Bad performance on 3235?
Hi,
I've started to really work on build 3235 this week. The same time my laptop broke down (not your fault:) and I had to switch computer to a desktop 1.7Ghz/1GB Ram machine.
I've noticed that everything was kind of slow, edition, compilation, dialogs etc so I figured out that the new PC was slow. Anyway I tested 4.5.4 on the same machine with the same project and I think I got a 200% boots in performance, again IDEA was snappy.
I known that this is a EAP, but I really hope that you do some CPU profiling while working towards 5.0
Thanks,
Børge
Please sign in to leave a comment.
I notice a noticeable difference in the Editor performance between 4.5.4 and Irida (using 3242 now).
The main pain points are:
1) Reformating is taking longer. I don't mind so much when reformatting the entire file. But sometimes I just hit SHIFT+TAB to reformat a few lines, and that takes noticeably longer.
2) Basic and SmartType Completion takes noticeably longer. Before it seemed instantaneous but now there is a delay.
Sometimes I'm just typing it out because it's faster than waiting for the completion.
Desktop: P4 2.8Ghz, 1GB memory, Windows XP SP 2
Codebase: 300 packages, 500k loc, about 40 3rd party libraries
The 3rd party libraries in sum probably dwarf our own code.
Hopefully the slowness is just because Irida is running in debug mode or something..
Sigh... I don't really care about reformatting taking longer. I do care about reformatting (and anything having to do with code placing) not working -- and I'm not talking about new features, but about working the same way it did on Pallada.
Formatter was rewritten recently and the work is still in progress, sorry
for inconvenience.
"Alex" <no_mail@jetbrains.com> wrote in message
news:12081050.1109893896467.JavaMail.itn@is.intellij.net...
>I notice a noticeable difference in the Editor performance between 4.5.4
>and Irida (using 3242 now).
>
>
>
>
I'm waiting for the new formatter to stabilize, with all my patience. It's just that... sometimes it gets to me, specially early in the morning :)
Anyways, it fathoms me why the reformatter was rewritten, given the very few new formatting options (well, there wasn't much to add anyway). We can endure buggy code when we see something new in the horizon, but with the reformatter, we were just seing broken functionality for no apparent reason.
Are seeing, you mean.... I just right now tried to format an XML file, with, er, interesting results.
Like you say, I could endure this much more easily if it didn't work basically perfectly beforehand.