Archive-TarGzip
view release on metacpan or search on metacpan
t/Archive/Data/SecsPack.pm view on Meta::CPAN
In other words, you do not see them flying out the door at
the local Barnes and Nobles. In fact, you will not even
find them inside the door.
It very easy to run these non RFC standard protocols over the net.
Out of 64,000 ports, pick a port of opportunity
(hopefully not one of those low RFC preassigned ports)
and configure the equipment
and host to the same IP and port.
Many times the software will allow a remote console that
is watch only.
The watch console may even be a web server on port 80.
If there is a remote soft console, you can
call up or e-mail the equipment manufacturer's engineer in
say Glouster, MA, USA and tell him the IP and port so he can watch
his manchine mangle a cassette of wafers with a potential
retail value of half million dollars.
SEMI E5-94 and their precessors do standardize the endian-ness of
floating point, the packing of nested data, used in many programming
languages, and much, much more.
The endian-ness of SEMI E5-94 is the first MSB byte,
floats sign bit first.
Maybe this is because it makes it easy to spot numbers in a packed data
structure.
The nested data has many performance
advantages over the common SQL culture of viewing and representing
data as tables. The automated fabs of the world make use of SEMI E5-94 nested
data not only for real-time communication (TCP/IP RS-2332 etc)
between machines but also for snail-time processing as such things as logs
and performance data.
Does this standard communications protocol ensure that
everything goes smoothly without any glitches with this wild
mixture of hardware and software talking to each other
in real time?
Of course not. Bytes get reverse. Data gets jumbled from
point A to point B. Machine time to test software is non-existance.
Big ticket, multi-million dollar fab equipment has to
work to earn its keep. And, then there is the everyday
business of suiting up, with humblizing hair nets,
going through air and other
showers with your favorite or not so favorite co-worker
just to get into the clean room.
And make sure not to do anything that will scatch a wafer
with a lot of Intel Pentiums on them.
It is totally amazing that the product does
get out the door.
=head2 SECSII Format
The L<Data::SecsPack|Data::SecsPack> suroutines
packs and unpacks numbers in accordance with
L<SEMI|http:E<sol>E<sol>www.semi.org> E5-94,
Semiconductor Equipment Communications Standard 2 (SECS-II),
avaiable from
Semiconductor Equipment and Materials International
805 East Middlefield Road,
Mountain View, CA 94043-4080 USA
(415) 964-5111
Easylink: 62819945
http://www.semi.org
The format of SEMI E5-94 numbers are established
by below Table 1.
Table 1 Item Format Codes
unpacked binary octal hex description
---------------------------------------------------------
T 001001 11 0x24 Boolean
S8 011000 30 0x60 8-byte integer (signed)
S1 011001 31 0x62 1-byte integer (signed)
S2 011010 32 0x64 2-byte integer (signed)
S4 011100 34 0x70 4-byte integer (signed)
F8 100000 40 0x80 8-byte floating
F4 100100 44 0x90 4-byte floating
U8 101000 50 0xA0 8-byte integer (unsigned)
U1 101001 51 0xA4 1-byte integer (unsigned)
U2 101010 52 0xA8 2-byte integer (unsigned)
U4 101100 54 0xB0 4-byte integer (unsigned)
Table 1 complies to SEMI E5-94 Table 1, p.94, with an unpack text
symbol and hex columns added. The hex column is the upper
Most Significant Bits (MSB) 6 bits
of the format code in the SEMI E5-94 item header (IH)
In accordance with SEMI E5-94 6.2.2,
=over 4
=item 1
the Most Significat Byte
(MSB) of numbers for formats S2, S4, S8, U2, U4, U8 is
sent first
=item 2
the signed bit for formats F4 and F8 are
sent first.
=item 3
Signed integer formats S1, S2, S4, S8 are two's complement
=back
The memory layout for Data::SecsPack is the SEMI E5-94
"byte sent first" has the lowest memory address.
=head2 IEEE 754-1985 Standard
The SEMI E5-94 F4 format complies to IEEE 754-1985 float and
the F8 format complies to IEEE 754-1985 double.
The IEEE 754-1985 standard is available from:
IEEE Service Center
445 Hoe Lane,
( run in 1.875 second using v1.01-cache-2.11-cpan-39bf76dae61 )