[Toybox] [PATCH] ifconfig.test: deflake.

[enh]

or, if you were actually asking "why didn't you *also* change \: to
just :?", well, the \:\: was already in both lines, and it wasn't
relevant to the change, and i thought it would confuse what the patch
was actually fixing.

i'd volunteer to tidy this whole test file up, but in my copious free
time i've actually written a readelf(1) that i need to make a last
couple of changes to and send your way, plus we're coming up on
another christmas and -- thanks to the other changes that happened in
the meantime -- i'm even further away from sending you my man nroff
subset than i was this time last year. and i still need to unpick my
hexedit changes and send those and...

On Wed, Dec 4, 2019 at 9:21 AM enh <enh at google.com> wrote:
>
> On Wed, Dec 4, 2019 at 9:01 AM Rob Landley <rob at landley.net> wrote:
> >
> > On 12/3/19 11:02 AM, enh via Toybox wrote:
> > > ::2 can legitimately be part of an IPv6 address, causing the test to
> > > fail because the grep matches an unrelated part of the output. Be more
> > > explicit about what we're actually searching for (as the previous test
> > > already was) to avoid flakiness.
> >
> > If you have an english description of the test that's well under 80 characters,
> > why not just have that be the test description instead of a comment?
>
> i didn't write these tests, i'm just fixing the flake :-)
>
> > I tend to indent continuation lines.
> >
> > I'm not sure what this is trying to accomplish:
> >   $ echo grep 10\.240\.240\.240
> >   grep 10.240.240.240
> >
> > The shell eats the backslash and it's still a wildcard for grep. (This isn't a
> > quoting thing, this is an eval thing.)
>
> yeah, i removed a useless \/ in this patch --- i know it doesn't
> strictly belong, but since i was making the next test look the same, i
> didn't want to have to copy & paste a meaningless \.
>
> > Hmmm, I'd apply the attached (and continue on) except I don't have a qemu test
> > instance with the dummy driver set up. I should spend a half an hour doing that.
>
> note that this is a *flaky* test. i've never seen it fail before, but
> randomly the test device had an IPv6 address something like
> "fe80::241e:8519:4a19:1603", which caused the test to accidentally
> match the unrelated "inet6" line, something like this:
>
>         inet6 fe80::241e:8519:4a19:1603  prefixlen 64  scopeid 0x20<link>
>
> > (Sigh. It's a pity Google never wanted to hire me to work on the test suite back
> > when I had spare time. I'm off doing j-core stuff again, I _think_ I fly back to
> > tokyo in a day or two? Waiting to hear how a planning meeting goes this evening...)
> >
> > Rob
> >
> > P.S. my blog is caught up to present! https://landley.net/notes.html has
> > yesterday's entry already for the first time in _years_. I'd like to thank the
> > academy, by which I mean the https://patreon.com/landley people who made my
> > lowest goal level, which was "get my development blog up to date"... You may
> > have noticed me posting slightly less here, since my constant toybox redesigns
> > and code frogging have been going there instead.
> > _______________________________________________
> > Toybox mailing list
> > Toybox at lists.landley.net
> > http://lists.landley.net/listinfo.cgi/toybox-landley.net