2016-01-15 14:02 GMT+01:00 Tudor Girba <tudor@tudorgirba.com>:
Hi,

Oh, I got it. Your test does not test inside a class, but the top search.

I now committed GT-SpotterExtensions-Core-TudorGirba.187 to only search for exact references. Could you check?

Yes! Much better :)

��

Before, searching for Morphs took 58s:



Cheers,
Doru


On Jan 15, 2016, at 1:33 PM, Tudor Girba <tudor@tudorgirba.com> wrote:

Hi,

Nice catch!

I looked at it, and I cannot understand where the slowness comes from because there is essentially no��change in the processors of a Behavior. Now I am intrigued.

Please open an issue.

Cheers,
Doru



On Jan 15, 2016, at 10:32 AM, Nicolai Hess <nicolaihess@gmail.com> wrote:

since pharo 50524 spotter is much slower.

Most probably related to search results from classes with many subclasses
Morph / Object / Error / Announcement.

For example, search for Announcement
and dive into the result for the Announcement class.
It takes about ~ 10 seconds to build the result page (instance methods / super instance methods....)

Testcase

GTSpotterStepTest>>timeClassTest:aClassName
[ self basicSearch:aClassName ] timeToRun logCr.

"Pharo 50523"
timeClassTest:'Announcement'
0:00:00:00.539
timeClassTest:'Morph'
0:00:00:01.448

"Pharo 50524"
timeClassTest:'Announcement'
0:00:00:06.936
timeClassTest:'Morph'
0:00:00:49.044

Or open SUnit Testrunner and run tests for package:
GT-Tests-Spotter-Scripting

nicolai

--
www.tudorgirba.com
www.feenk.com

"Every successful trip needs a suitable vehicle."






--
www.tudorgirba.com
www.feenk.com

"What we can governs what we wish."