[webkit-dev] Why are we running Sputnik?

Adam Barth abarth at webkit.org
Tue Aug 10 14:00:58 PDT 2010

On Tue, Aug 10, 2010 at 12:16 PM, Alexey Proskuryakov <ap at webkit.org> wrote:
> 10.08.2010, в 11:51, Eric Seidel написал(а):
>> Why do we want to be running these 6000 tests and slowing down our
>> builds.  I was talking with jamesr, and he seemed to think it adds
>> little value to run it every time?  (It was supposedly written as more
>> of a development tool for V8?)  But maybe I'm missing something?
> Just like any test suite, it catches regressions. We saw that it was effective when we introduced certain JSC regressions that it could have prevented. Since some of the errors Sputnik catches are platform specific (32 bit vs. 64 bit, JIT vs. interpreter etc), it is particularly important to have it on build bots.
> Some subtests are less interesting (e.g. many Unicode tests basically check ICU behavior), so we have the slowest of those disabled already.
> One possible way to speed up Sputnik tests would be to run them as part of run-javascriptcore-tests, avoiding web page overhead. Someone proficient in scripting languages just needs to improve run-javascriptcore-tests to the point when it can do that.

FWIW, a Sputnik test caught an error in the new HTML parser (related
to the interaction between \r\n collapsing and line numbering in
JavaScript) that otherwise might have gotten through.

A better long-term fix might be to finish new-run-webkit-tests so we
can run the tests in parallel.


More information about the webkit-dev mailing list