I'm really curious to know if either of you have looked at trails,
which builds on tapestry, hivemind and (what else, Chris?) to enable
very rapid application development. For the 1000 guys out there who are
building simple crud apps, it's /definitely/ a good place to start.
As far as "tapestry getting too complex"... while I'm not a tapestry
committer, I've watched the 4.0 developments with considerable interest.
>From where I sit, it seems like Howard et. al are pushing to make things
more simple, rather than the other way around. The improved listener
methods are a great example, as is providing the means to defer listener
invocation until the end of form-rewind (now the default; it will help
avoid many of the form "rewind" issues newbies experience). Another
example is the replacement of "conditional" with the if/else components
(currently available on t-deli), which also do a lot to keep
form-rewinds happy. That's less "blood and guts" you have to worry about.
That's not to say they haven't done a lot of neat things for the "power
user" as well... adding multiple persistence strategies, for instance.
But, as for me, I feel like tapestry has made a lot of things possible
(and easily possible) that seemed next-to-impossible before, and made a
lot of easy stuff even easier. For me, the biggest "learning curve" was
not tapestry, itself, but changing my paradigm on webapps. When I
stopped trying to treat webapps like a text-processing problem, and
started treating them like /applications/, with objects, state, visual
components, etc., as though I was building a swing or swt app, a lot of
things just sort of... fell into place.
That's not to say there isn't room for improvement... there certainly
is. But I don't see myself switching to any other java framework
anytime soon.
Robert
Danie Honig wrote:
> I am in agreement wholeheartedly with what you stated. As tapestry
> matures we need to find ways to keep people out of the blood and guts of
> Tapestry if they don't need to. This is what commercial tool vendors do
> with high level frameworks such as .NET or even the case of SAP's
> WebDynPro, to offer an alternative example. I think its vital that the
> Tapestry community stay focused and committed to Tapestry in order to
> avoid the scenario that you just described.
> It is harder for an open source project such as tapestry to create all
> of the nice GUI RAD tools that other environments have. WebObjects had
> Builder, .NET has Visual Studio and the WebDynPro environment has a set
> of wizards tools that allow you to rapidly assemble an application and
> do a very good job of keeping you out of alot of the blood and guts of
> the application. As Tapestry matures we should see the same thing and
> not the reverse. We cannot expect the core tapestry team and project to
> do everything, but if we had good tools that help ease the complexity of
> building a simple application then we have the best of both worlds, and
> it would make Tapestry a much more compelling framework. I have heard
> good feedback on Suns Visual Studio for JSF, what is the most promising
> tool for Tapestry support, the myeclipse integration? What do folks
> feel is the ideal concept for a Tapestry 4.0 and beyond development
> tool? Spindle is great, but merely scratches the surface of what is
> possible....
>
>
>
> Patrick Casey wrote:
>
>>
>>
>>> -----Original Message-----
>>> From: Danie Honig [mailto:[EMAIL PROTECTED]
>>> Sent: Tuesday, July 19, 2005 2:56 PM
>>> To: Tapestry users
>>> Subject: Re: Tapestry starting to look too steep
>>>
>>> Too early to criticize 4.0 and contrast it with .NET. Any new
>>> platform/paradigm requires a commitment to learn it. If your interested
>>> in solutions that require very little time to learn and are immediately
>>> productive but limit your ability to scale across a wide range of uses
>>> and offer substandard performance, I highly reccomend the microsoft
>>> approach...Historically this is what they are good at.
>>>
>>>
>>
>>
>> That's what I was trying to get at earlier I guess. There are some
>> tools which make the easy stuff easier and are ideal for relatively
>> simple
>> projects. From a reality check standpoint though, I think *most* projects
>> are simple. For everyone writing a 10,000 concurrent user e-business app
>> running atop a multiplexed server farm, there are probably 1,000 guys out
>> there writing simple little CRUD apps to people can update the corporate
>> phone list over the web.
>>
>> So it's a matter of choosing the right tool for the job. If tapestry
>> becomes more and more of a "high end" tool, I think it's going to have a
>> narrower and narrower user pool. At some point it may cross the threshold
>> where my projects usually sit and then I'll be in the unfortunate
>> position
>> of thinking "now why would I use that atomic bomb to kill a mouse?".
>> --- Pat
>>
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [EMAIL PROTECTED]
>> For additional commands, e-mail: [EMAIL PROTECTED]
>>
>>
>>
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
