[PJUG Javamail] PHP < -- > EJB?

Steve Mayzak stevemayzak at gmail.com
Fri Oct 23 18:17:03 EDT 2009


Richard,

If you are seriously considering Option 1, I would have a look at Grails or
Spring ROO.  Grails would allow you to take your existing PHP app and port
the pages to GSP's (similar to JSP) in a hurry.  I did this conversion for a
site with around 20-25 pages and it only took 2 days of half learning and
half doing the work.  Since Grails is built using Groovy which runs on Java,
you can always write things in Java if you don't want to use Groovy.

As for Spring ROO, it keeps you in java but still gives you that "convention
over configuration" model so you can get things up and running quickly.

However, both ROO and Grails are all about DDD so if your domain model is
already written, I am not sure how easy it will be to use either of these.
Conceptually I think it would work just fine, generating your controllers
and views from the Domain classes that is, but I haven't tried that in
practice with EJB3 Entities.

Just found this doing a quick google search
http://www.infoq.com/articles/grails-ejb-tutorial

My 2 cents,
Steve Mayzak

On Fri, Oct 23, 2009 at 2:59 PM, Richard Johnson <richardj at lingosys.com>wrote:

>  Great observations Gary,
>
> For the benefit of all the professionals, here's a case study in how things
> spin out of control in a hurry on a Friday.  :-)
>
> The deliverable date was cast as fiat by some executive (either in Chicago
> or Bay Area) who has only managed simple value-added services before--never
> manufacturing even, let alone software.  It was out of the hands of even my
> boss's boss.  We're doing our best.
>
> I explained to them that some ideas are conceptually simple and
> technologically hard.  For example, a light bulb.  You can make a light bulb
> by putting a wire across the terminals of a car battery.  Conceptually
> simple.  However making the wire produce more light means making it smaller,
> but then it burns out quickly.  And that's just the first hurdle.
> Technologically hard.
>
> It might indeed be that I think they want a lot more than they do, but I
> suspect not.  Current plan is to spend the weekend coming up with the spec I
> can and give them the cost/timeline if we do:
>
>     (1) rewrite the UI in Java
>    (2) use something like the PHP-Java bridge
>    (3) generate XML and write a program to use it
>    (4) upgrade the app and write PHP/SOA hooks
>    (5) hire a contractor to write it.
>
> I honestly think the delivery date is unachievable, but I think (at the
> moment) that some visible and important piece of the deliverable IS
> achievable by the end of next week.  What that is, and how much it will do
> remain to be seen.  I like incremental releases when we can make them work.
> --
> Richard
>
>
> _______________________________________________
> Web Site - http://www.pjug.org/
> Javamail mailing list
> Javamail at pjug.org
> http://www.pjug.org/mailman/listinfo/javamail
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.pjug.org/pipermail/javamail/attachments/20091023/6a4fa782/attachment.html 


More information about the Javamail mailing list