31 Dec
2012
31 Dec
'12
5:59 p.m.
Johnny Billquist <bqt at softjar.se> writes:
On 2012-12-31 17:42, Bob Armstrong wrote:
But I totally stand by my claim that this is broken. :-)
Well, since VMS and VMS RMS have no problems with this file, it's arguable
about which end is "broken" ... However, I have converted this file to a
more conventional format for the lesser-abled operating systems out there
:-)
Try it again.
Looks pretty now.
Yeah, VMS have no problems with it. VMS is somewhat forgiving internally
when parsing that file, it would appear.
But you got to admit that Stream_CR on a file with not a single CR in it
smells broken.
I doubt that is the case. You didn't see it on your side after copying it
but, at that point, the file was probably hopelessly horked by RSX.
When the file was served with FAL to my RSX system, I just ended up with
an extra LF per line, which is totally understandable, as there is an LF
between each line in your file. The most interesting detail was that the
LF was still interpreted as a record terminator, even though the format
was Stream_CR. But the LF was not stripped off, which seems reasonable.
I bet had you just changed the attribute to Stream_LF, it would have
worked just fine as well.
I'm not convinced that a '$ CONVERT/FDL="record; format stream_lf"'
would
handle that file's contents correctly.
But even in original, it was just a slight eyebrow-raiser to get the
extra LFs in there.
But they were probably free-of-charge! :)
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
Well I speak to machines with the voice of humanity.