Re: TN3270: potentially incorrect or corrupt datastream

[glen herrmannsfeldt]

Tim Brown wrote:

> My company and I have written a 3270 terminal emulator. Which for the most
> part works. However, we are getting the following in our data stream (HEX):
>
> 11 C9 7B 1D 60 11 4A C4 1D F0 00 02 E8 00 00 00 00 01 32 40
>
> 11 C9 7B : Set buffer address ...
> 1D 60 : Start field (field is protected / alphanumeric ... unmodified
> 11 4A C4 : Set buffer address ...
> 1D F0 : Start field (field is unprotected / numeric ...
>
> 00 02 E8 00 00 00 00 01 32 : Rubbish
>
> 40 : Space character
>
> I believe the segment 00...32 is invalid, and it is causing my emulator to
> lock up with an X PROG (my reaction to an invalid character in the data
> stream).
>
> 1. Can anyone confirm that this segment is no good, or explain to me why
> and how it is good TN3270 data stream code.
>
> 2. Is this the best place to ask the question? Can someone point me out
> to a more appropriate newsgroup or community?

How about comp.protocols.ibm (added)

-- glen

[Don Poitras]

What was the AID received? If it was x'88', then you're receiving structured
fields, in which case you need to parse according to the lengths of each
field. Even assuming that though, the part below doesn't look valid (2 is
not a valid length for a structured field). Are you sure you're looking
at the data sent though and not just a buffer that had old trailing data?

--
Don Poitras - Sas