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.
Create module project, then use the New Wizard wizard. In the 1st panel, type '1000' (or some other high number, I don't know which number is the lowest acceptable number) in 'Number of Steps'. Click Next. Wizard freezes, IDE hangs, big deadlock, and then I needed to force the IDE to close.
It is a real deadlock or just a lot of work needed to be done in the EDT? If later, than P4. 1000 is quite a lot.
Well, I just created without any problem a wizard with 50 steps. I don't think there are many users (maybe even none) who will want as many as 50 steps. So, for me, the problem is not that there aren't going to WANT to have 1000 steps, but maybe some who will accidently put too many 0's (instead of typing in '10' they might type in '100' or '1000'). This is when the problem might happen. But, if you think this isn't very likely, then this could also be a P4.
I think that this "typo" is almost impossible. Actually typo itself is possible, of course. But than you click "Next" button. The time to get to the second panel is always the same, no matter how much steps you have entered. Then you enter the class name prefix and... ups. Neverending list of panel classes is generated into textarea of created files. So here everybody would probably stop. Yes, except you, me, Petr Zajac, ... who are testing it :) But definitelly would be nice to fix.
So please don't open all the generated files in editor. BTW it is responsivness problem when you generate wizard for more steps (10 ..).
> So please don't open all the generated files in editor... responsiveness problem True and good idea. I'll constraint it to magic number 10... However I will not constraint the user in what he is doing.
Fixed. Checking in ui/wizard/BasicWizardIterator.java; 1.16 --> 1.17
Getting NPE when setting # fo steps to e.g.20:( I'll attach...
Created attachment 26625 [details] nPE
Has nothing to do with apisupport (we tried it with Tomas and cannot reproduce with fresh build).
Yes,NPE was consequent of P1 issue. This was fixed ok. Verified.