Check if a nightly build for the given date actually exists before attempting to install

[jimmycuadra]

$ rustup --version
rustup 0.6.3 (a0e2132 2016-08-28)

$ rustup toolchain install nightly-2016-10-13
info: syncing channel updates for 'nightly-2016-10-13-x86_64-unknown-linux-gnu'
error: no release found for 'nightly-2016-10-13'

$ rustup toolchain install nightly
info: syncing channel updates for 'nightly-x86_64-unknown-linux-gnu'
info: downloading component 'rustc'
 37.2 MiB /  37.2 MiB (100 %)   1.4 MiB/s ETA:   0 s
info: downloading component 'rust-std'
 44.5 MiB /  44.5 MiB (100 %)   1.6 MiB/s ETA:   0 s
info: downloading component 'rust-docs'
  9.2 MiB /   9.2 MiB (100 %)   1.4 MiB/s ETA:   0 s
info: downloading component 'cargo'
  3.9 MiB /   3.9 MiB (100 %)   1.1 MiB/s ETA:   0 s
info: installing component 'rustc'
info: installing component 'rust-std'
info: installing component 'rust-docs'
info: installing component 'cargo'

  nightly-x86_64-unknown-linux-gnu installed - rustc 1.14.0-nightly (098d22845 2016-10-13)

The most recent nightly I'm able to install by date is 2016-10-11.

[jonas-schievink]

I don't think the nightly-2016-10-13 nightly exists. See http://rusty-dash.com/nightlies, there are a few build failures for the recent nightlies. nightly-2016-10-17 works fine for me.

[jimmycuadra]

But if I just installed "nightly" without a date, 2016-10-13 is the one it installed. That's shown in the output above. I was also able to install 2016-10-13 on macOS.

[steveklabnik]

The dates can be off a day; it's possible that that's the one from 2016-10-14. Install it as well and check.

[cardoe]

@steveklabnik Is there a way to sync that up at all? Just got caught up by this as well and found this issue.

[Diggsey]

@cardoe There's an open issue for it: rust-lang/rust#36773 - the problem is that the dates reported by rustc and cargo are the dates of the most recent commits to each of those repositories, and not the actual build date. Apparently it's not an easy thing to fix.

[cardoe]

ah. I appreciated the background @Diggsey. And also good thing to get the issue linked.