VMS Kermit Beware File ---------------------- VMS Kermit 3.2.075 only works with VMS Version 4.0 and above. VMS Kermit 3.1 or earlier may be used with VMS Version 3. ------------------------------ Date: Friday, 6 Dec 85 07:37:12 PST From: vax135!cornell!uw-beaver!tektronix!teklds!scottt@ucbvax.berkeley.edu Subject: Error notice for VMS Kermit. I am letting a friend ask via this account that works for the City of Portland Oregon in the transportation engineering bureau... Mark L. Millard 8415 SE Stephens St. Portland, Or 97216 Home: (503)253-4545 Work: (503)796-7234 I have run into a problem with VMS kermit V3.1.066 and C3.0.051. I do not know if this is how you want bug reports. Just leave me mail if it is not. Reply to tektronix!teklds!scottt and I will get the reply. It goes like this for machines V (VMS) and M (masscomp): (0) V context: VAX port has alternate typehead so that 9600 baud does not have problems over running the usual 8 character over run area. The buffer is 200 with alarm (^S) point at 64. The port is used to log in on M from V and use it directly. It is also used to transfer data to and from V. At 9600, the effective rate is about 2400 and it takes hours to transfer one EMME2 (transportation flow model) database. I do not want to use a slower line. The problem only occurs using kermit on the VAX as a virtual terminal. No problems have occured with kermit running on both machines (file transfers). (1) M eventually sends sufficient text to fill V's typehead buffer past the alarm (^S) point so V sends ^S. (2) Kermit on V does not empty the typeahead buffer, instead it leaves some character(s) in the buffer and waits for some condition to start its read. Now M is waiting for V while V is waiting for M because... (3) VMS does not send ^Q to allow M to continue until the typeahead buffer is empty and... (4) Kermit waits at LEF for something to come from M but nothing comes because M is waiting for a ^Q before continuing. (5) A way to force this to occur is to turn on Read_Sync on the port before starting Kermit (Kermit leaves this characteristic alone). Then Kermit gets stuck immeadiately. (6) I have to stop/id the process that got stuck in Kermit. When I rerun Kermit and connect, the output starts where it left off when Kermit hung. In other words, M was waiting to output and running Kermit on V caused V to send M the required ^Q. (7) Kermit on V does not respond to typing on V's side when it is stuck waiting for M. This makes me expect that the wait is specificly for an event from M's side. M_M ------------------------------ Date: Fri, 13-DEC-1985 15:48 MST From: Eric McQueen Subject: VMS Kermit-32 version 3.1.066 bugs and suggestions (in decreasing order of importance): In the filename exchange, this Kermit always quotes multiple characters in the filename with "~" despite the setting of the REPEAT_QUOTE character. Some versions of Kermit do not support quoting of repeated characters in the filename exchange so files with repeated characters in the name must be renamed before being transfered. When receiving a FIXED file, if the final record is not exactly 512 characters, then the record should be padded with nulls and written and a warning issued. Currently an error occurs and the receiving Kermit croaks. It is impossible to get lowercase letters in the SET PROMPT command. Have you considered using the CLI$DCL_PARSE system service to handle command-line parsing? This would also allow a user to choose what base s/he wished to use when entering numeric values (%X, %O, and %D), unfortunately the default would have to be decimal (versus the current octal for some commands). A VERSION command could be added. Thanks much, Bitnet: ERICMC@USU Eric McQueen Office of Computer Services Utah State University Logan, Utah 84322-3700 (801) 750-2414 ------------------------------ Date: Tue, 28 Jan 86 20:48:37 est From: biomed!simon%wjh12.UUCP@harvard.HARVARD.EDU (Simon Rosenthal) Subject: problem with issuiung remote host commands from VMS kermit I am having some trouble with VMS Kermit (3.1.066), running under VMS 3.7. When using it in local mode to talk to C-Kermit running in server mode on a Masscomp system - (basically SYS V Unix), VMS Kermit insists on uppercasing the text of any command sent to the server using REMOTE HOST, according to the debugging log. Needless to say, the Unix end doesn't like this at all ... Is there a Kermit-32 command I'm missing which will force VMS Kermit to send out exactly what I type, or is this a misfeature I'll have to live with ? Simon Rosenthal ------------------------------ 9-Apr-86 14:34:15-EST,1913;000000000001 Return-Path: Received: from MARLBORO.DEC.COM by CU20B.COLUMBIA.EDU with TCP; Wed 9 Apr 86 14:34:11-EST Date: 9 Apr 1986 1417-EST From: B.Eiben LCG Ext 617-467-4431 To: sy.fdc@cu20b Subject: ["BERNIE DTN:297-5664 , AT&T:617-467-5664 LDP WORKSTATIONS" : forward] Message-ID: <"MS11(5146)+GLXLIB0(4)-4" 12197508925.15.249.94566 at MARLBORO.DEC.COM> I'm aware of a 'bug-fix' version of Kermit from Stevens - although it didn't come here [Brian Nelson got it] - want me to get the files over here, so that You can get them to Stevens ?? Rgds, Bernie. - - - - - - - Begin message from: "BERNIE DTN:297-5664 , AT&T:617-467-5664 LDP WORKSTATIONS" Sender: EIBEN@LSMVAX Date: 9 Apr 1986 1415-EST From: "BERNIE DTN:297-5664 , AT&T:617-467-5664 LDP WORKSTATIONS" To: EIBEN@MARKET Subject: forward Mailed to: MARKET::EIBEN From: GUIDUK::STEBBINS "Gary Stebbins @SEO,206/462-2419 08-Apr-1986 1406" 8-APR-1986 17:16 To: LSMVAX::EIBEN,STEBBINS Subj: Fixed bug in VMS Kermit Bernie, I have copied the files VMSTRM.BLI and VMSMIT.EXE to UPLOAD: on PAR5::. Kermit was not setting the TT2$M_PASTHRU bit on the terminal used fortransfer and connect. This was causing problems if control characters were sent by the other system. The biggest place I saw it was when using programs that displayed VT100 line-drawing set. The Shift-in/Shift-out characters are ^O and ^N. ^O sent caused the line to echo *OUTPUT OFF*. This usually triggered all kinds of junk flowing back and forth between the machines... VMSMIT.EXE is a new Kermit with this fix, and version number 3.1.066a ("a" so I would know for sure I was testing the right one...). Please pass the file(s) on to Columbia, if appropriate. thanks -gary- -------- - - - - - - - End forwarded message --------