"mark.supinski" wrote:
Wow do I feel ripped off. I actually purchased the subscription to the JB Dev Studio because I believed the marketing and wanted to avoid the wasted hours of work trying to get a working development environment. Stupid me -- little did I realize I would get a half-baked environment where NOT A SINGLE example application works. None. Not one.

I can't possibly be setting my expectations any lower here -- all I want to do is see one of the example apps run.

Step 1: Install Java -- version 6 in my case.
Step 2: Install JBDS 2.0.0 -- allow it to use all default values.
Step 3: Try installing *any* of the seam example apps (booking, dvdstore, etc) from within the environment

Result is always the same -- the apps don't work. Datasources are almost always the problem -- despite letting the apps use the default hypersonic built in db.

Can anyone claim they can get the apps to run out of the box with no mods / changes? What is the point in this distribution when it doesn't actually work...

I know exactly how you feel I also purchased JBDS and bought Seam in Action I was seriously eager until about a month later when I almost defected to Ruby on Rails. I was so pissed at how seemingly fragile and convuluted this technology was at first that I was going to ask for a full refund. Every time I tried to deploy a application in JBDS, my JBoss server would crash, or say a port was in use or maybe it had the swine flu. To this day I'm still not too sure about the details but it was simply ridiculous. I asked myself how the hell anyone can use this in a production environment without something falling apart and going postal.

Well I discovered that using seam anything less than seam 2.0 was big trouble as seam 1.2 was utterly horrible and unreliable. As with any new framework you gotta do some reading on available documentation. I'd like to think I did all the necessary research before investing in yet another Java Web Framework but it seems that one could spend their entire life just researching Java Web Technology and not build a single product or complete a single project because their RTFM(s). I have high hopes for Seam and I really want to thank the the guys at JBoss for coming up with this technology. But man I tell ya it's not their fault really but 1. You guys really need decent tutorials or fool proof quick start manuals at least for your paying customers. And no the Charlie Chaplin fast forwarded flash tutorials do not cut it. 2. Maybe have some micro prizes on who can come up with the cool richfaces components or seam components that will prevent developers from going postal and might, just might actually make programing fun again.

I've had problems with the program examples on and off for the past 5 months. I tried to use ant build then I tried downloading the examples via JBDS project interface from the SVN. It worked great for a couple weeks but then just stopped working. I asked the guys at JBoss forums about this and they say their having problems with their servers. 5 months later and it looks like project example repository is having one hell of a DDOS attactk or they can't cover the bandwidth fees.

Seam is a interesting technology and a hell of a improvement over struts anyday. But it sure is not perfect, it has some seriously odd quirks and there is some frustrating ambiguity in terms of how to use Seam-Gen properly in building entities and there is also entity-query and searchable session-beans list (don't let me get started on that). I'd actually like to thank all the Struts, Seam and other Java Web Dev out there as the more I start using Java Web frameworks the more I realize "There has got to be a hell of a better way to do this.". And now I'm taking that to heart and I'm getting out of that Java vision tunnel mind set and I'm looking at other technologies that 1. Have "complete documentation in relative size to it's complexity" 2. Does not suffer from "magic-code" or "blind-code" phoenomenon. and oh yeah 3. Has a active developer/user community where if you ask a question you can get a reasonable answerable within oh say a "month". (Try asking a legitimate question in SeamFramework forums and time how long it takes for someone to acknowledge you even exist).

 

"nmatrix9" wrote:

"mark.supinski" wrote:
Wow do I feel ripped off....

I know exactly how you feel I also purchased JBDS and bought Seam in Action I was seriously eager until about a month later when I almost defected to Ruby on Rails. I was so pissed at how seemingly fragile and convuluted this technology was at first that I was going to ask for a full refund. ...

Don't defect to Rails, defect here: http://www.atk-framework.com/

They're at version, what, 6.4.2? defect? nah. We've got to write portlets...

Rails doesn't really compete with SEAM, it's too low level, but a Rails project called Streamlined does, but that project seems to have fizzled.

I haven't deployed SEAM but like JBDS bundle and think it's worth it having tried to configure an eclipse bundle. I'll have to check out SEAM, but currently my main focus would be Spring and services, with hand rolled ajax.

Have deployed ATK, and the server-side bound widget idea is ok as far as it goes. What happens is you end up filling in cracks, fixing bugs in the framework, adding javascript enhancements that aren't in the framework. In the end you wish you had something simpler and never gone down that road. Just well-factored page templates and a decent template system that allows plugins and modifiers should be good enough.

Speaking of complexity, there is a visual framework built on top of SEAM, "Taylor" http://taylor.sourceforge.net/index.php/Overview

They talk again of competing with "Rails productivity", but this really a visual layer on a crud layer(SEAM) already beyond Rails. "Rails" means a lot of different things to a lot of different people.

 

"kucerar" wrote:
Spring and services, with hand rolled ajax.

By "hand-rolled ajax" I mean having a decent javascript templating system and in-page javascript SQL data source to leverage:

http://code.google.com/p/trimpath/wiki/TrimPathEnterprise

I've only seen other javascript DOM templating and that was bound up with ExtJS. JavascripMVC might have something good, but you have multiple serverside files to work with there rather than putting it right there in the single page. I don't want a separate file for every little snippet.

Have done a couple read-only look-in portlets(jboss) with "TrimPathEnterprise". I just read all the data into the page. According to tests there's 20 meg of storage available to javascript in the browser without needing to resort to gears storage or start making calls to the server. That's plenty of room for most things.

But, time to have a look at SEAM lol...