5-Apr-84 01:00:16-EST,4812;000000000001 Return-Path: Received: from CU20B by CUCS20 with DECnet; 5 Apr 84 00:59:51 EST Date: Thu 5 Apr 84 00:59:16-EST From: Peter G. Trei Subject: Re: Fix for Apple-DOS Kermit... To: info-kermit@CUCS20 I here repost two letters relating to using Kermit under Apple-DOS with the Super-Serial Card. The address specific stuff applies only to the unmodified 'official' version currently distributed. Peter Trei oc.trei%cu20b@columbia-20 ---------------------- People trying to use Kermit-65 (Apple DOS Kermit) Version 1.1 with a Super Serial Card have been running into problems. Here is how to make it work. 1. Before you start up Kermit, send the SSC the following string: ^AZ (thats Control-A, followed by Z). This will disable the SSC's command recognition. The SSC usually looks for ^A in the terminal input, and strips it out. It then looks at the next character, and if it is a valid SSC command, strips it out as well and performs the command. Trouble arises from the fact that Kermit uses ^A to announce the start of each packet. Typing ^AZ disables the SSC from seeing further ^A commands. If you really need to have access to the SSC commands again before you turn off the Apple, type ^A^W instead, which will change the command prefix to ^W, which should not appear during Kermit file transfer. There is a bug in the code to support the Super Serial Card, which must be fixed before it will work at all. If you look in the source code for Kermit-65 (APPLEK.M65 in , and search for the label TL2CP:, two lines further down you will see a line which reads: AND #$04 At this point, Kermit is ANDing a status register with a bitmask. If the result is non-zero, a character has been received from the modem. the problem is that 04 is the wrong mask; it should be 08, according to page 54 of the SSC manual. To fix this, you can either alter the source, recompile, and upload the new version, or much more quickly you can patch the binary version you already have. Here's how to do the patch from Applesoft: ]BLOAD APPLEK.BIN (or whatever you are calling your copy). ]POKE 8665,8 (thats a decimal address) ]BSAVE NEWKER,A$800,L$4900 Thats all. The new version contains the patch. With this, file transfer using the Super Serial Card has been done at 1200 baud. [This bug has been fixed in all the copies I hand out now. ] 3. Those of you who use 1200 baud modems will have noticed that you loose characters at the beginning of each line when the screen is scrolling. This is not Kermits fault, but rather the slowness of the software used to scroll the screen image in the Apples memory. According to the SSC manual, you can eliminate this by slightly narrowing the scroll window. The following poke does it: ]POKE 35,22 This will make line 22 the bottom of your scroll window, which is enough. I would be interested in hearing from anyone on the list who is using Kermit-65. Peter Trei, OC.TREI%CU20B@Columbia-20.Arpa Here is Richard garland's method for using the Videx and the super-serial card. Return-Path: Received: from CUCS20 by CU20B with DECnet; 27 Jan 84 21:07:10 EST Date: Fri 27 Jan 84 21:06:45-EST From: Richard Garland Subject: Apple with SSC & Videx 80 col. card To: info-kermit@CUCS20 Using Peter Trei's suggestions (in yesterday's Info-Kermit) we now have Apple Kermit working nicely with the SSC (Super Serial Card). We have no problem connecting to and doing file transfers with a VAX running VMS and Steven's VMS Kermit at 1200 baud. Mark Paczkowski here has worked out how to get the Videx 80 column card working under Kermit with the SSC. The Videx must go in slot 3. Assume the SSC is in slot 1. The following sequence gets the whole thing going: 1) Boot the Apple 2) Type "IN#1" <== this wakes up the SSC 3) Type "A3S" <== chain SSC to Videx 80 col. card 4) Type "AZ" <== turn off SSC's interception of ^A's 5) Type "PR#3" <== turn on Videx 80 col card 6) Type "BLOAD KERMIT" <== load kermit (patched as per Peter Trei) 7) Type "CALL 7855" <== Start up Kermit Then you are off and running. The 80 col card has faster screen handling and so Peter Trei's suggestion about reducing the scrolling region to 22 lines is unnecessary. The BLOAD is needed rather than the usual BRUN so that the chaining stuff you set up in the previous steps won't get reset. During the above sequence you will get various prompts from the system and from the cards. The screen will do various wierd things but in the end it will all be ok. [Now back to my Rainbow ...] Rg -------