You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
meta-issue: should RFC header include author name and/or contact info? #449
I have often relied on github metadata to identify immediately the RFC author. But this does not work without some amount of effort on the part of people maintaining the RFC repository to try to preserve author identification in the commits.
Should we add an Author: "Your Name here" <email@domain.here> to the RFC header? (And I guess give people the option of deleting the line from their RFC draft if they do not want to be personally identified with the RFC forever, though I am not sure whether trying to accommodate that request makes sense...)
@steveklabnik Maybe I did not write the first paragraph of my description clearly enough.
Look at the recent commits. Such as commit bc5b034 or commit 336ccaa . Perhaps more pertinently, look at the text associated with each of those commits:
In each of those cases, the author according to the git metadata is the person who merged the RFC. At best you might be able to go back through the history of the merge to discover that the branches originated from niko and nick, respectively, not brian and aaron.
I attempted to preserve the author metadata when I merged RFC #378. But its not necessarily part of everyone else's workflow. And in any case, if we ever have a broad revision of the RFC directory structure on par with what I did in PR #367, then we again lose the metadata.
I guess yes, if people mess with the metadata, it's not reliable then. The merge is --no-ff, so one of the two parent commits should be a commit from the actual RFC itself.
Activity
steveklabnik commentedon Nov 7, 2014
This information is already included in the git commit, though.
pnkfelix commentedon Nov 7, 2014
@steveklabnik Maybe I did not write the first paragraph of my description clearly enough.
Look at the recent commits. Such as commit bc5b034 or commit 336ccaa . Perhaps more pertinently, look at the text associated with each of those commits:
In each of those cases, the author according to the git metadata is the person who merged the RFC. At best you might be able to go back through the history of the merge to discover that the branches originated from niko and nick, respectively, not brian and aaron.
I attempted to preserve the author metadata when I merged RFC #378. But its not necessarily part of everyone else's workflow. And in any case, if we ever have a broad revision of the RFC directory structure on par with what I did in PR #367, then we again lose the metadata.
steveklabnik commentedon Nov 7, 2014
I guess yes, if people mess with the metadata, it's not reliable then. The merge is
--no-ff, so one of the two parent commits should be a commit from the actual RFC itself.pnkfelix commentedon Nov 11, 2014
withdrawing suggestion; (leaving out an author better reflects the community-driven nature of the process).