[PATCH 1/6] Allow running tests in a build directory different from source directory.

dormando dormando at rydia.net
Thu Nov 8 17:52:48 UTC 2007


I assume this is the final patch series?

Tomash, thank you for all your work here! Please let us know if you wish 
to continue the gperf work. It might make a little more sense once the 
binary protocol into mainline, to reimplement the text protocol using it.

If no one merges these by tomorrow evening, I'll be testing them and 
start merging some now. promise :)

Can't personally test 'noreply' speed on any of the sites here, so 
someone will have to do a synthetic benchmark to test. I don't see 
reason to exclude the last patch series if it doesn't break in testing 
though...

A bit confused by all of the arguments:

- Both sides have a number of good points. Though it's a bit silly when 
it would be easier to extend the binary protocol we'd all like finished 
anyway.
- 'append' and 'prepend' aren't released yet. Since CAS is a non issue, 
can't we just fix those for Brian's thing? (caveat being that I haven't 
looked to confirm if they _are_ an issue yet).
- Numbers are good. The burden is on the maintainers to prove incoming 
patches don't adversely affect performance. Since a lot of things _look_ 
good in theory, but tend to behave differently, it's worth confirming 
rather than simply stating why something's superior for non functional 
changes.

-Dormando

Tomash Brechko wrote:
> ---
>  trunk/server/Makefile.am            |    2 +-
>  trunk/server/t/lib/MemcachedTest.pm |   11 ++++++++---
>  2 files changed, 9 insertions(+), 4 deletions(-)
> 


More information about the memcached mailing list