| |
PA2AGA > HDDIG 27.02.00 21:44l 173 Lines 6418 Bytes #-9565 (0) @ EU
BID : HD_2000_57D
Read: GUEST
Subj: HamDigitalDigest 2000/57D
Path: DB0AAB<DB0ZKA<DB0CRL<DB0TTM<DB0SWR<DB0HBN<DB0SON<DB0ERF<DB0BRI<DB0SM<
PI8DAZ<PI8GCB<PI8HGL<PE1NMB<EA7URC<PE0MAR<PI8VNW
Sent: 000227/1706Z @:PI8VNW.#ZH2.NLD.EU #:54705 [HvHolland] FBB7.00g24
From: PA2AGA@PI8VNW.#ZH2.NLD.EU
To : HDDIG@EU
Received: from pa2aga by pi1hvh with SMTP
id AA31037 ; Sat, 26 Feb 00 19:31:33 UTC
Received: from pa2aga by pa2aga (NET/Mac 2.3.70/7.5.3) with SMTP
id AA00018141 ; Sat, 26 Feb 2000 15:34:00 MET
Date: Sat, 26 Feb 00 15:33:05 MET
Message-Id: <hd_2000_57D>
From: pa2aga
To: hd_broadcast@pa2aga
Subject: HamDigitalDigest 2000/57D
X-BBS-Msg-Type: B
course they could have no relevance or utility.
>
> HF protocols are good enough for the limited channel they have to
> operate in. Bits, bytes, who really cares anymore?
Hams do. Too bad Steve doesn't. (In a way... In another way, it's just as
well.)
Steve has done a great job here, as usual, in descibing the intellectual
limitations and character flaws inherent in the LandLine Lid community.
Thanks, Steve!
--
73 DE Charles Brabham, N5PVL
N5PVL @ N5PVL.#NTX.TX.USA.NOAM
http://www.texoma.net/~n5pvl
>.
------------------------------
Date: Fri, 25 Feb 2000 16:59:33 +0200
From: Paul Keinanen <keinanen@sci.fi>
Subject: What is a good TNC?
On Fri, 25 Feb 2000 11:32:07 GMT, nomail@rob.knoware.nl (Rob Janssen)
wrote:
>Paul Keinanen <keinanen@sci.fi> wrote:
>>I much rather use byte oriented protocols to transfer octet aligned
>>data, since much less processing have to be done in software (the
>>situation may be different e.g. when transferring 11 bit image samples
>>or other variable bit length data, in which case a bit oriented
>>protocol is referable).
>In reality, there is usually no difference. Bit-oriented protocols are
>usually handled by communication controller chips (like the SCC) that
>internally handle 8 bits at a time, and store the bits as bytes in memory.
>The only viewable difference is that the last byte stored need not have 8
>bits actually valid, so there should be a "bit count in last byte" dragged
>along with the data through all subroutine interfaces. This is normally
>not done, and an 8-bit multiple is implicitly assumed.
>The software that wants to send 11-bit samples usually has to put these
>in memory in a contiguous fashion, and they will be sent just like any
>8-bit data. No difference between bit and byte oriented protocols....
When 8 or 16 bit characters or 16 or 24 bit samples are to be
transferred, these entities can be very effectively constructed from
8 bit wide octets in all computer architectures, since most
architectures are byte addressable, not bit addressable.
However, if some strange character/sample length (say 11 bit, or even
worse varying according to some predetermined system) is to be
transferred efficiently in a packed form, thus, when the receiver chip
stores this streams into 8 bit bytes the sample boundaries are
different from byte to byte. In order to be able to postprocess these
samples, they must be converted to one sample/word, which typically is
16 - 32 bits. Thus, the 8 bit bytes have to be split into bits before
being reassembled to 11 bit samples, each of which is to be stored
into a 16 or 32 bit computer word. Thus, it would be preferable to
have the input data directly stored by the receiver as 1 bit in each
byte, in which case the bit stuffing removal can as well be made in
software.
Paul
>.
------------------------------
Date: Fri, 25 Feb 2000 18:11:30 GMT
From: "Ron" <nospam@wanted.com>
Subject: X1J4 question
You didn't tell us exactly which TNC you are using? MFJ1270C Rev 11 ?
Ron
"Rob Dover" <rdover@bclc.stamp.out.spam.com> wrote in message
news:38A86DCE.CD0CF759@bclc.stamp.out.spam.com...
> Greetings.
> I have a number of X1J4 nodes that I look after which all exhibit the
> same intermittant problem. Periodically the radio port will lock in
> transmit. The hardware watchdog timer kicks in and shuts the transmitter
> down but the TNC is still locked up. This isn't critical as long as
> there is another TNC in the node stack which allows me access through
> the serial side so I can send a reset. If the TNC is standalone, a site
> visit is required. Does anyone else see this or know if there is a fix?
>
> And before you comment UZI, yes I have tried to get Flexnode but they do
> not seem to want to respond to the e-mail listed on their web site! If
> there is another way to get an evaluation copy please let me know.
> Thanks 73 -Rob-
>
> Reply to: is spamblocked. Remove the obvious to reply by e-mail
>
>.
------------------------------
Date: Fri, 25 Feb 2000 13:31:22 -0800
From: Rob Dover <rdover@bclc.stamp.out.spam.com>
Subject: X1J4 question
It varies. We have MFJ1270b's and C's, MFJ1274's, PacComm Tiny 2's and
Spirits. All exhibit the same problem. The only commonality is that they are
all running X1J4. I don't even know what the trigger is. The node will be
happy as can be for weeks or months and then locks up. Strangely though, when
this does happen it usually happens in bunches. i.e. the network will be fine
for a period, then we get a bunch of nodes act up over a few days and then
after resetting them everything goes back to normal. It is almost like
something traveling through the network upsets the balance.
-Rob-
Ron wrote:
> You didn't tell us exactly which TNC you are using? MFJ1270C Rev 11 ?
>
> Ron
> "Rob Dover" <rdover@bclc.stamp.out.spam.com> wrote in message
> news:38A86DCE.CD0CF759@bclc.stamp.out.spam.com...
> > Greetings.
> > I have a number of X1J4 nodes that I look after which all exhibit the
> > same intermittant problem. Periodically the radio port will lock in
> > transmit. The hardware watchdog timer kicks in and shuts the transmitter
> > down but the TNC is still locked up. This isn't critical as long as
> > there is another TNC in the node stack which allows me access through
> > the serial side so I can send a reset. If the TNC is standalone, a site
> > visit is required. Does anyone else see this or know if there is a fix?
> >
> > And before you comment UZI, yes I have tried to get Flexnode but they do
> > not seem to want to respond to the e-mail listed on their web site! If
> > there is another way to get an evaluation copy please let me know.
> > Thanks 73 -Rob-
> >
> > Reply to: is spamblocked. Remove the obvious to reply by e-mail
> >
>.
------------------------------
End of Ham-Digital Digest V2000 #57
******************************
You can send in your contribution to this digest by
sending an e-mail to: hd-group@pa2aga.ampr.org
or (via BBS-net) to: hdaga@pi8vnw.#zh2.nld.eu
Read previous mail | Read next mail
| |
