Some thoughts about our tests and the build time
bjoern.michaelsen at canonical.com
Tue May 17 08:17:11 UTC 2016
On Tue, May 17, 2016 at 09:31:54AM +0200, Miklos Vajna wrote:
> Also, I didn't do measurements, but my impression is that the current
> hot path in 'make check' are still the Java tests, so if we want faster
> 'make check', then perhaps a way towards that would be to accelerate the
> conversion of the majority of those Java tests to in-process ones.
since Markus said he ran "make" (not "make check") twice, none of the above
number include anything Java-ish (those are in "make check" and "make
subsequentcheck" but not in "make" or "make slowcheck unitcheck"). So, if
anything, these numbers show that the non-Java tests grow and will dominate
even "make check" soon.
More information about the LibreOffice