id	summary	reporter	owner	description	type	status	priority	milestone	component	version	resolution	keywords	cc
355	Problem in card reader when reading absentee job	dean@…	Eric Swenson	"When reading a card deck in via the reader, in certain cases it appears that an extraneous newline is being received by Multics. I discovered this when I was adding to my absentee job for assembling the FPGA T&Ds.

The symptom: when I submit the absentee deck, at character position 836 it appears a newline is inserted somewhere in the path. This newline does not exist in the input deck but appears in the absentee output.

Why it might be Multics: at first I suspected the simulator, however, I used the same mechanism to send the MAP355 source code and it does not get an extraneous newline inserted at the character position (or any other). I verified this by submitting the source code to Multics and punching out a deck from the source and performing a comparison. There were no differences in the two files (except that the punched deck was space padded out to 80 characters due to the way cards are formatted).

This is an extract from the absentee output (I added the line numbers at the beginning of the lines for reference):

{{{
01: coremif tand
02: r 14:54 1.599 30
03: 
04: eor tand.list
05: 1 request submitted; 0 already in printer queue 3.
06: r 14:54 0.090 7
07: 
08: -pr -q 1
09: Segment -pr not found.
10: r 14:54 0.008 0
11: 
12: eor tand.mif -punch -q 1
13: 1 request submitted; 0 already in punch queue 1.
14: r 14:54 0.016 0
15: 
16: logout
}}}

The ""eor"" command on line 4 was supposed to have ""-pr -q 1"" on it (and does in the source deck). Instead the extraneous newline was inserted and line 04 was split into line 08.

If I insert a bunch of newlines in the input file so the extraneous one is inserted in a benign place, it works fine.

I have replicated this on the Quick Start system. 
"	defect	new	major		Administration