[rb-general] [PATCH] Document timestamp clamping

Daniel Kahn Gillmor dkg at fifthhorseman.net
Fri Nov 4 15:32:37 CET 2016


On Thu 2016-11-03 17:07:00 -0400, Ximin Luo wrote:
> On a separate note however: I think the scope of the specification
> document as it stands, is only good for one purpose - dictating what
> the specification is, to people that already agree with us, or that
> want to automatically follow us for whatever reason. This is a purpose
> that is very narrow in scope, and also very weak.

it's still an important purpose, and in terms of "telling agreers what
to do" i think we need to include the concept of clamping.  If someone
wants to propose a narrower edit, and/or wants to take any of the text
i've proposed and put it in the wiki, i have no problem with that.

I do have a problem with leaving clamping out of the spec itself,
though.  That suggests that we're now expecting multiple tools to do
something that is subtly and *explicitly* different from what the
specification says to do with this environment variable, which means
either the spec is wrong, or we're wrong to suggest that people use
SOURCE_DATE_EPOCH in different ways.

I think clamping is the right thing to do, which means i think the spec
is currently wrong.  We should fix it.  Just because it's a rare case
doesn't mean it should be ignored.

        --dkg


More information about the rb-general mailing list