has kindly volunteered to coordinate these and has the support of the TSC. Mickey takes the action to communicate this to members via email.

The TSC feels bitbake 1.8 should be an option for the upcoming OE December stable release as switching to 1.10 before the release is too soon. After the release OE.master should switch to 1.10. For reference, standalone python binaries from Yocto can be used (https://lists.yoctoproject.org/pipermail/yocto/2010-November/000176.html) on systems which don't have a recent python version.

The TSC considered the issue of staggered election of its members. It felt that with members serving a one year term, staggering this could lead to a lot of "paperwork" without any real net benefit. If the OE members wish the TSC to do this, they are happy to do so though, they would just like to see member representation of this.

For the fetch/unpack issues, moving the unpack code into bitbake and calling into the fetchers to run the unpack command is looking like the preferred option. Using the opportunity to clean up the fetcher code is desirable with a "fetch2" implementation in bitbake being acceptable. The TSC supports the idea in principle subject to final review of the patches.

FILESPATH - This doesn't scale well with layers and the TSC supports the idea of reducing the number of directories listed. The exact solution still needs to be determined.

@@ Line 8: / Line 8: @@
 * The TSC considered the issue of staggered election of its members. It felt that with members serving a one year term, staggering this could lead to a lot of "paperwork" without any real net benefit. If the OE members wish the TSC to do this, they are happy to do so though, they would just like to see member representation of this.
+* For the fetch/unpack issues, moving the unpack code into bitbake and calling into the fetchers to run the unpack command is looking like the preferred option. Using the opportunity to clean up the fetcher code is desirable with a "fetch2" implementation in bitbake being acceptable. The TSC supports the idea in principle subject to final review of the patches.
+* FILESPATH - This doesn't scale well with layers and the TSC supports the idea of reducing the number of directories listed. The exact solution still needs to be determined.