This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
There is a >15% regression in startup. Measured with build 200501051900 (w/ netbeans.conf updated and memory toolbar turned off) I see 16-19% slower start on my workstation. It is Dell P3/800MHz/512MB/JDk5.0fcs/Linux fedora core3. Similarly automated tests report 18-23% regression on Win2000.
Is ide component right for such issue?
any better suggestion? the problem can go across the whole IDE
Yes I know, but this category is kind of trash bin :-) Don't know what's better. Maybe reassign when you know the cause of the regression.
I compared builds from Jan 9 and Jan 12 to measure impact of editor split. I run the startup tests on my workstation and checked the number fromautomated test suite. There is ~2% regression on the first start and on the start with empty userdir. The difference in numbers from complex startup is not significant.
After the editor split the o.n.editor.Settings.reset() needs to be called from each new submodule (java/editor, html/editor, editor/plain) and also the localizers are being added to o.n.e.LocaleSupport. We plan to get rid of the localizers completely. Regarding Settings.reset() I guess there is something asking for a particular setting by Settings.getValue() after the reset which causes (re)initialization of the settings maps. Not sure whether this could be eliminated but we may try at least.
Here is comparision of NB4.0FCS and 4.1 build 200503232205 w/ release settings on my Dell Precision220 (800MHz/512MB/Linux FC3 or Win2K) and Ultra 10 w/ Solaris 10. Measured is startup with jedit project and 10 opened files (project explorer nodes are colapsed) till IDE window is shown and painted, classpath scan is finished and UI is stable, everything is done including warmup. Linux 4.0 22.5 33.0 35.5 25.9 32.8 35.2 26.1 34.3 36.1 dev 26.0 32.1 39.3 25.4 36.7 40.0 25.9 30.9 39.9 Solaris 4.0 33.5 51.8 59.7 37.1 48.9 59.7 31.8 50 57.4 dev 35.3 49.0 57.8 35.2 53.6 56.9 34.8 50.2 56.2 Win2k 4.0 19.6 32.2 37.7 20.0 33.3 37.3 19.8 32.4 36.1 dev 21.6 30.7 37.1 22.2 31.4 37 22.5 30.9 37.3
Your numbers show that we are now below the 15% boundary. Linux +3.6% (1st), +5.8% (next) Solaris +2.9% (1st), -1.1% (next) W2K +11.6% (1st), -2.0% (next) Automatic test results also show <15%. I think we can close this bug for now. If you don't want to keep it open for tracking of the other issues as an umbrella... Then I think we could downgrade this to P3.
There is a regression but we have acceptable numbers now. There is no plan for further changes in 4.1 release