Re: deployment/debugging

I am going to weigh in before this degrades into another "my tool is better than your tool" free for all.

Joe, I have to side with Jon on the "stack trace doesn't make any sense" argument. I have been doing a lot of Java programming recently and have forgotten how weird and difficult it is to unwind the stack as a relative newbie but even now, if the stack trace doesn't include MY code, only the stack above my code, I can sometimes sit there scratching my head trying to figure out just where my code went south. Scrolling down until I find my exact line of code and then looking at the stack above it *usually* helps me quickly resolve it. But, like I said, I have had my head in Java for the last couple of months and so I really feel comfortable with unwinding the stack trace. For a newbie, it is daunting and I wish there was a way to throw an error that says: "Dude! You forgot to initialize that variable and so I ended up with a null value!" Instead of...... Well, I'd post a stack trace, but my code has gotten so good I don't have one at the moment.... :-) In any case, just like CL and RPG (and MCH) errors are baffling to new RPG programmers, the Java programmer gets it in spades and EGL carries some of that baggage. It takes a while to use that stack trace to your advantage.

Not to put words in Jon's mouth, but he was simply commenting on the difficulty in unwinding the stack trace. At least that was the way I read it. My response to you Jon is, "Yeah, it can be a little daunting to deal with" You'll see those kind of errors thrown by RDi and WDSC occasionally even when doing a simple upgrade and like all "new" things it takes some help and experience in sorting them out. I DO remember just staring dumbly at my first stack trace and wondering "Man.Three lines of code and I get THAT kind of error? Yikes!"

I didn't see anything in Jon's post that was disparaging to EGL but Jon doesn't need me to defend him. EGL is a different beast and I will agree, Jon, that the errors and debugging takes some getting used to. Whether or not a programmer used to RPG and new to EGL can deal with the transition is something yet to be clearly established. EGL seems to be holding it's own and with each iteration, gets better. It is good to have RPG folks give feedback on what they find difficult in EGL because that is the only way it will get better.

Thats my 2 cents worth.

Pete


Jon Paris wrote:

I should have known that I couldn't say a thing on an EGL list without Joe turning it into a personal attack.

Note - I didn't say a thing about PHP being easier - better - or anything else - I didn't mention PHP.

I have ALWAYS said that I like EGL and think it has potential. I have recommended it to several of my clients and will continue to do so.

But I remain concerned over the ability of small shops with zero Java/ App server etc. experience to deploy it successfully.

Period - end of story.


Jon Paris

www.Partner400.com
www.SystemiDeveloper.com



On 10-Jul-09, at 6:18 PM, Joe Pluta wrote:

Jon Paris wrote:

I looked back in the archives and don't see as many as I remembered
(mismembered?) so I have to assume that some were on other lists.

This recent one http://archive.midrange.com/egl-i/200906/ msg00009.html
is an example of the kind of thing I mean though. And here's another http://archive.midrange.com/egl-i/200906/msg00000.html
.

Another example of what it seems should be simple is this one which
even the IBMers seemed unable to resolve http://archive.midrange.com/egl-i/200901/msg00006.html

As I noted before - many of the problems can be resolved if you have
enough experience in the App Server/Java arena - but most RPGers don't
have that.

Jon Paris

"so I have to assume some were on other lists". <chuckle> That speaks
volumes.

Anyway, AFAIK, Pete's problem has been resolved. Pete has a tendency to
push the edges of technology, and to his credit typically works through
those issues.

The other two issues were Tim's. One was from January which has
evidently been resolved since he asked other questions in June. The
more recent issue is an environmental issue in one PC works but not
another. Environmental issues are just as painful in PHP as they are
anywhere else - for example, trying to run MySQL with the DB2 storage
engine on Zend.

So in the end, we've got two resolved issues and an environmental
problem. And for that, you insist that EGL is too hard for RPG
programmers. Yet you seem to think that PHP, with its arcane syntax and
Web 1.0 functionality is a good fit for RPG programmers.

You have a unique perspective, Jon.

And if you don't think PHP is severely limited functionally, just try to
get Dojo and PHP to work together. Here's an example of the "easy" PHP
language:

http://www.dojotoolkit.org/forum/dojo-core-dojo-0-9/dojo-core-support/bringing-php-mysql-and-dojo-together


Joe

P.S. Just so you know, I haven't looked at a line of Java generated from
EGL for over a year now. The EGL debugger is phenomenal. It's really
nice to work at a business level, and not have to putz around in Java.
Of course even debugging Java beats debugging PHP.
--
This is the EGL on and around the IBM i (EGL-i) mailing list
To post a message email: EGL-i@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/egl-i
or email: EGL-i-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/egl-i.