This is an old revision of the document!
Project administration
Mailing lists
ipxe-devel
Member-only posting
To reduce spam, the ipxe-devel mailing list is configured to hold mails sent by non-members for moderator approval.
The iPXE policy for moderator approvals is straightforward:
The goal is to eliminate all spam via the list, while minimising the burden of moderation.
github integration
A github user ipxe-devel
is configured to receive github notifications at the address ipxe-devel@lists.ipxe.org, in order that github pull requests end up being forwarded to the ipxe-devel mailing list. The list is configured to automatically accept mails with a sender address matching ^[^@]+@reply\.github\.com$
, to avoid the need for every github-originated message to be manually moderated.
Commit log messages
The iPXE policy for commit log messages is:
Messages start with a header line comprising a [subject] tag and a brief description of the change
The [subject] tag represents the driver or subsystem being modified. In most cases, this can be deduced by looking at git log for the file(s) being modified by this patch.
The brief description should be written in the imperative mood (e.g. “Add PCI device IDs for Intel I218-LM” rather than “Added PCI device IDs for…” or “Adds PCI device IDs for”), and should have no terminating punctuation.
Trivial patches or patches which simply add a new feature may have an empty log message body.
Non-trivial bugfix patches should have a log message explaining the problem and the chosen solution.
All commits must include a Signed-off-by: line
Credit should be given to all participants in the work leading to the patch, including bug reporters (Reported-by:), testers (Tested-by:), original authors of a completely reimplemented patch (Originally-implemented-by:) etc. Since many contributions come via
IRC, the
nickmap page can be used to look up full names and e-mail addresses suitable for inclusion in the commit log.
Patches which fix a regression should refer to the commit which introduced the regression.