On Jan 2, 2013, at 5:25 AM, Jordi Guillaumes i Pons wrote: El 02/01/2013, a les 11:04, Johnny Billquist <bqt at softjar.se> va escriure: Does anyone have an RSTS/E system (or TOPS-20), and could report how the file appears to be, from their point of view? I don't know how to use RMSDSP against a full DECNET filespec, but a NFT DIR/FU shows this (in RSTS/E): Directory 2.1::SYS$SPECIFIC:[FAL$SERVER] Name .Typ Size Prot Access Date Time UIC INFO .TXT;2 8/10P < 42> 31-Dec-12 31-Dec-12 07:20 [000376,000373] RF:VAR FO:SEQ USED=8:150 RECSI=174 CC:IMP INFO .TXT;1 8/35P < 42> 27-Dec-12 08-Jul-10 10:54 [000376,000373] RF:UDF FO:SEQ USED=8:64 RECSI=174 CC:IMP Total of 16/45 blocks in 2 files. The ;1 version lists as RD:UDF. I guess that is "record format: undefined". Yes, that's what it means. That's different from stream, never mind stream_cr or stream_lf. RSTS should understand stream format -- after all, it exists partly because that's the native format of RSTS (though I think it was actually introduced to support Unix better). A network capture would be very interesting, since the evidence so far is all over the map. paul