NONNUMLNG
Richard Maher
Really? 320? In 2009? - Why???
Regards Richard Maher
jbriggs444
wrote:
> %COBOL-E-NONNUMLNG, Nonnumeric literal longer than 320 characters
>
> Really? 320? In 2009? - Why???
>
> Regards Richard Maher
Non-numeric... That could include a quoted text string.
So, got any unterminated quoted strings?
Richard Maher
> So, got any unterminated quoted strings?
No, but I definitely have/had a string literal over 320 bytes. (With
continuation hyphens over multiple lines.)
For our upcoming Tier3Client Example I have a web-page with a "Selection
Criteria" DIV on the left and an "Employee Details" DIV on the right. Once
an individual Employee is located, my COBOL wants to use a $fao control
string containing a plethora of HTML tags and FAO lexicals directives
"<table><tr><td>Name:</td><td>!AS</td></tr></table> and so on. . .". The
result is then injected into the destination div with "detailsDiv.innerHTML
= myBloodyBigString;".
As with Perl (or whatever) it's nice to have a free hand to edit and format
a page/div as a String. I don't want to break it up into 100 lines and then
measure each segment for a PICture clause every time I make a change. I had
PIC X(somthingTooBig) and then before the SYS$FAO I was gonna STR$TRIM it.
Anyway, it looks like it can't be done but I wanted to go on the record that
320 is an absolute joke of a limit. Are you implying that limit does not
exist and that I have done something else wrong?
Regards Richard Maher
"jbriggs444" <jbrig...@gmail.com> wrote in message
news:98171c4b-6354-4c33...@u16g2000pru.googlegroups.com...
Jeffrey H. Coffield
Richard Maher wrote:
> For our upcoming Tier3Client Example I have a web-page with a "Selection
> Criteria" DIV on the left and an "Employee Details" DIV on the right. Once
> an individual Employee is located, my COBOL wants to use a $fao control
> string containing a plethora of HTML tags and FAO lexicals directives
> "<table><tr><td>Name:</td><td>!AS</td></tr></table> and so on. . .". The
> result is then injected into the destination div with "detailsDiv.innerHTML
> = myBloodyBigString;".
>
This is where I would use an Ajax call (from JQuery) to get the XML data
and let the client side build the table. This appears to be a case of
having the server need to know so much about the html. Take a look at
www.keepmeuptodate.net for examples of building HTML tables from XML data.
Jeff Coffield
www.digitalsynergyinc.com
www.keepmeuptodate.net
Richard Maher
"Jeffrey H. Coffield" <jef...@digitalsynergyinc.com> wrote in message
news:hef6nv$jh1$1...@news.eternal-september.org...
Don't know how much effort I should expend on rebutting someone who seems
incapable of stringing an Ajax call together without 120KB of bloatware but,
for the greater good, here goes. (And judging by the recent discussions in
comp.lang.javascript they're now desperate to bring in some sort of
lazy-loading in to reduce the footprint and load-overhead of yet more jQuery
bollocks.)
Tier3 is all about choices and user-impowerment. If they want to send binary
data down the line then thats fine with me; CSV data? No worries; XML
Tag-Soup? More power to ya! What my Employee Lookup "example" seeks to do is
showcase just some of the options available to the application developer.
[A]synchronous sends, generic auto-suggest functionality, cookieless SSO
across any number of concurrent tabs, and, yes, HTML tag injection.
Personally, I would use the technique sparingly and much prefer the
size-delimited binary throughput that only a full-blown (as opposed to
half-pregnant HTML5 web-sockets) TCP/IP Sockets can provide. Having said
that I certainly don't believe I am in a position to lecture to every system
designer without at least knowing some of their business requirements. The
innerHTML tag injection can be a very powerful tool in the box and the fact
that one programmer has chosen to paint themselves into an XML-only corner
is nether here nor there.
But let's talk apple 'n' apples. My example is an employee lookup on the Rdb
MF_PERSONNEL database. If your willing to host it I'm happy to put it up on
your site so people can compare it like for like with your lovely "I would
use an Ajax call (from JQuery) to get the XML data" example. What say you
Jeff? Probably early January before we're shipping the kit so that's gotta
be more than enough time for the killer combo of Jeff & jQuery.
BTW. I have a second little screen-saverish web-page on the drawing-board
that will just randomly pick an employee and put it up somewhere in a
floating div. This is purely to show the bytes/sent & rcvd counters ticking
over in the AWT console/dialog box while you've got another tab on top in
the browser. (You can skip that one)
Having said that, what are you recommending for single sign-on? What's the
current codepath from user-click to data retrieval? How many server
processes per users? How are you controlling server-affinity and ensuring
that the server processes has only those rights and privileges as the
requesting client?
Anyway I'm really, really looking forward to the performance comparisons and
metrics! How 'bout you? The KB transfer requirement will also be revealing
(and I've got an Applet!)
>
> Jeff Coffield
Regards Richard Maher
PS. Good luck with jQuery and IE8!
Richard Maher
For those VMSers who are finally beginning to tear themselves away from
DEC's wonderful FIMS-compliant IFDL, and are looking at dipping a toe into
Javascript, may I recommend www.w3schools.com or specifically
http://www.w3schools.com/ajax/default.asp if you're confused about Ajax. One
problem I had with it was pulling stuff out of the XML DOM with extremely
verbose referencing like: -
xmlDoc.getElementsByTagName("compname")[0].childNodes[0].nodeValue;
To get around that I went looking for a DOM walker that would deliver a
Javascript-object version of the XML and found this site
http://goessner.net/download/prj/jsonxml/ which is about the best
XML-to-JSON solution I could locate. Does anyone have a better example that
still uses the browser to parse the XML?
Here's a little potential use example: -
<script type="text/javascript" src="file://c:/xml2json.js"></script>
<script type="text/javascript">
text2='<?xml version=\"1.0\" encoding=\"utf-8\"?><BLAH
xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\"
xmlns:xsd=\"http://www.w3.org/2001/XMLSchema\"
xmlns=\"http://tempuri.org/\"> <FirstName>Fred</FirstName>
<LastName>Nurks</LastName></BLAH>';
// Or use whatever TAG-soup comes down the line from Ajax and your server
var json;
if (window.DOMParser)
{
parser=new DOMParser();
xmlDoc=parser.parseFromString(text2,"text/xml");
}
else // Internet Explorer
{
xmlDoc=new ActiveXObject("Microsoft.XMLDOM");
xmlDoc.async="false";
xmlDoc.loadXML(text2);
}
json = xml2json(xmlDoc," ");
alert(json);
eval("var obj1 = " + json);
</script>
We found it useful with JSON callbacks when overcoming the browser's
cross-domain restrictions. You'll also find this useful if your application
supplier has you accessing their "cloud" to performance-poor servers in
America and you want field-level validation/lookup direct to your local
servers/data.
> PS. Good luck with jQuery and IE8!
Actually, I just tried you ironically titled "keepmeuptodate" site with IE8
and, as expected, it doesn't work. Don't worry, just tell your customers
that it's jQuery's fault and nothing to do with you. I'm sure one of the
thousand or so contributors to the source will get around to fixing it
within SLA, eventually, until the next time :-(
Regards Richard Maher
PS. Did I happen to mention that with Tier3 you can also send JSON down the
line? (In addition to XML, Binary, Comma-Delimited, Size-Delimited, and blob
formatted data)
Did I also mention that server-affinity could be maintained across multiple
message exchanges on a case by case basis, or that the server can return
multiple message responses (eg: result-set) as and when they are created
rather than delaying the transmission until they can all be packaged up in
some monolithic XML-tag fest? Can you see the parallelism opportunities
between client presentation and server manufacture?