Re: What can NetBeans do better for educators than it does now?

Anton Wallace wrote:

> I have a few suggestions that could make net beans more usable in a teaching environment
> (these suggestion would also apply to the workplace).
>
> I hope I haven't suggested any features that I just haven't stumbled across yet !
>
>    * A project template would be very useful.  I find myself constantly remounting common
>      / shared classes and folders for each project I set up.  Given that many teaching

>      examples are brief this can be quite laborious.

Hi, Anton,

Thanks for replying!

Hmmm...a couple of thoughts:
Project Manager | Save As should allow you to duplicate an existing project.

If students are running an install of the IDE on a shared drive, someone running the
IDE as root should be able to create mount points that will be visible to all users (on
Unix, make sure the file permissions are correct so they will actually be able to access
these).  That is something that needs to be made easier in the IDE.  An earlier
implementation had a concept of project inheritance, which seemed to me brilliant,
but created some worry that it was too confusing.

Since I came to work at NetBeans (the Czech company NetBeans, pre-Sun), I've
seen project support reimplemented four or five times, and the next reimplementation
is underway now - the docs on it are at http://projects.netbeans.org/ps/index.html

It might be a good idea for some of us to participate in the development discussion
for the new projects implementation, and make sure that it maps well to an
educational setting.

>    * A java decompiler so students can see how the jdk has been implemented (I currently
>      have a couple of these but it would be nice to have an integrated one).

That's definitely a cool idea.  I can't say I've looked at these much, but do you know of
an open source one out there that we might be able to integrate?

>    * A tool that identifies similar / identical algorithms - to identify students poaching
>      others work.

:-)  When the metadata repository really gets there, I think that would significantly
help with this, since there will be metadata about constructions existing in sources
to a much more fine-grained level.  Definitely something to look at doing!

How do you envision the UI for something like that working?  I could imagine
something like selecting a region of code, some kind of pattern extraction analysis
over it, and then selecting a set of files and examining them for similar patterns...
Probably you have a more concrete idea of what this would look like and how it
would work than I do.

I've cc'd Petr Hrebejk, architect of the metadata repository on this.

FYI, there's not much documentation on the Metadata Repository out there yet,
but we do have a white paper published in Dr. Dobbs journal this month, the
second half of which covers it, and a pdf will be up on the website in a day or
two.

Anybody else have thoughts on this or other suggestions?

Cheers,

Tim