The language does not matter I just want to know what do you look at when you search for an IDE? Do you look for great debugging tools, do you look at the screenshots or do you download it and testrun it for a week? Do you look at videos of the product or do you ask on Stackoverflow what IDE you should look at?
- read which features it's got
- view some screen-shots or videos presenting these features
- read users' opinions
- try it
BTW. http://en.wikipedia.org/wiki/Comparison_of_integrated_development_environments
Take in account, that some characteristics can be very important for some, and completely irrelevant for others.
Even though i might be slammed i'm a big fan of simple IDE's that don't get in my way such as notepad++
How often do you pick up a new IDE? For me it is fairly rare.
Unless I'm picking up a new language I tend to only try new IDEs if there is a short coming with my existing tool. I'll then try to find an IDE that specifically addresses that issue. Then it is a matter of trying it and seeing if it is better.
In general though I tend to rely on the IDEs that are most prevalent for each language, typically they will receive the most support in the long run.
You can watch videos and read comments on the web to get the general idea, but you must proceed with the hands-on to determine if the IDE is good.
The look and feel, the responsiveness, the level of customization (or lack thereof), the features while working with a real, huge project, the ability to use plug-ins, the integration with the major external tools (code repositories, bug repositories), the support for different languages, file types... It takes a lot of time.
I started working with Eclipse six years ago and never let it go. JBuilder, IDEA, NetBeans, Emacs... I tried them all, but still favor Eclipse. It has its shortcomings, but Ultraedit can fix almost everyone of them.
At the end of the day it is largely a matter of personal taste: an IDE is something you should be comfortable with, just like your chair.
Unless you have limited choices, that is (like working with C#... it's pretty unreasonable looking for alternatives to Visual Studio, there).
There are other things to be established first, then the IDE.
I would think first for what technology(ies) I need (possible what programming language(s)), what platform I have to work with (or should it be cross platform work), then to establish what are the needs to be satisfied by an IDE, then pick the best one that suits your needs.
I tend to prefer anything that's open source, including IDEs. I'm happy to use an IDE that is backed by an active community of developers, with good documentation, frequent updates, support, and a large number of fellow developers using it. Eclipse always comes to mind :)
The best way to evaluate it is to try it yourself. But if there is a large community around it, you can be assured that it is well worth looking at.
Of course if the language/technology in question is owned/promoted by a vendor along with an IDE, such as Visual Studio, then choice is pretty obvious if you are looking for productivity.
Depends on your situation.
At work, I'm using C# so I use Visual Studio. Not much decision to make there.
And it helps that it isn't a bad choice - and throw in ReSharper and it just gets better. So on that basis, plugins would be important.
As long as it's the same one I've always been using, I'll take the IDE that is most suited for the task.
Last time I had to make such a decision, I read the respective posts on SO and tried two IDEs. After a week of testing both I knew which one I needed/wanted (Aptana).