hivemind-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "James Carman" <>
Subject [Fwd: Hivemind doc.s]
Date Thu, 20 Apr 2006 10:35:45 GMT
Some interesting feedback from a tapestry user about HiveMind:

Hi James,

Excuse the rough email. My Apple Mail client decided to be difficult
overnight and I've had to switch to Thunderbird.

Yes, good idea about my giving you more feedback. I'll try to put aside
some time on Saturday afternoon to give you some constructive criticism
about the Hivemind doc.s. I do remember becoming frustrated that the
documentation was not simple. For example, if I want to learn about a
configuration I have to read through this:

"A central concept in HiveMind is /configuration extension points/. Once
you have a set of services, it's natural to want to configure those
services. In HiveMind, a configuration point contains an unordered
collection of /elements/. Each element is contributed by a module ...
any module may make contributions to any configuration point visible to it."

This is quite frustrating to read. I don't really care about elements,
at this stage, I just want a simple definition of how a configuration
works in Hivemind and what I can do with it. The element, contribution,
and module definitions which are thrown in, and themselves require
definition, make it really complex. The documentation then trails off
into definitions of everything else but a configuration point. Still,
what does a configuration point do? What kind of options do I have to
configure services? What are some common needs for this sort of thing?

I general description of my impression - documentation *should* give a
quick overview of key concepts. The detailed stuff is perhaps good but
it's mixed in with an introduction to the basic concepts and, as a
result, the reader walks away feeling confused.



If you can point to any specific issues with the HiveMind documentation that
made it difficult for you, I'd be glad to try to rectify them.  I'm one of
the committers on the HiveMind project.  By the way, have you read my
article on HiveMind at
(  I'm not
saying it's the best article in the world or anything, but it does give a
good introduction to HiveMind and I have received a lot of positive feedback
from folks about it.  It doesn't cover configurations, but I plan on doing a
follow-up article to address that.  I'm sorry to hear that your experience
with HiveMind has been so painful, but we need feedback like yours to make
it better!  Thanks and I hope that article helps.


-----Original Message-----
From: Nick Faiz []
Sent: Tuesday, April 18, 2006 10:42 PM
To: Tapestry users
Subject: Re: tapestry for a webwork developer

Ive been using webwork 2 previously to this, my first project with
Tapestry (4.0).

I think the weirdest thing was accepting the idea of using ASO's
(Application State Objects), whereas in webwork 2 we coded to a
stateless model. I still have to see how it scales but we're running
with it for now.

I'm also still not sure how, or even if, Tapestry meets the needs of
interceptors for pages and services alike. This was very clear in an
xwork.xml file, thanks to interceptor stacks. If anyone can point me
to an example of an interceptor which can be function within
Tapestry's framework I would be grateful (please don't reply and
suggest I use a ServletFilter - I mean Tapestry's framework).

Besides that there are differences - Tapestry is excellent at the
modular component thing, Webwork 2 is a little messier but makes for
a faster development path. I find that I'm often fussing about with
bindings and the like in .page and .jwc files. On the other hand, the
configuration effort might be worth it, given how easily I can drop a
component in somewhere.

I find Hivemind to be very verbose and complex compared to Spring.
The documentation for Hivemind is sometimes lengthy and indirect and
can be frustrating. Hivemind is running well within our project, but
whenever I seek to learn more about it's workings I find myself
preferring to read the source code.

On the other hand, I'm really enjoying using Tapestry and Hivemind.
If I work with it again I'd like to experiment with configuring
Velocity or Freemarker at the template level and see how else I could
customize it. I might, perhaps, try and involve Spring, rather than
using Hivemind as the IoC framework.


On 15/04/2006, at 6:24 AM, Payne, Matthew wrote:

We can use Tapestry 4.
It looks like I need to take a look at the specifics of applying
these services.



James Carman, President
Carman Consulting, Inc.

James Carman, President
Carman Consulting, Inc.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message