| |
ZL3AI > APRDIG 04.04.03 10:59l 239 Lines 10698 Bytes #999 (0) @ WW
BID : 1113-ZL3AI
Read: GUEST
Subj: TAPR Digest, Mar 31, 2/6
Path: DB0FHN<DB0RGB<DB0MRW<DB0ERF<DB0FBB<DB0GOS<ON0AR<ON0AR<IK1ZNW<RZ6HXA<
SP7MGD<VK7AX<ZL2BAU<ZL3VML
Sent: 030403/0948Z @:ZL3VML.#80.NZL.OC #:63198 [Chch-NZ] FBB7.00i $:1113-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: RE: Solution for adding more serial ports
From: "Gregg G. Wonderly" <gregg@skymaster.cytetech.com>
Date: Mon, 31 Mar 2003 08:28:06 -0600
X-Message-Number: 7
>I remember putting out a query over the sig
>last Oct but other than changing OS and changing APRS program suggestions
>(which I did) I didn't receive any other
>suggestions. I realize I may be pushing the limit and it is very difficult to
>work this type of problem from long range.
....
>There is undoubtably a better solution than what I'm using but I don't think
>a multiple MULTIPLEXER is the answer.
Microsoft has been very clever and deliberate in getting people to pay the
most for using their software. The windows 9X family (up through and
including windows ME), have had design decisions involving deliberatly
restricting the amount of system resources that are available to
applications. The GDI resources (graphics device interface) are notoriously
limited. This means that you can't run very many applications at the same
time, specifically those with complex graphical interfaces.
But it goes beyond that. Internet resources and memory usage and threading
are all designed to keep big business from being able to buy cheap
software. MS wants all the businesses to but windows-NT family OSes because
that is where they make a lot of money. Windows NT4 was a big success for
most shops, because it stabilized that platform at a reasonable mix of
compatibilities between win 9X and win NT. Windows 2K came out, and then
XP soon after that, and microsoft had to deprecate NT4 support to get
people to move off of it.
There is no competition in that market. It is monopolistic and very
focused on MS making more money. If you think the pricing is fair, and
that MS has done a good job, then you need to look at Mr. Gates worth and
decide if anyone really needs to be 'worth' that much at the expense of all
the other software companies in the world who have much better
technologies.
Consider that Mac OS-X is now running on top of BSD UNIX. And look at what
windows XP has finally become. It is a multitasking, multi-user operating
system, that has APIs that are incompatible with other platforms (directly,
sure you can delegate through #define and/or libraries like cygwin). But,
it now does what UNIX has done for decades.
Back in the mid 1980's, there was considerable exchange between the MSDOS
crowd and the UNIX zealots on the comp.os.* Usenet news groups about MSDOS
needing real resource tracking, and real scheduling. The MSDOS crowd just
said again and again, "Programmers need to write bug free code, they're
just lazy." And now, lo and behold, everyone has discovered that software
is not perfect, that people are incapable of doing things correct in
software, and its is actually best if you do have the OS protect the user
from misbehaved software etc...
It has taken about 15 years for the MSDOS crowd to learn this.
UNIX was publically visible in the 1971 time frame, and the mid 1980's
would be about 15 years later. The UNIX crowd became educated to the
benefits of their environment through the same period as it took the MSDOS
crowd to grow up their OS.
Live and learn...
So, I can only tell you this. Computers crash because the software stops
making the hardware work in a predictable fashion. The way to make this
happen less is through smart OS design, and deliberate separation of
application and OS responsibilities. The less the OS is responsible for,
the more it can get 100% correct. That way, user level code crashes, and
one thing stops working, and can be restarted through simple measures. If
the OS stops working correctly, some resource will most likely become
unavailable. Typically, on the windows family of OSes, it is the processor
that is not longer usable :-)
For any application such as APRS, that you expect to run for a long period
of time, you need to use an OS and related software, that is engineered to
work in that environment.
Windows has historically been a single user machine, and it stops working,
one person is inconvenienced. When file servers were added, then file
shares were inaccessable, but that was usually worked around by doing
something else locally.
UNIX grew up in the terminal connected to computer environment. When a
UNIX machine crashed, everyone stopped working. This was BAD. When file
sharing via NFS was added, workstations booted off of NFS shares, and used
NFS shares for everything. When an NFS server crashes, everyone stops
working! This is BAD!
But, what happened is that UNIX was tuned for a 24x7 and limited impact to
the users environment, because of the age that it was developed in. Where
as, windows 9x has never been tuned for multi-user environments, and this
shows through in its continued troubles with running more that one thing at
a time.
People jump up and down here about using UNIX/Linux/BSD for APRS. When you
see that happening, you should pay attention. These are people that have
the experience and the knowledge, and have come to realize where the real
problems are!
-----
gregg@cytetech.com (Cyte Technologies Inc)
----------------------------------------------------------------------
Subject: Re: XASTIR help needed
From: Gerry Creager N5JXS <gerry.creager@tamu.edu>
Date: Mon, 31 Mar 2003 08:38:25 -0600
X-Message-Number: 8
USGS mapping standards include a requirement for a minimum metadata
dataset. the issue is not that the data isn't supplied, but rather that
most of the general users don't get the associated metadata. This is
NOT simply an amateur radio problem. It's a problem we've been
wrestling with for years in the GIS community.
"If you don't have the metadata, you don't have the data. Period."
I'm on a State Agency-level council in Texas where we've engendered a
GIS rule requiring a minimum set of metadata (consistent with USGS) that
must be supplied whenever an Agency makes the data available. It's
pretty sparse, yet, a lot of GIS analysts think it's too much work.
Then, they criticize colleagues and other users when their data is
misrepresented.
73, gerry
----------------------------------------------------------------------
Subject: Re: Need microphone extension cable for D700A
From: radio_q@juno.com
Date: Mon, 31 Mar 2003 08:25:23 -0700
X-Message-Number: 9
On Sun, 30 Mar 2003 20:59:37 -0800 Richard Parry <rich@hercomputer.net>
writes:
>Thanks for your quick replay. I gave it a try with cable and coupler
>that I
>have, but it didn't work. I think I need either a CAT 5 that
>reverses the
>wires or a coupler that reverses the wires. In either case, I know
>what to
>do thanks to you.
>
>thanks again,
Hi Rich
Do not get a 'reverser' on either the cable or the coupler!
You may have a bad coupler as a friend of mine bought
one at radio shack and we fiddled for ever to figure out
why it didn't work. We finally went in the garage and
got the coupler out of my car and it worked fine. He took
the RS one back and they replaced it and Presto, all was fine.
There were some comments about needing shielded cable
and some details of how to shield the cable etc. I worried
about that when I did mine the first time but it was never
a problem for me. I have extension cables on the IC-706
HF rig and the IC-2710 in my RV. I have them on the
two IC-2700's in my cars and one on the D700 here in
the room next to my HF radio. No problems with any of them.
Good luck
Jim B
----------------------------------------------------------------------
Subject: Re: APRS development?
From: "Curt Mills, WE7U" <hacker@tc.fluke.com>
Date: Mon, 31 Mar 2003 09:16:23 -0800 (PST)
X-Message-Number: 10
On Fri, 28 Mar 2003, Jim Duncan wrote:
>It occurs to me that we seem to go through this same debate every year
>about this time. Hasn't anyone ever noticed that there are ALWAYS major
>APRS releases in conjunction with Dayton?
Plus Xastir releases happen roughly every two weeks. There should
be another one tomorrow or soon after.
--
Curt Mills, WE7U hacker_NO_SPAM_@tc.fluke.com
Senior Methods Engineer/SysAdmin
"Lotto: A tax on people who are bad at math!"
"Windows: Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me: I picked the coordinate system!"
----------------------------------------------------------------------
Subject: FW: RE: Solution for adding more serial ports
From: "Ray McKnight" <shortsheep@worldnet.att.net>
Date: Mon, 31 Mar 2003 09:25:38 -0800
X-Message-Number: 11
Well, your experience has probably killed any hope of considering another
attempt.
I never suggested that ALL multiport serial servers worked well. And I
certainly don't endorse those that are merely port multiplexers. Ask anyone
who's used RocketPorts - I doubt you'll be able to find once unhappy
person, nor any compatibility complaints. They work better than just about
ANY computer accessory I've ever used. Period. And I've been around since
the days of loading 4KB of program space with front panel toggle
switches... I've max'ed out servers with 128 ports using Rockets and never
have seen a single glitch or lost byte.
I just added one to my tired old PII-400 for the ham shack. It installed in
about 2 minutes, no drivers to load (under Win2000), and all eight ports
popped right up, just where I wanted them (I disabled my 2 internal ports
in favor of letting the RocketPort provide COM 1-8 - it's ports are far
superior to the internal 16550 UART).
The wide variety of motherboards and accessories available is responsible
for the majority of hardware compatibility issues. Have you bothered to
look at Microsoft's compatibility list to ensure your system has been
certified for Windows? And how about every card you stick into an expansion
slot? Even then it's a crap shoot with many things. I doubt that serial
card you added was as well tested. If it was a multiplexer, well, that's a
bad design from square one.
Several people have knocked the RocketPort approach on the SIG, and I will
bet money that none of them have ever used one. If they had, they would be
singing its praise from every rooftop. I keep hearing the same old story -
"add more ports using USB converters" - okay, now we listen to msg after
msg of people who tried that and about 50% of the time it doesn't work. And
who wants 5 or 6 USB dongles strewn about the desk - it's sloppy and a
xxxxxxx approach to a system that is proven and well documented in many
professional and scientific circles.
[commercial content deleted]
----------------------------------------------------------------------
Read previous mail | Read next mail
| |
