[Glass] [3.1.0.4] WAGemStoneRunSmalltalkServer

[Dale K. Henrichs]

Sebastian,

That is suspicious ... halt should halt if a halt is encountered, unless there is an exception handler on the stack ... 

So you are saying that when you explicitly start ZnServer from a GemTools workspace the halt in the accept method is "swallowed" whereas the same expression in SUnit causes a debugger to pop up?

If this is the case then I would be inclined to halt immediately and step through the code to get to the bottom of things...

What expression do you use to "manually start" Zinc ... I will try to find time to take a look today ...

Dale
 
----- Original Message -----
| From: "Sebastian Heidbrink" <sebastian_heidbrink at yahoo.de>
| To: glass at lists.gemtalksystems.com
| Sent: Monday, August 12, 2013 9:37:17 AM
| Subject: Re: [Glass] [3.1.0.4] WAGemStoneRunSmalltalkServer
| 
| Hi Dale,
| 
| well that is not the case.
| When I get such issues the SUnit tests show the same errors and fail.
| If I put a halt into the accept method of GSSocket I get different
| behavior with the SUnits-tests and my manual start of the ZnServer.
| The SUnit tests fire the halt and my manually started server does not
| :-/
| 
| Sebastian
| 
| Am 11.08.2013 08:39, schrieb Dale K. Henrichs:
| > Sebastian,
| >
| > I don't know if this is a factor or not, but if you don't take
| > great care in making sure that the sockets are all closed (by
| > using stop server before starting another one) then you might
| > start getting odd results ... especially during the phase where
| > the bugs in the start/stop logic may not be completely worked out
| > ...
| >
| > Dale
| >
| > ----- Original Message -----
| > | From: "Sebastian Heidbrink" <sebastian_heidbrink at yahoo.de>
| > | To: "Dale K. Henrichs" <dale.henrichs at gemtalksystems.com>
| > | Cc: glass at lists.gemtalksystems.com
| > | Sent: Saturday, August 10, 2013 4:28:44 PM
| > | Subject: Re: [Glass] [3.1.0.4] WAGemStoneRunSmalltalkServer
| > |
| > | Hi Dale,
| > |
| > | thank you for the detailed information,... I tried to start my
| > | server
| > | this way and I found out that in
| > | "ZnRequestLine readFrom: stream " there's always only an empty
| > | string
| > | resulting from "(ZnLineReader on: stream) nextLine"
| > |
| > | This does not apply for the Unittests. Within those the read
| > | lines
| > | are
| > | always properly filled.
| > |
| > | I also have to start my request prior to starting the server....
| > | otherwise nothing happens.
| > |
| > | I'll keep searching for some more needful details.
| > |
| > | Sebastian
| > |
| > | Am 09.08.2013 10:20, schrieb Dale K. Henrichs:
| > | > Sebastian,
| > | >
| > | > To debug the setup, you should run:
| > | >
| > | >    WAGemStoneRunSmalltalkServer startGemServer: '<servername>"
| > | >    On:
| > | >    <port number>.
| > | >
| > | > in your GemTools image. Be aware that the above statement will
| > | > block until you hit an error of some sort and you need to read
| > | > the
| > | > section on GemTools and processes[1] and some tips on working
| > | > with
| > | > servers[2]. For debugging out of memory problems, read the post
| > | > on
| > | > that topic[3].
| > | >
| > | > If you do the above, the GemTools debugger will come up when
| > | > you
| > | > hit a problem either out of memory or an error ... doing the
| > | > above
| > | > you can set breakpoints or halts and the debugger will come up
| > | > and
| > | > you can track things down using normal development techniques
| > | > ...
| > | >
| > | > Once you get to the point where the server starts and runs
| > | > normally, you will need to install some strategically placed
| > | > error
| > | > handlers so that you can arrange to snap off continuations and
| > | > stash them in the Object Log for later debugging ... this is
| > | > already done for you in Seaside...
| > | >
| > | > Dale
| > | >
| > | > [1]
| > | > http://gemstonesoup.wordpress.com/2009/04/15/glass-beta-update-working-with-soap-preview/#GsProcesses
| > | > [2]
| > | > http://gemstonesoup.wordpress.com/2009/04/15/glass-beta-update-working-with-soap-preview/#soaphttpdev
| > | > [3]
| > | > http://gemstonesoup.wordpress.com/2008/11/19/gemstone-101-managing-out-of-memory-situations/
| > | > ----- Original Message -----
| > | > | From: "Sebastian Heidbrink" <sebastian_heidbrink at yahoo.de>
| > | > | To: glass at lists.gemtalksystems.com
| > | > | Sent: Friday, August 9, 2013 8:51:17 AM
| > | > | Subject: Re: [Glass] [3.1.0.4] WAGemStoneRunSmalltalkServer
| > | > |
| > | > | Hi Dale,
| > | > |
| > | > | I tried it and I get the same result....
| > | > |
| > | > | Meanwhile I realized that my script only works if I commit my
| > | > | transaction prior to starting the gem like this:
| > | > |
| > | > | "
| > | > | WAGemStoneRunSmalltalkServer addServerOfClass: ZnServer
| > | > | withName:
| > | > | 'myserver' on: #(9001).
| > | > | .... commit the transaction ...
| > | > | WAGemStoneRunSmalltalkServer startGems:
| > | > | (WAGemStoneRunSmalltalkServer
| > | > | serverNamed: 'myserver').
| > | > | "
| > | > |
| > | > | Otherwise the (WAGemStoneRunSmalltalkServer  serverNamed:
| > | > | 'myserver')
| > | > | is
| > | > | nil in WAGemStoneRunSmalltalkServer startServer: serverName
| > | > | On:
| > | > | portInteger.
| > | > |
| > | > | Since the remote debugging does not work with the currently
| > | > | delivered
| > | > | runSmalltalkerver scriptsfor 3.1.0.4 I have to work with
| > | > | "halt"
| > | > | and
| > | > | looking into the logs.
| > | > |
| > | > | I already tried to change the script this way: see attached
| > | > |
| > | > | It seems the listenloop is running, but I never end up in
| > | > | "handleRequest" :-/
| > | > |
| > | > | Sebastian
| > | > |
| > | > |
| > | > | _______________________________________________
| > | > | Glass mailing list
| > | > | Glass at lists.gemtalksystems.com
| > | > | http://lists.gemtalksystems.com/mailman/listinfo/glass
| > | > |
| > | >
| > |
| > |
| >
| 
| _______________________________________________
| Glass mailing list
| Glass at lists.gemtalksystems.com
| http://lists.gemtalksystems.com/mailman/listinfo/glass
|