Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Stupid DCL Tricks revisited - Bar and Pie Charts in DCL

39 views
Skip to first unread message

Peter Weaver

unread,
Jul 15, 2009, 5:48:05 PM7/15/09
to
A few years ago I wrote up a procedure for creating Pie and Bar graphs
using 100% DCL. The DCL generated Scalable Vector Graphics code that
could be interpreted by your browser as long as you had a SVG viewer.
For details you can see
http://groups.google.ca/group/comp.os.vms/browse_thread/thread/e0d3de43978ca3cc/db8645fa7652910b?

Occasionally people stumble on that message and the send me requests
for the code followed by complaints that they cannot get SVG working
for some reason or another. So here is another attempt at generating
graphs in pure DCL but this time using Google’s Charting API. You can
see the new example at http://www.weaverconsulting.ca/cgi-bin/system_info,
the original SVG example is still at http://www.weaverconsulting.ca/cgi-bin/get_disks
if you want to compare. If anyone wants the code for either then let
me know.

BTW: I just noticed that the previous post was 3 years ago today, must
be a quiet time of the year for me…

Peter Weaver
www.weaverconsulting.ca
Winner of the OpenVMS.org Readers' Choice Award for System Management/
Performance
http://www.linkedin.com/in/peterweaver

Jan-Erik Söderholm

unread,
Jul 16, 2009, 2:53:48 AM7/16/09
to
Peter Weaver wrote:
> A few years ago I wrote up a procedure for creating Pie and Bar graphs
> using 100% DCL. The DCL generated Scalable Vector Graphics code that
> could be interpreted by your browser as long as you had a SVG viewer.
> For details you can see
> http://groups.google.ca/group/comp.os.vms/browse_thread/thread/e0d3de43978ca3cc/db8645fa7652910b?
>
> Occasionally people stumble on that message and the send me requests
> for the code followed by complaints that they cannot get SVG working
> for some reason or another. So here is another attempt at generating
> graphs in pure DCL but this time using Google�s Charting API. You can

Nice. Would like to look at the DCL sourcec for that. Was it available
for download someplace or was you going to send it it each one asking
for it ?

> the original SVG example is still at
> http://www.weaverconsulting.ca/cgi-bin/get_disks
> if you want to compare.

That give me a message saying "This XML file does not appear to have
any style information associated with it. The document tree is shown
below." followed by some XML code. This is in each of the 7 frames
that I guess shoulod have a graph displayed. I tried to download
and install the "SVG-viewer 3.03" from Adobe but that didn't help.

If anyone wants the code for either then let
> me know.
>
> BTW: I just noticed that the previous post was 3 years ago today, must

> be a quiet time of the year for me�

Peter Weaver

unread,
Jul 16, 2009, 10:39:12 AM7/16/09
to
On Jul 16, 2:53 am, Jan-Erik Söderholm <jan-erik.soderh...@telia.com>
wrote:
>...

>
> Nice. Would like to look at the DCL sourcec for that. Was it available
> for download someplace or was you going to send it it each one asking
> for it ?

You can pick up the .COM at http://www.weaverconsulting.ca/cgi-bin/system_info.zip.

>
> > the original SVG example is still at
> >http://www.weaverconsulting.ca/cgi-bin/get_disks
> > if you want to compare.
>
> That give me a message saying "This XML file does not appear to have
> any style information associated with it. The document tree is shown
> below." followed by some XML code. This is in each of the 7 frames
> that I guess shoulod have a graph displayed. I tried to download
> and install the "SVG-viewer 3.03" from Adobe but that didn't help.

>...

SVG looked like a great idea when I first played with it, but Adobe
has retired the viewer now so I guess it is time to give up on it. I'm
not sure what platform you are using but if you have another utility
that understands SVG (like GIMP) then you can view the graphs.

The Google Chart API is platform neutral so you should see the graphs
on any system that has a browser (what am I saying, this is COV - I'll
get 10 people responding that it doesn't work if you use Lynx, then
the thread will morph into an argument over whether I meant 10 decimal
or 10 binary…) but it does have a lot of limitations. For example when
are using "Text Encoding with Scaling" you have a range of (+/-)9.999e
(+/-)100 but you can only have 4 non-zero digits in the number,

MetaEd

unread,
Jul 16, 2009, 10:54:43 AM7/16/09
to
On Thu, Jul 16, 2009 at 9:39 AM, Peter Weaver<info-
v...@weaverconsulting.ca> wrote:
> SVG looked like a great idea when I first played with it, but Adobe
> has retired the viewer now so I guess it is time to give up on it. I'm
> not sure what platform you are using but if you have another utility
> that understands SVG (like GIMP) then you can view the graphs.

Google Chrome will render SVG. The feature is native---no add-on
needed.

--
What part of "Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn"
don't you understand?

Richard B. Gilbert

unread,
Jul 16, 2009, 11:12:56 AM7/16/09
to

"In his house at R'lyeh great Cthulhu lies sleeping."

See H.P. Lovecraft's Cthulhu Mythos

So you see that at least two of us understand. . . .

ken

unread,
Jul 16, 2009, 12:42:03 PM7/16/09
to
On Jul 15, 5:48 pm, Peter Weaver <info-...@weaverconsulting.ca> wrote:
> A few years ago I wrote up a procedure for creating Pie and Bar graphs
> using 100% DCL. The DCL generated Scalable Vector Graphics  code that
> could be interpreted by your browser as long as you had a SVG viewer.
> For details you can seehttp://groups.google.ca/group/comp.os.vms/browse_thread/thread/e0d3de...

>
> Occasionally people stumble on that message and the send me requests
> for the code followed by complaints that they cannot get SVG working
> for some reason or another. So here is another attempt at generating
> graphs in pure DCL but this time using Google’s Charting API. You can
> see the new example athttp://www.weaverconsulting.ca/cgi-bin/system_info,
> the original SVG example is still athttp://www.weaverconsulting.ca/cgi-bin/get_disks

> if you want to compare. If anyone wants the code for either then let
> me know.
>

I wrote something similar. It uses PHP, DCL, Javascript, (jQuery,
AJAX), and Google Gauges (http://code.google.com/apis/visualization/
documentation/gallery/gauge.html) to display the info. A snapshot of
the screen can be seen at http://www.rbnsn.com/vms/

The screen updates every 10 seconds. I wrote it to monitor our 3 node
cluster, but probably could be modified for more nodes.

Ken

VAXman-

unread,
Jul 16, 2009, 1:03:48 PM7/16/09
to
In article <c7cbe694-b793-4df0...@a26g2000yqn.googlegroups.com>, Peter Weaver <info...@weaverconsulting.ca> writes:
>On Jul 16, 2:53=A0am, Jan-Erik S=F6derholm <jan-erik.soderh...@telia.com>

>wrote:
>>...
>>
>> Nice. Would like to look at the DCL sourcec for that. Was it available
>> for download someplace or was you going to send it it each one asking
>> for it ?
>
>You can pick up the .COM at http://www.weaverconsulting.ca/cgi-bin/system_i=
>nfo.zip.

Internal Server Error

The server encountered an internal error or misconfiguration and was unable to complete your request.

Please contact the server administrator, ad...@weaverconsulting.ca and inform them of the time the
error occurred, and anything you might have done that may have caused the error.

More information about this error may be available in the server error log.

Apache/2.0.52 (OpenVMS) mod_ssl/2.0.52 OpenSSL/0.9.7d PHP/5.2.6 Server at www.weaverconsulting.ca Port 80

--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG

http://www.quirkfactory.com/popart/asskey/eqn2.png

"Well my son, life is like a beanstalk, isn't it?"

Peter Weaver

unread,
Jul 16, 2009, 2:29:22 PM7/16/09
to
On Jul 16, 1:03 pm, VAXman- @SendSpamHere.ORG wrote:
>...
> >You can pick up the .COM athttp://www.weaverconsulting.ca/cgi-bin/system_i=
> >nfo.zip.
>
> Internal Server Error
>...

http://www.weaverconsulting.ca/system_info.zip should work better.

Peter Weaver

unread,
Jul 16, 2009, 2:40:29 PM7/16/09
to
On Jul 16, 12:42 pm, ken <kenrb...@gmail.com> wrote:
>...

> I wrote something similar. It uses  PHP, DCL, Javascript, (jQuery,
> AJAX), and Google Gauges (http://code.google.com/apis/visualization/
> documentation/gallery/gauge.html) to display the info. A snapshot of
> the screen can be seen athttp://www.rbnsn.com/vms/

>
> The screen updates every 10 seconds. I wrote it to monitor our 3 node
> cluster, but probably could be modified for more nodes.
>
> Ken

Nice. I was looking at Google Gauges the other day and I was thinking
that it might be fun to use the Word Cloud to show how many users are
on a system, or batch vs. interactive .vs. network processes or even
the mode of the processes.

My original goal was to create some simple graphics for Kostas
Gavrielidis' VMS_CHECK
(http://h71000.www7.hp.com/openvms/journal/v7/vms_check_tool.pdf) and
Kostas said that it had to be something that would run on any VAX/
Alpha/IA64 machine without installing any additional packages so that
is why I stuck with 100% DCL (plus it is fun to see what can be done
in DCL). I don't know if Kostas ever put my code into VMS_CHECK or if
VMS_CHECK is even around these days.

George Cook

unread,
Jul 16, 2009, 5:09:18 PM7/16/09
to
> On Jul 16, 2:53=A0am, Jan-Erik S=F6derholm <jan-erik.soderh...@telia.com>

> wrote:
>>...
>>
>> Nice. Would like to look at the DCL sourcec for that. Was it available
>> for download someplace or was you going to send it it each one asking
>> for it ?
>
> You can pick up the .COM at http://www.weaverconsulting.ca/cgi-bin/system_i=

> nfo.zip.
>
>>
>> > the original SVG example is still at
>> >http://www.weaverconsulting.ca/cgi-bin/get_disks
>> > if you want to compare.
>>
>> That give me a message saying "This XML file does not appear to have
>> any style information associated with it. The document tree is shown
>> below." followed by some XML code. This is in each of the 7 frames
>> that I guess shoulod have a graph displayed. I tried to download
>> and install the "SVG-viewer 3.03" from Adobe but that didn't help.
>>...
>
> SVG looked like a great idea when I first played with it, but Adobe
> has retired the viewer now so I guess it is time to give up on it. I'm
> not sure what platform you are using but if you have another utility
> that understands SVG (like GIMP) then you can view the graphs.

FWIW, the next release of VMS Mosaic displays SVG images. The HTML
height and width specs on your page causes Mosaic problems, but the
images looked okay once I edited the HTML.

I should have had the next release out last year, but back surgery,
etc. has prevented most work on it.


George Cook
WVNET

VAXman-

unread,
Jul 16, 2009, 6:53:36 PM7/16/09
to
In article <f89bc8a0-e58c-4335...@c1g2000yqi.googlegroups.com>, Peter Weaver <info...@weaverconsulting.ca> writes:

>On Jul 16, 1:03=A0pm, VAXman- @SendSpamHere.ORG wrote:
>>...
>> >You can pick up the .COM athttp://www.weaverconsulting.ca/cgi-bin/system=
>_i=3D

>> >nfo.zip.
>>
>> Internal Server Error
>>...
>
>http://www.weaverconsulting.ca/system_info.zip should work better.

It's a miracle! (In an echoing voice from a monty python sound bite...)

Peter Weaver

unread,
Jul 16, 2009, 7:47:25 PM7/16/09
to
On Jul 16, 5:09 pm, c...@wvnvms.wvnet.edu (George Cook) wrote:
>...

> FWIW, the next release of VMS Mosaic displays SVG images.  The HTML
> height and width specs on your page causes Mosaic problems, but the
> images looked okay once I edited the HTML.
>
> I should have had the next release out last year, but back surgery,
> etc. has prevented most work on it.
>...

Sorry to hear about the back surgery, hope you are feeling better.

If you have a chance to try it then give http://www.weaverconsulting.ca/cgi-bin/get_disks
a shot with Mosaic and see if my code works.

Anyone else who tried the SVG version and/or downloaded the SVG.ZIP
file before the timestamp on this post should try it again. The
original SVG worked fine with the Adobe viewer, but it did not work
with Chrome until just a few minutes ago.

urbancamo

unread,
Jul 17, 2009, 4:18:12 AM7/17/09
to
Works in Opera under NetBSD. That's probably a recommendation for
widespread acceptance.

George Cook

unread,
Jul 17, 2009, 5:37:51 PM7/17/09
to
In article <f9ecb493-2633-44ad...@p15g2000vbl.googlegroups.com>, Peter Weaver <info...@weaverconsulting.ca> writes:

> On Jul 16, 5:09=A0pm, c...@wvnvms.wvnet.edu (George Cook) wrote:
>>...
>> FWIW, the next release of VMS Mosaic displays SVG images. =A0The HTML

>> height and width specs on your page causes Mosaic problems, but the
>> images looked okay once I edited the HTML.
>>
>> I should have had the next release out last year, but back surgery,
>> etc. has prevented most work on it.
>>...
>
> Sorry to hear about the back surgery, hope you are feeling better.

Let's just say I've lost what little respect I had for doctors.

> If you have a chance to try it then give http://www.weaverconsulting.ca/cgi=


> -bin/get_disks
> a shot with Mosaic and see if my code works.

Displayed with no problems. The earlier HTML problems I had were
Mosaic bugs (WIDTH=xx% and HEIGHT=xx% are not very common in image
tags) which have now been fixed.


George Cook
WVNET

Peter Weaver

unread,
Jul 20, 2009, 9:00:40 PM7/20/09
to
On Jul 17, 4:18 am, urbancamo <m...@wickensonline.co.uk> wrote:
> Works in Opera under NetBSD. That's probably a recommendation for
> widespread acceptance.

The SVG version even works on the iPhone with Safari, you can even
zoom in on the "font-size: 4" text and it looks crisp and clear. But
for some reason the spacing between graphs is off and you have to
scroll down quite a ways to see all the charts.

I'm surprised at the amount of interest this thread has generated,
about 80 unique IP addresses downloaded the ZIP file with the Google
Charts example and about 20 downloaded the SVG example code. I didn't
bother to check for addresses of people who downloaded both.

Jan-Erik Söderholm

unread,
Jul 21, 2009, 6:01:49 AM7/21/09
to
Peter Weaver wrote:
> On Jul 17, 4:18 am, urbancamo <m...@wickensonline.co.uk> wrote:
>> Works in Opera under NetBSD. That's probably a recommendation for
>> widespread acceptance.
>
> The SVG version even works on the iPhone with Safari, you can even
> zoom in on the "font-size: 4" text and it looks crisp and clear. But
> for some reason the spacing between graphs is off and you have to
> scroll down quite a ways to see all the charts.
>
> I'm surprised at the amount of interest this thread has generated,

I'm not. :-)

I think/guess that this is the kind of stuff many VMS sites
are looking for/at at these times. Getting the VMS systems to
co-exist in todays IT infrastructure (web, soap, graphics and
so on) is probably one of the major issues to get the VMS
systems to survive...

I myself are currently testing interaction between Excel VBA
scripts and our VMS applications. Looks quite good. Using the
"Office 2003 Web Services Toolkit" I've been able to access
WSDL files and SOAP services through WASD on the VMS box.
It's realy easy on the Excel side, just import (point to
the URL of) the WSDL file and write (in the simplest case)
*one* single line of VBA code and it's done. In my case
two Excel cells are sent to a test service and the sum
is retured to a third cell. All with a simple call like :

ReturnedCell.value = myAdd_funktion(cellA, cellB)

The myAdd_funktion is automaticly created by the Toolkit
when the WSDL file is imported.

The SOAP service used is the Add sample in the
[samples.cobol] directory in the gSOAP/VMS kit.
The addition is in a simple COB file, just to show the
SOAP calls to a COBOL application.

This is to solve an actual business need at my client site.
A internaly developed "application" written in Excel/VBA
has to interact with the main applications on VMS...

I'm also doing some test with Silverlight where one have
similar functionality (with regard to the SOAP support).

Oops, got a bit carried away, sorry for that... :-)

Jan-Erik.


Arne Vajhøj

unread,
Jul 21, 2009, 8:29:39 PM7/21/09
to
Jan-Erik S�derholm wrote:
> Peter Weaver wrote:
>> On Jul 17, 4:18 am, urbancamo <m...@wickensonline.co.uk> wrote:
>>> Works in Opera under NetBSD. That's probably a recommendation for
>>> widespread acceptance.
>>
>> The SVG version even works on the iPhone with Safari, you can even
>> zoom in on the "font-size: 4" text and it looks crisp and clear. But
>> for some reason the spacing between graphs is off and you have to
>> scroll down quite a ways to see all the charts.
>>
>> I'm surprised at the amount of interest this thread has generated,
>
> I'm not. :-)
>
> I think/guess that this is the kind of stuff many VMS sites
> are looking for/at at these times. Getting the VMS systems to
> co-exist in todays IT infrastructure (web, soap, graphics and
> so on) is probably one of the major issues to get the VMS
> systems to survive...

There are practically none VMS-only sites today.

Lots of Windows stuff, 3-4 flavors of *nix, maybe some Mac's,
maybe some i or z/OS (OS/400 or MVS).

Interoperability is a must.

> The myAdd_funktion is automaticly created by the Toolkit
> when the WSDL file is imported.
>
> The SOAP service used is the Add sample in the
> [samples.cobol] directory in the gSOAP/VMS kit.
> The addition is in a simple COB file, just to show the
> SOAP calls to a COBOL application.
>
> This is to solve an actual business need at my client site.
> A internaly developed "application" written in Excel/VBA
> has to interact with the main applications on VMS...
>
> I'm also doing some test with Silverlight where one have
> similar functionality (with regard to the SOAP support).

Flash/Flex/AS2, Silverlight and Java/JavaFX applets all speak SOAP.

If one is willing to accept the overhead, then it gets a lot
done with with very little effort.

Arne

Jan-Erik Söderholm

unread,
Jul 22, 2009, 4:37:10 AM7/22/09
to
Arne Vajh�j wrote:
> Jan-Erik S�derholm wrote:
>> Peter Weaver wrote:
>>> On Jul 17, 4:18 am, urbancamo <m...@wickensonline.co.uk> wrote:
>>>> Works in Opera under NetBSD. That's probably a recommendation for
>>>> widespread acceptance.
>>>
>>> The SVG version even works on the iPhone with Safari, you can even
>>> zoom in on the "font-size: 4" text and it looks crisp and clear. But
>>> for some reason the spacing between graphs is off and you have to
>>> scroll down quite a ways to see all the charts.
>>>
>>> I'm surprised at the amount of interest this thread has generated,
>>
>> I'm not. :-)
>>
>> I think/guess that this is the kind of stuff many VMS sites
>> are looking for/at at these times. Getting the VMS systems to
>> co-exist in todays IT infrastructure (web, soap, graphics and
>> so on) is probably one of the major issues to get the VMS
>> systems to survive...
>
> There are practically none VMS-only sites today.

Correct, But sorry to say, some posters on c.o.v are having
a hard time accepting that.

>
> Lots of Windows stuff, 3-4 flavors of *nix, maybe some Mac's,
> maybe some i or z/OS (OS/400 or MVS).
>
> Interoperability is a must.
>
>> The myAdd_funktion is automaticly created by the Toolkit
>> when the WSDL file is imported.
>>
>> The SOAP service used is the Add sample in the
>> [samples.cobol] directory in the gSOAP/VMS kit.
>> The addition is in a simple COB file, just to show the
>> SOAP calls to a COBOL application.
>>
>> This is to solve an actual business need at my client site.
>> A internaly developed "application" written in Excel/VBA
>> has to interact with the main applications on VMS...
>>
>> I'm also doing some test with Silverlight where one have
>> similar functionality (with regard to the SOAP support).
>
> Flash/Flex/AS2, Silverlight and Java/JavaFX applets all speak SOAP.
>
> If one is willing to accept the overhead, then it gets a lot
> done with with very little effort.
>
> Arne

Yes, and "very little effort" is not that bad... :-)

B.t.w, the maintainer of gSOAP/VMS and the maintainer of WASD
have been looking at implementing FastCGI (where the server process
is keep alive between calls to save the image activation at each
CGI call) into the gSOAP port for VMS. That will greatly reduce
the overhead with SOAP calls through WASD. From what I understand
it's more or less ready and will be in the next gSOAP/VMS release.

I find WASD/gSOAP far easier to maintain and setup then to write
direct TCP-socket calls, even if that probably would have given
slightly better performance.

Jan-Erik.

Richard Maher

unread,
Jul 22, 2009, 6:22:22 AM7/22/09
to
Hi Arne,

"Arne Vajh�j" <ar...@vajhoej.dk> wrote in message
news:4a665d6d$0$48233$1472...@news.sunsite.dk...

And they all speak Ajax and all speak Sockets. (The recent Silverlight
embrace of the paradigm proving quite illuminating!)

>
> If one is willing to accept the overhead, then it gets a lot
> done with with very little effort.
>
> Arne

Cheers Richard Maher


Richard Maher

unread,
Jul 22, 2009, 6:57:10 AM7/22/09
to
Hi Jan-Erik,

"Jan-Erik S�derholm" <jan-erik....@telia.com> wrote in message
news:WcA9m.10411$U5.1...@newsb.telia.net...

Snip>


> Yes, and "very little effort" is not that bad... :-)

Depends how one defines "very little effort" I suppose.

>
> B.t.w, the maintainer of gSOAP/VMS and the maintainer of WASD

Strange bed-fellows indeed!

> have been looking at implementing FastCGI (where the server process
> is keep alive between calls to save the image activation at each
> CGI call) into the gSOAP port for VMS.

More cutting-edge stuff.

> That will greatly reduce
> the overhead with SOAP calls through WASD.

Maybe they should patent it - it might catch on?

No RESTful interest in there then?

> From what I understand
> it's more or less ready and will be in the next gSOAP/VMS release.

Good to see HP/VMS license-payer money can still be siphoned off by an
expert! Just a shame those pitiful VMS Engineers you all knew and loved, and
have just been sacked for cost-savings, simply didn't engage in the
Machiavellian pursuits of those who now gorge themselves at the trough.


>
> I find WASD/gSOAP far easier to maintain and setup then to write
> direct TCP-socket calls,

Oracle/SUN Java Sockets, Adobe Flex Sockets, Microsoft Silverlight Sockets,
and HTML 5 Bollocks Web Sockets - but don't let that get in the way of your
myopia. Still I have to agree that when it comes to the server-side
requirements of multi-threading, authentication, network communication,
client-impersonation, and transaction coordination, it certainly would be
nice to have some middleware to help you out.

Imaging just having to write 6 User Action Routines (in any VMS 3GL you
choose) and have the rest all done for you? A context-rich, full-duplex,
load-balanced, pipe as the main parameter between your client and server
code sound appealing?

No? How 'bout adding single sign-on, single TCP/IP connection serving
multiple disparate pages across multiple tabs in a browser instance?

Nah, didn't think so; funny how you'll love it on Linux :-(

> even if that probably would have given
> slightly better performance.

Then why bother with "Fast" CGI?

"Slightly" better?

>
> Jan-Erik.

Full steam ahead!

Regards Richard Maher
>


Arne Vajhøj

unread,
Jul 22, 2009, 9:37:48 PM7/22/09
to
Jan-Erik S�derholm wrote:
> Arne Vajh�j wrote:
>> Jan-Erik S�derholm wrote:
>>> Peter Weaver wrote:
>>>> On Jul 17, 4:18 am, urbancamo <m...@wickensonline.co.uk> wrote:
>>>>> Works in Opera under NetBSD. That's probably a recommendation for
>>>>> widespread acceptance.
>>>>
>>>> The SVG version even works on the iPhone with Safari, you can even
>>>> zoom in on the "font-size: 4" text and it looks crisp and clear. But
>>>> for some reason the spacing between graphs is off and you have to
>>>> scroll down quite a ways to see all the charts.
>>>>
>>>> I'm surprised at the amount of interest this thread has generated,
>>>
>>> I'm not. :-)
>>>
>>> I think/guess that this is the kind of stuff many VMS sites
>>> are looking for/at at these times. Getting the VMS systems to
>>> co-exist in todays IT infrastructure (web, soap, graphics and
>>> so on) is probably one of the major issues to get the VMS
>>> systems to survive...
>>
>> There are practically none VMS-only sites today.
>
> Correct, But sorry to say, some posters on c.o.v are having
> a hard time accepting that.

True.

>>
>> Lots of Windows stuff, 3-4 flavors of *nix, maybe some Mac's,
>> maybe some i or z/OS (OS/400 or MVS).
>>
>> Interoperability is a must.
>>
>>> The myAdd_funktion is automaticly created by the Toolkit
>>> when the WSDL file is imported.
>>>
>>> The SOAP service used is the Add sample in the
>>> [samples.cobol] directory in the gSOAP/VMS kit.
>>> The addition is in a simple COB file, just to show the
>>> SOAP calls to a COBOL application.
>>>
>>> This is to solve an actual business need at my client site.
>>> A internaly developed "application" written in Excel/VBA
>>> has to interact with the main applications on VMS...
>>>
>>> I'm also doing some test with Silverlight where one have
>>> similar functionality (with regard to the SOAP support).
>>
>> Flash/Flex/AS2, Silverlight and Java/JavaFX applets all speak SOAP.
>>
>> If one is willing to accept the overhead, then it gets a lot
>> done with with very little effort.
>

> Yes, and "very little effort" is not that bad... :-)
>
> B.t.w, the maintainer of gSOAP/VMS and the maintainer of WASD
> have been looking at implementing FastCGI (where the server process
> is keep alive between calls to save the image activation at each
> CGI call) into the gSOAP port for VMS. That will greatly reduce
> the overhead with SOAP calls through WASD. From what I understand
> it's more or less ready and will be in the next gSOAP/VMS release.

Interesting.

> I find WASD/gSOAP far easier to maintain and setup then to write
> direct TCP-socket calls, even if that probably would have given
> slightly better performance.

It could give a lot better performance.

But not everyone needs 4-6 digit TPM.

And the API centric view of SOAP/HTTP is a lot faster to develop
against compared to the protocol centric view of a custom
binary format over plain sockets.

Arne

Arne Vajhøj

unread,
Jul 22, 2009, 9:41:26 PM7/22/09
to

The ability to speak SOAP over HTTP implies the ability
to use POX or JSON over HTTP and use sockets.

But you need some special requirements to justify implementing
your own protocol.

Arne

Arne Vajhøj

unread,
Jul 22, 2009, 9:49:25 PM7/22/09
to
Richard Maher wrote:
> "Jan-Erik S�derholm" <jan-erik....@telia.com> wrote in message
> news:WcA9m.10411$U5.1...@newsb.telia.net...
> Snip>
>> Yes, and "very little effort" is not that bad... :-)
>
> Depends how one defines "very little effort" I suppose.

If the toolkit generates a stub, then calling a web service
is as simple as calling a local function/method.

>> have been looking at implementing FastCGI (where the server process
>> is keep alive between calls to save the image activation at each
>> CGI call) into the gSOAP port for VMS.
>
> More cutting-edge stuff.
>
>> That will greatly reduce
>> the overhead with SOAP calls through WASD.
>
> Maybe they should patent it - it might catch on?

Using FastCGI to reduce activation overhead is not exactly new.

> No RESTful interest in there then?

You can use REST with FastCGI, but gSOAP most likely only
supports SOAP.

>> From what I understand
>> it's more or less ready and will be in the next gSOAP/VMS release.
>
> Good to see HP/VMS license-payer money can still be siphoned off by an
> expert!

Hm.

Are gSOAP for VMS being done by HP ??

>> even if that probably would have given
>> slightly better performance.
>
> Then why bother with "Fast" CGI?

X > Y does not imply that Z <= X.

Arne

Jan-Erik Söderholm

unread,
Jul 23, 2009, 3:43:54 AM7/23/09
to
Arne Vajh�j wrote:

> If the toolkit generates a stub, then calling a web service
> is as simple as calling a local function/method.

That's how it works, yes.

>
> Using FastCGI to reduce activation overhead is not exactly new.

No, it's not. But it wasn't used by the the gSOAP/VMS kit.
There are examples of running as an "module" within
Apache, and as far as I understand using FastCGI is
quite similar to that. And I prefer WASD over the Apache
port.

> Hm.
>
> Are gSOAP for VMS being done by HP ??

Not up to today. But as far as I understand it will be
(more) officialy supported by HP/VMS soon. gSOAP is in
VMS roadmap now (FWIW). This is of course an important
issue for some sites/customers, included the one I'm
working with.

Jan-Erik.

Richard Maher

unread,
Jul 23, 2009, 6:10:24 AM7/23/09
to
Hi Arne,

"Arne Vajh�j" <ar...@vajhoej.dk> wrote in message

news:4a67bfbe$0$48233$1472...@news.sunsite.dk...

Fascinating!

So XML via HTTP bloatware is kosher but XML over Sockets is "special" and
"your own protocol"?

So JSON via a poorly performing session-hijackable piece of shite does not
offend Arne's sensibilities, yet JSON over a socket is anathema?

I'm afraid your bigotry and myopia on this issue does you an injustice and
you'd be much better off answering questions about ThreadDeath Errors in
c.l.j.p :-)
>
> Arne

Cheers Richard Maher


Jan-Erik Söderholm

unread,
Jul 23, 2009, 6:20:34 AM7/23/09
to
Richard Maher wrote:
> Hi Arne,
>
> "Arne Vajh�j" <ar...@vajhoej.dk> wrote in message
> news:4a67bfbe$0$48233$1472...@news.sunsite.dk...
>
>> The ability to speak SOAP over HTTP implies the ability
>> to use POX or JSON over HTTP and use sockets.
>>
>> But you need some special requirements to justify implementing
>> your own protocol.
>
> Fascinating!
>
> So XML via HTTP bloatware...

I don't think Arne thought it is (or mentioned) "bloatware".
And WASD definitely isn't "bloatware". Can't talk for
the Appache/VMS port, but WASD is definitely not. And I very
much prefer having the same admin tools and config files for
setup of SOAP services as for the other HTTP pages and services.

Anyway...
The point is that if the actual business need is solved, the
customer is happy with the result and my family gets food
on the table as a result, I'm happy also. Nothing else counts.

Jan-Erik.

Richard Maher

unread,
Jul 23, 2009, 6:26:59 AM7/23/09
to
Hi Arne,

"Arne Vajh锟絡" <ar...@vajhoej.dk> wrote in message
news:4a67bee5$0$48233$1472...@news.sunsite.dk...
> Jan-Erik S锟絛erholm wrote:
> > Arne Vajh锟絡 wrote:

And how would you rate your development quantum whem compared to any LAMP
application? Speed of the essence? Hmmm?

No PHP, Perl or Python to overcome Arne's development-time hurdle? How DRY
was it?

Anyway, I for one am glad to hear about the evolution of ADBC "Arne's Data
Base Connectivity protocol"! No more "custom" or "binary format" shenanigans
in Arne-World; no from now on when you say SELECT SALARAY FROM EMPLOYEE:
you'll get: -

<ns: escoteric; xml: insecure minus 1; SessionId: "CopyThis">
<employee charset sweedish case insensitive 'cos no one knows why>
<salary exponential depending on source>
TooLittle.3333333333E+01
</salary>
</employee>
</My WSDL is still barfing what version was it again>

Much better! You should work for HP/VMS!

Imagine opening a file and reading a binary record, or adding two numbers
together, or not having to serialize/de-serialize every time you scratched
your arse - how crap is that eh?

>
> Arne
>

Cheers Richard Maher


Arne Vajhøj

unread,
Jul 23, 2009, 8:56:43 PM7/23/09
to

POX requires you to write your own formatter and parser.

With SOAP you get that part for free.

BTW, there are more overhead in the SOAP part than in the HTTP part
(assuming keep-alive).

> So JSON via a poorly performing session-hijackable piece of shite does not
> offend Arne's sensibilities, yet JSON over a socket is anathema?

JSON is practically always used over HTTP, because it is used in
client side JavaScript.

But I really don't like JSON. Executable code is not a good
wire format.

Arne

Arne Vajhøj

unread,
Jul 23, 2009, 8:59:59 PM7/23/09
to
Jan-Erik S�derholm wrote:
> Arne Vajh�j wrote:
>> Using FastCGI to reduce activation overhead is not exactly new.
>
> No, it's not. But it wasn't used by the the gSOAP/VMS kit.
> There are examples of running as an "module" within
> Apache, and as far as I understand using FastCGI is
> quite similar to that. And I prefer WASD over the Apache
> port.

FastCGI and Apache module are different but solves the same
problem.

Apache modules run code in process, FastCGI is a persistent
separate process used for multiple requests. Both reduce
activation overhead.

>> Hm.
>>
>> Are gSOAP for VMS being done by HP ??
>
> Not up to today.

So none of Richard VMS license money has been used so far.

> But as far as I understand it will be
> (more) officialy supported by HP/VMS soon. gSOAP is in
> VMS roadmap now (FWIW). This is of course an important
> issue for some sites/customers, included the one I'm
> working with.

But will now, when there are real customers wanting it.

Arne

Arne Vajhøj

unread,
Jul 23, 2009, 9:06:34 PM7/23/09
to
Richard Maher wrote:
> "Arne Vajh�j" <ar...@vajhoej.dk> wrote in message
>> And the API centric view of SOAP/HTTP is a lot faster to develop
>> against compared to the protocol centric view of a custom
>> binary format over plain sockets.
>
> And how would you rate your development quantum whem compared to any LAMP
> application? Speed of the essence? Hmmm?
>
> No PHP, Perl or Python to overcome Arne's development-time hurdle? How DRY
> was it?

For complex SOAP work a static typed language is actually more
productive than a dynamic typed language.

> Anyway, I for one am glad to hear about the evolution of ADBC "Arne's Data
> Base Connectivity protocol"! No more "custom" or "binary format" shenanigans
> in Arne-World; no from now on when you say SELECT SALARAY FROM EMPLOYEE:
> you'll get: -
>
> <ns: escoteric; xml: insecure minus 1; SessionId: "CopyThis">
> <employee charset sweedish case insensitive 'cos no one knows why>
> <salary exponential depending on source>
> TooLittle.3333333333E+01
> </salary>
> </employee>
> </My WSDL is still barfing what version was it again>

Are you aware that MS SQLServer, Oracle DB and IBM DB2 all
support returning SQL output as XML ?

:-)

> Much better! You should work for HP/VMS!

Not likely.

> Imagine opening a file and reading a binary record, or adding two numbers
> together, or not having to serialize/de-serialize every time you scratched
> your arse - how crap is that eh?

The best solution is the solution that solves the problem
the cheapest way.

Arne

Richard Maher

unread,
Jul 25, 2009, 2:26:18 AM7/25/09
to
Hi Jan-Erik,

"Jan-Erik S�derholm" <jan-erik....@telia.com> wrote in message

news:_wU9m.10438$U5.1...@newsb.telia.net...


> Arne Vajh�j wrote:
>
> > If the toolkit generates a stub, then calling a web service
> > is as simple as calling a local function/method.
>
> That's how it works, yes.
>
> >
> > Using FastCGI to reduce activation overhead is not exactly new.
>
> No, it's not. But it wasn't used by the the gSOAP/VMS kit.
> There are examples of running as an "module" within
> Apache, and as far as I understand using FastCGI is
> quite similar to that. And I prefer WASD over the Apache
> port.
>
> > Hm.
> >
> > Are gSOAP for VMS being done by HP ??
>
> Not up to today.

I'll leave the veracity of that complete bollocks as a judgement for the
reader.

> But as far as I understand it will be
> (more) officialy supported by HP/VMS soon. gSOAP is in
> VMS roadmap now (FWIW). This is of course an important
> issue for some sites/customers, included the one I'm
> working with.
>
> Jan-Erik.

Regards Richard Maher

Richard Maher

unread,
Jul 25, 2009, 2:27:14 AM7/25/09
to

"Jan-Erik S�derholm" <jan-erik....@telia.com> wrote in message
news:SPW9m.10440$U5.1...@newsb.telia.net...

Richard Maher

unread,
Jul 25, 2009, 2:30:31 AM7/25/09
to
Hi Jan-Erik,

"Jan-Erik S�derholm" <jan-erik....@telia.com> wrote in message
news:SPW9m.10440$U5.1...@newsb.telia.net...

That's the spirit ;-)

It's certainly got us this far; and there is a growing school of thought
that tomorrows are over-rated.

>
> Jan-Erik.

Regards Richard Maher


Richard Maher

unread,
Jul 25, 2009, 3:22:01 AM7/25/09
to
Hi Arne,

"Arne Vajh�j" <ar...@vajhoej.dk> wrote in message

news:4a6906c2$0$48246$1472...@news.sunsite.dk...


>
> POX requires you to write your own formatter and parser.

I'll hazard a guess that there are one or two XML parses already available,


>
> With SOAP you get that part for free.

Or, to put is another way, "With SOAP it's mandatory, inescapable baggage
that blows out your betwork bandwith and saturates both ends of the network
pipe with assemble/disassesmble or deserialize/serialize operations. We say
it's nice 'cos that's the shit we *have* to do".

You really are the spin-doctor from hell. Alistair Campbell has nothing on
you!


>
> BTW, there are more overhead in the SOAP part than in the HTTP part
> (assuming keep-alive).

Yes, sticking your head in a boiling chip-pan is more painfull than just
banging your head up against a wall. Why do either? HTTP is very good at
delivering up web-pages, trying to shoe-horn it into a middleware back-bone
makes about as much sense as SMTP.


>
> > So JSON via a poorly performing session-hijackable piece of shite does
not
> > offend Arne's sensibilities, yet JSON over a socket is anathema?
>
> JSON is practically always used over HTTP, because it is used in
> client side JavaScript.

And "client side Javascript" can't use sockets now - is that your contention
du jour?

Do yourself a favour. Go out there and see what Silverlight is doing with
Sockets and the HTML Bridge. See what Adobe is doing with Flex Sockets and
the FABridge. Try to let go of your poorly-performing, insecure,
context-devoid, HTTP security blanket.


>
> But I really don't like JSON. Executable code is not a good
> wire format.

I certainly don't like it, but Yahoo seem to be big fans and there's always
stringify() and parse() for the squeemish.

> But you need some special requirements to justify implementing
> your own protocol.

Who says? What absolute and utter bollocks! And what "protocol"?

(Also, I can only presume you're complaining to HP about the money recently
wasted on the RTR and ACMS protocols then?)

With sockets whatever *you* choose to send up and down the line is your
business. No restrictions, no degradation.

Binary? Same endian? ASCII? CSV? XML? JSON? HTTP? RTMP? ODBC? The world is
yours!

Still maybe you work on a lot of applications where the data types, lengths,
and precision can all change at the drop of a hat and supporting numerous
WSDL and Server endpoints is common place? I hope that one day you get to
write a SOAP client to a something like Amadeus' e-ticketing system and they
ask you to cater for infants :-( Honestly, the amount of client-based
hard-wiring and ideosyncratic business-exception handling is a revelation
(at least to me!). Then turn trace on and and sit back in awe at how much
crap is going up and down the line!

But this is where they make additional money with the consultancy services
to write the client for you. Sounds like a business plan - just make it so
bloody difficult that they'll pay you to develope the intercae to your
systems :-(
>
> Arne

Regards Richard Maher


Richard Maher

unread,
Jul 25, 2009, 11:54:10 PM7/25/09
to
Hi Arne,

"Arne Vajh�j" <ar...@vajhoej.dk> wrote in message

news:4a690911$0$48234$1472...@news.sunsite.dk...


> Richard Maher wrote:
> > "Arne Vajh�j" <ar...@vajhoej.dk> wrote in message
> >> And the API centric view of SOAP/HTTP is a lot faster to develop
> >> against compared to the protocol centric view of a custom
> >> binary format over plain sockets.
> >
> > And how would you rate your development quantum whem compared to any
LAMP
> > application? Speed of the essence? Hmmm?
> >
> > No PHP, Perl or Python to overcome Arne's development-time hurdle? How
DRY
> > was it?
>
> For complex SOAP work a static typed language is actually more
> productive than a dynamic typed language.

Certainly can't argue with those statistics, case studies, and empirical
evidence.

But I too am a big fan of the anecdotal so - "No it isn't".

>
> > Anyway, I for one am glad to hear about the evolution of ADBC "Arne's
Data
> > Base Connectivity protocol"! No more "custom" or "binary format"
shenanigans
> > in Arne-World; no from now on when you say SELECT SALARAY FROM EMPLOYEE:
> > you'll get: -
> >
> > <ns: escoteric; xml: insecure minus 1; SessionId: "CopyThis">
> > <employee charset sweedish case insensitive 'cos no one knows
why>
> > <salary exponential depending on source>
> > TooLittle.3333333333E+01
> > </salary>
> > </employee>
> > </My WSDL is still barfing what version was it again>
>
> Are you aware that MS SQLServer, Oracle DB and IBM DB2 all
> support returning SQL output as XML ?
>
> :-)
>
> > Much better! You should work for HP/VMS!
>
> Not likely.

Never say "never" Arne! If HP decides to support every piece of dodgy
freeware that their employees have used as a pretence to manufacture a job
for themselves (in their "freetime" of course) then surely there'll be a
huge recruiting drive to back-fill for all of the other critical, billable,
customer-essential, work they've been performing during "business hours".
N'est-ce pas?

And *don't forget*, taking someone else's code and compiling it is much more
difficult than writing an operating system. These sub-continental types may
be clever enough to support the image-activator and the rest of the VMS
kernel, but getting the compiler VMS qualifiers right on Linux-freeware is a
tricky business.


>
> > Imagine opening a file and reading a binary record, or adding two
numbers
> > together, or not having to serialize/de-serialize every time you
scratched
> > your arse - how crap is that eh?
>
> The best solution is the solution that solves the problem
> the cheapest way.

Fabulous! So you'd be recommending replacing all VMS with LAMP servers as
quickily as possible then? Hold on, you're a big SaaS fan IIRC; perhaps we
should shift all our IT Infrastructure to the cloud and really cut our
costs? You should bring up the option in the London Stock-Exchange thread,
look how much they'd save!
>
> Arne

Regards Richard Maher


0 new messages