home *** CD-ROM | disk | FTP | other *** search
- ************
- Topic 27 Wed Jul 25, 1990
- H.HERMAN1 (Forwarded)
- Sub: A place to ask lingering questions...
-
- Here's the place to ask those lingering questions, that you've always wanted
- answers for, but were afraid to ask...
- 9 new messages.
- ************
- ------------
- Category 8, Topic 27
- Message 1 Wed Jul 25, 1990
- H.HERMAN1 (Forwarded)
-
- A few questions for the Sysops.... And, everyone else.... About
- things I've been wondering about:
-
- [1] 7-E-1 seems to work here, as well as the recommended 8-N-1. Simce
- 7-E-1 uses one bit less, wouldn't this actually speed up transmissions by
- 1/8th, or about 12%?
-
- [2] A file after being transferred by modem probably has acquired added
- padding. One dose of "padding" can't add too much, but if the file has
- seen its way through a few transfers, it may grow considerably larger than
- its original. Before releasing files are they first stripped of possible
- accumulated extra padding?
-
- [3] I sent a check to the author of a 128 mode shareware term. It has
- now been upgraded for SwiftLink, and the new release is also shareware. I
- do not know if I am supposed to pay again.
-
- Howie
-
- P.S. That 128 mode term is actually in pre-release form only for
- purchasers of SwiftLink. It will be made available generally, very soon,
- with some more additions.
- ------------
- Category 8, Topic 27
- Message 2 Wed Jul 25, 1990
- ED.BELL [* sysop *] (Forwarded)
-
- If you strip the hi bit of your transmissions, the only thing you would have
- is 7 bit values. That MAY be ok for text files, though
- *I* wouldn't do it (especially for PETSCII files), but for any kind of
- executable file, you would end up with garbage.
-
- As for padding, Xmodem protocols pad to some multiple of 128 bytes. Under a
- worst case scenario, the most accumulation of padding that could occur is 1027
- bytes. If you upload a file with Xmodem and it is padded out to the nearest
- 128 bytes, then download it using Ymodem, the 128 byte rounding could do 1 of
- 2 things. It could end up on an even 1K boundary, in which case the Ymodem
- would add no more (GEnie only, I'll explain in a moment) or it could lap 1 or
- more 128 byte blocks into an even 1K block. make sense. If it lapped, the
- most cumulative padding that could ever occur would be 1024 bytes - 128 bytes
- * number of blocks lapped. Check this out...
-
- XXXXX 5 Block file uploaded in Xmodem X's are 128 byte packets O-------
- Same file downloaded with Ymodem. The O's are the actual Ymodem packets. The
- '-' characters represent what those packets cover in terms of Xmodem packets.
- With this example, you can see another 3 128 byte packets could be added to
- the original file. Some terms do strip this stuff. I don't because the way
- most terms do it (Desterm, Novaterm, Wizard) are using an assumption that is
- contrary to Ymodem specification and under certain circumstances could strip
- away valid file data. Here on GEnie, Chuck Forsberg chose not to step down
- block size for the last 7 blocks of a file, meaning more padding for most
- transfers than necessary. But the scenario you mention has a limit. If all
- terms in the picture use only Xmodem, even if none of them strip padding, the
- program can get no bigger than the original upload. In the diagram above, the
- first 4 X's are all file data, and the 5th one would most likely be part file
- data and part file padding. So a Ymodem download would add its padding to
- this, but that would be the extent of the padding added, because in all cases,
- the file would now be an even multiple of 128 bytes.
-
- As for the Shareware, I don't think the author expects you to register twice.
- Registered users (Desterm) are automatically sent the next release. I think
- Matt asks $10 if you would like a laser printed manual.
- ------------
- Category 8, Topic 27
- Message 3 Thu Jul 26, 1990
- H.HERMAN1 (Forwarded)
-
- Ed, Thanks for the explanations. I quess that I just had a lot of
- mis-information about those things.
-
- Howie
-
- ------------
- Category 8, Topic 27
- Message 4 Thu Jul 26, 1990
- DIGITAL.DOC (Forwarded)
-
- Howie,
- Perhaps I have the answer to the 7-E-1 verses 8-N-1 question. I don't
- think you would see any speed up at all. In both cases, you are effectively
- dealing with 9 bits. 7-E-1 will send 7 data bits plus 1 parity bit plus 1
- stop bit: total 9. 8-N-1 will send 7 data bits, no parity bit, and 1 stop
- bit: total 9. To see the possible speed up you suggest, you would have to
- deal with 7-N-1, which would be 8 bits and should transfer data in 8/9 of the
- time. (Subject to the problems you would find dealing with program files)
- <Grin>
- <Doc>
- ------------
- Category 8, Topic 27
- Message 5 Fri Jul 27, 1990
- ED.BELL [* sysop *] (Forwarded)
-
- I believe that 8N1 is actually 10 bits/byte, making baud rate in terms of
- bytes a multiple of 10's. I believe there is a bit you are not counting in
- there, but overall, the only place it could be factored is in things like the
- BB, where at worst you would lose your capital letters. In file transfers,
- you would be in a lot more trouble, except, again, for text files.
-
- ------------
- Category 8, Topic 27
- Message 6 Fri Jul 27, 1990
- WC.COLEMAN [*Sysop*] (Forwarded)
-
- In both cases you are dealing with 10 bits - Doc forgot about the start bit!
- The only real difference between 8N1 and 7E1 is that 7E1 trades in the high
- bit for a parity bit - now that's what I call a waste! -WC
- ------------
- Category 8, Topic 27
- Message 7 Fri Jul 27, 1990
- ED.BELL [* sysop *] (Forwarded)
-
- I didn't know about the 7e1, but I did know it was 10 for 8n1. Thanks Bill!
- ------------
- Category 8, Topic 27
- Message 8 Fri Jul 27, 1990
- WC.COLEMAN [*Sysop*] (Forwarded)
-
- All RS232 bytes have 1 start bit, x stop bit(s), x data bits, and an optional
- parity bit. So 8N1 is 1+1+8+0 = 10 and 7E1 is 1+1+7+1 = 10.
- ------------
- Category 8, Topic 27
- Message 9 Sat Jul 28, 1990
- R.RANDALL5 [Zeroy] (Forwarded)
-
- It is only because of the serendipitous fact that most asynch transmissions
- are 10 bits per character that the term "bps" is equivalent to "baud".
- ------------
-