| There is definitely some advanced developments work in progress from TBG in order
to have a generic ASCII AM.
Various alternatives are currently being evaluated :
1-a 100% home-developed tool based on the concepts of the current Data
collector, but generating OSI alarms against user-defined object classes.
The parsing functions will take advantage of LEX and YACC.
2-Integration of NetExpert from OSI into DECmcc/TeMIP. There is already
a prototype working.
3-Integration of tools developed by other third parties. Negociations are
in progress with a UK-based software house, familiar with DECmcc
environment. Another company will probably been approached soon.
TBG produced a "product requirements" document as well defining the scope of the
current investigaton.
JJ
|
|
This is a very important and critical application. But you probably
won't get very much higher level support for it. Mainly because it's
an architectural abomination. ASCII protocol is just not cool (from an
architectural viewpoint). So this is something that someone will just
have to do, and then ask for forgiveness. I have worked with the
Galileo parser before, which takes BNF input, and generates the source
code (in C) for a parser, and I think it is perfect for the job. In
this case, the tool would end up generating both the ASCII parsing AM
and the MS file for the specified BNF input file. With the TBG
advanced developement work already layed out, I don't think it's very
much more work. But, with the downsizing going on, I would be
surprised if it gets done anytime soon.
-Matt.
|
| > Galileo parser before, which takes BNF input, and generates the source
Do you really believe that you can describe the ASCII "user
interfaces" of today's boxes in BNF? I've seen many of them and
almost all look like inconsistent hackery which is extremely
difficult to parse. The command input usually has kind of a
syntax but the output surely has not. And it is mainly this
output that you are interested in...
Just my 2�
*Robert
|