Using Irida (3431) I am seeing some sizable changes in
performance between Irida and 4.5.x; In some cases Irida
is slower and in other cases it is faster.
For example, in another post I mentioned that Find Usages
on a particular public static method was taking ~15 seconds
in Irida ( 3431) versus ~9 seconds in 4.5.x.
Just now, I was doing some string searches (No regex, case
sensitive, scope all project files), and Irida was taking
just 3.5 seconds versus 9 seconds for 4.5.x, so Irida was
almost 200% faster in this case!
It would be great if JetBrains published some performance
statistics showing sample performance on a large frozen
workspace, e.g. IDEA 4.5.4 source code. Stuff like reformat
a large java file, find usages, find in path, rename (
involving renaming references in many files), etc. I am
esp. interested in the performance of operations which
depend on the size of the workspace.
I noticed Eclipse 3.1 has started to do this (although only
on a limited set of metrics).
I view performance as a main feature of every release,
whether or not marketing states it as such in the "what's
new in this release" list.
I've never used Eclipse for that long to know how it stacks
up against IntelliJ as far as speed and responsiveness. If
IntelliJ is faster than Eclipse, then why not publish some
benchmarks using a large open source codebase showing
IntelliJ kills Eclipse on speed?