Hi,
We wrote version 2.0 of our product using Struts with many developers(~9).  We 
started with what we called an engine that needed a GUI, which we built in 
Struts from scratch.  2.0 was a much simpler product (SAML 1.1 vs. SAML 2.0).  
During the development, we had almost all 8 developers working on the GUI at 
some point in time.  Started in July, ended in December.

In the end, it was hard to debug and maintain.  There were conflicts with 
sharing some of the key Struts files.  There isn't a component architecture 
available, as far as I know.  It was very hard to re-use any of the code and we 
ended up with scriptlets and such embedded in the .jsp's.  JSP's are a 
nightmare to debug and also hard to work with.  In addition, every time my HTML 
Expert needed to re-write the L & F, I had to "port" all of his changes, even 
using tiles.

So, let's fast-forward to 3.0.  Started in February, ended in November.  First, 
one of our developers went to QA.  Second, one of our developers went to Sales. 
 Third, we fired one and one quit.  So, now, we have five (5) developers.  Two 
spent the entire time building a new engine.  One spent a good portion of his 
time writing requirements and also working on the GUI.  The last two developers 
did nothing but GUI.

The only file that is really shared by developers is the .application file.  
So, far no problems.  We have a component architecture with over 120 
components, some of which are generic and re-usable.  The rest are swappable 
through out the project, re-usable with some minor changes.  Only OGNL in the 
HTML, moving that out over time.  And when the CEO and the HTML Expert decided 
to redo the entire L & F, NO developer envolvement.  Debugging usually means 
putting a breakpoint in java.  Hot deploy works fine and no run-time or JSP 
compile errors.  We use Spindle which highlights and warns us when we make 
mistakes.

When it came time to ship the product, we spent the last couple of weeks fixing 
bugs.  The last two days, we had less than 8 minor bugs in the bug list.  We 
shipped on time with few bugs.  Most of the developers were bored during the 
last week because we had stopped making changes and since the bugs were minor, 
we really had nothing to do.

Since starting 3.0, two other projects were started and they are using 4.0.  
One is shipping soon, the other is just starting but will be built on Tapestry.

All-in-all, I don't think I will ever willing do Struts again, not if I am 
given a choice ;-).

regards,

Mark

-----Original Message-----
From: spm [mailto:[EMAIL PROTECTED]
Sent: Tue 1/24/2006 11:24 AM
To: [email protected]
Subject: RE: Tapestry Benefits
 

Richard Clark wrote:
> 
> This strikes me as a "troll" -- a statement intended merely to  
> provoke people via annoyance. The simple fact is, your previous  
> message asked an incredibly broad question and that's exactly the  
> type of question that tends to get ignored on mailing lists. (And, if  
> you haven't noticed, most of the people posting here are actively  
> developing, so their free time is at a premium.)
> 
> If you have a few focused, specific questions, you'll get much better  
> results.
> 
>   ...R
> 



Ya, I guess you're right. I probably asked too many questions. Btw, ASP.NET is 
being pitched to my management...so half a troll...reality.

I am sold on Tapestry. I need to sell it to my mgmt. I was hoping to get a few 
answers here...was just disappointed in no answer at all.

Okay the important ones I have.
1. Any references of Tapestry in production?
2. What is the developer productivity Tapestry vs Struts? Estimate?

Thanks!


-------------------- m2f --------------------

Sent from www.TapestryForums.com

Read this topic online here: <<topic_link>>

http://www.tapestryforums.com/viewtopic.php?p=13664#13664

-------------------- m2f --------------------



Reply via email to