Stop 2nd Arev running (AREV Specific)
At 18 FEB 2000 05:35:54PM Barry Stevens wrote:
Could someone tell me the technique/Code for ensuring that a workstation doesn't try to run a 2nd copy of AREV.
T.I.A.
Barry
At 18 FEB 2000 05:41PM [url=http://www.sprezzatura.com" onMouseOver=window.status=Click here to visit our web site?';return(true)]The Sprezzatura Group[/url] wrote:
At login take the unique portion of the network user id. Place a lock against VOC. If the lock fails you're already logged in.
World Leaders in all things RevSoft
At 19 FEB 2000 04:21AM Warren wrote:
Now if it were OS/2 you'd just change the behavior of the program object to "Open Existing Window" rather then "Open New Window". But then OS/2 is soooooooo inferior to Windows…
At 19 FEB 2000 04:56AM [url=http://www.sprezzatura.com" onMouseOver=window.status=Click here to visit our web site?';return(true)]The Sprezzatura Group[/url] wrote:
We loved Betamax…
World Leaders in all things RevSoft
At 19 FEB 2000 12:56PM rayc@symmetryinfo.com wrote:
There are some cool things about OS/2 . Periodically, I have to walk over to our OS/2 PC and say "hey that's pretty cool." Then I go back to my Windows NT. I'm starting to forget OS/2 things…For now, MS is king of the desktop.
BTW, I also have an old Betamax machine if anyone's interested.
rayc@symmetryinfo.com onmouseover=window.status=imagine … ;return(true)"
Ray Chan ~ Symmetry Info
At 20 FEB 2000 11:35PM Jonathan Bird wrote:
I get each user to lock a record in a "USERDATA" table as they log in. If this is locked, it means they are already logged in. A few problems exist with this, not the least is with "unlock all" statments. The method is reasonible OK, though. There is also a programme around called "EXEONE" meant to control the execution of programmes under Windows. Can't remeber wher I got it from, but it never seemed to work for me, but you may like to find/try it.
Regards,
J
At 21 FEB 2000 06:35PM Barry Stevens wrote:
]]At login take the unique portion of the network user id. Place a lock against VOC. If the lock fails you're already logged in.«
I have done this via a startup prog at logon, but lock dissapears after the menu comes up. If I run the startup prog again (via TCL) the lock is placed.
Any clues or Hints.
At 21 FEB 2000 07:08PM [url=http://www.sprezzatura.com" onMouseOver=window.status=Click here to visit our web site?';return(true)]The Sprezzatura Group[/url] wrote:
There is an interesting bug in probably all versions of ARev related to initial logon and TCL levels. Usually this manifests itself with a corruption with certain @VARS, like @RECORD and @DICT and @ID. Usually this will happen if you try and run an R/LIST report off a softkey.
The basic fix for this bug is to have the startup options do an E RUNMENU menu or whatever it is needed to fire up the application. This seems to place the system in a pseudo level 2 which solves these interesting level 1 problems. END from the debugger works more consistantly this way as well. Perhaps this will also solve your problem.
As an aside, and I'm sure you've already done this, but disable the network cache within ARev and remove the 020 from REVBOOT. We've been told that this is no longer a requirment with the NLM and NPP, but it has caused so many other problems in the past, it can't hurt to try it.
World Leaders in all things RevSoft
At 22 FEB 2000 07:09AM Barry Stevens wrote:
]]The basic fix for this bug is to have the startup options do an E RUNMENU menu «
Was already starting the menu this way. The logon is called then the menu is run. I will try call logon routine after menu run
E;E open_common_files;runmenu mainmenu
Thanks
Barry
PS Knew about the last bit.
At 22 FEB 2000 10:05AM Victor Engel wrote:
If you are using the NLM, you can make calls to get a list of logged users. Check the source code to NLM_STATS as a guide.
At 22 FEB 2000 05:21PM Warren wrote:
As I recall this 'bug' appeared in version 2.0. Seems like CATALYST and many of the other system subroutines no longer saved off (protected) many of the @variables, most notable @DICT, @RECORD and @ID.
Version 1.15 and 1.16 (and probably earlier) behaved differently in this respect. I had to recode several modules to store and restore these @vars in order for an app to behave as it did prior to upgrading to 2.0x.
RTI probably stopped 'protecting' these @vars because 2.0x ate up so much more string space than v1.16.
At 22 FEB 2000 07:58PM [url=http://www.sprezzatura.com]The Sprezzatura Group[/url] wrote:
push.session and pop.session are still always called. It's just that for some reason, TCL level 1 is behaving differently. Aaron never found out why, though he spend some time looking for it.
[/i]World leaders in all things RevSoft[/i]
At 23 FEB 2000 02:51PM Warren wrote:
I don't exactly recall when ARev 2.0 was released - 1989 seems about right. As I recall (and I can look at a running copy of the app some time but not in the near future)
Softkey calls from Windows had to be changed. These Windows were called 2 or 3 menu levels down - certainly not from TCL-1. I don't recall exactly but I think 'E' or 'X' calls had to be changed to 'X' (in the case of 'E') calls or 'S' calls (I didn't develop the app).
In one case the Window soft key would call a subroutine and the subroutine would do a catalyst call to another window/and or popup (don't remember which). After the window/popup was executed and control returned to the calling routine calculates that worked in v1.16 no longer would - I traced this to @DICT and @RECORD being changed from the catalyst call.
I also documented cases where @DICT would get changed/corrupted when free string space would drop below a certain threshold.
Again none of these problems existed in v1.15/1.16. All this was documented along with code examples that reproduced the errors and sent to RTI.
At 23 FEB 2000 04:06PM Barry Stevens wrote:
Sprezz Idea now works fine.
In the user setup shell command:
E;E LOCKFOR2NDAREV;RUNMENU MAINMENU
Lockfor2nd arev just locks VOC using field(@station,"*",2)
In the program that is run during logon I check if record is locked and logoff with msg if true.
Thanks all
Barry