[pskmail] Re: The hardware modem idea (sri, long rant)

  • From: John Douyere <vk2eta@xxxxxxxxx>
  • To: pskmail@xxxxxxxxxxxxx
  • Date: Wed, 2 May 2012 16:32:44 +1000

Hi Jack,

Interesting points.

Actually I have been asked for a "push email" type of alert from Daniel
(VK2FWW) here in Sydney.

I think that with the advent of mobile devices people are becoming to
expect that new mail arrival will be at least flagged if not downloaded
automatically on their devices.

Although our bandwidth may not accommodate automatic downloads from a group
of people, even if the emails are size limited, I think it would be of
value to extend a feature like the Link-to function to allow for at least
the email count to be reported.

The auto-link function (also available in the Android version) sends a link
request at set intervals (typically 20 to 30 minutes) to maintain a
keep-alive flag at the server. At present this is only used for APRS
messages forwarding from server to client, but it could be used for new
mail alert too since a client is linked with one server at a time.

This would make marginal impact on the reply size of the Link-to request
and provide valid information to the client.

It also avoids frequency clogging by trying to push emails "in the dark"
hoping for the client to listen as only clients sending link requests would
get their new mail count.

This may not be a mailbox system as you are used to, but at least it can
alert (in may ways) the operator of new mails waiting. It is then simply a
matter of connecting and doing a QTC to get the email list and downloading
the desired ones.

I would not advocate the automatic email download personally simply because
of the potential for QRMing and clogging the HF frequencies (it is probably
OK on VHF at high speed).

Regarding your stripped-down platform point, since we have the choice of
phone or tablet (the 7" ones are pretty light too) you can have basically a
low power device with andPskmail running in the background and performing
regular link-to requests (and maybe position beacons too as I do).

I like the all-in-one aspect and you can look at it as having a keyboard
and display "attached to the modem". Otherwise you have to have another
device (like a phone or tablet or some sort of terminal to connect to the
modem).

On my HTC Desire I get a total concumption of 280mA from the battery when
running the andPskmail application with the screen off (Bluetooth, WIFI and
GPS OFF, but 3G ON). I think that can be classified as a low power device
even if you count 20% extra for the recharge circuit losses.

I also run my topographic mapping applications like Androizic (for
OziExplorer maps) and OruxMap. So I get a lot of value from my phone in
portable and low power situations.

Client to client exchanges are already built in jPskmail but disabled in
the phone version. They allow ARQ message exchange and file
up/download (the receiving end can be unattended) and that with automatic
mode adjustment like with the servers.

It would not be hard to re-enable this in the Android client.

73, John

On Wed, May 2, 2012 at 8:39 AM, Jack Chomley <radio@xxxxxxxxxxxx> wrote:

> Hi John,
>
> Thanks for your insight.
> I guess my reasoning for the hardware solution is because for many years
> that has been my main interest, from the early Packet days. Being able to
> run a Mailbox and Beacon, with no PC running at all. Being a
> mobile/portable operator meant that I could receive mail at any time while
> the PC was off and not having to switch it on, to check for mail......just
> look for a blinking status light on the box to know I had mail :-)
> secondly, I could load mail into the box and turn off the PC and IF anyone
> connected, they could collect that mail.
> In answer to your question about Pactor and Winlink.......several things,
> again It's the old mailbox problem, the Server is the mailbox and you need
> to log on to it all the time to check for mail, bit like having to go to
> the Post Office to collect mail, because you have no mailbox at home!
> In the days of FBB BBS operations, it would automatically forward to your
> mailbox and........you could reverse forward back to FBB.
> SCS have been asked on a number of occasions to add features to their
> Pactor firmware, one option was a Beacon mode, another has been an more
> friendly TNC 2 type mailbox, not to mention the latest DR-7800 modem has NO
> mailbox at all :-(
> Many years ago I asked  for a mailbox in SCS DSP Tracker TNC that has
> Packet and Robust Packet, I felt that it would be very useful for Ham
> portable operation, due to size and features etc.
> SCS refused on the basis that their products sales were only 5% to ham
> operators so they would not look at those features in any further firmware
> release.  People asked for Beacon mode in the PTC
> modems.............nothing :-(
> So John, Pactor does not hold all the answers for the minimalist mobile
> portable station, on a low power budget. Again, speed is nice as in P2,3,4
> but........it's not deal breaker!
> I favour lower speeds, in an effort to get Comms over more difficult link
> paths, with low power radios. With that comes the benefit of no cost
> Pskmail.
> As you can see Pactor does not hold all the answers that Pactor operators
> seek, mainly because the box maker is not flexible in looking at our
> firmware needs.
> As time goes on, our next big problem is bandwidth.......we need to get
> mean and lean ! As spectrum occupancy increases......the room to move,
> decreases. Bandwidth and filtering will soon be a big priority if links are
> to work with good propagation and high occupancy.
> The concept and operation of Pskmail is a great one.....it really is -:)
> I guess what I am looking for is how to get the most from it, without
> becoming entangled in a constant debug scenario, to make it work properly
> :-)
> Maybe the Android phone Is the answer.........would be good to gut the
> unwanted phone firmware out of it and just have it as a dedicated Pskmail
> Terminal -:)
> Client to client Comms with Android based Pskmail terminals would be
> great, too.
> Leave the Android box in sleep mode, could make it the mailbox, we seek -:)
> Again......just one mans opinion, in trying to contribute to progress -:)
>
> 73,
>
> Jack VK4JRC
>
> Sent from my iPad
>
> On 01/05/2012, at 12:53 PM, John Douyere <vk2eta@xxxxxxxxx> wrote:
>
> Hi Jack,
>
> I appreciate your input.
>
> From a purely personal point of view I have serious reservations about
> doing hardware development for anything if possible.
>
>  When I look at the trend of ever more powerful hardware in ever
> decreasing prices I really have difficulties justifying small volume
> hardware development. Just looking at the cost of a Nue-Psk (a very well
> designed system for its purpose IMHO) versus an Android phone or tablet,
> the cost benefits in my opinion put the mass-produced items well in front,
> even without considering the additional features of a smart phone.
>
> Yes a hardware solution provides a better control of all variable that are
> critical to top performance but I really think that we have ample speed and
> link (not system) reliability right now and certainly even more after we
> release the new modes.
>
> I agree with you Jack that some people will always want more throughput
> but we are not that far from the limit of throughput in a given s/n and
> bandwidth in fact.
>
> I agree that Pskmail user friendliness could be improved. I believe we
> should have a "Simple" and "Expert" mode where both functions and setting
> would be either hidden/at default or visible to the user. Simple mode would
> have only email, Ping and Beacon for example but not any of the other
> functions and most settings would be to default.
>
> Regarding reliability maybe we need to be more careful with what we call
> stable and really have two streams, one stable and one development in
> separate sections so that new users are not tempted to use the development
> path.
>
> We may also have to have a slower release path for Fldigi too so that new
> modem features do not interfere with the reliability of the overall system.
>
> Again personally speaking, I don't see why we should deviate from one of
> the key initial objective of Pskmail: to provide an open source and free
> solution when all (most?) others are paying and/or require proprietary
> hardware. I really feel that is a cornerstone of what Pskmail is.
>
> But still, I have a genuine question I would really like to have an answer
> for, and believe me absolutely exempt from ulterior motives: for those of
> us who have used/are using Pactor modems and Winlink why would we be
> interested in Pskmail? There must be some reason?
>
> All the best,
>
> 73, John
>
> On Tue, May 1, 2012 at 8:23 AM, Jack Chomley <radio@xxxxxxxxxxxx> wrote:
>
>> Hi Everyone,
>>
>> For a long time I have sat and watched the progression of Pskmail, and I
>> have to say that everyone involved with its development do a very dedicated
>> job.
>> But in my opinion, its development path has seen a major priority to
>> provide more and more features, and this comes with a
>> price.......increasing complexity :-)
>> It also means a constant flow of problem solving work, just look in the
>> past at all the posts, where this or that won't work......and it's only on
>> maybe some people's PC, or version of O/S.
>> This means that it's hard to provide a stable release of Pskmail that
>> works reliably.
>> I have had numerous tries at making it go, last week I put a stable?
>> version on a MacBook Air, but something is not quite right.
>> For those who are in the development circle, the resolving of problems is
>> not so hard as they are very familiar with the program and it's progress
>> along the way.
>> For an outsider, trying to climb onboard it becomes a big task to get it
>> all up and running.
>> IF Pskmail is to get more acceptance, it needs to be easier to set up and
>> get running.......
>> Whilst people generally look for feature laden software, I never
>> have......... I have always gone in search of reliability and simple to
>> operate setups.
>> If I can have a Pskmail like system that has very narrow bandwidth, the
>> speed of PSK31 and the reliability to decode very weak signals with ARQ
>> like Pactor that can run on a small hardware box, driven by an iPod Touch,
>> Android Phone,,or similar........then my portable, low power budget needs
>> are met. Because the hardware box can run as a standalone unattended
>> Mailbox, with radio and if required the PDA/phone can be used only to
>> read/send mail when needed.
>> However, the next person may want screens of swimming pools and movie
>> stars software :-)
>> No automatic mode changing to up the speed, if a signal level comes
>> up.........speed is not critical, reliability is..........
>> Yes, I agree that SCS modems are way too expensive and out of reach of
>> many people.
>> We see many low cost processor platforms now on the market, maybe we need
>> to look at some of those to embed the Pskmail modems and Mailbox/Beacon
>> functions into..........
>> Right now, my weapon of choice is an iPod Touch, driving a PTC-IIex with
>> FT-817 or Icom 703.
>> A great system using the simple Get Console software for the iPod Touch
>> or iPad that works well and has a Mailbox, but no Beacon. The big bugbear
>> is.......the cost of the modem and the fact that there are no cheaper
>> options for Pactor.
>> Yes, I also have the NUE-PSK modem which is a great unit, with a low
>> power budget, little bit heavier to carry, with keyboard but does a great
>> job, for what it is and is better cost-wise than Pactor and a far simpler
>> system.
>> Look, you cannot please all of the people, all of the time :-)
>> We all have different ideas on how something should be done........maybe
>> the greatest potential lies with John's Android Pskmail ?? Take an Android
>> phone with a spare battery, along with a pocket solar charger and radio,
>> that will do the job well.
>> The fact that Pskmail is free is to be applauded, but maybe it's reached
>> a crossroad? Does a software modem platform need to be used, that costs
>> money, or a piece of hardware?
>> What now, if a change away from FLdigi? Can someone come up with a simple
>> software ARQ modem, that does not require lots of processing power? Me
>> thinking of the trials and tribulations the WINMOR team had.
>> There are many considerations, and a hardware solution is just but one of
>> them :-)
>> Please understand this just my opinion, just one opinion of many and not
>> intended to denigrate anybody, or their ideas.
>>
>> 73,
>>
>> Jack VK4JRC
>>
>>
>>
>

Other related posts:

  1. Home
  2. pskmail
  3. Archive
  4. 05-2012