Is this possible? I really hope it is.

You see, some of my design decisions is based on the procedure where I
need the actual J Window showing so I can step through the code line by
line. Query the current values and understanding what the problem is
with the data.

Around 98% of problems that I encounter in production systems are bad
data. There are even cases that users are even able to enter data that
they should not be able to enter. Unfortunately, without J Window in
your control ... it's next to impossible to do so.

Being able to debug remotely without stopping the server and being able
to pick which instance to debug is something that would be very-very
helpful.

:)


-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Oleg Kobchenko
Sent: Friday, February 16, 2007 2:32 PM
To: General forum
Subject: [Jgeneral] Remote J Debugging

That brings about an interesting idea:
Remote J Debugging. Useful for command-line
and headless jconsole processes like CGI.

GUI Debugger is implemented in J. There is also
J remoting interface via sockets. So it would be
possible to debug a headless instance of J with
a GUI instance of J communicating with this remoting.

Would be also a good "real-world" example of remoting
J programming.

Implementation note: existing debugger is split into
two parts: client and server. Client sets 
debugging mode and when debug condition occurs, 
attaches (starts or resumes) to the server, the
server showing the GUI. They communicate with a
protocol to update GUI and pass GUI events that control
debugging settings and commands like "step" to
the client.


--- Alex Rufon <[EMAIL PROTECTED]> wrote:

> Arrrgggghhh! Old habits really die hard. My J601 workspace has the
> allowed x. and y. enabled. You see, I still have running installation
of
> J504 in Vietnam, Saipan, Taiwan, Dongguan and the Philippines that I
am
> supporting. I really want to upgrade them but the Powers That Be is
> playing the Don't Fix the Not Broken card. :)
> 
> Sorry about the RAR thingy. I was having trouble uploading and I don't
> want to clog Chris email that I double compressed it.
> 
> About the Session Window, this is something that I picked up from
Chris
> Burke. I believe that this is how they envisioned J will be used
during
> testing/development cycle. I actually find this handy. You see, I have
> implementation of J running the Web, App, database servers (3-Tier
> Systems) and on the Client. I designed my J .NET wrapper for servers
in
> such a way that when I need to debug a code in production, I can
easily
> switch the whole server side system to run on a single instance of J
...
> instead of the normal multiple instances. This way, I can actually
catch
> all calls and execution on one Session Window. 
> 
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of David Mitchell
> Sent: Thursday, February 15, 2007 10:38 PM
> To: General forum
> Subject: Re: [Jgeneral] J.NET Guide
> 
> Alex Rufon wrote:
> > The file download for the "Real World Example" is now working. :)
> > 
> 
> I downloaded the zip and (after a brief tussle with unrar) installed
it.
> I 
> tried running the program 
> D:\Projects\J.NET\timedeviceviewer\bin\Debug\timedeviceviewer.exe and
> got:
> 
> |spelling error
> |   10000 100 100 base y.
> |                      ^
> |   dateencode=:    3 :0
> |[-343] D:\Projects\J.NET\timedeviceviewer.ijs
> 
> I do like the Session Server window.  It makes debugging simpler.



 
________________________________________________________________________
____________
Want to start your own business?
Learn how on Yahoo! Small Business.
http://smallbusiness.yahoo.com/r-index
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to