Area 44 ought to be reachable.
-----Original Message-----
From: Mark Wickens <mark at wickensonline.co.uk>
Sender: owner-hecnet at Update.UU.SE
Date: Wed, 20 Jun 2012 14:37:25
To: hecnet at Update.UU.SE<hecnet at Update.UU.SE>
Reply-To: hecnet at Update.UU.SECc: hecnet at Update.UU.SE<hecnet at Update.UU.SE>
Subject: [HECnet] Psilo status?
Is everything ok in hecnet land? I can get out from SLAVE in Area 4 to anywhere else. My bridge to Psilo is up and running. I can see another local node on the same circuit (tiger)
Thanks, Mark
Sent from my iPad
On 20 Jun 2012, at 12:32, Johnny Billquist <bqt at softjar.se> wrote:
On 2012-06-20 12:27, Sampsa Laine wrote:
Does anyone have an outline of how the MAIL-11 protocol works? Toying with the idea of writing a Python library to send MAIL-11 messages using the DECNET-linux stack..
Let me get back to you in a while. I wrote up a document that I think I have on MIM, as I was writing my own MAIL-11 implementation for RSX...
Busy the next three hours, though...
Johnny
Title VMAIL - MMAILR / TOPS-20 Mail-11 Listener
Subttl LCampbell/GStevens/DDyer-Bennet/MDLyons/JHTorrey/PLBudne/MRCrispin
Search Macsym,Monsym
.require Sys:Macrel
Sall
.directive Flblst
.text "/NOINITIAL"
.text "VMAIL/SAVE"
IFNDEF OT%822,OT%822==:1
; This program is derived from the VMAIL distributed on the
; "integration tools tape". It listens on DECnet object 27 for mail
; from the VMS MAIL Utility. When a VMS user says "MAIL NODE::USER" A
; connection is established then and there, and text is sent one line
; at a time. A problem is that this server is single thread, and one
; VMS user can tie up the port, making all other mail to the '20
; fail!!
; At the current time the addressee must be a valid TOPS-20 username,
; since the destination is stored as a user number. I never finished
; an interface to MMAILBOX to vaildate full MMAILR addresses.
; Another interesting problem is that there may be no reply path for
; mail. I had thought of taking the VMAILR program and hacking it to
; be the delivery agent for a #Special domain. At BU we run software
; tools mail on most VMS systems, so this is not a problem.
IFNDEF FTRCVD,FTRCVD==1 ;INCLUDE RECIEVED: HEADER
IFNDEF MBXF,MBXF==0 ;INCOMPLETE!!
IFN MBXF,<
.REQUIRE VMAMBX ;INTERFACE TO MMAILBOX
EXTERNAL MBXFRK,MBXVFY
;MBXF
T1==1
T2==2
T3==3
T4==4
T5==5
P1==6
P2==7
P3==10
P4==11
P5==12
Ptr==13 ;Global Byte Pointer To Receive Mail
Cnt==14 ;Global Byte Count For Same
Cx==16
P==17
.ver==6
.mvr==1
.edt==^D82
Loc 137
Exp <.ver>B11+<.mvr>B17+1B18+.edt
Reloc
Define Jerr(String),<
Xlist
IFJER.
Hrroi T1,[Asciz /VMAIL error: /]
Esout
Hrroi T1,[Asciz /String/]
Psout
Hrroi T1,[Asciz / because: /]
Psout
Movx T1,.priou
Hrloi T2,.fhslf
Erstr
Erjmp .+2
Erjmp .+1
Call Dtstmp ;; Log This Lossage Also
Log <String>
Log < Because: >
Move T1,LogJFN
Hrloi T2,.fhslf
Erstr
Erjmp .+2
Erjmp .+1
Jrst Fatal
ENDIF.
List
Define Log(String),< ;; Put Message Into Log File
Xlist
Hrroi T1,[Asciz \String\] ;; So It Can Type Slashes
Call Logmsg
List
Define Debug(String),<
Skipn Dbugsw
IFSKP.
Hrroi T1,[Asciz /String/]
Psout
ENDIF.
Define Debstr(String),<
Skipn Dbugsw
IFSKP.
Hrroi T1,String
Psout
ENDIF.
Define Clrbuf(bufnam,Buflen),<
Setzm Bufnam
Move T1,[Bufnam,,Bufnam+1]
Blt T1,Bufnam+buflen-1
Define Nchar,<
Move T1,NetJFN
Bin
Move T1,T2
Define Nrecord(Buffer,Nchar),<
Move T1,NetJFN
Hrroi T2,Buffer
Movni T3,Nchar
Setz T4,
Sinr
Jerr <SINR failed at Nrecord>
Define Vaxsuccess,<
Move T1,NetJFN
Hrroi T2,[Ascii//]
Movei T3,-4
Setz T4,
Soutr
Define Vaxerr(Errmsg),<
Move T1,NetJFN
Hrroi T2,[Ascii//]
Movei T3,-4
Setz T4,
Soutr
Move T1,NetJFN
Hrroi T2,[Asciz /Errmsg/]
Setzb T3,T4
Sout
Hrroi T2,Atmbuf
Setzb T3,T4
Sout
Hrroi T2,[Asciz/ At Node /]
Setzb T3,T4
Sout
Hrroi T2,ournam
Setzb T3,T4
Soutr ; force string transmission
Hrroi T2,[0]
Movei T3,-1
Setz T4,
Soutr
Define Die(String),< ;; Fatal Internal Error
Xlist
Jrst [ Hrroi T1,[Asciz /VMAIL fatal internal error: /]
Esout
Hrroi T1,[Asciz /String/]
Psout
Hrroi T1,[Asciz /
/]
Psout
Call Dtstmp ;; Time Stamp It
Hrroi T1,[Asciz /Fatal error: /]
Call Logmsg
Hrroi T1,[Asciz /String/]
Call Logmsg
Jrst Fatal]
List
Define Herald(Ver,Edt),<
Xlist
; Tmsg <VMAIL version Ver(Edt) running>
Hrroi T1,[Asciz /VMAIL version Ver(Edt) running/]
Call Logmsg
List
Define Log(String),<
Xlist
Hrroi T1,[Asciz \String\]
Call Logmsg
List
;Storage
Tmplen==500 ; Temporary Storage
Natmbf==100 ; Length Of Atom Buffer In Words
Bbflen==300000 ;[154] Length Of Big Mail Buffer
Nfrmbf==70 ; Length Of Sender Name Buffer
Timen==^D<10*60*1000> ; Milliseconds Before Sender Declared Tardy
Stklen==200 ; Size Of Stack
Dbugsw: 0 ; -1 If Debug
Atmbuf: Block Natmbf ; Atom Buffer
Subbuf: Block Natmbf ; Subject Buffer
Toocnt: Block 1 ; Count of recipients per line in TOOBUF
Tooptr: Block 1 ; Pointer to current position in TOOBUF
Toobuf: Block Natmbf ; To Names Buffer
Bigbuf: Block Bbflen ; Where It All Is Combined To
Nodstr: Block ^D200 ; String space for recipient nodes/USERS
NODPTR: BLOCK 1 ;[BUDD] BP INTO NODSTR
Ulist: Block ^D200 ; Where To Store Mailbox Directory Numbers
Frmbuf: Block Nfrmbf ; Where To Put Sender'S Name Plus Host
Frmnam: Block Nfrmbf ; Where To Put Sender'S Name
Ournam: Block 2 ; Our Host Name
Hstnam: Block 2 ; Host name we are sending to
Usrnum: Block 1 ; User number we are sending to
MYPID: BLOCK 1 ;[BUDD] OUR IPCF PID
BUFFER: BLOCK 10 ;[BUDD] IPCF BUFFER
IPCBLK: BLOCK 4 ;[BUDD] IPCF DESCR BLOCK
OURJOB: BLOCK 1 ;[BUDD] OUR JOB NUMBER
QUEJFN: BLOCK 1 ;[BUDD] MMAILR QUEUE JFN
Filnam: Block 1 ;[BUDD] File name for MMAILR QUEUE FILE
Temp1: Block Tmplen
Temp2: Block Tmplen
Stack: Block Stklen ; One Stack For Each Fork
NetJFN: Block 1 ; Network File JFN
LogJFN: Block 1 ; Log File JFN
LOGP: BLOCK 1 ; Non-zero if keeping logs
Ntime: Block 1 ; Time Receipt Of Mail Initiated (For Status)
Elptim: Block 1 ; Elapsed Time For Receipt Of Mail
Kepliv: Block 1 ; Keep alive count for dead mans timer
Bytcnt: Block 1 ; Length Of Mail In Bytes
Capenb: Block 1 ; Saved Capabilities
IFN MBXF,< ;[BUDD]
GOTMBX: BLOCK 1 ;[BUDD] -1 IF HAVE MBX FORK
;[BUDD] MBX
Pc1: Block 1 ; Pc Save Location For Psi Level
Pc2: Block 1
Pc3: Block 1
Levtab: Pc1
Pc2
Pc3
Chntab: 2,,Conect ; Connect Initiate On Level 2
1,,Timout ; Timeout Psi On Level 1
Xlist ; Nothing Else
Repeat ^d34,<Exp 0>
List
VMAIL:: Reset
Move P,[-Stklen,,Stack]
GJINF ;[BUDD]
Movem T3,OurJob ;[BUDD]
Movx T1,.ndgln ; Get Local Node Name Function
Move T2,[Point 7,Ournam]
Movem T2,1(P) ; Put Pointer On Stack
Movei T2,1(P) ; And Point To It
Node ; Get Node Name
ifjer.
tmsg <?No DECnet in this monitor>
haltf%
endif.
IFN MBXF,SETZM GOTMBX ;[BUDD] CLEAR MBX AVAIL FLAG
Movx T1,.fhslf ; This Process
Move T2,[Levtab,,Chntab]
Sir ; Init Psi System
Eir
Call Opnlog ; Open Log File
Movem T1,LogJFN ; Save JFN
Call Dtstmp ; Time Stamp It
Herald \.ver,\.edt
Log< on node >
; Tmsg < on node >
; Hrroi T1,Ournam
; Psout
; Tmsg <
;>
Hrroi T1,Ournam
Call Logmsg
VMAIL0: Call Opnlsn ; Open Connection And Set Interrupt Up
Move T1,LogJFN ; Close Log File For Perusers
Closf
Erjmp .+1
Wait ; For Connect Initiate
;Here When Connection Initiated
Conect: Move P,[-Stklen,,Stack] ; Reset Stack
Call Timeit ; Time This Guy
Call Opnlog ; Open Log File
Movem T1,LogJFN
Call Dtstmp ; Time Stamp This Transaction
Log <----Connect from >
Debug <----Connect from >
Call T4nhst ; Type Foreign Host Name At Log File
Clrbuf Subbuf,Natmbf
Move T1,NetJFN ; Accept Connection
Movx T2,.mocc
Setzb T3,T4 ; No Additional Data
Mtopr
Jerr <Couldn't accept net connection>
Conct1: Move T1,NetJFN ; Get network link status
Movei T2,.morls
Mtopr
Jerr <Couldn't get link status>
TXNE T3,MO%ABT ; [154] Has the link been aborted?
JRST DMPLNK ; [154] Yes, get rid of it.
Txne T3,Mo%con ; Skip if link not connected
Jrst Conct2
Movei T1,^D1000 ; Wait a second and try again
Disms
Jrst Conct1
Conct2: Movx T1,.hpelp ; Elapsed Time Since System Startup
Hptim ; Snarf It
Jerr <HPTIM failed>
Movem T1,Ntime ; Remember Time This Reception Started
Call Parse ; Parse The Mail
Jrst Errxit ; Failed, Quit Now
Call Dtstmp ; Time Stamp Log
Log <Message from >
Hrroi T1,Frmbuf ; Sender'S Name
Call Logmsg ; Log It
Log < received >
Call Lstats ; Log Statistics
Call Mailit ; Send The Mail Off
Die <Failure returned from MAILIT>
Errxit: Call Clznet ; Close And Reopen Net Link
Call Cncltm ; Cancel Timeout Request
Call Dtstmp
Log <----Connection closed>
Move T1,LogJFN
Closf ; Close Log File For Perusers
Erjmp .+1
Debrk ; Return To Background
; Parse Mail Received. Place Sender Name In Frmbuf, Recipient Directory
; Numbers In Ulist, Terminated With A Zero Entry
; Headers Must Appear In The Following Order.
; From, To, Cc
; Returns +1: Failure
; +2: Success
;
;PROGRAM FLOW DESCRIPTION NOT ALL ITEMS IN FLOW ARE IN THIS ROUTINE BUT
;IT DOES REPRESENT THE PROCEDURE TO SEND TO A VAX WHICH IS WHY IT IS
;INCLUDED HERE
;
; RECEIVE FROM FIELD FROM VAX
; PARSE FROM FIELD CONVERTING IT TO MS TYPE FIELD IN FRMBUF
; REPEAT UNTIL NULL RECEIVED
; : RECEIVE A RECIPIENT NAME FOR VERIFICATION
; : IF NULL RECEIVED
; : : THEN
; : : : EXIT REPEAT LOOP
; : ENDIF
; : PARSE USER NAME AND NODE
; : IF NODE SAME AS THIS NODE
; : : THEN
; : : : IF USER IS ON THIS SYSTEM
; : : : : THEN
; : : : : : SEND SUCCESS CODE TO VAX
; : : : : : PUT USER NUMBER INTO ULIST
; : : : : ELSE
; : : : : : SEND FAILURE CODE TO VAX
; : : : : : SEND ERROR MESSAGE TO VAX
; : : : : : SEND NULL TERMINATING ERROR MESSAGE TO VAX
; : : : : : RETURN FROM ROUTINE
; : : : ENDIF
; : : ELSE
; : : : SEND SUCCESS TO VAX (MESSAGE WILL BE QUEUED)
; : : : PUT -1 INTO ULIST
; : : : PUT NODE NAME INTO NODLST
; : ENDIF
; END REPEAT
; RECEIVE TO FIELD FROM VAX
; PARSE TO FIELD CONVERTING IT TO MS TYPE FIELD IN TOOBUF
; RECEIVE SUBJECT FIELD FROM VAX
; BEGIN FORMATING MESSAGE INTO MS TYPE MESSAGE
; REPEAT UNTIL NULL RECEIVED
; : RECEIVE A LINE FROM VAX
; : IF NULL RECEIVED
; : : THEN
; : : : EXIT REPEAT LOOP
; : ENDIF
; : OUTPUT TO MS MESSAGE BUFFER
; END REPEAT
; REPEAT UNTIL NULL DETECTED
; : GET FIRST ITEM IN ULIST
; : IF FIRST ITEM IN ULIST = -1
; : : THEN
; : : : GET NODE FROM NODLST
; : : : PUT MAIL INTO FILE FOR DMAILR
; : : : SET FLAG IN DECNET-FLAGS SO FILE GETS SENT
; : : ELSE
; : : : PUT MAIL INTO USERS MAIL FILE
; : : : SPLAT OBNOXIOUS MESSAGE ACCRESS USERS SCREEN
; : ENDIF
; : IF NO ERROR
; : : THEN
; : : : SEND POSITIVE ACKNOWLEDGEMENT TO VAX
; : : ELSE
; : : : SEND NEGATIVE ACKNOWLEDGEMENT TO VAX
; : : : SEND ERROR MESSAGE TO VAX
; : : : SEND NULL TERMINATOR TO VAX
; : ENDIF
; END REPEAT
; RETURN +2
Parse: Clrbuf Frmnam,Nfrmbf
Move T1,NetJFN ; Save It
Movei T2,.morss ; Read Max Record Size
Mtopr
Jerr <Couldn't read max record size> ; This could fail...
Nrecord <Frmnam>,<Nfrmbf*5-1> ; Read From Field
Hrroi T1,Temp1 ; Setup Default Host
Hrroi T2,Hstnam
Setzb T3,T4
Sout
Hrroi T1,Frmbuf ;Parse from field, results to FRMBUF
Move T2,[Point 7,Frmnam]
Call Prsnam
Setzm Toobuf ; Clear first location of TOOBUF
Hrroi T1,Toobuf ; Setup pointer to TOOBUF
Movem T1,Tooptr
Movei T1,3 ; Setup count of recipients per line in TOOBUF
Movem T1,Toocnt
MOVEI T1,NODSTR ;[BUDD]
Movem T1,NodPTR ;[BUDD]
Movsi P1,-^D100 ; Maximum Of 100 Names In List
Parse3: Clrbuf Atmbuf,Natmbf ; Clear receive area
Nrecord <Atmbuf>,<Natmbf*5-1> ; Receive recipient from VAX
Aos Kepliv ; Increment keep alive count
Skipn Atmbuf ; Skip if not end of list
Jrst Parse6 ; End of recipient list
Call Prsusr ; Parse recipient
Jrst [ Call Dtstmp ; None Found, Complain
Vaxerr <%Network mail error: No such user >
Log <%Network mail error: No such user >
Hrroi T1,Atmbuf ; Also Log Losing Name
CallRet Logmsg]
Movem T1,Ulist(P1) ; Save number returned for mailing
Vaxsuccess ; Send VAX the success code
Aobjn P1,Parse3 ; Jump if not too many recipients
Call Dtstmp ; Woops, Too Many
Hrroi T1,Atmbuf ; Also Tell Log File
CallRet Logmsg
Parse6: Setzm Ulist(P1) ; Tie Off Recipient List
;
; Now Get Mailed To Field With Node Name And Subject
;
Clrbuf Atmbuf,Natmbf
Nrecord <Atmbuf>,<Natmbf*5-1>
Nrecord <Subbuf>,<Natmbf*5-1>
; Now Conbine It All Into Bigbuf
Setzm Bytcnt
Hrroi T1,Bigbuf
IFN FTRCVD,< ;[BUDD]
Hrroi T2,[Asciz /Received: from /] ;[BUDD] now, write Received line
Setz T3, ;[BUDD]
SOUT ;[BUDD]
Hrroi T2,HSTNAM ;[BUDD]
SOUT ;[BUDD]
HRROI T2,[ASCIZ ' by '] ;[BUDD]
SOUT ;[BUDD]
HRROI T2,OURNAM ;[BUDD]
SOUT ;[BUDD]
Hrroi T2,[Asciz ' using MAIL-11 '] ;[BUDD]
SOUT ;[BUDD]
Seto T2, ;[BUDD] output current date/time
MovX T3,OT%DAY!OT%SPA!OT%TMZ!OT%SCL!OT%822
ODTIM ;[BUDD] RFC 822 standard date
JERR <ODTIM in recieved line failed> ;[BUDD]
Hrroi T2,CRLF0 ;[BUDD]
Setz T3, ;[BUDD]
SOUT ;[BUDD]
;IFN FTRCVD
Hrroi T2,[Asciz /Date: /]
Setz T3,
Sout
Seto T2,
MovX T3,OT%DAY!OT%SPA!OT%TMZ!OT%SCL!OT%822
Odtim
Jerr <Couldn't get current date-time at PARSE6+n>
Hrroi T2,[Asciz/
From: /]
Setz T3,
Sout
Hrroi T2,Frmbuf
Setz T3,
Sout
Hrroi T2,[Asciz/
To: /]
Setzb T3,T4
Sout
Hrroi T2,Toobuf
Setzb T3,T4
Sout
Hrroi T2,[Asciz/
Subject: /]
Setzb T3,T4
Sout
Hrroi T2,Subbuf
Setzb T3,T4
Sout
Hrroi T2,[Asciz/
Mailed-to: /] ;[BUDD]
Setzb T3,T4
Sout
Hrroi T2,Atmbuf
Setz T3,
Sout
Hrroi T2,CRLF0
SOUT
Hrroi T2,CRLF0
SOUT
Move T5,T1
Repeat: Clrbuf Temp1,Tmplen ; Clear storage area
Nrecord <Temp1>,<Tmplen*5-1> ; Get a message line from the VAX
Aos Kepliv ; Increment keep alive count
Camn T3,[-<tmplen*5-1>] ; Blank line ?
Jrst Crlf ; Yes, output CRLF
Skipn Temp1 ; End of message ?
Jrst Done ; Yes
Move T1,T5 ; Output message line to BIGBUF
Hrroi T2,Temp1
Setzb T3,T4
Sout
Jerr <SOUT failed at REPEAT+a few> ; [151] Add error check
Move T5,T1
Crlf: Move T1,T5 ; Output CRLF to BIGBUF
Hrroi T2,[Asciz/
/]
Setzb T3,T4
Sout
Jerr <SOUT failed at CRLF+3> ;[152] Add error check
Move T5,T1
Jrst Repeat
Done: Move T1,T5 ; Close off message
Hrroi T2,[Asciz /
/]
Setzb T3,T4
Sout
Call Gtbfsz ; Get Buffer Size Into Bigbug
Debug <
Debstr <Bigbuf>
Retskp ; All Done!
;
;Store Size Of Bigbuf Into Bytcnt
;
Gtbfsz: Setz T3,
Setz T2,
Move T1,[Point 7,Bigbuf]
Geta: Ildb T2,T1
Jumpe T2,Go
Addi T3,1
Jrst Geta
Go: Movem T3,Bytcnt
Jrst R
;Routine to parse a node and user name and convert it to a TOPS-20
;string compatable with MS
;
;CALL:
; T1 = STRING POINTER TO THE DESTINATION
; T2 = STRING POINTER TO FIELD RECEIVED
; CALL PRSNAM
;
;VARIABLES RETURNED ON SUCCESSFUL COMPLETION
; DESTINATION CONTAINS THE MS STRING
; TEMP1 = HOST
; TEMP2 = USER NAME
; T1 = UPDATED STRING POINTER
;
;RETURNS:
; +1 ALWAYS
PRSNAM:
ACVAR <SPTR,DPTR> ; STORAGE FOR SOURCE/DEST. POINTERS
MOVEM T1,DPTR ; SAVE DESTINATION POINTER
MOVEM T2,SPTR ; SAVE SOURCE POINTER
PRSNM1: MOVE T1,[POINT 7,TEMP2] ; GET POINTER TO WHERE TO STORE STRING
PRSNM2: ILDB T3,SPTR ; GET CHARACTER
JUMPE T3,PRSNMD ; NULL, DONE
CAIN T3," " ; SPACE SEPARATOR?
JRST PRSNM3 ; YES, LOOK FOR PERSONAL NAME
CAIE T3,":" ; END OF NODE SPECIFIER?
IFSKP.
ILDB T3,SPTR ; YES, EAT NEXT COLON
SETZ T3, ; TERMINATE STRING
IDPB T3,T1
HRROI T1,TEMP1 ; UPDATE ORIGINATING HOST
HRROI T2,TEMP2
SOUT
JRST PRSNM1
ENDIF.
IDPB T3,T1 ; SAVE CHARACTER IN TEMP2
JRST PRSNM2
PRSNM3: SETZ T3, ; TERMINATE STRING
IDPB T3,T1
PRSNM4: ILDB T3,SPTR ; GET CHARACTER FROM INPUT STRING
JUMPE T3,PRSNMD ; STRING EXHAUSTED
CAIE T3,42 ; IS CHARACTER START OF PERSONAL NAME?
JRST PRSNM4 ; NO, LOOP TILL STRING EXHAUSTED OR '"' FOUND
MOVE T1,DPTR ; GET DESTINATION POINTER
MOVE T2,SPTR ; GET SOURCE POINTER
SETZ T3,
MOVEI T4,42 ; TILL '"'
SOUT
MOVEI T2," " ; CHANGE '"' TO SPACE
DPB T2,T1
MOVEI T2,"<" ; OUTPUT START OF USERNAME at NODE STRING
IDPB T2,T1
CALL PRSNMS ; CALL ROUTINE TO OUTPUT USERNAME at NODE
MOVEI T2,">" ; OUTPUT END OF USERNAME at NODE STRING
IDPB T2,T1
SETZ T2, ; TERMINATE THE OUTPUT STRING
IDPB T2,T1
RET ; SUCCESS RETURN
PRSNMD: IDPB T3,T1 ; TERMINATE STRING
MOVE T1,DPTR ; BUILD STRING IN DESTINATION BUFFER
CALLRET PRSNMS
ENDAV.
;Routine to create user name at node string in output buffer.
;
;CALL:
; T1 = STRING POINTER TO WHERE TO STORE THE OUTPUT
; TEMP2 = USER NAME
; TEMP1 = NODE NAME
; CALL PRSNMS
;
;VARIABLES RETURNED:
; T1 = UPDATED STRING POINTER
;
;RETURNS:
; +1 ALWAYS
PRSNMS: HRROI T2,TEMP2 ; OUTPUT USER NAME
SETZB T3,T4
SOUT
;; HRROI T2,[ASCIZ/ at /] ; OUTPUT SEPARATOR
;; SOUT
MOVEI T2,"@"
BOUT
HRROI T2,TEMP1 ; OUTPUT NODE NAME
SOUT
HRROI T2,[asciz/.#DECnet/]
SOUT
RET ; SUCCESS RETURN
;Routine to parse addressing of VAX mail and build TO string in TOOBUF
;
;CALL:
; ATMBUF = ADDRESS STRING RECEIVED FROM THE VAX
; CALL PRSUSR
;
;VARIABLES RETURNED ON SUCCESS:
; TEMP1 = NODE OF RECIPIENT
; TEMP2 = NAME OF RECIPIENT
; T1 = USER NUMBER IF ON CURRENT NODE OR
; PTR TO NODE // '\0' // USER FOR 4N HOST
;
;RETURNS:
; +1: ERROR, MAIL WAS ADDRESSED TO THIS NODE AND USER WAS UNKNOWN
; +2: OK, ALL RETURNED VARIABLES VALID
PRSUSR:
MOVE T1,TOOPTR ; GET POINTER TO TOOBUF
SKIPN TOOBUF ; SKIP IF TOOBUF NOT EMPTY
JRST PRSUS1
SOSE TOOCNT ; SUBTRACT FROM COUNT/LINE - SKIP IF .NE. 0
IFSKP.
HRROI T2,[BYTE (7) ",", "M"-100, "J"-100, "I"-100]
SETZ T3,
SOUT
MOVEI T2,3 ; RESET COUNT OF USERS PER LINE IN TOOBUF
MOVEM T2,TOOCNT
JRST PRSUS1
ENDIF.
HRROI T2,[ASCIZ/, /] ; OUTPUT SEPARATOR
SETZ T3,
SOUT
PRSUS1: MOVEM T1,TOOPTR ; SAVE STRING DESTINATION POINTER
HRROI T1,TEMP1 ; SETUP DEFAULT HOST
HRROI T2,OURNAM
SETZ T3,
SOUT
MOVE T1,TOOPTR ; GET DESTINATION POINTER
MOVE T2,[POINT 7,ATMBUF] ; GET POINTER IN INPUT STRING
CALL PRSNAM ; GET TOPS-20 MS STRING
MOVEM T1,TOOPTR ; SAVE POINTER TO TOOBUF
HRROI T1,OURNAM ; GET POINTER TO THIS SYSTEMS NODE NAME
HRROI T2,TEMP1 ; GET HOST NAME FROM FIELD
STCMP ; IS MESSAGE FOR THIS HOST ?
JUMPE T1,PRSUS2 ; JUMP IF FOR THIS HOST
;MAIL IS FOR A REMOTE HOST. GET HOST,,USER STRING POINTERS
HRRZ T4,NODPTR ;GET POINTER TO STRING SPACE
HRRO T1,T4 ;PUT NODE NAME INTO NODE STRING SPACE
HRROI T2,TEMP1
SETZ T3,
SOUT
IDPB T3,T1 ;GET TERMINATOR
HRROI T2,TEMP2 ;GET USER
SOUT
IDPB T3,T1
HRRZM T1,NODPTR
MOVE T1,T4 ;GET STRING POINTER
RETSKP ; RETURN ADDRESS
;MAIL IS FOR THIS SYSTEM. CHECK TO SEE IF USER NAME IS VALID.
PRSUS2:
IFE MBXF,<
HRROI T2,TEMP2 ; POINT TO USER NAME STRING
MOVX T1,RC%EMO ; EXACT MATCH ONLY
RCUSR ; IS THIS USER NAME VALID ?
ERJMP R ; NO, ERROR
TXNE T1,RC%NOM ; SKIP IF USERNAME FOUND
RET ; NO SUCH USER - ERROR
MOVE T1,T3 ; RETURN USER NUMBER IN T1
;IFE MBXF
IFN MBXF,<
PRINTX NEED CODE AT PRSUS2
;IFN MBXF
RETSKP ; RETURN SUCCESS
; Here To create MMAILR queue file
; Returns +1: Problems Of Some Sort
; +2: Ok
Mailit: Hrroi T1,FILNAM
Hrroi T2,[Asciz 'MAILQ:[--QUEUED-MAIL--].NEW-']
Setz T3,
SOUT
Move T2,T1
GTAD
Exch T1,T2
Movei T3,10
NOUT
Trn
Hrroi T2,[Asciz '-VMAIL-J']
Setz T3,
SOUT
Move T2,OURJOB
Move T3,[3,,^D10]
NOUT
Trn
Hrroi T2,[Asciz '.-1;P770000']
Setz T3,
SOUT
Idpb T2,T3
Movsi T1,(GJ%SHT+GJ%FOU)
Hrroi T2,FILNAM
GTJFN
JERR <Could not get queue JFN>
Movem T1,QUEJFN
Move T2,[fld(7,OF%BSZ)+OF%WR]
OPENF
JERR <Could not open queue JFN>
Hrroi T2,[Asciz ' =NET-MAIL-FROM-HOST:']
Setz T3,
SOUT
Hrroi T2,HstNam
SOUT
Hrroi T2,CRLF0
SOUT
Hrroi T2,[Asciz ' =RETURN-PATH:']
SOUT
Hrroi T2,FRMBUF
SOUT
Hrroi T2,CRLF0
SOUT
;; This is not documented -- but it makes MMAILR splat your TTY right!
Hrroi T2,[Asciz ' _']
SOUT
Hrroi T2,HstNam
SOUT
Hrroi T2,[ASCIZ/.#DECnet
/]
SOUT
Move T3,[Point 7,FrmNam]
DO.
Ildb T2,T3
Jumpe T2,ENDLP.
Cain T2," "
EXIT.
BOUT
LOOP.
OD.
Hrroi T2,CRLF0
Setz T3,
SOUT
Setzb P1,P2 ; Init Index And Failure Flag
Mailt1: Move T1,Ulist(P1) ; Get Next Recipient
Jumpe T1,Mailt5 ; End Of List
Call Sendit ; ADD NAME TO TOP OF QUEUE FILE
TRN
Mailt2: Vaxsuccess
Aoja P1,Mailt1
Mailt5: Jumpn P2, Rskp ; Don'T Log Success on error
;; Skipe P1 ; Anything Sent?
;; IFSKP.
;; Call Dtstmp ; Yes, Loc Lack Of Local Users
;; Log <No local electronic recipients>
;; ENDIF.
Move T1,QUEJFN ; Get MMAILR queue JFN
;;[76] Hrroi T2,CRLF0
;;[76] Setz T3,
;;[76] SOUT ;Sendit now has been fixed!!
Movei T2,"L"-100
BOUT
Hrroi T2,CRLF0
SOUT
Move T2,[Point 7,Bigbuf] ; Shove message into queue file
Movn T3,Bytcnt ; Get negative byte count
SOUT
Erjmp .+1
CLOSF
TRN
CALL WAKEUP
Retskp
; Append Mail To User'S Mail File
; Call With User Number Of Recipient In T1
Sendit: Stkvar <Usrno>
Movem T1,Usrno ; Save Recipients User Number
Jumpg T1,Quefil ; 4n host
Move T1,QueJFN ; Get QUEUE file
Movei T2,"L"-100 ; Get <FF>
BOUT
Hrroi T2,OurNam
Setz T3,
SOUT
Hrroi T2,CRLF0
SOUT
Move T2,UsrNo
DIRST
ERJMP .+1 ;[76] DIRST% always gives +1 return
Hrroi T2,CRLF0
SOUT
Retskp
;Sending to a remote host.
Quefil: MOVE T1,QUEJFN
MOVEI T2,"L"-100
BOUT
HRRO T2,USRNO ;GET HOST BP
SETZ T3,
SOUT
MOVEM T2,USRNO ;SAVE BP TO USER
HRROI T2,CRLF0
SOUT
MOVE T2,USRNO ;GET USER BP
SOUT
HRROI T2,CRLF0
SOUT
RETSKP ;all done
;Here to copy (and quote) "from" string into area pointed to by T1
; Quotes all characters (to save trouble of checking need for it)
QUOTE: MOVE T2,[POINT 7,FRMBUF]
TLC T1,-1 ; lh of byte pointer all ones?
TLCN T1,-1 ; ..
HRLI T1,(POINT 7,) ; yes, make real byte pointer
MOVEI T4,<24*5>-1 ; maximum characters allowed in string
QUOTE1: MOVEI T3,"V"-100 ; quote character
IDPB T3,T1 ; stuff it
ILDB T3,T2 ; next char of source string
IDPB T3,T1 ; stuff it
JUMPE T3,[MOVNI T2,1 ; if zero, back up over last ctrl-V
ADJBP T2,T1 ; ..
DPB T3,T1 ; wipe it out with null
RET] ; and return
SOJGE T4,QUOTE1 ; insure no overflow
DIE <QUOTE overflow>
;Open Log File
Opnlog: MOVX T1,.NULIO ; No logs unless
SKIPN LOGP ; LOGP is set non-zero
RET ; Oh well
Movx T1,Gj%sht ; Try Logical Name First
Hrroi T2,[Asciz /MAIL:VMAIL.LOG/]
GtJFN
Erjmp Opnerr
Movx T2,<070000,,0>+Of%app
Openf ; Open For Append
Erjmp Opnerr
Ret
Opnerr: Hrroi T1,[Asciz /VMAIL: Can't open log file because: /]
Esout
Movx T1,.priou
Hrloi T2,.fhslf
Erstr
Erjmp .+1
Erjmp .+1
Jrst Fatal
;Time Stamp Log File
Dtstmp: Move T1,LogJFN
HRROI T2,[ASCIZ/
/]
SETZ T3,
SOUT%
Jerr <Can't write to log file>
Seto T2, ; Current Time
Odtim
IFJER.
Hrroi T1,[Asciz /VMAIL: Odtim Failed: /]
Esout
Movx T1,.priou
Hrloi T2,.fhslf
Erstr
Erjmp .+2
Erjmp .+1
Tmsg <
DTSTMP called from >
Movx T1,.priou ; Type Pc Of Caller On Terminal
Hrrz T2,(P)
Movx T3,^D8 ; In Octal
Nout
Erjmp .+1
Jrst Fatal ; Go Fire Up The World Again
ENDIF.
Movei T2," " ; Space
Bout
Ret
;Write Asciz String Pointed To By T1 To Log File
Logmsg: Move T2,T1 ; Copy String Pointer
Move T1,LogJFN
Setzb T3,T4
Sout
Jerr <Can't write to log file>
Move T1,T2
Ret
;Write Statistics To Log File
Lstats: Stkvar<Elptm0>
; Move T1,LogJFN
; Move T2,Elptim ; Elapsed Time For Mail Receipt
; Fltr T2,T2 ; Flost It
; Fdvr T2,[100000.0] ; Compute Seconds
; Movx T3,<1b1+Fl%one+Fl%pnt+3b23+3b29>
; Flout ; Type Seconds
; Erjmp [Haltf] ; Never Happens
; Movem T2,Elptm0 ; Save Time
; Log < seconds, >
Log < : >
Move T1,LogJFN
Move T2,Bytcnt ; Byte Count
Movx T3,^D10 ; Base 10
Nout
Jerr <NOUT failure>
Log < chars
; Move T1,LogJFN
; Fltr T2,Bytcnt ; Float Byte Count
; Fdvr T2,Elptm0 ; Compute Bytes Per Second
; Movx T3,<1b1+Fl%one+Fl%pnt+5b23+3b29>
; Flout
; Jerr <FLOUT failure>
; Log < chars/sec/
;>
Ret
;Close Net Connection And Reopen It. Re-Enable For Interrupts
; On Connect Initiate Messages
CLZNET: MOVEI T1,^D4000 ; Give pipe four seconds to empty
DISMS ; ..
MOVE T1,NETJFN ; normal close
CLOSF
IFJER.
CALL DTSTMP ; We should complain about these
LOG <%Close error for net link: >
MOVE T1,LOGJFN
HRLOI T2,.FHSLF
ERSTR
ERJMP .+1
ERJMP .+1
MOVE T1,NETJFN
TXO T1,CZ%ABT ; Try real hard to close it
CLOSF ; so we don't eat all job 0 JFNs
ERJMP .+1
MOVE T1,NETJFN
RLJFN
ERJMP .+1
ENDIF.
CALL OPNLSN ; open connection again
RET ; return
;Open The Net Connection And Listen For Connect Initiates
Opnlsn: Movx T1,Gj%sht
Hrroi T2,[Asciz /Srv:27/] ; Magic Number For Vax Mail Server
GtJFN
Jerr <Can't get net JFN for server>
Movx T2,Of%rd!Of%wr!<100000,,0>
Openf
Jerr <Can't open net JFN>
Movem T1,NetJFN
Movx T2,.moacn ; Enable For Psi On Network Transitions
Movx T3,0b8+<.mocia>B17+<.mocia>B26 ; Channel Zero
Mtopr
Jerr <Can't enable for PSI on network transitions>
Movx T1,.fhslf
Movx T2,1b0 ; Activate Channel Zero
Aic
Ret
;Log Name Of Foreign Host
T4nhst: Setzm Hstnam ; Zero This String
Setzm 1+Hstnam ; ..
Move T1,NetJFN ; Get Net JFN
Movx T2,.morhn ; Return Host Name
Hrroi T3,Hstnam ; Where To Put It
Mtopr
IFJER.
Hrroi T1,[Asciz /???/]
Callret Logmsg ; Log confusion
ENDIF.
Hrroi T1,Hstnam ; Copy Name To Log File
Skipe Dbugsw
Psout
Hrroi T1,Hstnam ; Copy Name To Log File
Call Logmsg ; ..
Ret
;Set Up To Time Out If Network Too Slow
Timeit: Move T1,[.fhslf,,.timel]
Move T2,[Timen] ; Milliseconds To Allow
Movei T3,1 ; Channel One
Timer
Jerr <Can't time myself>
Movx T1,.fhslf ; Activate Timer Channel
Movx T2,<1b1>
Aic
Ret
;Cancel Above Timer Request
Cncltm: Move T1,[.fhslf,,.timal] ; Remove All Pending Timer Requests
Movei T3,1 ; For This Channel
Timer
Jerr <Can't remove pending timer request>
Ret
;Here On Timeout
Timout:
; Call Dtstmp
Skipn KEPLIV ; Skip if still alive
Die <Timeout Occured> ; [153] No activity, dead
Setzm KEPLIV ; Clear keep alive flag
Push P,T1 ; Save ACs before calling Timeit
Push P,T2
Push P,T3
Call Timeit ; Start new timer
Pop P,T3 ; Restore ACs before resuming
Pop P,T2
Pop P,T1
Debrk ; Allow things to continue
;Here If Net Link Dies While Outputting To It
Dmplnk: Cis ; Zap Things
Movx T1,Cz%abt ; Abort The Net JFN
Hrr T1,NetJFN ; ..
Closf ; ..
Erjmp .+1 ; Don'T Care
Call Dtstmp
Log <----Connection aborted
Movx T1,.fhslf ; Deactivate Connect Initiate Channel
Movx T2,<1b0> ; ..
Dic ; ..
Call Cncltm ; Cancel Pending Timer Requests
Move T1,LogJFN ; Close Log File
Closf
Erjmp .+1
; Jrst VMAIL0 ; Go Wait For New Mail
Jrst VMAIL ; Restart on connection abort
;Here On Fatal Wipeout (Jsys Which Can'T Fail Does, For Instance)
Fatal: Movx T1,.fhslf
Dir ; Disbale Interrupts
Cis ; Clear Interrupts
; Remove out-of-synch message on fatal error
;; Move T1,NetJFN ; Type A Record To Force Net Buffers Out
;; Hrroi T2,[Asciz /
;;?VMAIL Internal Error/]
;; Setzb T3,T4 ; Add Question Mark So Mail Isn'T Requeued
;; Soutr ; ..
;; Erjmp .+1
;; Movei T1,^D5000 ; Wait Five Seconds
;; Disms
skipe t1,logJFN
Closf
Erjmp .+1
setzm logJFN
Movx T1,.fhslf!cz%abt ; Abort All JFNs
Clzff ; ..
ERJMP .+1 ; Ignore errors in the fatal error routine
Call Opnlog ; Reopen Log File
Movem T1,LogJFN
Call Dtstmp
Log <Error restart...
Tmsg <VMAIL error restart...
Move T1,LogJFN
CLOSF%
ERJMP .+1
Movei T1,^D5000 ; Wait Some More
Disms
Jrst VMAIL ; And Fire Up The World Again
;Disable Capabilities So Quota-Checking Happens
Capoff: Push P,T1 ; Don'T Clobber
Movx T1,.fhslf ; Get My Caps
Rpcap
Movem T3,Capenb ; Remember For Later
Setz T3, ; No Caps At All
Epcap
Pop P,T1 ; Restore
Ret
;Re-Enable Caps
Capon: Push P,T1 ; No Clobberage
Movx T1,.fhslf
Move T3,Capenb ; Caps We Had Before
Epcap
Pop P,T1
Ret
CRLF0: ASCIZ /
/
; FROM DMASER.MAC
A==1
B==2
C==3
WAKEUP: SKIPE T2,MYPID ; have a PID already?
TDZA T1,T1 ; yes, use it
MOVX T1,IP%CPD ; no, create a PID
MOVEM T1,IPCBLK+.IPCFL
MOVEM T2,IPCBLK+.IPCFS ; PID to use if one there
SETZM IPCBLK+.IPCFR ; send to INFO
MOVX T1,<.IPCI2+3,,BUFFER> ; length of INFO msg,,where INFO msg is
MOVEM T1,IPCBLK+.IPCFP
MOVX T1,.IPCIW ; return PID associated with name
MOVEM T1,BUFFER+.IPCI0
SETZM BUFFER+.IPCI1 ; duplicate copy not needed
DMOVE T1,[ASCII/[SYSTEM]MM/] ; 1st part of PID to look up
DMOVEM T1,BUFFER+.IPCI2
MOVE T1,[ASCII/AILR/] ; 2nd part of PID to look up
MOVEM T1,BUFFER+.IPCI2+2
MOVX T1,.IPCFP+1 ; length of block
MOVEI T2,IPCBLK ; get MMailr's PID
MSEND%
ERJMP R ; looks like INFO isn't there
MOVE T1,IPCBLK+.IPCFS ; get the PID I made
MOVEM T1,MYPID ; remember it for next time
DO.
SETZM IPCBLK+.IPCFL ; no flags
SETZM IPCBLK+.IPCFS ; any sender
MOVE T1,MYPID ; I'm the receiver
MOVEM T1,IPCBLK+.IPCFR
MOVX T1,<10,,BUFFER> ; place to put the reply
MOVEM T1,IPCBLK+.IPCFP
MOVX T1,.IPCFP+1 ; length of block
MOVEI T2,IPCBLK ; get reply from INFO
MRECV%
ERJMP R ; failure irrelevant here
LOAD T1,IP%CFC,IPCBLK+.IPCFL ; see who sent message
CAIE T1,.IPCCC ; from <SYSTEM>IPCF?
CAIN T1,.IPCCF ; no, from <SYSTEM>INFO?
IFSKP.
LOOP. ; no, get another message
ENDIF.
ENDDO.
JN <IP%CFE,IP%CFM>,IPCBLK+.IPCFL,R ; give up if undeliverable
SETZM IPCBLK+.IPCFL ; no flags
MOVE T1,MYPID ; I'm the sender
MOVEM T1,IPCBLK+.IPCFS
MOVE T1,BUFFER+.IPCI1 ; MMailr is the recipient
MOVEM T1,IPCBLK+.IPCFR
MOVX T1,<1,,BUFFER> ; one word from BUFFER
MOVEM T1,IPCBLK+.IPCFP
MOVX T1,'PICKUP' ; magic word to wake up MMailr
MOVEM T1,BUFFER
MOVX C,^D20
DO.
MOVX T1,.IPCFP+1 ; length
MOVEI T2,IPCBLK ; send wakeup to MMailr
MSEND%
IFJER.
MOVEI T1,^D1000 ; failed, wait a bit
DISMS%
SOJG C,TOP. ; try a few times
RET ; failed, give up
ENDIF.
ENDDO.
MOVX T1,.MUQRY ; query function for MUTIL%
MOVEM T1,BUFFER
MOVE T1,MYPID ; query packets for our PID
MOVEM T1,BUFFER+1
MOVX C,^D20 ; number of retries
DO.
MOVX T1,.IPCFP+2 ; number of words to return
MOVEI T2,BUFFER ; argument block in BUFFER
MUTIL%
IFJER.
MOVEI T1,^D1000 ; wait a bit
DISMS%
SOJG C,TOP. ; retry a few times
RET
ENDIF.
ENDDO.
DO.
SETZM IPCBLK+.IPCFL ; no flags
SETZM IPCBLK+.IPCFS ; sender is filled in by monitor
MOVE T1,MYPID ; I'm the receiver
MOVEM T1,IPCBLK+.IPCFR
MOVX T1,<10,,BUFFER> ; where MMailr reply will go
MOVEM T1,IPCBLK+.IPCFP
MOVX T1,.IPCFP+1 ; size of block
MOVEI T2,IPCBLK ; get reply from MMailr
MRECV%
ERJMP .+1 ; error uninteresting here
LOAD T1,IP%CFC,IPCBLK+.IPCFP ; get sender code
IFN. T1 ; special sender?
CAIE T2,.IPCCF ; from <SYSTEM>INFO
CAIN T2,.IPCCP ; or private <SYSTEM>INFO?
LOOP. ; yes, try for another message
ENDIF.
ENDDO.
RET
End VMAIL
Is everything ok in hecnet land? I can get out from SLAVE in Area 4 to anywhere else. My bridge to Psilo is up and running. I can see another local node on the same circuit (tiger)
Thanks, Mark
Sent from my iPad
On 20 Jun 2012, at 12:32, Johnny Billquist <bqt at softjar.se> wrote:
On 2012-06-20 12:27, Sampsa Laine wrote:
Does anyone have an outline of how the MAIL-11 protocol works? Toying with the idea of writing a Python library to send MAIL-11 messages using the DECNET-linux stack..
Let me get back to you in a while. I wrote up a document that I think I have on MIM, as I was writing my own MAIL-11 implementation for RSX...
Busy the next three hours, though...
Johnny
On 2012-06-20 12:27, Sampsa Laine wrote:
Does anyone have an outline of how the MAIL-11 protocol works? Toying with the idea of writing a Python library to send MAIL-11 messages using the DECNET-linux stack..
Let me get back to you in a while. I wrote up a document that I think I have on MIM, as I was writing my own MAIL-11 implementation for RSX...
Busy the next three hours, though...
Johnny
There's a VERY rough reverse-engineering of it in the mail part of DECnet for Linux. I didn't document it, sorry, because I was implementing it as I went. But it might be helpful to you.
http://linux-decnet.cvs.sourceforge.net/viewvc/linux-decnet/dnprogs/mail/
Chrissie
On 20/06/2012 11:27, Sampsa Laine wrote:
Does anyone have an outline of how the MAIL-11 protocol works? Toying with the idea of writing a Python library to send MAIL-11 messages using the DECNET-linux stack..
Sampsa
Does anyone have an outline of how the MAIL-11 protocol works? Toying with the idea of writing a Python library to send MAIL-11 messages using the DECNET-linux stack..
Sampsa
On 06/19/2012 02:04 PM, Johnny Billquist wrote:
On 2012-06-19 16:54, Dave McGuire wrote:
On 06/19/2012 10:03 AM, Saku Set l wrote:
Found 1 DELQA and 4 DEQNA (condition unknown), and also have 1 broken
DELQA
(sorry for chiming in late)
I have quite a few of both here; if you have a hard time getting one
to Johnny, let me know and I'll send one over. Shipping from USA will
take awhile though.
Sounds like there are some options nearer, so I think you can rest.
With shipping from the US, however, it might also be worth talking to me
if you live close to San Francisco or Mountain View, as I might get
someone to carry it over for me in those cases... But for now, I
think/hope we can solve this a little more "local".
I am fairly close to the East coast, unfortunately. :-( In any event,
let me know if you get into a bind for that or any related stuff.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
On 2012-06-19 16:54, Dave McGuire wrote:
On 06/19/2012 10:03 AM, Saku Set l wrote:
Found 1 DELQA and 4 DEQNA (condition unknown), and also have 1 broken DELQA
(sorry for chiming in late)
I have quite a few of both here; if you have a hard time getting one
to Johnny, let me know and I'll send one over. Shipping from USA will
take awhile though.
Sounds like there are some options nearer, so I think you can rest.
With shipping from the US, however, it might also be worth talking to me if you live close to San Francisco or Mountain View, as I might get someone to carry it over for me in those cases... But for now, I think/hope we can solve this a little more "local".
Johnny
On 2012-06-19 17:13, Paul_Koning at Dell.com wrote:
On Jun 19, 2012, at 10:52 AM, Dave McGuire wrote:
On 06/19/2012 10:27 AM, Johnny Billquist wrote:
Found 1 DELQA and 4 DEQNA (condition unknown), and also have 1 broken
DELQA
I recall that the DELQA is more "modern" and better i some way, are they
also more prone to failure?
No. From a reliability point of view, I've not heard that either should
be better or worse. However, the DEQNA is rather buggy in its general
performance.
Side-by-side, the DELQA is quite a bit faster than the DEQNA, the
DELQA-YA ("Turbo DELQA") even more so.
That said, the DELQA has a "DEQNA compatibility mode"...whether that
makes it as slow and/or buggy as the DEQNA, I don't know.
I don't think so.
QNA compatibility mode means it works like a QNA from the driver point of view. That was never the issue with the QNA -- if it had worked according to the specification, it would have been quite a good device. The problem was that it never worked correctly, not even after 12 revisions. Finally when rev L didn't work adequately either, the DEC software teams said enough is enough, we're dumping the QNA.
The issues were most obvious on VMS; I'm not sure if they were visible enough on the PDP11 OSs to cause concern. Maybe with LAT, probably not or less so with DECnet. I don't remember the details, other than that Local Area VAXclusters were the ones that tended to run into trouble, since those protocols were particularly unforgiving.
I know that the QNA was not dropped from the PDP-11 OSes, but there are hacks in drivers to work around some issues, such as the receiver (I think it is) that sometimes stops working and needs to be kicked.
The LQA is better, but I've seen issues with that one too.
Johnny
On 06/19/2012 11:13 AM, Paul_Koning at Dell.com wrote:
That said, the DELQA has a "DEQNA compatibility mode"...whether that
makes it as slow and/or buggy as the DEQNA, I don't know.
I don't think so.
QNA compatibility mode means it works like a QNA from the driver point of view. That was never the issue with the QNA -- if it had worked according to the specification, it would have been quite a good device. The problem was that it never worked correctly, not even after 12 revisions. Finally when rev L didn't work adequately either, the DEC software teams said enough is enough, we're dumping the QNA.
The issues were most obvious on VMS; I'm not sure if they were visible enough on the PDP11 OSs to cause concern. Maybe with LAT, probably not or less so with DECnet. I don't remember the details, other than that Local Area VAXclusters were the ones that tended to run into trouble, since those protocols were particularly unforgiving.
I remember hearing from 25 years ago the bugs in the DEQNA, but never
knew what they were/are. I wonder if that information is archived
anywhere. In some Micronotes somewhere maybe?
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
