OK, I had this crazy idea and I need another head to discuss it with so =
I'm throwing it out and see how you like it...
I was browsing around on and off other SourceForge projects with similar =
intents as ours. I'm doing tha to get a feel for what would make our IDE =
application unique and stand out amongst the rest. OK, I know, being =
DynAPI compliant is one major, but still we could just as well =
concentrate on making plugins for another app like jEdit or netBeans or =
something and not bother at all with starting from scratch. I just want =
to make sure we're not onto just creating the wheel all over again if =
you get what I mean.
Anyway, after have browsing around a couple of sessions now, seen =
numerous of IDE application projects - alot of them done in Java too - I =
started going back to see what features are in commercial applictions I =
find useful, that might be usefulfor this project. Thing is commercial =
applications still have the best features in many cases, due to the =
resources they're able toput on the development. My favorite =
applications that I use include Paintshop Pro, Winamp and Dreamwever =
(Ultradev). Now the one of those that are most close of an IDE =
application is of course Dreamweaver. So what of Dreamweavers features =
is it that I think is really useful, and could do being implemented in =
this IDE? Well, here's something interesting: Dreamweaver is extensible =
by Javascript(and HTML)
So what is good about that and why does it fit this project so well? =
You're probably already guessing, but I'm going to tell you anyway:
+ Javascript is simple to learn, many webdevelopers already knows it and =
knows it well, not least the target audiance of the DynAPI IDE
+ It will enable users to pretty quick pick up on onfiguring and =
customising if they can do it with the language that is simple and that =
they already are developing with
+ They can learn from developing extensions how to code javascript for =
their pages and vice versa
+ They can configure, customise and develop the extensions for the IDE =
in the IDE itself!
+ If we're using Rhino for the javascript engine (which is the best =
looking alternative so far) we have it ready for implementation and =
saves us a lot of hazzle coming up with an all new API
The only downside I can see so far is that it might cut down on =
performance of the app to have it parse javascript instead of Java code. =
Beware I have no indication of wheter this would be the case, it just =
seems like a possible scenario. OTOH it all depends on how well Rhino =
runs and how much javascript we use.
An additional thought is that if we go along these lines, we should =
concentrate on making the javascript extending as similar to the DynAPI =
syntax as possible, meaning try and implement it the way syntax is used =
in existing DynAPI code and widgets. For example we might use an almost =
identical "object" for application menus as the menu widgets avilable =
for DynAPI (see =
http://www.resass.f2s.com/dynapi/Ken_Ono_Examples/Pulldown_Menu.html for =
example). Not that we use the widget itself, but that we create an Java =
object that interfaces the same methods, have the same properties and =
names, even if we might add and subtract some. From what I understand, =
this is very simple to accomplish when implementing Rhino.
So are there any further thoughts? Or do you adapt the idea? If so we =
might as well, implement it into the framework from the begining, don't =
you think?
BTW I'm getting an app from Richard (owner and maintiner of =
Richardinfo.com ) which more or less automates getting methods and =
properties definitions from the dynAPI.
BTW2 I seem to remember Dan Steinman, the originator of DynAPI, to have =
begun on a DynAPI IDE in Java. I'll get aound to ask him for any sources =
or materials he might have lying around and if he'd want to contribute =
it.=20
Henrik V=E5glin [ hv...@ya... ]
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
|
