Re: best prototyping tool

Excellent refs Shawn. Thanks. I guess I'm looking for a Web/J2EE prototyping
tool and none exists. None usable by a graphics designer or non-technical
analyst that is.

The gotcha for us is we used Struts, Taglibs, and Tiles so heavily and they
are integrally tied to the display. E.g. internationalization is done with
Tiles and that is not portable to Flex/Flash for instance. I guess our
strategy will be to frontend the Struts Action classes with WebServices or
such so we can move to jMaki-based development. I think that's where we are
headed.

Borland. Nice flow pathing and documentation tool. Does not appear to
produce any code.
http://www.borland.com/media/en/edm/caliber_analyst_product_demo/main.html
IBM: spec tool. That's not too useful. People already can do specs fine. The
time loss in the SW dev process is re-making the wheel as the work product
from specs is not reusable to the actual product development. Maybe IBM has
something that does WYSIWIG and generates J2EE compat code. Did not say so
on their web page. Too many tools listed to evaluate each one. Bottom line,
in a pre-existing large project, there's no need to define a new
architecture and try to spit it out. We need something that works with the
existing architecture. I'm down on IBM as RAD does not even understand
Struts or Tiles enough to let one jump to references in these. Its sad for a
company claiming to be a leader in Java development. I did not complain
about the weak support in Eclipse for the fad frameworks but IBM should do
alot better for what we pay for RAD. We are on 7. Maybe they have it in a
newer version.
Adobe:  Looks like they've given up on HTML design altogether and are
pushing hard for people to do Flash. Not bad but not a fit for us.
http://www.adobe.com/designcenter/cs3/pdfs/CS3_Web_Workflow_Guide.pdf   I
actually like the Flex-Flash approach. I think DataServices is a bad,
proprietary, non-stanrad idea  but Flash (with or without Flex) to WS is a
fine, extensible, manageable, and standard architecture. Talk about
performance all you want but with downloading a Flash verus a simple (think
Yahoo's pages) web page means Adobe is the pot calling the kettle black.
Mockedup Screens. Documentation showed promise but Crippleware Trial was
like a weak DreamWeaver web page layout. Did not see how to hook up
scenarios. http://mockupscreens.com/index.php?page=Free_Download_2


   As Dr. Duncan Clarke was wont to say, - if you do a complete
specification then you've written the software. That is, the implementation
is the final specification. This is what UML architects fail to realize and
why the "formal" process is often scoffed by people who get things done.
What is needed is a decent prototyping tool which is easy enough to use that
screens can be created quickly but versatile enough it produces something
truly useful to the final product. I'm not sure any such thing exists for
our project. Sure, if we re-architect the UI to use x,y placement of
components then we can probably use one of these. NetBeans in its present
incarnation is probably still too complex for the average analyst... a
screen could be laid out but the linking ot other screens and the population
with sample data seems out of reach.

   Run away from anyone who tells you they can do a complete specification
without prototyping. ("Software Requirements Management or Software
Requirements Engineering tool" as standlaone professions are a joke. These
are pretend work.) You'll never know the real requirements until you build
the first version. This is the idea Agile is trying to capture. I've never
met one analyst yet who could list off the complete requirements. I think it
is because nobody really understands the business data and intricacies until
they define them (build the software) and, most of all, the software
redefines the business process. You may not want it to but it does. And then
after a decade of coding trying to implement every aspect of the business
then we get spaghetti.... :-)

Thanks for your feedback,
I'm guessing the industry is still maturing in this area. I really liked the
DENIM project you referenced. If that was finished out into a real product
complete with ability to morph pen drawings into real pages then it would
rock.("

Tim
It's odd IBM and Adobe do not have Demo Videos. How three years ago.
P.S> I do not represent my employer here. These are just my opinions. I'd
love to see some good demos.

On Thu, Sep 25, 2008 at 10:10 PM, Shawn Hartsock <[email protected]> wrote:

> You know, the Serena product isn't even in the same class as NetBeans or
> Flex Builder. NetBeans and Flex Builder produce code on some level. Serena
> produces a collection of documents... unless I completely missed a step in
> their demo where they said "compile" or something.
>
> This is a class of product called a Software Requirements Management or
> Software Requirements Engineering tool. The goal is to relate user stories
> to a requirements document. Some systems go as far as relating requirements
> to code... Serena calls their documentation a "prototype" and that's a tad
> misleading. I would call what their tool does a "mock-up" But, that might be
> the uniqueness of their product. It is both a mockup tool and a requirements
> management tool.
>
> Without thinking too hard Serena probably competes with:
>
> Borland Caliber Analyst
> http://www.borland.com/us/products/caliber/index.html
>  * (see story boarding)
> http://www.borland.com/media/en/edm/caliber_analyst_product_demo/index.html
>
> Rational software
> http://www-01.ibm.com/software/rational/offerings/reqanalysis.html
>  * http://www-01.ibm.com/software/rational/announce/rrc/ ... as usual IBM
> sucks at marketing their tools.
>
> ... and googling around on software requirements prototyping and rapid web
> prototyping gets...
> http://www.rallydev.com/products/demo_center/
> http://www.analysttool.com/
> http://www.mockupscreens.com/
> http://articles.techrepublic.com.com/5100-10878_11-1058664.html
> http://www.adobe.com/devnet/fireworks/articles/rapid_prototyping.html
>
> Unless you are planning on doing "proto-permanent" (where you just deploy
> the prototype and call it good) I can't imagine that J2EE or Struts2 would
> make a difference. The most important point to remember is that you have
> three distinct activities... producing a protoype/mockup... managing
> requirements documentation... producing an application. These are
> complimentary but distinct goals with separate artefacts.
>
> On Thu, Sep 25, 2008 at 4:04 PM, Tim Jowers <[email protected]> wrote:
>
>> Does anyone have a recommendation for a prototyping tool?
>>
>> I'm research prototyping tools for my company and have been advised to
>> look at
>> http://www.serena.com/products/prototype-composer/home.html
>>
>> Serena looks nice: it is wysiwig. drag-n-drop, with many field types such
>> as date input, and, most of all, an abliity to track requirements as screens
>> are prototyping.
>>
>> OTOH, netbeans lets you drag and drop jMaki components so could effect
>> something similar sans requirements.
>>
>> OTOH, Flex Builder lets you drag and drop Flash components so gives you
>> the best GUI experience possible to date. Yet no requirements tracking.
>>
>> The latter two could possibly produce real, usable artifacts for the dev
>> team.
>>
>> My big challenge is our product gets embedded into corporate webapps so
>> must be vanilla, customizable technology. Flash is out. Today is it <table>
>> based so x,y positioning (Div, wysiwig drag-n-drop) is hard or at least a
>> sea of change.
>>
>> Does anyone have a recommendation for a prototyping tool which makes sense
>> for a very large (>2M LOC) J2EE project which is still moving to Struts 2?
>>
>> Thanks!
>> tim
>>
>>
>> _______________________________________________
>> Juglist mailing list
>> [email protected]
>> http://trijug.org/mailman/listinfo/juglist_trijug.org
>>
>>
>
>
> --
> /** Shawn.Hartsock http://hartsock.blogspot.com/  //*/
>

_______________________________________________
Juglist mailing list
[email protected]
http://trijug.org/mailman/listinfo/juglist_trijug.org