Hi,
yeasterday I finished a version of SEASON7 that uses a (Phoenix)-activated
Original Skycard instead of the (again) defunct decode() - function.

It uses one COM-Port to communicate with the decoder (eg. COM2) and the other
Comport (eg. COM1) to talk to the original card to retrieve a valid 78 result.

Start it with "season7 c<n> <oldoptions>" where <n> is the COM-Port a original
09 card is connected to. <n> has to be different from the COM-Port that is
used to talk to the decoder (not checked).
Note that with normal PC enviroment COM3 and COM1 (also COM4 and COM2) can
not be used together, since they share the same interrupt. (IRQ 4 and 3 resp.)
If you use IRQ 5 or IRQ 9 (alias IRQ 2) you will have to modify the *.asm
files.

The card is shielded from all unneccassary card traffic; only those 74 messages
that will be followed by a 78 messages are passed to the card. (First Byte
of 74 has bit 3 set to 1), no other commands.

Also, the card adressing scheme currently (August 1st 94) used is analysed,
and the card shielded from those 74 messages containing the cards packet-ID
(byte 8-10). This is a little too careful, perhaps, since the 74 message does
not neccassarily contain the final ID-byte in the byte 11 - 26 range).
If the packet-ID is recognized, the remaining 74, that has not yet been
tranfered to the card is set to 00s, destroying the 74 message/signature.

The recognition of the card's packet-ID will result in a ON-SCREEN-Message and
2.5 to 7.5 seconds of scrambled picture, which will clear subsequently.

Currently, this procedure would not really required, since only those 74 msg
seem to contain authorization information, that are NOT followed by a 78
message. These are not sent to the card anyway.
It is unclear though, if this will remain the case. Also funny, that the E8
messages still contain the packet-ID, but no backlist-info (so am told).

The current code is a quick-and-dirty solution and should be replaced as
soon as the author decides to give SEASON7 a general overhaul/extension etc.
I found it rather hard to put my code in the right places; somehow SEASON7
has been growing extensivly in it's features (DDT-support, ecomomy mode etc.)
and like any program that multiplies in size, it becomes harder for
any non-programmers to make extensions.
But several people asked me for it, so it is.

I took quite a number of indecent shortcuts to achieve my goal:
- duplicated async.asm into async1.asm to achieve dual COM-support. Real
  ugly. I would offer to modify async.asm to address more than one COM,
  if no one else has done it yet. async.asm is used by season7.c, async1.asm
  uses slightly diffrent labels and is used by usecard.c

- I have had to modify the command()-function, that communictes with the card
  to allow reception of data from the decoder at the same time. Unfortunately
  the 74 command to the card has to be started, while the PC is still
  receiving data from the decoder. Otherwise the decoder times-out.
  command_mt now returns on all possible occasions to the caller and indicates
  thru a "stage"-parameter if it is finished or where it left off.

- Modified the DDT-"o"-command, to allow data-output, even though SEASON7 is
  actively talking to the decoder.

- Modified the economy-mode to remove screen messages (eg. "PACKET KILL").

- Made the byte delay 00, as required by the old 10 Mhz AT, I have dedicated
  to this task. The wa0 parameter somehow was not accepted by the system
  (didn't check why). Use the wa500 or wa1000 parameter as required by your
  faster system. Just barely runs with this antique.

Also I:
- fixed a bug, where vclog would be all 00's, after dumped with the 'l'-key.

- I used the table posted via Markus, that contained all the XOR-values for
  the packet-ID. There was an error in the second row (starting with 06).
  The last byte should be 4c instead of 4a. Noticed it, when backlist addresses
  were "out-of-order".

- designed a "makefile"", that will compile and assemble all required files
  (and link them to a season7.exe). Using Microsoft-C simply change to
  the src/ directory and type "nmake". All files should work with Borland C
  but I haven't tried this.

Note that this is not an offical release of SEASON7. I have called it SEA7_13C
to indicate, that the card now takes the 'c' parameter followed by a COM-Port
digit, where it talks to the card.

Hardware is the same as adapter.txt has described it, but I used a seperate 
modul on each the decoder, and card side. This is not neccessarily required,
to prevent switching off, but makes it easier to shield the card from
the decoder and to enable on-the-fly card-reactivations in the future.

Note, that the card has to be previously enabled either by SKY or the
Phoenix-program. This programm does NOT enable your original VC SKY
card!

Again, this is a quick-and-dirty solution and is really below my (anyone's!!!)
programming standards. Sorry, but I have little time, and just wanted to make
this work. I would like this feature to become part of next SEASON7 version.

Gratitude belongs to the author of SEASON7, DECOEM and PHOENIX and the nice
fella, who has sent me a qs-card. Thanks!

BTW, today SKY uses a VERY diffrent encoding on TP 47 (preview channel).
It had no recognizable card addresses sequence, and byte 2 was > 0x80. Hmmm.
No picture descrambled with my SEASON7/C / 09 card combination either.
Could be, that Sky is testing encrypted card addressing schemes for future use.
This could mean, that we can no longer shield the card from kill-commands.
Perhaps no problem, since I have heard reports, that Phoenix can be used
>>10 times per card. Has anybody else "finished off" a card this way?
If re-enabling can be repeated indefinitely, SEASON7/C would simply
re-enable the card automatically, when the card has been disabled by a
74 command. Not sure, this would work, though.
 
