To Everyone:
I agree with Stephen's reading of the Spec that no column (typ, min, max) of
any table (V/I tables or Waveform tables) can begin or end with NA's unless
all entries of that column are permitted to contain all NA entries (this
excludes the typ column) and actually contain all NA entries.
This was NOT what I meant because such a rule is in violation of the
existing IBIS_CHK parser operation for Version 1.1. It did not check for this
rule for the min and max columns of V/I Tables.
However, reasons for the rule apply equally well for V/I min and max columns. I
do not know of any IBIS models that currently violate the rule. Therefore I
recommend that the IBISCHK2 parser retain the check for IBIS compliance, but be
corrected so that the min and max columns can contain all NA entries.
Bob Ross
Interconnectix, Inc.
> Hello All:
> My comments are preceded by >>>
> >From Kellee
> To Stephen Peters:
> You asked is the whole column NA, Answer is YES the whole columns are NA.
> I believe this is valid, and Bob's reply is correct. This was not meant to
> apply to V/I tables.
> >>> In my reading of the spec the 'can't start/end a column with NA' rule
> >>> applies to BOTH the V/I tables and the waveform tables. Bob, is that
> >>> what you meant?
> I believe the parser needs to be fixed!
> >>> Yes, the parser needs to be fixed, but the fix involves allowing NA
> >>> to start or end a column if the whole column is NA. The fix is NOT
> >>> to stop checking the V/I tables.
> Stephen Peters
> Intel Corp.
> Have a good one
> Kellee...
Received on Thu Jan 19 14:11:31 1995
This archive was generated by hypermail 2.1.8 : Fri Jun 03 2011 - 09:52:28 PDT