[PATCH 1 of 6 RFC] radixlink: introduce a new data structure

Jun Wu quark at fb.com
Wed May 24 17:20:51 EDT 2017


Excerpts from Augie Fackler's message of 2017-05-24 17:11:51 -0400:
> [..]
> 
> does "source of truth offset" mean "number of bytes in the source of
> truth file as of when this index was valid"?

Yes.

> [...]
> 
> If we're gonna have a new test, let's use our silenttestrunner instead
> of .py and .out files. It's a lot easier for newcomers to
> understand. :)

Sure.  I'll figure out how to write tests properly. This RFC is just my
weekend effort trying to prove this direction is promising.

> [...]
>
> nit: `raise RuntimeError('unexpected')` is a more idiomatic spelling
> (does that not work for some reason?)

Will rewrite the test using a proper unittest framework.


More information about the Mercurial-devel mailing list