Brian,
Please don't misunderstand me. I am very appreciative of the work
done on RadioMobile and the fact that it's free! And, having worked
with RF for a long time, I'm well aware of how complex the issues
are. Which is why searching a Yahoo group doesn't strike me as an
ideal way to learn the program.
Some of what concerns me about RM are my general concerns with most
"free" software, and this is especially true for Linux stuff, where
"free as in speech" is a religious argument that often doesn't result
in this much of a product. The RF core of the RM package is
wonderful. I can just see a lot of ways in which the whole thing
could be even nicer if the UI were updated. Decisions that got the
job done up front may not be ideal in the long term. (You can see a
lot of that in how TCP/IP itself was written.) Commercial vendors
generally work on these issues -- somebody is paid to do the boring
crap -- because it helps them sell.
So if I am making suggestions, it is in the spirit of constructive
criticism, not putting down anyone's efforts. As a user, I can see
ways that it could be better. I am not a coder and don't know how
much work any such changes would be to implement.
But -- as a user -- wouldn't it be nice if it had a zoom function,
for instance, so you could blow up a detail of a combined-cartestion
coverage map or show-networks view without having to redraw the whole
thing? This might be practical if the computations were done on SRTM
points (to make the underlying map) but then to display it, those
results stayed put while you moved around and rescaled your view of
it. Lots of picture-manipulating programs do this. Heck, I'd pay to
have it, though I guess Roger's employment situation makes it
impossible to sell a premium version.
I hate how many Linux coders, when confronted with a problem in the
code, whine, "you have the sources, luser, fix it". Mozilla, at
least, pays rewards for some bugs, and it's constantly improving
(argh, another "important" update this morning!). Commercial
software companies pay for ideas to improve the product. I'm not
whining; I just spend a lot of time with RM (I've been using it for
several years) and just have ideas how it could be even better. As
no doubt do others. ;-) I'm sorry if you think I'm sounding
unappreciative of the work that went in to it.
At 7/22/2010 02:00 AM, Brian Webster wrote:
I'd like to chime in as one who does professional RF engineering for
a living. I have worked with Roger for the last 10 years helping to
improve radio mobile. There are so many features that the program
supports now compared to when I started it just boggles the mind.
That being said, it should be put in to perspective that commercial
software packages for RF engineering cost from $15,000 to $70,000
dollars and do not perform any better than Radio Mobile. As a
professional RF Engineer who has had to learn these commercial
tools, I can tell you it takes a huge amount of time to learn any RF
propagation software package and that there are no shortcuts or
dummies guides. There is no substitute for time and persistence in
learning any RF software package. Much of this will be spent in
trial an error and reading the archives and tutorials. There are no
shortcuts, mouse over or quick start guides. If RF design was as
easy as an idiot proof software package, then anyone with a mouse
could do it, there would be no need for expertise in the field.
Spend the time to read the archives in the yahoo group and the
tutorials. Impatience will not speed up anything and whining about a
set of documentation or quick tools tips will not make you and RF
design expert. Just because a person can point and click does not
mean you can generate RF coverage maps. Some knowledge of RF theory
helps so that one may understand the methods and settings within a
program. If you want hand holding, cough up the 35 grand or so to
purchase a commercial software package. I'm sure that is easier than
reading a little bit
.. If I seem sarcastic it is because I really
get annoyed at people who complain about a free lunch
..This
software package is free, quit complaining to the author about your
lack of ambition to read the many manuals available on line. For
those of us who understand the value of this free product, we
dislike those who would say things to possibly discourage the author
who donates his time and effort for free giving a product that would
otherwise cost in the thousands of dollars
.those who think they are
entitled just because they sit behind a keyboard and can complain,
need a reality check. If you don't like the product that you paid
nothing for
simply move on and be quiet. This was never meant to be
a for profit program
..accept it for what it is and be thankful
rather than complain. Any other options are far worse
Brian
From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org]
On Behalf Of Fred Goldstein
Sent: Wednesday, July 21, 2010 3:25 PM
To: WISPA General List
Subject: Re: [WISPA] Ping --- Radio Mobile Hates Me.
At 7/21/2010 12:37 PM, Scott Reed wrote:
If you have, join the Yahoo RM group. Lots of help there. Plus
links to at least 2 tutorials.
I've been there, and it helps. But it is not a substitute for a
good collection of documentation. There's useful stuff on Roger's
web site too, but it isn't always easy to figure out certain things,
like when to use which mode to use for a network (spot, accidental,
broadcast...) and what settings make the most sense. I use MapInfo
a lot and it has thick manuals, the unabridged one being PDF
only. Yes, it's expensive commercial software. I'm
spoiled. ;-) I suppose a wiki might be a way for the community to
collect its thoughts.
I did see some interesting discussions on the Yahoo group about the
nodes, and about the land cover. I roughly doubled the forest loss
numbers, from Roger's default. This still might not be adequate,
though, since it makes it seem *possible* to blast 5.8 GHz through
the woods. Is 180 a good setting for most forests?
Roger does this for a living and his employer sells a very nice
commercial package. They have been nice enough to allow him to to
RM for free, so we get a super program at no cost. This also means
that Open Source is out, as I am sure the source is too similar to
their commercial package. If you want the pay version, I am sure an
e-mail to him would get you company contact information.
What is the commercial product? He certainly hides any mention of
it. If it's reasonable, I might look. I remember seeing an add-on
for MapInfo, though. The price was roughly similar to the price of
the local calling area database license. My car cost less, new.
SPLAT looks to be a somewhat similar open source program, but much
more limited in scope and not nearly as well updated. This is
complicated stuff, I know. About 3/4 of the confusion might be
solved by having a "mouse-over help" function, where you could
right-click on a box and pop up a tutorial on what the values mean
and how to set them. That could be an interesting volunteer
project. Of course Roger's primary market is 2 meter repeaters, so
the parameters we use in the WISP bands are a bit different...
...The only time I would see a need for antenna patterns is if you
have a fixed-base AP and mobile CPE. If both are fixed-base, I am
not sure what the patterns will gain you. I do the same thing; I
have a 5.8 network, a 2.4 network and a 900 network. Most of my
POPs are setup with 3 120* sectors, so all POPs are setup with an
omni of the same gain as the sector antenna. In my experience so
far, the results are fairly accurate when there is clear
line-of-sight. If there are a significant number of trees in the
path, it obviously is not so good. I suppose if you have 2 90*
sectors trying to cover 360* you would want patterns to find the
nulls and edges, but if you have antennas for full coverage, the
pattern probably is not so important. For point to point links,
antenna pattern does not matter , assuming you are planning to aim
the antennas directly at each other as that is the assumption RM makes.
Not all of the sectors need full-circle coverage, so I was thinking
about using the model to see how it looked with partial coverage on
some poles. This would save radios and antennas... In fact, with
three sector radios and two backhaul radios (not to mention needing
three backhaul radio "degrees" at mesh junctions), that exceeds the
four-slot maximum of any one Routerboard, right? So do you often
put back-to-back radios in one box?
I think the only way to do sectors in RM is to treat them as
separate radios, So if Unit 10 was three sectors, it might end up
as say Units 10, 91, and 92, in the access network, right?
The Yahoo group has also had discussions about exports and
imports. There are several things you can do. Again, check out the tutorials.
I would have to disagree about the need for many
improvements. Granted, I have been using it for over 5 years, but I
find everything to be where expected and do what it should. Roger
is open to suggestion, though. Let him know what you would like to see.
I don't want to disparage Roger and his great work; it's just little
things. I just hate drop-downs, which RM's UI makes me use too
often, especially for selecting radios. But also the fact that
adding a radio requires going to both the unit properties and then
the network properties is counter-intuitive and a bit clumsy. These
sorts of things aren't show-stoppers, just places where it helps
reduce one's sanity just a bit more. Which can be in short supply...
Fred Goldstein wrote:
At 7/21/2010 11:41 AM, MarlonS wrote:
Radio Mobile hates everyone that doesn't use it every day.
It's a great tool, but boy is it frustrating! Roger has done a
wonderful thing by putting this out there for free and improving it
as he has. But there are so many things that could be done to
improve it, especially the clumsy user interface. If it were an
open source project, then more people could contribute to the
effort. If he had a premium payware version, then he'd have
incentive to at least prettify the pay version.
Documentation wouldn't hurt either...
My current project has set up three "networks" using the same batch
of nodes. One is 5.8 GHz backhaul. One is 900 MHz backhaul, for
heavy-forest paths. One is 5.8 GHz access. When it does the "show
networks", it doesn't seem to find the best path, but it's not
terribly predictable as to which common "network" it's using. So I
end up having to do path-by-path comparisons anyway.
My next chore is to add antenna patterns. I think this means taking
each node and turning it into two or three nodes, if it has two or
three separate sectors. I can "save network" as a CSV, but that
seems to only save the node locations. Copying network parameters
between projects seems impossible. :=(
grin
marlon
----- Original Message -----
From: <mailto:robert.w...@just-micro.com>Robert West
To: <mailto:wireless@wispa.org>WISPA General List
Sent: Tuesday, July 20, 2010 7:38 PM
Subject: [WISPA] Ping --- Radio Mobile Hates Me.
Ping.
(Had to)
Bob-
Still fighting the animal that is Radio Mobile.
--
Fred Goldstein k1io fgoldstein "at" ionary.com
ionary
Consulting <http://www.ionary.com/>http://www.ionary.com/
+1 617 795 2701
--------------------------------------------------------------------------------
WISPA Wants You! Join today!
http://signup.wispa.org/
--------------------------------------------------------------------------------
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
--
Fred Goldstein k1io fgoldstein "at" ionary.com
ionary Consulting http://www.ionary.com/
+1 617 795 2701
--------------------------------------------------------------------------------
WISPA Wants You! Join today!
http://signup.wispa.org/
--------------------------------------------------------------------------------
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/