Discussion Forum Postings for June 1998
Infobase Search ~ ISSUG Core Page
========
Glenn Smyth posted the following article on the ISSUG Bulletin Board
Dated : 01 June, 1998 at 04:25
Subject: Re: What's the meaning of missing flags
There is one dis-advantage of removing them. IT's mpatc depends on whether or
not the EAB is used in a loop or in batch/online. Basically once you turn off
the missing flags option High Performance View Matching will be blown. This
has a significant impact on the amounbt of code COmpsoer has to generate to
call your EAB. There are numerous discussions on HPVNM on the archives.
========
Murat Gungor posted the following article on the ISSUG Bulletin Board
Dated : 01 June, 1998 at 07:46
Subject: Re: Cool:Gen Application: Is Client Manager Running?
Hi Larry,
Sorry, the cool:gen code in my previous posting does not work! Window function
finds the current and child window only! You have to use an external C code
or other OLE utilities such as the WinFun.
Murat
========
Kenn Griessel posted the following article on the ISSUG Bulletin Board
Dated : 01 June, 1998 at 18:53
Subject: Robo-Help & COOL:Gen
Some time ago there was a thread on this BB re: Cool:Gen and on-line help. The
notes are no longer on the active board, and are not yet in the archive. Is
there a way for me to access that thread?
Does anyone have any hints/tips/advice about linking Robo-Help files with a
Cool:Gen application? Some of my questions are:
- What files does Cool:Gen automatically generate (*.hpj, *.hlp, etc.) and what
file(s) need to be replaced with the Robo-Help file(s)?
- Is there a need to define Help-IDs in Cool:Gen?
- How are the executables then linked with the correct help file?
We require window, field, and control level help.
Thank you.
========
cc posted the following article on the ISSUG Bulletin Board
Dated : 01 June, 1998 at 19:05
Subject: Multi-use report needed
I am wanting to create a report that can be viewed or downloaded via the internet.
I need this report to be created at run-time. A specific web page will kick
of a procedure that creates a report. The user can then select to view or download
the report. The big problem is that the server procedures and database will
reside on an AIX box within our firewall, but the report that is created needs
to reside on an NT box outside the firewall where the ASP code resides.
Anyone have a solution or suggestion to head me in the right area?
========
george simpson posted the following article on the ISSUG Bulletin Board
Dated : 01 June, 1998 at 23:08
Subject: Re: Robo-Help & COOL:Gen
Kenn,
1- Cool:gen's generated help is cheap to do.
2- I'm no expert but I'm not sure how much value
tools like robohelp are, our coders write the
stuff in MS Word, they compile it with forhelp,
but I'm not sure why.
3- they are linked via the some suffix (load module name)
4- help id's change any time you add or subract a window
control or add or subtract a command from the business system
To resolve the shifting help id problem in the intdbclm sort
the header .h file using the sort utility then call a program
to renumber all helpid's, yep name em, not starting in a RP or
CI, then send a quarter to the posting addiction fund ( I need help )
Here is my c code,, if you need more help I, you can email me
or post,, here's my c code (hope it goes through,, if not
I'll email you) - george
include
include
include
include
main (int argc, char * argv[]) {
FILE *fp, *fp2, *dbg;
char buf[122];
short x;
int rcid=3000;
fp = fopen("gws.srt", "r");
strcpy(buf, argv[1]);
strcat(buf, ".h");
fp2 = fopen (buf, "w");
memset(&buf, ' ', sizeof(buf));
while (fgets (buf, 121, fp) != 0) {
if ((buf[8] == 'R' && buf[9] == 'C')
|| (buf[8] == 'P' && buf[9] == 'I')
|| (buf[8] == ' ' && buf[9] == ' '))
fprintf(fp2 , "%s", buf);
else {
x = 10;
while (buf[x++] != ' ')
;
buf[x] = 0;
fprintf(fp2 , "%s %d\n", buf, rcid++);
}
memset(&buf, ' ', sizeof(buf));
}
fclose(fp);
fclose(fp2);
}
========
Preben Nilsson posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 08:49
Subject: Re: Robo-Help & COOL:Gen
I think that I was responsible for bringing the subject of online help up previously.
To summarize, here is what you do:
- Create help identifiers for each item you want to document. Multiple items
(things on your GUI) can have the same help identifier.
- Generate the window manager (to get the .H file)
- Run a small macro in RoboHELP to create default topics (I made this macro.
Please mail me directly, if you are interested. Or catch me in Berlin...)
- Make your help project (in RoboHELP)
- Copy the .HLP file to the same directory as the .EXE and .DLL
As George mentioned, then you have to re-make you help project with the new
.h file every time you change anything on the GUI. I find it safest to re-make
the help project every time I build a new production copy of the window.
BTW. If anyone is interested in using RoboHELP to create help for OS/2 systems,
then pls. mail me directly.
========
Arun Kalavath posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 09:36
Subject: Totally different problem with help system
Hi,
What I am trying to do is this - create one central help file, let composer
create one dummy help file per loadmodule and then put a macro jump in these
help files (jumpid macro) to jump to a given help topic in the central help
file.
I have done all these and it pretty much works except the following case -
If I press F1 - help works fine and the relevant topic from the central help
file is displayed
But if I press the help button (which executes the special event - help system),
the central help file is opened with the message that the help id not found.
The .h file of my central help file contains help ids for the help topics in
that file only (not for every control on every windows that jumps to the central
help file).
Any ideas?
Thanks
Arun
========
Christopher Macias posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 09:46
Subject: Re: Robo-Help & COOL:Gen
Kenn,
> What files does Cool:Gen automatically generate (*.hpj, *.hlp, etc.)
From memory, it creates a .h, .hpj, .gml and .hlp file for each load module.
The .h file includes all the help ids for the loadmodule, but also has
the contents of the wrc.h file appended to it when it is created. The
wrc.h file contains pre-defined resource ids and help ids for standard
COOL:Gen window elements. You should check this file and decide which
of these "standard" help ids you will want to create your own help
for.
> and what file(s) need to be replaced with the Robo-Help file(s)?
It depends on how you organise the work. Assuming you work on
the help files outside of the \c subdirectory (highly recommended)
the file exchange can be as simple as "Developer gives the help author
the .h file. Help author returns the developer a .hlp and perhaps a
.cnt file, which replace the equivalent files in the \c subdir or in
the production directory."
- Is there a need to define Help-IDs in Cool:Gen?
If you will be developing help ouside of COOL:Gen (i.e. with RoboHelp)
then yes. Technical considerations aside, you will need a good help
naming scheme to keep your help authors from going nuts.
- How are the executables then linked with the correct help file?
A load modue (executable) named ABC123.exe will need a .hlp file
named ABC123.hlp. The executable 'calls' the help engine with the
help id you defined, and the help engine then looks for that topic in
ABC123.hlp and displays it.
-------
Note that, since you are using RoboHelp, you may have to worry
about the Map IDs (the numbers assigned next to each help id
in the .h file). It is these map ids (or 'context ids') that can change
when you add or alter things on the window, even though the help
id you assigned does not change. We use an unorthodox approach
here that means we don't use the map ids generated by COOL:Gen,
so I'm not sure how difficult these are to manage in 'typical'
organisations.
By the way, although George is right and COOL:Gen help is quick,
it is also very basic. If you are doing a quick and dirty help
system with no graphics, index, or other embellishments, then
RoboHelp is definitely overkill. If you *do* want to include those
features, you should use RoboHelp or something similar. It will
save your help authors hours and hours of work. It will also
allow you to use non-programmers as your help authors with minimal
training.
Regards,
Christopher Macias
========
Preben Nilsson posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 12:12
Subject: Re: Totally different problem with help system
From Microsoft Help Workshop (part of C++ package) you can tell it to view WinHelp
Messages. This actually debugs all messages going through the WinHelp engine.
Regards/Preben
========
Dave Pember posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 12:18
Subject: Referential Integrity
Can anyone relay any 'real-life' experiences on RI, particulalry performance
and management issues? I am considering which option is best, i.e. to enforce
on DBMS or Implement via COOL:Gen. The main factors I am looking at are that
DBMS Constraints increase the work on the DB server and COOL:Gen RI Triggers
will increase my Oracle SQL*Net traffic. Are there any other factors that forced
your decision one way or another?
Thanks,
Dave
========
Andy Baker posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 12:58
Subject: Re: Robo-Help & COOL:Gen
Does anybody know of a why to combine help files for several load modules into
the same .hlp file?
========
Jean-Franois THILS posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 13:06
Subject: Re: Referential Integrity
We choose to implement RI on the DBMS for the following reasons :
First, the DB server are optimized to handle Constraints and Cool:Gen only has
the same mechanism for all the plateforms (is it optimized ?).
Second, with our choice, the physical DB schema reflects the data model of Cool:Gen.
Third, we can use the database with non-Cool:Gen application and be sure that
such an application will not corrupt the coherence of the DB.
Fourth, on Informix, the RI triggers seems to be slower than Constraints.
Jean-Franois
Luxembourg.
========
utpal patel posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 14:29
Subject: Re: Cool:Gen Unix Applications- Installation on production Environment
Hi Ana,
I have bigger problem than you have. I have to work on production testing, i
worked in cellular phone testing area but I didn't work much on application/production
testing of software in UNIX environment.
I will appriciate, if you can give me idea - how to get started? what were your
experience and problems, when you started?
thanks and regards
utpal
========
Christopher Macias posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 15:06
Subject: Re: Robo-Help & COOL:Gen
Andy,
Briefly, you need to edit the .RC file created
for each load module. Towards the top of this
file is a line beginning:
RCTL_HELP, "loadmodulename.HLP"
Edit this line and change the load module name
to the name of the .HLP file you want the .EXE
to use.
The complication, of course, is that the .RC
file can be overwritten when you regenerate the
load module. You will probably want to make
this edit occur at the time the application is
generated, which may involve calling an editing
program from the generation scripts.
I hope that was of some help.
Regards,
Christopher
========
Marie-Agnes Pilon posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 15:28
Subject: Re: Dll sizes and excutable sizes/ Geoff Lee
Geoff lee, you send me a email however I didn't get a returning address.
Small dll is around 37kb
Medium size dll is between 100 to 312kb
Lagest dll I've got is 1,352kb
Sorry about the delay of the response
========
g posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 17:17
Subject: Re: Multi-use report needed
cc,
this is not pretty,, but
I THINK :
that the microsoft iis cd has the ftp service which you can install
allowing anonymous logins,, etc.. then you could ftp or rpc in an external
hope something better comes along..
========
Andrea Motta posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 19:19
Subject: Composer4/Office97
Hi,
I have an application (Composer 4) that generate a report to the Excel (version
7.0 - Office 95), it works, but with the Execel from Office 97 it is not working.
How can I do to generate to Excel of Office97?
Thanks,
Andrea Motta.
========
Mike posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 20:05
Subject: Re: any reason to upgrade to cool:gen 5 ?
Your right George. The Knowledgeware experience should have helped them
develope a better customer focus. Sterling has stated they have a 10-year
plan for Cool:gen, their start product, but at around $30K per seat how
many more can they hope to sell? They have sadly lost too much mind share
in corporate circles (US), and users are looking elsewhere.
========
jg posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 21:34
Subject: Re: Multi-use report needed
'cc' -
Adding to what 'g' said, supposing that your reporting software is the Crystal
Reports ActiveX control, I believe that there is a higher-performance server
reporting piece that Seagate Software sells (no, not by the seashore, 'g')that
could generate the report on AIX or at least on an NT box within the firewall,
to then be ftp'd across the filewall as an .rpt file containing the just generated
data, to a prearranged location. The part I just described would be spawned
(and asynchronous). The location where the .rpt file should be placed is then
polled. Once there, the report is run.
ps - xyz pdq
========
Sean Cohan posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 21:49
Subject: Password Security
Anyone have any slick algorithms for encrypting a password
before storing it in the database? I have an 8 character
password entry field on my window that I think I will store
as an integer of length 8.
Thanks.
========
Bogusslav posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 21:49
Subject: Re: Dynamic SQL
That's not boguslaw, which sounds something for picnic. BOGUSSLAV!!
========
JoJo Calora posted the following article on the ISSUG Bulletin Board
Dated : 02 June, 1998 at 23:17
Subject: Re: Composer4/Office97
Composer or CoolGen statements coded for Office 95 will not work for Office
97 and vice versa. This is because there are differences to the programmatic
interface for all Office products between Office 95 release and Office 97.
You can make your application work with Excel 97 by making the necessary coding
changes. When making the coding changes, make sure that you are using the correct
Type Library (example: Microsoft Excel 8.0 Object Library (1.2)) so you have
the right methods(functions) available in Excel 97.
========
Edorta Simn posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 10:51
Subject: Sheridan's DateCombo Control
Has anybody use the DateCombo control from Sheridan?
I've tried to use it (COOL:Gen) and everything seems to be right until I try
to test or execute the application. When I try to interact with the control,
it ends in a Win'95 exception. I have simplified my code to the point I'm sure
it's a matter of the control itself.
Does anybody know where I can get another version of this OCX or another one
with the same funcionality?
Thank you very much,
Edorta Simn. Madrid, Spain
========
Sean Cohan posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 12:34
Subject: Re: Password Security
Or I may just store it as characters, whichever works better.
========
Nigel Hall posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 13:35
Subject: Re: Totally different problem with help system
We have got one help file for our application and found the same problem with
the Help Button in Cool:Gen. We have reported it to Sterling but they have not
provided a solution. We amened our program that edits the .RC file to replace
the help id with the one for the window.
Nigel
========
Nigel Hall posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 13:45
Subject: Re: Robo-Help & COOL:Gen
We use RoboHelp and have one help file for our application instead of one per
load module.
We put help context strings on all Windows and some Controls in Cool:Gen.
From RoboHelp we generate .H files for the Help File.
In the build script INTDBCLM.SCR we append the RoboHelp .H file to the load
module .H file. This redefines the help id numbers used to access the help file.
In the script we run a program to edit the .RC file. It replaces the load module
help file name with the name of our help file. And defaults the context id for
all controls with a CoolGen default id to the one for the window.
Nigel
========
Ganesh Nagasamy posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 16:44
Subject: Extra first character in EAB
I have an EAB which accepts a data field of length 128(concatenated commarea)
and the corresponding COBOL code calls an assembler and passes this commarea.
While testing with CEDF, The commarea always has a first blank character which
does not match with the commarea on the other end. Is this because of the missing
flag option in the EAB??
========
Ganesh Nagasamy posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 16:52
Subject: Information on the CE
I need technical information on the CE for MVS/DB2 and UNIX/Oracle. Does anybody
out there know where I can find some details like the architecture of the CE,
the components, how I can directly access the CE using SQL queries, How are
the objects stored etc etc....
Thanks
========
CCarter posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 16:53
Subject: One app with server procedures on 2 different platforms
I have a web based application that uses ASP located on an NT box. I would like
to place my report building PrADs and PADs on the same NT box, but place all
other PrADs and PADs on a Unix box.
Do you know if this is possible, or if there are any problems with it?
Also, what if I use a GUI client rather than a web app?
========
Larry Barreca posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 18:35
Subject: Re: Cool:Gen Application: Is Client Manager Running?
Thanks for the information. I did try to use the window function but could not
get it to find the client manager window. I will use external C code.
========
Siraj Konkader posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 20:25
Subject: Word wrap - 3270
Does anybody have any tips on building a 'word wrap' routine for output directed
to a printer? Env. is Composer, CICS, COBOL, DB2.
========
Julie Chapman posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 20:33
Subject: Re: Window look in Cool:gen
Dear Ms. Pilon,
Please allow me to introduce you to Johnston McLamb CASE Solutions, Inc. Johnston
McLamb is a steadily growing, high quality consulting firm with job opportunities
both in the continental U.S. and in Europe. Our corporate headquarters is in
Chantilly, Virginia which is near Washington, D.C. At JMCS, we specialize in
full-life-cycle rapid systems design, development, and deployment. Our reputation
continues to grow as we enter new geographic regions and customer lines of business.
Our track record at delivering tangible results to large Fortune 500 customers
is our testament to breakthrough levels of productivity and quality. Our motto:
develop applications better, faster, and cheaper. We use leading development
tools and define specific approaches to use them effectively. Many of our employees
are principal consultants who are recognized as industry experts in their fields
and come from elite organizations such as big six accounting firms, software
product vendors, and Fortun!
e 500 companies. You can view our home page at www.jmcsinc.com.
We have a unique corporate culture that provides technical consultants with
unlimited earning potential and compelling strategies for project success. We
are currently growing at over 50% per year compounded and have been growing
at that pace since our founding in 1991. This creates a dynamic company environment
where your skills can be applied to their fullest. If your interests include
technical leadership, project management, customer relationship management,
or technical staff mentorship, there are many opportunities through growth at
JMCS for you to leverage and refine those skills.
Our preference is to hire qualified individuals as full-time JMCS employees,
however, we would consider hiring the right individuals as independent subcontractors.
Independent subcontractors must provide documentation of being a legal corporation.
For full-time employees, JMCS offers a competitive benefits package and a unique
opportunity to achieve high levels of compensation through our innovative bonus
program. JMCS believes in every employee participating in the overall management
and direction of the company. Annually, JMCS conducts a two-day Planning Session
that every employee is invited to attend at company expense.
JMCS is a unique and exciting place to work. Our clients and projects afford
our employees the opportunity to use their advanced skills on real business
challenges and to develop new skills. JMCS compensation plan, flexible scheduling,
and employee participation creates a superior environment where motivated people
can achieve their professional and personal goals. If this style of consulting
appeals to you, please call us to arrange an interview.
The company contact information is:
Johnston McLamb CASE Solutions, Inc.
4443 Brookfield Corporate Drive, Suite 105
Chantilly, VA 20151-4021
Phone: 703-502-0901 Fax: 703-502-0905
Corporate E-mail: corpmail@jmcsinc.com
Home Page: http://www.jmcsinc.com
JMCS is an equal opportunity employer with an affirmative action plan. If you
wish to speak to our Human Resource department directly, please contact Dorothy
Burnett or Julie Chapman at 703-502-0901.
Were looking forward to hearing from you soon.
Sincerely,
Julie A. Chapman
Technical Recruiter
Johnston McLamb
========
Sean Cohan posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 20:46
Subject: Enter Key
Is there any way to trap when the enter key was pressed
on an entry field in a primary window? A Keypress event
doesn't recognize the enter key. I've also tried a hidden
pushbutton as the default pushbutton, but it doesn't work
immediately after returning from a flow. Any suggestions?
Thanks.
========
gsimpson posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 20:53
Subject: Re: One app with server procedures on 2 different platforms
Howdy do there neighbor ?
Part 1 I do not know, maybe that rascal jg will!
Part 2 I do know. Use the directory services exit in client manager.
The sample code supplied by Sterling is overly complicated. We just
use a naming standard, and redirect based on the first two characters
of the program.
hth,
george
========
Puggy posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 22:11
Subject: Composer Key
I don't understand one thing.How does Cool:Gen 4.1 a work without key here in
USA? B'cos when i worked in my home country(India) we used the key(sentinel
scribe).
Thanks .Bye
========
Loren Froomin posted the following article on the ISSUG Bulletin Board
Dated : 03 June, 1998 at 22:41
Subject: CBUF Error
What is a CBUF (Common Format Buffer) Error 632 or which document can be referenced
to locate the error text?
Thanks
========
X posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 02:55
Subject: Re: Extra first character in EAB
Yes
========
Derrick Ackermann posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 07:26
Subject: Re: Information on the CE
Sterling do provide an excellent reference containing most of what you're after.
It comes as a set of 4 Windows help files, and is apparently available on the
CG 4.01a installation CD.
The files are called csref.hlp, heref.hlp, Odrpta.hlp, Odrptb.hlp.
We are still on C4, which does not provide them, but I got hold of them elsewhere.
If you are stuck, I can mail them to you.
Regards
Derrick
========
Doug Michael posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 08:20
Subject: Re: Sheridan's DateCombo Control
We had some problems with this too. You can download an update from http://www.shersoft.com
- it worked for us
Doug
========
Doug Michael posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 08:26
Subject: Re: Composer4/Office97
... or use ExcelFun (see thread below) and avoid these problems completely for
current and future Office versions. Then you don't need to worry about what
mix of versions your Users have - the component interrogates the system and
does the right things for the version it finds.
Doug Michael
Rainier Software
========
Andy Baker posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 09:02
Subject: Re: One app with server procedures on 2 different platforms
It looks as if you need the ODC add-on to Composer. This allows you to use either
MQSeries or DCE to connect to servers, rather than Client Manager.
Using DCE allows your client application to access server load modules on any
number of platforms.
========
Andy Baker posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 09:05
Subject: Working Directory
Does any body know who to pick up the current working directory within Composer
4 (NT/Windows 95)?
Thanks,
Andy.
========
Carsten Olsen posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 09:32
Subject: Re: Referential Integrity
DBA maintain the datamodel and do not feel responsible for programs. RI-triggers
are programs - not DBAs problem..
Developers code procedures and PADS. They did not code and do not feel responsible
for RI-triggers which are auto-generated from the datamodel.
It is hard to make anyone take responsibility for these trigger-programs.
The DB2-system knows all DB2-tables that it holds and can handle the RI from
top to buttom.
Usually you will have several models each containing a part of the entire datamodel.
Some entities/tables will exist in more than one model.
From which model should the trigger be generated ?
========
Andy Baker posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 09:45
Subject: Re: Robo-Help & COOL:Gen
Thanks Christopher,
But how can I stop Composer allocating the same Help ID numbers (in the .h files)
for each load module?
Andy.
========
Patrick Hahn posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 10:05
Subject: Re: Enter Key
The keys ENTER, TAB, stop working after return from a flow.
This is solved with one PTF. I applied RTN4115.
========
Ganesh Nagasamy posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 10:46
Subject: Re: Information on the CE
Derrick, Currently we're working with OS/2 IEF5.2 and MVS/DB2/Composer 3 CE
so I do not have any documentation here. I'm also interested in the UNIX/Oracle
CE option and since we do not have this configuration currently, I'm out of
information. If it would be possible for you to mail me the info, I'll be much
obliged.
Ganesh.
========
Martin Schneidewind posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 10:59
Subject: RI-Cascade-Performance
In our DB2-Data-model we have several Tables with DBMS-RI. Some of them have
a hirarchy
of 3-4 Cascade deletion in Sequence. The problem is now,that we can't proceed
a Delete of such constellations in a "normal" time with DBMS-Enforced
RI.
So I wonder how it would function when I change the RI-Control from DBMS to
"Enforced by COOL:gen" ?
Is someone expierienced or able to say how this would change the performance
??
========
Jesus Marina posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 11:44
Subject: DROP DOWN LIST
An easy questions for yours (not for a COOL:Gen new user as I)
How can I use a DROP DOWN LIST "dinamic"? (for example, permitted
values list filled by a SQL statement).
I need fill the drop down values list with content of one table (for example,
table of telephone numbers)
I think in use a new OLE object, but which?.
thanks.
Jesus Marina. Madrid. Spain
========
Niall McMullan posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 12:18
Subject: Re: DROP DOWN LIST
Jesus,
You do not need to use an OLE object. You may set the properties of a list-box
to be a drop-down list. You are limited to one data field and the selection
character in the group view. If you need to display more than one field, you
could concat the fields into a presentation field, or then use an OLE control.
The group view associated with the list-box should be populated from a READ
EACH
statement triggered by a command or event.
========
Richard van der Meer posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 12:32
Subject: High Performance View Matching
Hello Everybody,
Does anyone know what the advantages or disadvantages are when using
the option High Performance View Passing.
I generated some MVS-COBOL-code in both situations (on and off) and found
the only difference in the LINKAGE SECTION where every import and export view
has become a 01-level instead of one 01-level per import and export area.
Kind Regard
Richard van der Meer
========
Chris Uttley posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 14:18
Subject: Re: High Performance View Matching
Here is my understanding of HPVP. I welcome any further input on this from everyone.
The preferred setting for HPVP is ON. However, read on...........
HPVP boosts the performance of USE statements. A USE statement is essentially
a call to a subroutine. When you have data to pass to a subroutine (and/or receive
back), there are basically two ways to do this.
One is to have all the data sitting some place in storage and you tell the subroutine
where that data is. This is called 'pass by reference'. The calling action block
simply passes a pointer to the called action block telling it where the import
views are. (Because only a pointer is passed, the structure of the views between
the calling and called action block must be absolutely identical or else pass
by reference cannot occur). HPVP is the same as 'pass by reference'.
The other is to have the subroutine have its own place to store its import views
and when you call it, the code has to move all the views from the calling action
blocks place to the called action blocks place. Same thing for export views.
So the first case is much faster, because views don't have to be moved around.
The performance benefits to an online transaction are probably not noticeable,
but to a batch program it could be very significant.
If you were a new COOL customer and had nothing 'historical' to link in or USE,
you would just let COOL default the HPVP to on. When you create an action block,
the default is HPVP on.
Note that even if you have HPVP on for an action block, COOLGEN may decide to
not use HPVP "IF THE STRUCTURE OF THE VIEWS IS NOT AN EXACT MATCH."
I am unaware if there is a CBD 96 standard for this.
========
Christopher Macias posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 14:36
Subject: Re: Robo-Help & COOL:Gen
Andy,
Your best bet is probably to set the map ids in RoboHelp and use these, instead
of using the map ids generated by Composer. Check out Nigel Hall's message in
this thread for a way to do this. It's roughly the same as what we do.
Christopher
========
CCarter posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 15:03
Subject: Re: One app with server procedures on 2 different platforms
Would COM proxy accomplish this task?
========
Simon Connock posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 15:18
Subject: Re: One app with server procedures on 2 different platforms
Yes it should do. You effectively have one COM proxy for each transaction. The
routing for that transaction (i.e. how it finds the server) is controlled either
statically from the proxycfg.ini file or dynamically from your application (using
the ComCfg property - works from VB or ASP or whatever). Either way there's
nothing to stop you routing one proxy to a remote server and another to the
local machine.
If you using a COOL:Gen GUI client then you can use the COM proxy if you want
but probably easier to use Client Manager with Directory Services as George
suggested.
Hope this helps...
Simon
========
Karl Erik Bergaust posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 15:29
Subject: Oracle 8 with cool:gen NT DPS
Hello,
Does anyone know how to access Oracle 8 with DPS servers.
Geting the message that 'sqllib18.dll cannot be found' when running on a Oracle
8 machine.
Is it possible to relink DPS runtimes (mkexits.bat) with oracle 8 libraries?
Which library should be used?
Regards
Karl Erik Bergaust
========
george simpson posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 15:39
Subject: Re: Oracle 8 with cool:gen NT DPS
Karl Erik,
We are in production with O 8. sqllib18.dll will be found on the
O 7.3.3 CD. It is just the first Oracle DLL you're missing.
When we linked with the O 8.0.3 libraries in Unix we found
that an Oracle connect didn't reuse the connection that was open,
it opened a new connection, so that by the end of the day you had
tons of oracle connections, which are expensive. I'd imagine that
would eventually crash the box. I don't know if that is a bug or
if it is a new feature. Everything else worked great.
Hope all is well, and you're having fun,
-george
========
karl erik bergaust posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 16:06
Subject: Re: Oracle 8 with cool:gen NT DPS
George,
How did you proceed when you relinked for O 8?
(and did you try it on NT?)
Do you know the library name which should be used instead of sqllib18.lib?
We have this test environment:
- Win NT
- Personal Oracle 8
- SQL*Net for O 8.
Is it neccassary to install any O 7 software to make this work?
having fun.....
Karl Erik
========
george simpson posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 16:10
Subject: Re: Window look in Cool:gen
Marie-Agnes,
I wish I'd have known about the environment variable. It is Microsoft who
changed, not Sterling. Here's what got us on this one. When they went to
the 3d look they needed to more real estate on the Window. They couldn't
steal it from the outside, so they stole it from the inside. The default
in Window Designer's entry field autoscroll property is off. We had many
very tight fields that we could no longer type in the last digit. To solve
this we wrote some sql to go against the encyclopedia, and update all of
these properties true.
hth,
george
========
CCarter posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 18:55
Subject: Thanks, you've been a big help!
Thanks, you've been a big help!
========
george posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 19:46
Subject: Re: Oracle 8 with cool:gen NT DPS
Karl Erik,
We have both NT and UNIX. The NT is for sqlnet development only though.
We run using the 7.3.x precompiler and libs,, COOL:gen doesn't take
any advantage of 8 and told us they couldn't support us there, and we have
other projects back in 7.3 who would need to learn a few small rules so we
just put the data in 8, and left the code back in the old runtime. We did some
very extensive testing in UNIX with 8 and only found the one gotcha I mentioned
previously. It seems like that should be easy to get around, so no I don't
think you need 7 especially since they will support 8 in the fall. I don't
know why they are tieing support to a release. sqllib18.dll is a 7.3 release
dll. Did you change the precompiler in your build script ?
If you go our route, 7.3 in 8.0 here is the only thing that got us:
To run a batch job, or start the daemon you must point ORACLE_HOME to
/oracle/product/8.x.x The libpath still points back to 7.3
Again we're in 8.0.3. We had hoped to be in 8.0.4 because it fixes a ton
of bugs your DBA will be very interested in, though your coder's won't care.
On another note, we just got a DB monitor from Savant called Q www.savant-corp.com.
It is a 2 year old company. Their 1.1 product supports 8, it is cheap,
5000 $(us) per each 8 processor box and all development environments are FREE
!
Must say I am impressed with what they got already, online explains for the
developers, and a ton of information at your finger tips for your dba,,
I would like to make it my screen wallpaper, and wish they sold stock.
Take care,, mail me or post if I can help or don't understand,
George
========
Jim Davis posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 19:56
Subject: Delete Rules Report
Is anyone aware of a way to create a report of delete rules
by entity type for a given model?
========
Siraj Konkader posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 19:59
Subject: Re: Delete Rules Report
I would think that you could do this using the public interface.
========
Darius Panahy posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 21:44
Subject: Re: Oracle 8 with cool:gen NT DPS
sqllib18 is the library that is referenced for code that is pre-compiled with
Pro*C 2.2 I think, so you could need to use a different version of the pre-compiler
if you did not have sqllib18.
However have you tried copying sqllib18.dll from the Oracle 7.* ORANT\PRO22\LIB
directory. It should be upward compatible and work with Ora 8, but this is just
a guess based on sqllib17 working with 7.3.
Darius
========
Darius Panahy posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 21:48
Subject: Re: High Performance View Matching
Chris is correct, but beware...
We have found a number of problems with HPVP relating to corrupted data in the
views. They are not easily predictable, but appear to be related to group views
that contain a mixture of text and number. I never spent much time investigating
since switching HPVP off immediately cured the problem and I moved on to something
else! (This was with C3, may be fixed in CG 4.1)
You may want to do some LinkFAQ searches on HPVP.
Darius
========
Chris Uttley posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 21:58
Subject: Re: High Performance View Matching
Yes, there were problems with HPVP in C3, which forced us to turn it OFF everywhere.
However, we just built a large component using COOL:GEN 4.1 and all appears
well when HPVP is ON.
I would like to hear from others though to see if they consider the C3 HPVP
issue dead or not, now that COOL:GEN is here.
========
Chris Uttley posted the following article on the ISSUG Bulletin Board
Dated : 04 June, 1998 at 22:03
Subject: Re: Delete Rules Report
On the workstation, you can just run the entity definiiton report and ask it
to report on RELATIONSHIPS. The report then shows the delete rules for each
relationship of the entity. This is in COOL:GEN and C3 I think.
If you are looking for a way on the CSE or CE, not sure how to do that.
========
Y posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 00:07
Subject: Re: Extra first character in EAB
Thats a very detailed answer !!!!!!!!!!!! Thanks
========
rod maxwell posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 00:25
Subject: Re: Delete Rules Report
can probably help you out... got a user unknown on your email link...
please send me an email (click on my name) and we'll get something to you
========
Glenn Smyth posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 04:28
Subject: Re: High Performance View Matching
Your description of how HPVM works is not quite correct. Composer/cool:gen ALWAYS
passes by reference. This cannot be done using the structures (01 levels) from
the calling program (action block) if the view structures (order and contents)
are not identical. So to get around this if the views are not identical Composer
generates an additional structure (01 level) that does match the called action
block and then generates move statements that move the contents to (import)
or
from (export) these addtional structures from/to the structures in the calling
action blocks. So the impact is not just on performance but rather also on
memory usage. This is particularly insidious in the case of large group views.
As one example say an import group view had a cardinality of 100 with 10
attributes and the calling procedure had only had 9 of these attributes this
would result in the calling action block code containing 2 01 levels occurs
100, one
with all 10 attributes, one with 9. It would then move all the contents of one
to the other before the call = 9000 extra move statements executing at run-time
!, plus
all the extra storage required. If you redundantly added the extra attribute
to
the calling action block this would all be removed from the code. The best example
of this I know is at the UPS a number of years ago (actually before HPVM was
added
to Composer, see below) they reduced the execution time of a batch job - where
HPVM has its
biggest impact as it is typically with a loop (the impact of removing 9000 within
a loop processing a million records cannot be overstated) - from 28 hours to
20 minutes. This is an extreme example but shows the impact of this. In an online
transaction the idividual performance gain is not necessarily that great but
the
overall reduction in CPU and memory requirements for the system can be very
significant
(as we found at DEETYA).
A few more points :
1) All the above actually applied before HPVM was introduced except as all import
views were combined into a single structure (01 level) in order to achieve
perfect view matching (as it was called in the old days) every view had to be
identical
in contents and order. HPVM by spliting the views into separate structures meant
that the decision as to whether or not the structure from the calling action
block
could be used was moved down to the inividual view level and so you could have
some views
perfectly view matched and others not and still gain benefits.
2) The view structures in a PRAD are different from those in an action block
as
the PRAD contains and additoinal 6 byte field for all fields in import and export
views
and so perfect view matching has no effect and all the additional code described
above is generated UNLESS local views are used. This can lead to some interesting
discussions (for another day) on coding standards, particularly for batch.
3) All problems with HPVM are fixed in all Cool:gen versions. It was an enforced
standard at the last two sites I was at its impact of performance/memory use
was
so great. We had no problems at all (in fact with a few PTFs applied there was
only one outstanding problem with Composer 3 which could be avoided - I can't
remember
what it was though).
I have a detailed paper on this I wrote I can send to you if you e-mail me quickly
at this temporary address (soon to be moving to an as yet undetermined new site).
Glenn
PS George, I share your posting addiction/affliction and found a good holiday
on
an island without computers the only rememdy. However, not unlike the Betty
Ford
clinic it is so easy to relapse once you return, a case in point this 'binge'
long
posting. Boy, it feels good though :-)
========
Glenn Smyth posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 04:35
Subject: Re: RI-Cascade-Performance
Sorry, to say the performance would with 99% probability be worse. A couple
of sites I have been at
have done benchmark tests on this and found DBMS RI (depending on the nature
of
the rule) to be 2-6 times faster (this is in DB2 I hasten to point out). I
suspect that your case would be nearer to the later as the main impact is in
the
increased level of calls between the application and the DBMS. With DBMS cascade
rules all the work is done within the DB engine.
Glenn
========
Glenn Smyth posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 04:39
Subject: Re: CBUF Error
A mismatch between the views passed between a client and a server at run-time.
99% of the time caused by changing the views and not re-generating one or the
other.
At a technical level it is caused by a mismatch in the data types when the server
manager attempts to 'unpack' the data sent and place it in the target views,
e.g.
it found character data where it expected to find numeric data.
The TI books is the only place I can give you to find more info on this.
Glenn
========
Glenn Smyth posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 04:42
Subject: Re: Composer Key
I won't reveal the name of the DLL as it would enable the unscrupulous (never
could spell that one) to bypass the security but the US and South-east asian
versions are delivered with a different DLL, one that essentially does nothing
and one that looks for the sentinel (or Dongle as they are called in Oz).
TIs rationale for this was simply that software piracy is not as much of an
issue
in the US as in the Asian markets.
Glenn
========
Jesus Marina posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 06:51
Subject: Re: DROP DOWN LIST
Thanks Niall:
Its very easy.I began to think that it was imposible for COOL:Gen. (remember,
Im a begginner COOL:Gen user).
What happened? I didnt define the group view, which explains I couldnt select
in List Type properties window the option . "Enterable Drop Down List".
I looked for (and looked for ...) and I didnt never find it.
Thanks.
========
Jesus Marina posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 07:25
Subject: GUI statement "SORT"
I had troubles with the GUI statement "SORT"...
I used the next group view:
Group View export_g_fh (optional, 50, explicit, import only)
Work View export_i ief_supplied (optional, transient)
select char (optional)
Entity View export_i fh (optional, transient)
code
description
begin_time (field used to SORT the list, ascending order)
end_time
cost
Entity fh
*code (TEXT, 3, Mandatory, Basic)
description (TEXT, 15, Mandatory, Basic)
begin_time (Time, 6, Mandatory, Basic)
end_time (Time, 6, Mandatory, Basic)
cost (Number, 7 (4.3), Mandatory, Basic)
Sometimes COST One or More CALL
*Always IS_DEFINED_BY One Operator
Sequence:
1.- Group view is initially sorted by SQL statement (ORACLE) when it read from
te database (READ EACH .. SORTED BY ASCENDING fh begin_time...) -> Correct
Order
2.- Add a new row (item) to the list (at last)
3.- Sort the list with GUI statement "SORT" (SORT export_group_fh
BY "4 a")
ERROR What happens?
The first, second, third and fourth columns appear in correct oreder (in screen),
but the last column appear in the OLD (bad) order.
In bad order, what means that? Simply, this column dont sort.
An example:
1 JOHN 08:00 10:00 1.57
2 PAUL 12:00 18:00 12.34
add row
4 TOMAS 06:00 15:00 2.33
(third and fourth column Edit Pattern: HH:MM)
Before GUI statement SORT:
1 JOHN 08:00 10:00 1.57
2 PAUL 12:00 18:00 12.34
4 TOMAS 06:00 15:00 2.33
After GUI statement SORT: (look fifth column)
4 TOMAS 06:00 15:00 1.57
1 JOHN 08:00 10:00 12.34
2 PAUL 12:00 18:00 2.33
(note correct order, look fifth column)
4 TOMAS 06:00 15:00 2.33
1 JOHN 08:00 10:00 1.57
2 PAUL 12:00 18:00 12.34
What is the problem?. I need the list sorted by this field (begin_time)
Thanks.
========
Ganesh Nagasamy posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 11:54
Subject: Re: Extra first character in EAB
Hi Y,
Refer to the replies to the message posted by Richard Van de Meer on High Performance
View Matching. This is exactly to do with that although it does tend to be a
disadvantage when you concatenate the commarea without spaces between the attributes
of your commarea workset. Hope the information is useful to you.
Ganesh
========
Hannu posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 12:17
Subject: Re: Composer4/Office97
Even, if having other than english version of Office?
Word6 and Word7?
========
Hannu posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 12:31
Subject: Re: any reason to upgrade to cool:gen 5 ?
I've seen some foils as well and unfortunately to me it seems,
the development is being frozen.
The things, I've waited has been moved under section 'Looking to the Future',
which has earlier meant 'forgotten issues'.
They were CBD enhanced toolset and model support for components,
ency usability/performance (incremental subsetting), Web client user interface
design to name a few! 'CE98' is not even listed!
========
Jayson deVries - MCM Technology Inc. posted the following article on the ISSUG
Bulletin Board
Dated : 05 June, 1998 at 12:57
Subject: Re: Password Security
Sean,
I can't give you the exact algorithm that we use here at MCM Technology but
I can give you some idea on what we found is needed/important when encrypting
a password:
1) You don't necessarily need to decrypt - just compare the encrypted password
which was entered with the string in the database. Your algorithm only needs
to go one way - however the password-to-encrypted number may not be one-to-one
i.e. more than one password may give the same encrypted number, i.e. more than
one password will work . . not a good thing - try to minimize this.
2) Use a number of keys for the encryption - something that changes (like dates)
so if the username is JOHNS and the password is PASS, manipulate the two strings
but also include something like the date the password was set so that the next
time JOHNS sets his password to be PASS, the encrypted number is something different.
3) Make sure that the order of the characters changes the encrypted number i.e.
MARYG is different from GARYM.
4) If the end product is a number, you want a good distribution of numbers.
5) Use an EAB - more flexibility - easier to hide the code.
I hope this helps,
Jayson deVries
MCM Technology Inc.
Fredericton, New Brunswick, Canada
========
Jayson deVries - MCM Technology Inc. posted the following article on the ISSUG
Bulletin Board
Dated : 05 June, 1998 at 13:27
Subject: Intercepting the F1 key
I have a C3 model running in Windows95 and I need to be able to intercept the
F1 key to fire off my own help rather than going to the default Windows help.
I've tried to follow it through the generated C code and as far as I can tell,
there is a VK_F1, RCMD_HELP, VIRTKEY somewhere in WRG530N.dll or WRC530N.dll,
but I don't have the source so I can't be sure. Is the source for these available?
If not, can I somehow intercept the F1 key and fire my own event?
Thanks,
Jayson deVries
MCM Technology Inc.
Fredericton, New Brunswick, Canada
========
Glenn Smyth posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 14:12
Subject: Re: starting IMS conversational trx from cool:gen
From where do you need to start it ? From another Cool:gen generated IMS
Transaction ? From a Cool:gen GUI client (via the Comms-bridge or using a third
party tool e.g. MQSeries or DECMessageQ) ?
If you post (or e-mail me) more details I should be able to help.
Glenn
========
Doug Michael posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 17:05
Subject: Re: Composer4/Office97
Yep. Word 6, 95 or 97 and Excel 5, 95 or 97. Any language. The idea is that
should future versions have similar compatibility problems you would not have
to revisit your COOL:Gen apps, rather deploy a new version of the component
DLL which supports those new versions as well as all the older ones.
-Doug
========
Janet Evans posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 17:21
Subject: Printing
Has anyone out there managed to find an acceptable solution for printing things
like the ERD with Cool:Gen. Now I know that printing was never the easiest thing
to do with IEF/COMPOSER/Cool:Gen..... but it seems to be even more difficult
these days. We have to expand each of the entities if we don't want ugly black
boxes that don't allow us to see the names! And then the fonts get screwed.
etc. etc. Has anyone come up with a good way to print/plot?
========
george posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 17:24
Subject: re: Dodgy Funnel,, don't the hp customers want it fixed ???
Come on guys,
don't you want the funnel fixed ?
from what i hear you guys got the same bugs
========
karl erik bergaust posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 18:21
Subject: Re: Printing
We are plotting through HPGL/2 (HP) plotter driver.
We don't have a plotter ourselves, but print the diagrams to a file (you can
set up the HPGL/2 plotter to do that in NT).
Then walk over to the nearest printshop and plot a A0 datamodel.
(Black & White copies are very cheap to plot: in Norway about 60,- = almost
$10).
Regards
Karl Erik
========
Darius Panahy posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 21:59
Subject: Re: CBUF Error
I have found in Composer 3 that a numeric attribute of length 3 will cause a
CBUF. Changing the length to 4 cured it.
Darius
========
Chris Uttley posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 22:07
Subject: Re: Robo-Help & COOL:Gen
Heres some info I posted in a previous HELP thread. Perhaps it will trigger
some ideas. We use this technique with ROBO HELP.
What is CSFHelpFix?
CSFHelpFix is a Windows 95/Windows NT 32 Bit application that is used in conjunction
with Composer C generated code and Win Help Office to create consolidated help
files.
Prior to the development of CSFHelpFix each load module for a composer system
required a separate Windows help file. This limitation was cumbersome because
any overlapping or common help between load modules required a "cut and
paste" approach to developing the individual help files. This duplication
among multiple help files increased the chances for missed updates and inconsistencies.
The duplication also increased the footprint of the help files shipped with
a product, and did not allow end users to navigate using the "back/forward"
buttons in help since each request for help was likely to load a new help file
into the WinHelp sub-system.
How does WinHelp work, and what are Resource ID's?
The WinHelp sub-system is called each time the user requests help from any windows
based application. This system is passed two items with each request for help:
the name of the help file, and the help resource ID number for context sensitive
help on a field/control/dialog. These two pieces of information are stored in
the .Exe for the application. WinHelp takes these two pieces of information
and attempts to load the given help file and then search that file for the given
help resource ID. Prior to compilation these two pieces of information are stored
in two separate source file locations. The name of the help file is stored in
the .RC or Resource Definitions file under the RCTRL_HELP section. The resource
ID's name and value for each help topic are stored in the .H or header map file
for a load module.
What prevents us from using standard Composer load module source files pointing
to one central help file?
Composer generates the values for each help resource ID (and non-help resource
ID) for a given load module. These ID's are stored in the load module's C header
file. The system Composer uses for number these ID's is simple: start at 100
for each load module resource ID's value and increment by one for every resource
ID that follows. This system has an obvious flaw when applied to help ID's:
each load module has a set of help ID values which are only unique for that
load module. ID's across multiple load modules are almost guaranteed to have
conflicts with other ID's in other load modules. These conflicts translate to
incorrect links to help topics within a central help file. This limitation prevents
developers from using one central help file across multiple load modules.
What does CSFHelpFix do to allow for multiple load modules using a single help
file?
The CSFHelpFix overcomes the inability of Composer to use a central help file
by directly modifying the generated source code output of the composer system
prior to compiling the code into an executable form. CSFHelpFix modifies two
files for each load module: the .rc resource definitions file is changed so
that it points to a central help file; and the .h resource ID header file is
modified so that help ID values are reassigned to guaranteed globally unique
numbers.
What was CSFHelpFix developed in?
CSFHelpFix was created using Microsoft Visual C++ 5.0 and relied heavily on
the ClassWizard which automatically generated a set of stub functions for application
development. The application is based on the single dialog Application Wizard
model.
========
Glenn Smyth posted the following article on the ISSUG Bulletin Board
Dated : 05 June, 1998 at 23:31
Subject: Re: Attribute state flags in views
Be warned that turning them off immediatly means that HPVM will not be possible.
See separate thread for the performance implications of this.
Glenn
========
Phil Clarke posted the following article on the ISSUG Bulletin Board
Dated : 06 June, 1998 at 08:16
Subject: Re: Working Directory
First, set a character string to the Application Object, property FullName.
This gives the full pathname of the application.
To derive the working directory, find the position of the last backslash ('\')
in the string, then do a substring from 1 to that postion.
This will give you the working directory.
========
George Simpson posted the following article on the ISSUG Bulletin Board
Dated : 06 June, 1998 at 16:34
Subject: Re: Printing
Hey Janet,
You are right, plotting is an aquired art, which is OK with me, since I've
done that (don't tell anyone). It is very helpful if someone on a project does.
Two things I don't like ( I think ).
1- Scoped data model, you have to use the root subject area,
because you can't join outside your area.
2- Scoped data model, didn't follow IDEF 0
notation in that the identifier isn't notated,,
gots to run, regards, - george
========
Jeff posted the following article on the ISSUG Bulletin Board
Dated : 08 June, 1998 at 08:13
Subject: Re: A real simple Q : where do I find BMP's to my buttons
I also have the same question with you, can you give me some help about this
========
Rony Van Hove posted the following article on the ISSUG Bulletin Board
Dated : 08 June, 1998 at 08:28
Subject: Re: Oracle 8 with cool:gen NT DPS
Hi Karl-Erik,
First of all, it is important that starting from Composer 4/COOL:Gen 4.1a, you
can tailor almost any version of your DBMS to our product.
However it won't take advantage of your DBMS system.
For DPS:
Modify the user exit tiroconn.pc and recompile the stuff with mkexits
You should be aware that you also have to modify the build script in order
to include the correct libraries and precompiler stuff.
For NT Clients
Modify the stuboran.mak file and replace it with the correct build libraries.
Rebuild the stuff with: nmake -f stuboran.mak
Secondly, it is important to point out that you can modify the file stuboran.pc
in order to tailor everything to your taste. For example you can include your
own connection logic. This is a major change prior to COOL:Gen/Composer 4 software.
You can also upgrade your precompiler stuff, therefore, you need to tailor the
build scripts (the infamous .scr scripts).
This technique is also possible on Unix platforms.
Files are located in $IEFH/ae/... (don't remember the exact structure) for tiroconn.pc.
There is another place $IEFH/src for other useful user exits (most users aren't
aware of this!). Build make files are included.
Write if you need more information.
BTW, this tailoring technique doesn't limit to ORACLE. You can do the same stuff
with INFORMIX, INGRES etc.
PS: If you want to report problems with a newer version of a DBMS then you should
first test it with the original DBMS version to see whether this problem also
occured previously, otherwise we can't help you very much.
Best regards,
Rony
Member Technical Staff
Application International Division
Sterling Software Benelux (II) NV
========
Rony Van Hove posted the following article on the ISSUG Bulletin Board
Dated : 08 June, 1998 at 08:33
Subject: Re: DROP DOWN LIST
Nial/Maria,
Correction,
Nial suggestion is ok, as long as you use the a dropdown listbox, however if
you use a permitted value dropdown box (thus the values are already defined
in the attribute view etc) then this technique is not applicable.
Lucky for you, there is another way!!!!
The trick exists in using the COOL:Gen Application Library.
You can modify dynamically the number of items in a permitted value list of
a dropdownbox (not dropdown listbox) by issueing the appropriate Type Library
functions.
I have still an old demo app (need to figure out where I have stored that application)
that demonstrates this nice feature.
Best regards,
Rony
Member Technical Staff
Application International Division
Sterling Software Benelux (II) NV
========
Rony Van Hove posted the following article on the ISSUG Bulletin Board
Dated : 08 June, 1998 at 08:38
Subject: Re: Robo-Help & COOL:Gen
Hi Kenn,
I have set of scripts and programs that enabled a seamless integration of Robohelp
stuff with Composer 3 (never tried to tailor it for CG 4.1a).
The set was developed couple years ago.
Best regards,
Rony
Member Technical Staff
Application International Division
Sterling Software Benelux (II) NV
========
Rusty Thomas posted the following article on the ISSUG Bulletin Board
Dated : 08 June, 1998 at 09:02
Subject: Re: GUI statement "SORT"
There was/is a problem with COOL:Gen 4.1a and the GUI sort statement on columns
that are numbers.
Support just came out with a new PTF that might solve your problem. PTF RTN4119
could be of some use.
Try the site http://www.cool2.sterling.com/support/
Good luck!
========
Jesus Marina posted the following article on the ISSUG Bulletin Board
Dated : 08 June, 1998 at 12:09
Subject: Re: GUI statement "SORT"
Rusty Tomas:
Ive had good luck
Ok. The new PTF you said me solve my problem.
I had two issues:
-Issue 10067601: Last week, I solve it using a numeric attribute implemented
in C with decimal precisin.
-Issue 10068985: To solve above Issue I create a new problem. Today, Its solving
with the new PTF.
Regards and thanks.
========
Andy Baker posted the following article on the ISSUG Bulletin Board
Dated : 08 June, 1998 at 12:45
Subject: Re: Robo-Help & COOL:Gen
Thanks Chris,
Is CSFHelpFix freely available, or is it a licenced product?
Andy.
========
Andy Bisby posted the following article on the TISUG Bulletin Board
Dated : 08 June, 1998 at 13:50
Subject: COOL:Gen Books
Does anyone know of any third-party, off-the-shelf books or manuals about COOL:Gen that would be useful to someone relatively new to the product?
Thanks.
Andy Bisby.
========
Rony Van Hove posted the following article on the TISUG Bulletin Board
Dated : 08 June, 1998 at 15:24
Subject: Re: Dll sizes and excutable sizes
Hmmm....
The response from Henrik Meisner is accurate and correct.
Actually, it is maybe interesting to point out that you can put even more stuff in DLL's. However the software doesn't allow you that directly.
For example, you could create seperate DLL's from the procedures...
For example:
A Window Manager DLL can call multiple procedure step DLL's.
A Procedure Step DLL can call one or more Action Block DLL's.
To achieve this setup, you require to modify the INTDBCLM.SCR script file.
Especially, in CBD approach, you can achieve quite 'logic' components with it.
Hmmm... maybe time for CBD-AIK ???? Let's think about it. To be continued.
Best regards,
Rony
Member Technical Staff
Application International Division
Sterling Software Benelux (II) NV
========
Rony Van Hove posted the following article on the TISUG Bulletin Board
Dated : 08 June, 1998 at 15:40
Subject: Re: Attribute state flags in views
Folks,
Apparently some clarification is required... Let's light the torch in the darkness here.
Attribute State Flags (suffix _as).
Previously (pre-IEF 5.31) it was called 'Missing State' Flags (suffix _ms).
These flags are required for the Dialog/Window Manager to indicate whether a certain field was changed etc... This one byte field is generated for every attribute by default.
However this is not necessary for Action Blocks. Procedure steps requires this field (because the ProcStep is the one that is called form the DM/WM).
By turning off the attribute state flag generation, the attribute state flag will no longer generated. The switch was introduced rather than simply remove the entire generation of the attribute state flag: for backward compatibility.
This has couple of major advantages.
Especially for Legacy data (e.g. PL/1, COBOL etc) this can introduce a slight performance gain (long running batches).
For Tandem platforms, this will reduce the number of bytes per object (Tandem has a imposed limitation of 64K per object). It will also generate slightly smaller dll's/executables/load members.
This parameter can affect the High Performance View Passing mechanism. If you synchronise the paremeters then you can benefit additional performance gain in combination with the HPVP stuff.
Hopefully this clarifies little bit this parameter.
Best regards,
rony
Member Technical Staff
Application International Division
Sterling Software Benelux (II) NV
========
Ganesh Nagasamy posted the following article on the TISUG Bulletin Board
Dated : 08 June, 1998 at 16:26
Subject: Re: COOL:Gen Books
Andy, As far as I know, there are no specific books available on Composer/COOL:Gen. There are a lot of General books available on Information Engineering but these I believe are not the one's you're looking for. If anybody knows of any available books, I'd be interested to know.
Ganesh
========
Marie-Agnes Pilon posted the following article on the TISUG Bulletin Board
Dated : 08 June, 1998 at 18:32
Subject: Oracle connections and trapping db errors
I have an application that has a local database (Oracle) and also needs to update a remote database.
When a user tries to update the remote database in real time, my application verifies that the remote database is up and running by connecting to it. Once the connect has taken place I loose my local database connection...
I created a work around where I reconnect to my local and then loose the connection to the remote database. When the application actually updates the remote database (because I successfully connected to it earlier), I then have two connections. One to the remote database and the other to the local database (which the way it should be).
I personnaly do not like the work around. This application worked fine with Composer 3.0, where it would not loose the connection to the local database. Someone here (at work), has mention to me thta there is a way (supposedly) with Cool:gen to trap database error to would remove the need for an external to check if the database is there.
I have been searching all morning and cannot find any direction or commands that would let me perform such a test. Does any one know of the commands to use or example of codes I could look at?
The other question I have is why am I loosing my Oracle connections? Does Cool:gen have anything to do with that phenomona???
Thank you in advance for your help!
========
Glenn Smyth posted the following article on the TISUG Bulletin Board
Dated : 09 June, 1998 at 01:04
Subject: Re: Oracle connections and trapping db errors
You cannot directly trap 'DB not there' (or indeed most DB) errors directly
in action diagram logic. What the other person was probably refering to was the
error (termination) exits where you can trap this error but your choices for what you can
then do about it are limited. One thing you can't do is return to the pint in the
code where the error occurede and continue processing (which is what I suspect you
would like to be able to do). The best you can do is to supress the standard
'abend type' error and display a more meaningful message (in either a pop-up or
on the window).
I would hope that Cool:gen is not responsible for the DB connection being dropped
but if this problem only started with the move from Compsoer to Cool:gen it has
got to be prime suspect.
Glenn
PS There was a long thread previously where I posted details about the client and server
side termination exits but it mostly seems to be gone (curiously only the RE:
telecomunications errors final thank you message to me remains below as a high
level posting ???). Maybe the full thread is in the process of being moved to
the archives. There is also some info. on the exits in the electronic books.
========
Edorta Simn posted the following article on the TISUG Bulletin Board
Dated : 09 June, 1998 at 10:38
Subject: Simple Cool:Gen Installation
This question refers to the installation of the development tool itself, COOL:Gen.
We'd like to create a two-people team, both of them working under Windows'95. Is it possible for one of them to have phisically the Encyclopedia in his machine, or should we have to use Win-NT server for this purpose?.
And, looking at the installation CD, there seems to be (under the Win95 directory) a directory for each type of installation. Does anybody know which of them must I execute to manage the type of team we need?.
As you can see, we are a bit lost. Thank you very much,
Edorta Simn. Madrid, Spain
========
David Mueller posted the following article on the TISUG Bulletin Board
Dated : 09 June, 1998 at 11:09
Subject: Re: Simple Cool:Gen Installation
The Client Server Encyclopedia needs to run on a Win NT 4.0 pc.
Run the setup program in the Win95 directory and enter the product codes given to you by Sterling to enable you install your software.
To install the CSE run the setup program from the WinNT directory.
========
Edorta posted the following article on the TISUG Bulletin Board
Dated : 09 June, 1998 at 11:19
Subject: Re: Simple Cool:Gen Installation
Thank you very much, David
========
Kevin Bingham posted the following article on the TISUG Bulletin Board
Dated : 09 June, 1998 at 12:45
Subject: Cross platform server flow
Hi,
I am looking for specific info on the following scenario:
I have a GUI client (W95) and I flow to (or use) a Unix server PStep, from there,
I need to FLOW to an NT server Pstep, and then return back up the chain. I have
read the links concerning COM proxy, 2servers and 1 client, but think that this
is different to my needs, as there seems to be a return to the client in between
the server calls.
Has anyone done something like this before? More info available on request...
Thanks in advance
Regards
Kevin Bingham
========
Glenn Smyth posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 00:40
Subject: Re: Cross platform server flow
This is not possible in straight Composer/Cool:gen as it is essentially
a cross-platform server to server flow you are looking for. It used to be on
the enhancement request list, don't know if it still is and when it might be
coming. We did this at the Bank Of Ireland by developing an EAB interface to
DECMessageQ and at Allied Irish Banks (though there the servers were on the same
platform, different regions)using MQ_SERIES.
I can give you a solution & code for same platform server-to-server flows. If you are interested
(though it sounds like it won't meet your requirement) send me an e-mail.
A few weeks ago I would have given you the details of the other solutions but after
recent experiences I have had a re-think on my previous policy of giving all my
knowledge/tools away for free. It appears it is more and more important to have
access to inside knowledge that others don't so I am forced (relunctantly) to
protect knowledge that is 'very marketable'.
Glenn
========
Sean Hansen posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 11:25
Subject: Hiding Fields (Win 3.1 - 16 bit)
I've got the trick to hide text fields (white cat in the snow effect) using:
MAKE NORMAL REVERSE VIDEO and filling the field with ALT0160
and can make date fields appear the same:
MAKE DARK REVERSE VIDEO and populate the field with something (e.g. XXX or CURRENT DATE for a DATE field)
but when I try the trick with a numeric field, I end up seeing a few pixels of my entry field (no margin box, but they do have a yellow background)
Does anyone have a recommendation for solving my dilemma?
========
Philippe Meersschaut posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 13:48
Subject: HPVP
Glenn,
At our company (KBC Bank), we have a long track of experience with HPVP.
HPVP covers two things
1) The possibility of having the temporary storage and associated MOVE's
eliminated (as outlined in one of the previeuw messages).
2) passing through the 'globdata' to external action blocks,
(extremely usefull for manipulating exit state structure, userid,
error flags, ...)
We've had lots of problems with HPVP on different platforms (MVS, HP/UX
Tandem and OS/2), most of them were associated with faulty behaviour of
the TRACE-facility. At our knowledge only two problems stay open under
COOL:Gen 4.1a; a) use of HPVP & TRACE on Tandem/Cobol and
b) call of an EAB with HPVP=ON & Missing-flags=OFF on OS/2
(already fixed for Win NT & HP/UX).
The benefits of using HPVP are HUGE, both for online and batch,
in terms of the number of lines of code, CPU- and memory-consumption
at runtime. In one (timecritical) transaction, we noticed a performance
gain of about 30% using HPVP=ON.
I'm interested to hear some extra tips for getting better performance; maybe
there are some in the document you mentioned.
Thanks in advance,
Philippe
========
Chris Uttley posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 15:00
Subject: Using OLE/OCX requires CHANGE access to folder
We are using OLE/OCX controls for our COOL:GEN WINNT client procedures. We have found out that as we promote our generated applications from unit test to system test to production, we have to allow the users CHANGE access to the folder that contains the EXE, DLL, REG, OLE and SST files.
So when we get to production, we'll have to grant CHANGE access to our production folder that contains these things. This sort of goes against the usual thinking that production is a 'read-only' environment.
Anyone know why CHANGE access is required? Is this a COOL:GEN thing?
Secondly, has anyone developed a way to get around this requirement?
========
John posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 15:04
Subject: Calling Cool:gen Code from C++
Does anyone have advice or examples of how a C++ program can call Cool:gen action blocks.
========
ARISETTI AJAY KUMAR posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 15:56
Subject: Re: XPEDITER
pl. send me sample copy of xpiditor
========
Another CAB-er posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 16:21
Subject: Re: any reason to upgrade to cool:gen 5 ?
George,
I've heard that Composer 5 is gone in beta-test (I know one site)
So, I suspect there are specs, but I have not seen them.
From (very) high level presentations to our (top) management, we understand
that COOL:Gen is becoming SSW's cash-cow. The more 'exotic' targetplatforms
are dropped rapidly. Support is brought back at a minimum level. New
functionalities will only be implemented if it could help them sell you
all that new stuff. After several times 'being involved in the customer
advisory board', having seen several 'enhancement programs', it seems to
me that 'the 100% generated paradigma' is near to dead.
I hope I'm to pesimistic, what's your opnion ?
Another CAB-er from the TI-days.
========
David Mueller posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 16:43
Subject: Re: Calling Cool:gen Code from C++
Sterling Software has a white paper on calling Cool:gen code from c code.
The document is titled 'Composer Interfacing Guide' and was written by Rony Van Hove.
We received it by calling the Sterling Support Desk.
========
Sean Cohan posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 16:58
Subject: Deallocation Abend for Cooperative Server on DEC ALPHA
I'm getting the following error message trying to run my Coolgen 4.1 Win 95
client with a Dec Alpha Server:
TIRM030E: Application failed - Updates have been backed out
TIRM733E: Error deallocation abend
This occurs in a very simple model with a single procedure use statement
to my server, which sets a single ief suppied command and returns to the
client. I'm using Client Manager and the TE with TCP/IP. I've talked with
the Sterling Help Desk, and right now, they don't know why this is
happening. We've checked my ENV variables and the IT installation.
Everything looks ok. I've mailed them my .dat files; they can
get it running on their machines.
Any suggestions?
========
John McIntyre posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 17:00
Subject: Total cost of Composer/Cool ownership ???
We, in Telecom Ireland, are currently undergoing a process of
attempting to define a cost model for the Composer/Cool
development environment.
ie. what is the TOTAL cost of ownership?
Have you ever come across such a model or exercise ?
Are you also interested in such a model ?
Would you like to contribute with us?
If you answered even a small 'yes' to any of these questions, please
feel free to contact me at jmcintyre@telecom.ie (or in Berlin) or
here on the Bulletin Board.
Thank you
John McIntyre
========
Doug Michael posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 17:18
Subject: Re: Using OLE/OCX requires CHANGE access to folder
Chris,
It's the SST files that cause the problem. These contain the persistent info for the OCXs and the application must be able to update them in some situations.
You can put everything else in a network read only location and move the sst files somewhere else which has write access - this can be a local directory on the workstation which is in the path.
Hope this helps,
-Doug
========
Andrew Last posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 18:14
Subject: Re: Deallocation Abend for Cooperative Server on DEC ALPHA
This is the message when the server EXE cannot be run.
- Check the user running the EXEs (the user that starts AEFAD and AEFUF) has execute access (permissions on the files).
- Check the inqload directory is in the path when AEFAD and AEFUF are started.
- Lack or memory and EXE cannot be loaded ???
Hope this helps
Andrew Last
Orygen, Dublin, Ireland
========
Michael Watts posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 18:22
Subject: Changing bitmaps within an application
Does anyone know how to change a bitmap within an application? I have a need to show users a listbox containing information that has been sent to them. They would like a bitmap of an open envelope or a closed one to let them know what they have already looked at. Also can a bitmap be placed in a listbox?
Thanks for any help.
========
Johnny Serrano posted the following article on the TISUG Bulletin Board
Dated : 10 June, 1998 at 20:23
Subject: Re: Sheridan's DateCombo Control
Re-register with regsvr32.exe all the .ocx and .dlls just like a deployment
machine. Also be careful with the naming length on controls.
========
Andrew Last posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 08:41
Subject: Re: Changing bitmaps within an application
The standard list box does not support bitmaps. However if you look at Green Tree's List View OCX supplied with CoolGen then this supports bitmaps. The bitmaps can be changed at run time.
Hope this helps
Andrew Last
Orygen, Dublin, Ireland
========
Andrew Last posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 08:43
Subject: Re: Deallocation Abend for Cooperative Server on DEC ALPHA
Other reasons -
- Was the server installed (compiled) on the Alpha box and NOT on an INTEL
box. An INTEL EXE will not run on an Alpha box.
- Check the packaging (especially the environment settings and trancodes -
prior to generation and install). Must have a clear screen trancode and
dialog flow trancode otherwise the server does not know where to start.
========
Hannu posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 08:55
Subject: Using NT COOL:Gen toolset sw from the LAN
Hi!
We've had a concept in use with OS2 Composer 3, where one portion of the
developers have been using toolset from the lan drive instead of having a local
copy (was called 'shared license' or something). Sterling (TI Software on those days)
gave this oppurtunity and announced a separate document, which described,
how this was done. You had to have certain files on your PC.
We are busy to upgrade our toolset to Coolgen V41a and are going to jump into
Windows NT at the same time.
I haven't been able to find any document, how this could be done. Has anyone?
Or if anyone has examined this themselves, I'd appreciate to have that info.
Cheers, Hannu
========
Rusty Thomas posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 09:16
Subject: Re: Using NT COOL:Gen toolset sw from the LAN
The proper channel would be to ask your Sterling account representative for help
or have your local Sterling consultant ask for you. They will give you a form to sign
(the shared license you mention) PLUS the documentation and files you need to
configure your environment correctly. Without that document, Sterling support
will not answer any tickets dealing with this subject.
Have a nice day.
Rusty
========
Kevin Bingham posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 09:38
Subject: Re: Deallocation Abend for Cooperative Server on DEC ALPHA
Hi,
I had a similar problem that I made a posting on the BBS for quite some time
back, which can be found in the Archive, part 7, under following the heading
TIRM733E in Unix environment.
I am not sure if this will help....
Regards
Kevin
========
Christopher Macias posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 10:05
Subject: Re: Using NT COOL:Gen toolset sw from the LAN
Hannu,
To give you a less 'legalistic' answer... ;-)
The document Sterling provides describing the setup on NT server is a good starting point, but you will still have to do a lot of experimenting to get it working correctly.
Since you are moving from OS/2 to NT I have a few suggestions. First, if you have a lot of developers, consider alternatives to configuring each developer machine. If you are new to NT, mastering the updating of software across a large LAN is something you don't want to learn while you are *also* experimenting with switching OSes and moving to NT COOL:Gen. The simplest software distribution method is probably creating a master hard drive which is known to work, then duplicating that hard drive and swapping these copies in to each developer's machine each time the configuration changes significantly. Later you can move to something more sophisticated.
Secondly, Client Server Encyclopedias are great, but can be hell to maintain until you know what you are doing. If you start working with an NT CSE, have a good backup regime in place and TEST IT!
Finally, if you have to maintain some *users* on OS2 during the NT conversion, I can give you some suggestions about maintaining COOL:Gen apps that have to work on both platforms without having to maintain code in two models. Let me know if this applies to your situation and I can post or e-mail you the info.
Regards,
Christopher
========
Ken McIntosh posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 10:29
Subject: Re: Using NT COOL:Gen toolset sw from the LAN
Please could you send me a copy.
========
Hannu posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 11:46
Subject: Re: Using NT COOL:Gen toolset sw from the LAN
Thanks Christopher!
First of all, using toolset software from the LAN was legal (you still have to
pay for Sterling).
Second, I've noticed, what does it take to automize the installation from the
server by your own program. It can be done, but means efforts in every upgrade.
We've chosen your method to image hard drives, when rolling out the sw to
developers.
But, I was looking for details, how to setup your NT workstation to enable toolset
software usage from LAN drive, instead of your local hard drive. We've saved some
costs by this way of those developers, who aren't daily programmers!
Looks like you've been involved in the similar process than we over here. Yes,
if you have any documentation about how to maintain your model, when having two
client target envs, I'd be more than happy to have that. We've examined the issue
quite heavily, but the real work will begin on September. We haven't even thought of
situation having two different models (we don't love a song called 'Migration Blues'
too much).
Cheers, Hannu
========
Hannu posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 11:52
Subject: Re: Using NT COOL:Gen toolset sw from the LAN
Thanks Rusty!
I was going to ask a doc existence from support center, but didn't know,
that it is not public. Will do it, Cheers Hannu
BTW, for those interested, with this same method, you can create separate
development environments for different life cycles (unit test, system test ...)
Makes life much easier, since you don't have change the settings
(model paths etc.) And still you have only one copy of the software on your
PC.
========
Kevin Bingham posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 15:01
Subject: Calculations using floating point precision
Hi,
Picture this, you want to multiply 156,987.03 by 0.0123456 and be sure
(100% sure) that the answer is correct, especially if this where to be
in some kind of a loop where any intermediate error would be compounded
by each new iteration of the loop. What I have seen so far, is that CoolGen
tends to "loose" some of the "less important" digits in calculations of
this style.
Does anyone have a bullet proof method for performing these kinds of
large calculations?
Typically, the kind of calculation I'm envisaging involves numerous formulas,
powers, etc, and needs to be completely accurate, and able to reproduce the
same results each time. (It's money where are working with in the application.)
Any ideas will be most welcome.
Thanks
Kevin
========
Dan Atwood posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 22:05
Subject: Re: Changing bitmaps within an application
Just figured out how to add bitmaps to the GreenTree ListView control yesterday and was itching to show somebody.
The GreenTree ListView control uses an ImagePool (like 2 rolls of film) to manage two types of icons that can show up in the ListView control small and large. Once the size has been established it can't be changed until the ImageList is cleared. Since the GreenTree ListView can display either small or normal pictures next to each item, two ImagePools are needed to store either size.
INVOKE LISTVIEW.ListViewCtrl.ListItems.Clear()
+= FOR SUBSCRIPT OF group FROM 1 TO LAST OF group BY 1
| SET node gui gui_handle TO LISTVIEW.ListViewCtrl.
ListItems.Add(-32767, -32767,
textnum(export_gv customer id), 1, -32767)
| | SET node gui gui_handle SubItems(1) TO export_gv customer last_name
| | SET node gui gui_handle SubItems(2) TO export_gv customer first_name
| | SET node gui gui_handle TO NOTHING
| +--
| INVOKE LISTVIEW.ListViewCtrl.Refresh()
The statement that adds the bitmap is the "ListItems.Add" and is the 4th parameter. The "1" in this parameter pulls the first picture from the image pool.
Hope this helps
========
Dan Atwood posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 22:39
Subject: Re: Composer4/Office97
Will send you an email with the differences in Office95 and 97 OLE calls. Sorry it's written with Gen 4.1 in mind instead of Composer 4, but it should give you an idea of the differences.
There's a class called Desktop Integration that's available that gives a lot more detail.
Hope this helps,
Dan
========
Darius Panahy posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 22:49
Subject: Re: Calculations using floating point precision
There have been some previous postings on the BB on this (or similar) subject. You could check the archive to see if any of them are relevant.
Darius
========
ISSUG Webperson posted the following article on the TISUG Bulletin Board
Dated : 11 June, 1998 at 23:48
Subject: IMPORTANT MESSAGE ABOUT THIS WEBSITE!
Hi there,
I'm the guy who looks after this website and this is to let you know that
we ISSUG people have been quietly preparing a new and better set of facilties
for you. They are almost ready, so www.tisug.com will close down this weekend
(15th June) and a new site on our own server will open. When you come to this
site next week you will be automatically redirected to the new site.
You will find it very different and, we hope, better, more useful, and prettier.
We're using new bulletin board software (from the market leader in the field) rather
than the free Perl script which we currently run. But I won't give you details right
now. Come back next week and see for yourself.
One point, though. Because we're providing more facilities, we've had to push at
the limits of what some older web browsers can handle. We've tried to cover all
bases (we've written three different versions of the site for different browser types)
but the new site definitely works best with Internet Explorer 4 and Netscape 4,
simply because they handle scripts reasonably well. We know that nearly all of
you use one of those two browsers (we've been checking) but a few are still on v3.
Well, browser choice is a personal matter and we wouldn't dare suggest you upgrade!
We do have major problems with Opera, though, and can't figure out why.
One other point. The new site will give you cookies. Accept them. They are harmless
and their sole purpose is to make things easier for you.
We hope you have a smooth ride at the new site. If not, let us know and we'll try
to sort things out. Please bear with us. And let's hope for a problem-free transition!
Thanks to you all for supporting this site and we hope you'll want to carry on posting
on the new site.
See you there.
George H.
ISSUG webperson
========
Dan Atwood posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 00:30
Subject: Re: Grid-Control
Not sure if you're using guiobject handles or strictly dot notation, but know I was slow to learn to always set my guiobject handles to NOTHING to free up memory. Do it as soon as you can (preferably before leaving the event). If views are initialized each time you enter the logic, the guiobject handles (and memory) are left out there hanging.
Memory leaks have been worked on lately with the controls. Might check to see if you've applied all the way to RTN4119.
Hope this helps
========
Dan Atwood posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 01:02
Subject: Re: any reason to upgrade to cool:gen 5 ?
In Gen 5.0 no targeted platforms are going away. Same list of platforms are going to get generated for in Gen 5.0.
It's true the majority of revenue comes from COOL:Gen. That's why over half the Labs is dedicated to enhancements and its maintenance. It's not being put out to pasture as a "cash cow".
Probably telling you stuff you already know, but in COOL:Gen 5.0 focus is on:
1. Middleware (Tuxedo and Encina) for native messages, multiphase commits and data dependent routing.
2. Improving the Proxies (Java beans, security exits, C)
3. Database exception handling to capture / retry
4. Model manager to make it easier to share components across models
Unless you're using these specific features I guess this doesn't make it the most exciting release, but there needed to be a more stable code base to build on (tough to count all the PTFs in Gen 4.1a). COOL:Gen 5.0 provides that.
Hope this helps,
Dan Atwood
Member Technical Staff
Labs, Sterling Software
========
Elwin Tennant posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 09:06
Subject: Re: Thanks for the Paper!
Many thanks for your paper. This kind of support is what makes this
Bulletin Board as useful as it is. At the moment it's great - I just
hope it stays that way, and that people do not become reluctant to
publish information they have, for personal 'competitive advantage'.
At the end of the day, it is the success of projects using the COOL:Gen
tool which will determine whether it succeeds or fails in the
marketplace, against the increasingly competitive, cheaper and
'trendier' tools. Let's see a continuation of this openness, particularly
by Sterling, who have the most to gain!
========
Christopher Macias posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 12:02
Subject: Re: Using NT COOL:Gen toolset sw from the LAN
Hannu,
I only have the NT-on-a-server document in paper form, so you may have to get it from Sterling. See if somebody at your site has access to their technical support site, or if their is a representative who can get it for you.
The LAN people here are using Tivoli to manage software configuration. Microsoft's Systems Management Server (SMS) is, I'm told, too weak in its support of WANs to work well in a large business. Microsoft itself seems to have acknowledged this by buying or licensing a competing product to Tivoli and planning to integrate it with NT. I can hunt up the name of that product if you're interested, or you could check it out on the web.
As to the 'one model:two target environments' problem...
We followed up on a suggestion from another site that thought you could use the 'Dialects' feature for this. (Yeah, I never heard of it before either, but look for it in the tool and you'll see it.) Dialects allow you to attach multiple sets of windows to a single procedure. The idea was to support multiple languages by letting you create one set of windows for, say, English and another set for German or whatever.
I developed a procedure to use two dialects (one for OS2 and one for NT) to allow us to use a single code base for both environments until the OS2 clients are 'retired'. We've tested it and, although it has a few tricks you have to be careful of, it works. It is also MUCH less effort than splitting the models and trying to maintain the code base in parallel.
You can follow up that lead yourself or, if you want, I'll ask if I can send you a copy of the procedures I wrote.
Regards,
Christopher
P.S. You've got a 'pohjolagroup' e-mail address, so I assume you are in Helsinki. Strangely enough, I'll be there this weekend to visit a couple of friends of mine, one of whom (Rochelle Moore) works for the same company as you do (I assume). Maybe I'll see you around.
========
Preben Nilsson posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 12:04
Subject: Re: Calculations using floating point precision
Have U tried cliking the "Implement in C with decimal precision" flag in the TD properties of the attribute ?
========
Christopher Macias posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 12:07
Subject: Re: IMPORTANT MESSAGE ABOUT THIS WEBSITE!
George,
Greate news! But...
Can you tell us if the current message threads will migrate to the new server? What about the archives?
Thanks,
Christopher Macias
========
george posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 13:24
Subject: Re: IMPORTANT MESSAGE ABOUT THIS WEBSITE!
Hi Chris,
Archives - no problem. In fact you've got a lot more archived postings available
on the new site. It should be pretty well up to date, barring a few messages that got
lost along the way through server glitches, admin error etc.
Current threads - I'm moving what seem to be live threads (but it's a manual job
so I can't do many, just ones added to in the last few days) and adding the rest to
the archive. If I miss something important let me know.
George
George
========
Theresa Ellis posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 13:25
Subject: CICS Password expiration/change
We are deploying with Windows NT client and MVS/CICS/DB2 server and would like
to handle signon errors, invalid password, expired password, etc. From what
I've been able to find from Electronic books, CICS manuals, and such I THINK
that what we need to do is write an APPC requestor to interface with the CICS
PEM (Password Expiration Managment) Server. I've found the CIDEXIT program
for the client manager and am guessing that this is where the code would go. Has
anyone ever had any experience with this or have any other (better or correct)
ways of handling CICS signon/password errors from the Client side? I can't
imagine that this hasn't already been done somewhere else. (This is our first
application and we're relatively new at it)
P.S. If you e-mail after June 26,1998 send responses to christj@state.wi.us
========
Angus posted the following article on the TISUG Bulletin Board
Dated : 12 June, 1998 at 14:12
Subject: Popup (Whats This ?) Help and Cool:Gen
Has any succesfully managed to display context sensitive help on a Cool:Gen application in a popup window ?
If so how ?
Cheers ....
========
top
of page