What’s the reason to make it NOT depend on Spring? tapestry-acegi does NOT uses Spring IoC container, it uses T5 IoC for configuring the services, but it does use some Spring utility classes, which is in my opinion just fine (it certainly increases size of the final archive, but who bothers about this? :).

The bad thing about tapestry-acegi is that it is not very flexible, though. Recently I was trying to apply tapestry-acegi for CAS authentication (http://www.ja-sig.org/products/cas/) and found that I had to rewrite almost every line of its module class, which unfortunately indicates that tapestry-acegi is not very reusable. On the other hand, tapestry-acegi does have an added value above the core Acegi (for example, worker that allows applying the @Secured to the methods/pages), so I can’t just throw it away and write my own module file — that’s not enough, I need these helper classes.

yuan gogo wrote:
I don't understand u well, I think.

But I'm not making FUD here if you read the mail list before. Or my English
is so bad to understand u, if so, I apologize.

yes, security "without configuration" is impossible, by "without
configuration" I mean I hope there's something looks like "Native", if that
confused u, please blame my pool English at will.

After sereval months using Tapestry 5, I like it very much. I hope I can
work with tapestry only, at least no spring any more. But if I use acegi,
spring jar will be included too, that makes me unconfortable because I think
tapestry-ioc is good enough for my work and I don't need any other stuff
spring provides.

At least, if tapestry-acegi doesn't have a dependency on spring, that will
be very GOOD!

Just my own point of view.

Thanks!

2008/3/19, Christian Edward Gruber <[EMAIL PROTECTED]>:
I love it when someone invents new people to stage internet
conversations.

note the "just only" and other little language features that this
troll uses in all his posts, regardless of the story.

Even if I'm wrong in that supposition, this is open-source software.
If this was a real request, and not just another attempt to slander
Tapestry, then this person could make concrete recommendations, or
even better, contribute code.

But mostly I love the premises.  In a "real web framework" security is
"built in" and no configuration is even necessary.  Bullcrap.  Every
security system needs configuration, because no system can anticipate
all possible needs and use-cases and auto-configure for that magically
discovered state.  And why should T5 re-build what someone already
wrote and tested.  If acegi works, then why not integrate it?  Ooh,
wait!  I know, I know!  HLS sucks because he didn't create his own O/R
Mapping layer!  'Cause why use Cayenne/Hibernate/JPA/TopLink when you
can, without compelling necessity, write your own, untested system
from scratch!

Meh.


Christian.


On 18-Mar-08, at 23:51 , Joshua Jackson wrote:

I agree. I have requested this before for Tapestry to be a full web
framework since right now it's just only a plain web framework.

On 3/19/08, yuan gogo <[EMAIL PROTECTED]> wrote:
Tapestry is good.
But I personally think it's lack of something that should be built-
in.
e.gsecurity!

It's right that tapestry can work well with acegisecurity, but it's
not
native, is it?
And we programmer need time to write the configuration,
debug ........ it's
boring!

Even there has been tapestry-acegi, but 2 projects can not be
always be
synchronous. What's more, tapestry-acegi can not do all that acegi
does.

I was asp.net user once, even I think it's not so good, but it's
good for
"rapid" development, because some functions like security is built-
in.

thanks.

--
Let's show the world what we've got.

Blog: http://joshuajava.wordpress.com/

---------------------------------------------------------------------
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]





--
WBR,
Ivan S. Dubrov

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to