[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
Re: [eclipse-dev] Fw: [Bug 215464] Regression in org.eclipse.core perf test: StartupTest. testApplicationStartup (aka 'Core Headless Startup')
|
To set a comment on a test, one need to use the PerformanceTestCase
following method:
/**
* Set a comment for the scenario represented by this TestCase.
* Currently only comments with a commentKind of
EXPLAINS_DEGRADATION_COMMENT are used.
* Their commentText is shown in a hover of the performance summaries graph
if a performance
* degradation exists.
*
* @param commentKind kind of comment. Must be EXPLAINS_DEGRADATION_COMMENT
to have an effect.
* @param commentText the comment (shorter than 400 characters)
*/
public void setComment(int commentKind, String commentText) {
There's currently only one available comment kind: Performance.
EXPLAINS_DEGRADATION_COMMENT
As the constant name includes 'degradation', it makes me naturally think
that test result is expected to be negative.
So, the current behavior does not hurt me a lot...
I think that the behavior you're describing would need a new comment
constant which would use another color when displayed in the fingerprint...
However, my personal feeling about this case would be that if a test is
known to have wrong results even when positive, then it should be removed
temporarily from the test suite until it's fixed instead of being still
displayed in the fingerprint.
Cordialement/Regards,
Frédéric
Mike Wilson
<Mike_Wilson@ca.i
bm.com> To
Sent by: "General development mailing list
eclipse-dev-bounc of the Eclipse project."
es@xxxxxxxxxxx <eclipse-dev@xxxxxxxxxxx>
cc
01/17/2008 06:23 Subject
PM Re: [eclipse-dev] Fw: [Bug 215464]
Regression in org.eclipse.core
perf test: StartupTest.
Please respond to testApplicationStartup (aka
"General 'Core Headless Startup')
development
mailing list of
the Eclipse
project."
<eclipse-dev@ecli
pse.org>
Am I the only person who thinks this is odd?
To me, the grey coloring means "there is something wrong with the test (or
the system has changed in a way that invalidates it), so the result isn't
interesting" -- I don't actually care whether it's slower or faster. For
example, ongoing fixes elsewhere could cause the test to run faster, making
the test *look* like it's working again, even though it's still not
McQ.
Frederic Fusier
<frederic_fusier@
fr.ibm.com> To
Sent by: "General development mailing list
eclipse-dev-bounc of the Eclipse project."
es@xxxxxxxxxxx <eclipse-dev@xxxxxxxxxxx>
cc
17/01/2008 12:04 Subject
PM [eclipse-dev] Fw: [Bug 215464]
Regression in org.eclipse.core perf
test: StartupTest.
Please respond to testApplicationStartup (aka 'Core
"General Headless Startup')
development
mailing list of
the Eclipse
project."
<eclipse-dev@ecli
pse.org>
Note that when a degradation comment is set on a test, the fingerprint bar
becomes gray only if the test result is negative.
Cordialement/Regards,
Frédéric
----- Forwarded by Frederic Fusier/France/IBM on 01/17/2008 05:43 PM -----
bugzilla-daemon@e
clipse.org
To
01/17/2008 04:25 Frederic Fusier/France/IBM@IBMFR
PM cc
Subject
[Bug 215464] Regression in
org.eclipse.core perf test:
StartupTest. testApplicationStartup
(aka 'Core Headless Startup')
https://bugs.eclipse.org/bugs/show_bug.cgi?id=215464
Product/Component: Equinox / Framework
--- Comment #4 from Thomas Watson <tjwatson@xxxxxxxxxx> 2008-01-17
10:25:56 -0400 ---
I released a degradation comment for the core headless startup test
(StartupTest.java) which points to this bug to explain the degradation.
--
Configure bugmail: https://bugs.eclipse.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You reported the bug.
_______________________________________________
eclipse-dev mailing list
eclipse-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe
from this list, visit
https://dev.eclipse.org/mailman/listinfo/eclipse-dev
_______________________________________________
eclipse-dev mailing list
eclipse-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe
from this list, visit
https://dev.eclipse.org/mailman/listinfo/eclipse-dev