installation failure not respected?

[dabrahams]

installation failure not respected?

[dimitri]

change?

[dabrahams]

The 2nd parameter to el-get-start-process-list-sentinel

[dimitri]

  (if (not (eq 0 status))

I'm relying on this test instead. Shouldn't I?

[dabrahams]

Hmm, I didn't notice that.

The docs on sentinels are a little bit vague about the circumstances under which they'll be called:

two changes in status, coming in
quick succession, can call the sentinel just once. However, process
termination will always run the sentinel exactly once. This is because
the process status can't change again after termination.

and the docs on process-exit-status say that

If PROCESS has not yet terminated, the value is 0.

which makes it sound a bit like we could mistake some other change in status for success? Perhaps a nonfatal signal could be sent?

[dimitri]

I'm closing this issue in my clean-up frenzy tonight, because I really think that the exit status is the right way to handle that. And we're already protected against the mistake you're talking about:

(when (eq (process-status proc) 'exit)
(if (not (eq 0 status))

[dabrahams]

fine with me