|
|
| Author |
Message |
| 25 new of 306 responses total. |
keesan
|
|
response 220 of 306:
| 
May 31 20:03 UTC 1998 |
We looked up mv -i, which means to rename a file and have an interactive
promprt, which asks us whether to overwrite any existing file by that name.
We looked up sed, with is the stream editor, and the $ addresses the last
input file. We have not found '/^V//, but presume this means to delete all
those ^M's at the line ends. We are about to try out these commands.
See you soon!
|
rcurl
|
|
response 221 of 306:
|
May 31 20:36 UTC 1998 |
The file was permitted 644 in your directory. I think 600 would be the
permission of the unix file after it is decoded.
|
mcnally
|
|
response 222 of 306:
|
May 31 20:48 UTC 1998 |
yes, the 600 perms would apply to the *output* of uudecode, not the input..
^V means "take the next character literally" (i.e. instead of treating it
as a special character like a carriage return or backspace..) so
s/^V^M$// would "s"ubstitute a literal control-M at the end of the line
with nothing, i.e. would delete it..
|
keesan
|
|
response 223 of 306:
|
May 31 20:59 UTC 1998 |
Well, we followed the directions to use uudecode and delete the control M's,
and, once I figured out that I could not send a file to a directory that
already had a file with that same name (something to do with the Procomm
settings, it will not overwrite but aborts, I deleted the original files
there), we succeeded in sending the uudecoded three files (Ge9495a.cec, and
the b and d versions) to my computer. They can be read in DOS, just like the
automatically decoded MIMEd files, but are still gibberish in WP5.1, in fact
the uudecoded file looks identical to the mime-decoded file. The first line
reads something like:
Nm'hn{^S.)=5<h k^Q,%4#;"0{9',ik*/<2^\.
This is what I get trying to view the file in WP5.1, but if I try to download
I get the message 'incompatible file format'. SO now I have two ways to
decode to the point where I cannot read the file in WP5.1. What next?
What is now in my home directory is the uudecoded versions Ge9495a.cec etc.
and the decoded mimed versions mimea, etc., which seem to be identical.
I don't understand the 600 and 644 stuff, am getting pretty confused by now,
but does anyone still have WP5.1 for DOS, and could you possibly download oneo
f these decoded files and see if it is readable in WP5.1, and tell me where
I am going wrong? MANY, MANY THANKS. If the files are no longer accessible,
tell me how to make them accessible.
The first line of the files viewed in DOS is not actually text, but is stuff
like happy faces, hearts, and other upper ASCII characters, and the command
Prestige 12.
While sending via Procomm, with Superkermit, after typing set file type binary
on a separate line, then return, we were told that the file was being
transmitted as TEXT. Could the problem be with a Procomm setting? We have
had no problem uploading binary files with Kermit. I may try downloading the
same files with zmodem, but so far I have not been able to figure out where
the files go when I use zmodem, they seem to be sending but just disappear.
(Too much to learn all at once - new computer, new wordprocessor, new version
of Procomm, unix, e-mail, grex.... I am having trouble keeping the unix and
dos and wp commmands separate and tend to remove instead of delete, etc.)
|
rcurl
|
|
response 224 of 306:
|
May 31 21:05 UTC 1998 |
When I recoded the file on my computer it appeared as a WORD text document.
Are you sure it is from WP5.1? My system usually tells me it is translating
a WP file.
|
keesan
|
|
response 225 of 306:
|
May 31 22:03 UTC 1998 |
We had the same problem when we sent ourselves a known WP5.1 file (upload,
e-mail to Jim, view and save, download), so it cannot be WORD, we don't have
WORD in theh ouse anywhere.
We are attempting to download with zmodem, but when I type sz filename,
give the send command (PgDn), specify parameters (drive name and file name),
the system freezes up and I have to log off and on again. Whatever we are
getting after decoding certainly is not acting like a WP5.1 file at our end.
Jim thinks that Word may automatically convert WP documents to Word.
Well, back to the zmodem manual. We seem to be pushing grex (and a few other
programs) to beyond their normal limits. We are downloading a binary (8-bit)
file, but can't read it in WP5.1, just in DOS. Maybe we have to change the
settings on KERMIT? It is downloading 8-bit, but says it is sending text.
In the middle of this I got a chat request from Turkey, while using
the external 9600 baud modem, and the screen froze up, so it does not seem
to be a modem problem that freezes up the screen.
Davel, did you download the decoded file and then view it with WP5.1
for DOS? If so, how did you download (Procomm, Kermit, or how?).
Jim will set up a second computer the same as this one to see if that
also has the freeze-up problem. It is annoying but tolerable. The inability
to read decoded files may not be tolerable if people are not willing to both
convert their WOrd files to WP for DOS and then also convert to 7-bit transfer
format. I am probably driving quite a lot of translation agencies crazy by
using outdated software, but then there are not a lot of Slovene chemistry
translators around who know English as wellas I do, so they are patient.
|
keesan
|
|
response 226 of 306:
|
May 31 22:23 UTC 1998 |
We changed the kermit settings from TEXT to BINARY and downloaded the file,
and it is now readable in WP5.1. (I had thought the command Set file type
binary would have taken care of that. Do I continue to set file type binary?
Do I set it binary for text files?).
So now I can recognize and decode and read both uuencoded and mimed binary
files, as well as sending Mimed files. (I will not worry about learning how
to encode with uuencode.) We still don't know how to download with zmodem,
but will read the manual for hints.
I will now write out a check for $50 to the grex general fund, in honor of
DaveL for teaching me how to uudecode, Rane for proving that the text was
uuencoded and could be decoded, and everyone else who has helped me to make
some sense of unix and other computer stuff.
I would include here Marcus for his magnificent explanation of how binary
logic works, Mike McNally for trying to explain a few other things (never sure
whether I understand or not, but after a while it gets clearer), Jan for his
extract program, Valerie for her immediate and detailed answers to a lot of
beginner's questions, and anyone else I forgot to mention above. I will
probably continue to come up with questions, some challenging, some rather
basic (read stupid?), and be grateful for the answers.
I wonder why Kermit was uploading files as binary, but downloading as text.
|
keesan
|
|
response 227 of 306:
|
May 31 22:56 UTC 1998 |
I don't think I had used Kermit to upload anything since we switched from
Procomm to Procomm Plus (too many switches all at once -computer,
wordprocessor, communications program and protocol - had been uploading with
zmodem, and downloading with kermit but until now only text).
The check is signed and in a stamped addressed envelope.
My screen froze up this time while trying to access a website.
Is there some version of uudecode that could be innstalled on grex that would
take care of the line-end Control-M's, such as UUtool?
|
kentn
|
|
response 228 of 306:
|
Jun 1 01:16 UTC 1998 |
The dos2unix program is available for doing conversions. Use that in
combination with uudecode.
|
keesan
|
|
response 229 of 306:
|
Jun 1 03:26 UTC 1998 |
Ratheer than try to figure out how to use one more program, I will stick
with Davel's one line ctl-M remover.
I compared the first of the three files sent me with what I had sent to the
translation company. I had asked to see the copy after the chemistry editor
finished with it. I put my original file on screen 2 and the edited file on
screen 2 and kept switching back and forth every line. I found only about
three changes, such as steam for water vapor and a couple of typos. This did
not seem worth the work to compare. Now a friend just emailed me a uuencoded
zipped program for comparing files. I exported it from the message to my home
directory under a slightly different name than what was in the header, and
typed in the line with the ctl-M in it, and got out a file of the original
name (it may still have the one line text intro, which I hope I can manage
to delete) which looks decoded. For some reason, I also got the message:
uudecode: stdin: Short file
What does this mean?
Anyway, I have now been able to use uudecode on another binary file, in this
case a zipped program. (Still have to learn to use the program).
The check is in the mailbox waiting for the mailman to pick it up.
Jim also figured out that you could transfer the printer files from a 3.5 inch
floppy to two 5 1/4 floppies by zipping them, and we got me a printer
definition working. All that is left to do is to get zmodem to download,
figure out why my screen freezes up so often, and it would be nice to have
a monitor that can display graphics (Cyrillic alphabet) but first Jim wants
to use the Hercules graphics plus card to fix another broken one, so I may
get a VGA temporarily instead. The end is sort of in sight, and at least I
am prepared to send and receive binary files, which is what I originally
joined grex for. Many many thanks again to everyone.
|
mdw
|
|
response 230 of 306:
|
Jun 1 07:59 UTC 1998 |
"short file" means the file got truncated. The "uuencode" format works
something like this:
first line: "begin" mode filename
N lines: each containing a non-zero character encoded byte count,
followed by character encoded binary data.
A line with a zero byte count, and no data.
A line that contains the word "end", and nothing else.
Binary values from 0 to 63 are encoded as follows:
0 ` 1 ! 2 " 3 # 4 $ 5 % 6 & 7 '
8 ( 9 ) 10 * 11 + 12 , 13 = 14 . 15 /
16 0 17 1 18 2 19 3 20 4 21 5 22 6 23 7
24 8 25 9 26 : 27 ; 28 < 29 = 30 > 31 ?
32 @ 33 A 34 B 35 C 36 D 37 E 38 F 39 G
40 H 41 I 42 J 43 K 44 L 45 M 46 N 47 O
48 P 49 Q 50 R 51 S 52 T 53 U 54 V 55 W
56 X 57 Y 58 Z 59 [ 60 \ 61 ] 62 ^ 63 ^
Note there aren't any lower-case characters here. When uuencode was written,
there were plenty of systems around that automatically translated text from
lower-case into upper-case.
Since a binary byte contains 8 bits, but the above only encodes 6 bits,
it takes 4 of these printable characters to store the information from 3
bytes. The encoding looks something like this:
Binary data(hex): 4D 44 57
(in binary): 010011010100010001010111
aaaaaabbbbbbccccccdddddd
6-bit numeric values: 19 20 17 23
character-codes: 3 4 1 7
(This was probably not a particularly good example, as it is
"coincidence" that the 4 characters, 3417, all happen to be digits. All
those other characters, including punctuation and letters, are also
valid data.
Each of these N lines must contain at least 1 character of data, but
generally speaking, most lines will contain 45 bytes of data. These 45
bytes will expand out to 60 printable characters, and, with the byte
count, that means most lines in a uuencoded file are 61 characters long.
Each of these lines will also start with M, because M is the character
that encodes the byte value of 45. The line with the zero data count
will contain one printable character, `, because ` is the character that
encodes the zero byte count. If the file isn't a multiple of 45 bytes,
the last data line will contain fewer than 45 bytes. If the file isn't
a multiple of 3 in size, then the last byte of the file doesn't line up
on a nice character boundary. Now, it doesn't *really* matter what's
done here, because uudecode will ignore whatever's put there, but what
uuencode does is to pad the last character triplet out with "garbage"
(actually, data from the previous data line), then write out 4 character
values. So, most lines in a uuencoded file will contain 1 mod 4
characters. Also, because uudecode ignores extra trash on most lines,
there is only 1 line where a trailing control-M is bad -- the *last*
line! So, rather than running the whole file through dos2unix, you can
just use vi (or whatever) to delete the last control-M in the file.
uudecode can complain about the following:
uudecode: <file>: no "begin" line -if the first line
doesn't
start with "begin".
uudecode: <file>: illegal ~user. if the filename starts
with ~, but doesn't
have a /.
uudecode: <file>: no user <user>. if ~so-and-so isn't on
this system.
uudecode: <file>: path too long. # if the expanded
filename is
too long...
uudecode: <file>: <operating-system-error> if there's an error
creating
the file.
uudecode: <file>: short file. EOF before reading a
data line
with a zero byte count
(ie,
the file dones't contain the
line (`)
uudecode: <file>: no "end" line. the next line in the
file
after ` doesn't say
"end".
|
davel
|
|
response 231 of 306:
|
Jun 1 12:52 UTC 1998 |
Sindi,
(1) kermit usually needs to be set to binary on *both* ends - otherwise when
the two ends negotiate they choose the least-common-denominator setting & go
with text, & garbage your file. This is dumb, but is the way it is. So you
need to set Procomm to use binary - which you should be able to do as a
configuration setting & thereafter not need to do again, I think - and also
do "set file type binary" on Grex's kermit. (Or, to just send a file,
invoke it as "kermit -is yourfilename", I think. I don't use kermit much.)
(2) What all Marcus said is (I presume) correct, & is interesting & useful,
besides being way more than I ever knew about uuencode/uudecode. But in
your particular case I think you have one likely cause for the file truncation
which yielded the "short file" error. (I mentioned it earlier, but didn't
stress it enough, I think.) If the filename on the "begin" line in the
uuencoded file is the same file name as the uuencoded file itself - which was
the case with the one I looked at - the file will get overwritten before
uudecode is done reading it. That's why I put in the step to rename the file
to a different name.
(3) I'd recommend using the dos2unix program Kent suggested, instead of the
sed command I gave you. I was going to suggest flip (or its alias toix),
but it apparently got lost in coming to the Sun4, and I didn't know about
the (hopefully equivalent) dos2unix program. You should probably read
the man page on dos2unix (I just did so, & think you probably want the
-ascii option).
Better yet, you could write a script to rename your file, convert it, &
send it all with one command. Assuming that (as with the sample) the name
your file was saved to is always the same as the name on the "begin" line,
the following should work:
#!/bin/sh
# for each parameter, renames & uudecodes. Then sends the entire batch
# using kermit
for filen in "$@"
do
echo Going to convert ${filen} now
mv -i "${filen}" "${filen}.dos"
dos2unix -ascii "${filen}.dos" | uudecode
done
echo
echo "... Going to send the following files, kermit binary mode:"
echo "$@"
kermit -is "$@"
(You would still need to make sure that Procomm is set for binary kermit
first. Also, I'm pretty sure that kermit would send all the files on the
command line, but I haven't tried it to make sure.)
|
keesan
|
|
response 232 of 306:
|
Jun 1 14:12 UTC 1998 |
Thanks Dave, I did have to change the Procomm settings for Kermit from text
to binary before I got readable flie (I think I mentioned this, but I dont
write all that clearly), and I had saved the file from my e-mail message to
a new file name instead of the one it decoded as, rather than using your
mv command. What I did not do was to remove the first line of my e-mail
message saying something like : Here is a file I am sending you. This may
have confused uudecode. I will edit the message and reexport it (I had better
delete it from my home directory first). We will do that and see if we get
the short file message.
Is there any problem with uudecoding a file 140,000 bytes long?
We will also attempt to look at the manual and learn to write our
own unix programs, some day, or at least enough to understand what you wrote
for us above. (First we are attempting to figure out why my screen freezes
up by putting me back on my old commputer, which I bought new in 1985 and
which is known good, with the same new software, minus the shroom program,
and if it all works, add the shroom program and if that works it was a flaky
computer and I will use another one, we have a large choice of them).
Marcus, as always, many thanks for the detailed explanation (which must have
taken you at least an hour to write up typo-free, even if you knew this all),
which we will extract and print out for further reference. I vote that we
start a separate archive item for this explanation and for a few of Marcus's
other contributions, including the one on binary logic. (In fact I would be
happy to save all that Marcus writes, Marcus not only knows a lot about all
sorts of things but writes about them so even a beginner can understand, but
others probably don't find it boring).
Unfortunately we do not have time to read and understand all the above
right now, as we are supposed to be at Kiwanis at 9:00 fixing stuff.
The inability to download with zmodem (it freezes up the screen) may
be related to the other freezeup problem, we will try it on this computer.
WOuld it be possible to at least add to the uudecode part of the manual a
suggestion to use Dave's program (script) above if you get the error message
no end of line (or whatever it was)? Some day another person may try to use
uudecode on grex and start this same discussion again, otherwise. (And can
Marcus' explanation be part of the manual?)
Will report back as to whether we get this file to uudecode okay.
|
kentn
|
|
response 233 of 306:
|
Jun 1 23:28 UTC 1998 |
Re 231: I looked for flip, which I would also recommend (even over
dos2unix for most uses, since flip will do in-place conversions where
with dos2unix you need to redirect the output...although piping the
output through uudecode might work better with dos2unix) but I couldn't
find flip on grex, either.
In general, unix is all about learning new programs since huge monolithic
multi-functional apps are not welcomed, usually. You get by by stringing
two or more small commands together (and generally get more flexibility
that way, too). I wouldn't worry about learning the dos2unix command.
It's just a drop in the bucket of all the commands you could learn, and
if it helps you, do it. You will be ahead in the long run especially
if you use a unix system frequently.
|
senna
|
|
response 234 of 306:
|
Jun 2 02:03 UTC 1998 |
One wonders what item it is I'm in.
|
davel
|
|
response 235 of 306:
|
Jun 2 14:02 UTC 1998 |
Yes. Sindi, I really don't think these are Grex system problems.
The Info conference has been pretty near dead for quite a while now, but
that's been a really good place, in the past, for trying to debug problems
of this sort. I'd suggest entering a few new items there. (Or maybe I'm the
only one who still has that in a .cflist. <sigh>)
|
remmers
|
|
response 236 of 306:
|
Jun 2 15:31 UTC 1998 |
I have it in my .cflist too.
|
keesan
|
|
response 237 of 306:
|
Jun 3 18:54 UTC 1998 |
Thank you all, still trying to decode DELTA.ZIP, but I got a book on UNIX (for
dummies) from the library. Will ask further questions in Info.
|
keesan
|
|
response 238 of 306:
|
Jun 4 01:06 UTC 1998 |
I am happy to report that uudecode is working fine now, and that I may
have solved the problem of the screen freezing up by switching modems.
(The second modem kept disconnecting, but this one works okay). Still cannot
get zmodem to download, or even upload now, but Ymodem batch works, and
Kermit, in both directions. Time out to learn Unix for Idiots. Thanks to
all of you for all the lessons.
|
rcurl
|
|
response 239 of 306:
|
Jun 4 05:04 UTC 1998 |
You'll probably find that you already know more unix than does Dummies.
|
senna
|
|
response 240 of 306:
|
Jun 5 10:41 UTC 1998 |
My terminal's word wrap isn't working at the moment. Any idea what's going
on?
|
keesan
|
|
response 241 of 306:
|
Jun 5 22:11 UTC 1998 |
Wish I could help, Senna. Could it be some setting on your communication
program that needs changing? Rane, is a dummy or an idiotsmarter? This
book is for Idiots, not Dummies.
|
rcurl
|
|
response 242 of 306:
|
Jun 6 02:01 UTC 1998 |
Reasonable, if not good, question. A dummy just doesn't know much; an
idiot has incorrect ideas. Or, at least that's how I see it (at this
moment). It is harder dealing with idiots than dummies. Idiots probably
don't read many books; dummies might want to learn. So, what
sterling cutting-edge Unix knowledge is imparted to idiots?
|
mcnally
|
|
response 243 of 306:
|
Jun 6 05:28 UTC 1998 |
"Dummy" and "idiot" used to have fairly specific meanings that are
only partly related to their current usage (especially "dummy", which
referred to someone who could not speak, "idiot" at least still
connotes feeblemindedness, though generally of a different magnitude)
A similar generalization has taken hold of other related phrases (for
example "moron".) Does anyone know at what time the technical use
of terms like "idiot", "moron", etc. to describe levels of mental and
developmental retardation fell into disuse and whether it was
as a result of their being coopted as insults by the general population
or for some other reason?
|
rcurl
|
|
response 244 of 306:
|
Jun 6 16:32 UTC 1998 |
I think single terms like that cannot describe any complex mental deficits.
Medical terms today more and more refer to a syndrome that can display
a variety of features, or even more particularly to specific genetic
or congenital defects that result in a syndrome. So the improvements in
medical specificity obsoletes broad categorical terms.
However I don't think there is a specific syndrome or genetic defect that
makes learning Unix difficult.
|
