To all the denizens of the list who are now thoroughly confused over what this 
thread might have to do with what they understand as "USS", we actually 
happen to be talking here about the real and original "USS", namely VTAM's 
Unformatted System Services which has been around since the mid-seventies 
rather than what is usually implied by "USS" in this list, namely "UNIX System 
Services", very much a Johnny-come-lately.

I felt that was worth an airing given how infrequently we encounter a thread 
dealing with genuine "USS".

-

Yan Ying

Exploiting the 3270 data stream, which is what you are really asking about, 
can be a rather complex topic. It needs a quite detailed understanding of how 
the 3270 data stream works.

The first point to get out of the way is that the format of the USS messages 
about which you are asking is the 3270 format. I see that you include the 
@@@@@@@@@IPADDR text which implies that you are using the USS table 
in the context of the TN3270 server rather than purely VTAM. You have the 
opportunity these days to use USS messages in 3270 data stream format or 
SNA Character String (SCS)[1] format. It's only when you use the 3270 format 
that you can play with colours.[2]

I guess the second point to get out of the way is that the details of the 
formatting characters has very little - not quite nothing - to do with VTAM or 
the TN3270 server. Thus you will look in vain in Communications Server 
(CS) documentation for an explanation of how to enhance the presentation of 
your USS messages.

If you are very persistent in reading through the CS IP Configuration Guide 
manual - I used the V1R9 edition, SC31-8775-12 - you can find a reference to 
the manual John Chase suggested, namely the "3270 Data Stream 
Programmer’s Reference" manual. The following hierarchy of section titles 
takes you to the relevant text:

- Chapter 10. Accessing remote hosts using Telnet

- TN3270E Telnet server

- Using the Telnet Solicitor or USS logon panel

- Using the Telnet USS and INTERPRET support

- Creating a USS table:

<quote>

- Both the 3270 data stream and the SNA character stream (SCS) formats are 
supported. For more information, see 3270 Data Stream Programmer’s 
Reference and the table samples.

</quote>

-

Now a short introduction to what you want to do:

First of all, given the USS message 10 you put in your post, you should 
already be seeing colours in what is shown on your presumed TN3270 
emulator window. The following is the relationship between colours and 3270 
data stream field attributes:

protected normal: blue
unprotected normal: green
protected intense: white
unprotected intense: red

Thus "IBM z/OS SYSTEM V1R9", "IP Address    = @@@@@@@@@IPADDR", 
VTAM Terminal = @@LUNAME" and "STARTED DATE : 2008-03-27" will appear 
in white since the attribute byte which applies, positioned at row 24, column 
80, - after wrapping from the last character to the first character - 
is "protected intense". (I haven't checked X'C8' is indeed "protected intense" 
since I am relying on the comments.)

The only other attribute byte you have is the one positioned at row 23 column 
80 which is "unprotected normal" and so the command text which you key will 
appear in green.

Had I been faced with what you want to do using real 3270 display terminals 
and associated controllers, say, 20 years ago, the only enhancement 
regarding the use of colour I would feel was appropriate might be to set all 
the 
unchanging text to "protected normal", hence blue, and arrange for the 
variable text to be "protected intense", hence still white so that it stood 
out. 
Setting up some text in an "unprotected" field just so that it would appear in 
red is not a source of possible confusion I would impose on my end users.

The reason I would not be any more adventurous is the reason given by John 
Chase in that I would not have been certain that the combination of display 
and controller in the case of a CUT display or the display itself in the case 
of a 
DFT display supported anything other than the basic attribute bytes as 
described above.

Today with clever TN3270 client programs running on PCs and the like I would 
expect that you could step up to the next level of colour exploitation and use 
the extended attribute bytes and hence use any of the colours, red, green, 
blue, turquoise, pink, yellow and white, to your heart's content. Your 
organisation probably has a standard for TN3270 client and so you can 
probably simply check your USS messages with your own PC to be sure that 
the messages will appear correctly on any of your organisation's PCs.

It's now that I'm afraid I have to contradict Pat O'Keefe. As I, in effect, 
just 
said, when I exploited colours in my USS messages long, long ago, I limited 
myself to the colours which translate to the basic attributes. However, I had 
a colleague running another set of classes who was somewhat bolder. His USS 
messages - probably just his USS message 10 - was set up to exploit the full 
range of 7 colours. Thus, not only would I expect use of the extended 
attributes to be possible on theoretical grounds, I've seen it!

-

Incidentally, I do like your way of setting the buffer position number in 
your "set buffer address" sequences. This is so much better than having to 
look up all the addresses in the "3270 Buffer Address Codes" booklet. I believe 
it relies upon a more advanced option than was available to me with the 
devices I worked with long ago but it is surely valid with the 3270 emulators 
of 
today.

-

It's a shame that the TN3270E RFC, 1647, is deficient in not promoting a 
proper way of supporting the SSCP-LU session using the "SYSREQ Function". If 
it did I could offer you a way of much improving your help desk support based 
on suggesting that you add USS message 5 - at least - to your USS message 
10.

-

Another point you should watch out for is in the use of the "@@LUNAME" text 
substitution. According to the manual this text will not be substituted in the 
case of TN3270 as opposed to a TN3270E logic because of the difference in 
when the secondary APPL statement name is selected between the two cases.

-

I said earlier that "details of the formatting characters has very little - not 
quite nothing - to do with VTAM or indeed the TN3270 server". I found the 
following in the description of the USSTCP statement in the CS IP 
Configuration Reference:

<quote>

Telnet does not modify or translate the message text.

</quote>

This is (should be) almost correct. When you use the USSMSG TEXT operand 
as opposed to the BUFFER operand for specifying your USS messages - 
something that used to be necessary in order to benefit from variable 
substitution - the OPT operand applies. By default the OPT operand removes 
what it imagines are extraneous blank characters. When using 3270 data 
stream formatting of the message, buffer position 0, row 1, column 1, using 
the original method for the "set buffer address" sequence, was represented as 
two blanks. If the TN3270 server were faithfully to follow the rules of USS 
message handling it would have been obliged to reduce these two blank 
characters to one - and thoroughly mess up the formatting of the USS 
message! Scars like that are not forgotten!

Talking about extraneous blank characters, I see that you do have rather a lot 
of extraneous blank characters in your text. You might like to be more precise 
with your "set buffer address" sequences in order to reduce them. You do not 
always have to start in column 1 of the specified rows.

-

If you decide to be very colourful and you need more help, please post again. 
I used to know how to set up SFE and SA sequences so perhaps rereading the 
manual will bring it all back!

Chris Mason

[1] SCS can also be found expanded to SNA Character Set, the SNA Formats 
manual for example, or SNA Character Stream, the Communications Server IP 
Configuration Reference manual for example. The 3174 Establishment 
Controller Functional Description, GA23-0218-11, expands SCS to SNA 
Character String and so that seems the most appropriate interpretation given 
the context.

[2] Being a custodian of the "pure" language, that is, having as little as 
possible to do with the mangled version used across the Atlantic, you will note 
that I spell the word relating to the key feature of this thread correctly!

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to