GM John, I have done some testing with MFSK64 with fldigi. I have not extended that testing to pskmail. I found the MFSK54 mode to be much more robust than psk250. The speed is also better than any modes except MT63 and RFSM2400. It seems to hit a very good spot of compromise between speed and reliability. Please reconsider your position. We need to make the most of this mode rather than remove it. Howard K5HB ________________________________ From: John Douyere <vk2eta@xxxxxxxxx> To: pskmail@xxxxxxxxxxxxx Sent: Thursday, November 13, 2008 5:08:20 AM Subject: [pskmail] Re: [pskmail] RE: [pskmail] Notes from the Zweibrücken meeting Hi All, Reviewing the notes, I noticed the comment about the new modes. MFSK64 was tested here to allow another dimension to the modes available and that was a higher data speed than psk250. In practical terms, the wide bandwidth I agree is an issue as a) it goes against the initial objectives of Pskmail of narrow bandwidth and b) it prevents the used of a narrow filter that is useful for the performance of the server. On the other hand I can use Thor22 with a 500Hz filter quite successfully. So in summary, since I am the one who made the original proposal for Thor22 and Mfsk64 to be integrated in the code, i would recommend that we remove Mfsk64. Comments welcomed. I am happy to take the latest clients and servers codes and perform the change if agreed. 73s, John (VK2ETA) On Thu, Nov 13, 2008 at 9:06 PM, Sven <sven98de@xxxxxxxx> wrote: Hi all, DAMA could provide a solution to collisions, but I don't see how DAMA would work on HF, because to keep control, routes must be stable, and HF isn't. 73 Sven --- Pär Crusefalk <per@xxxxxxxxxxxx> schrieb: > Hi, > > I have not received any updates, apart from Thomas > (thank you), on the > meeting notes from the participants so I figured I > may as well post them > on the mailing list. I know there are lots of > subjects discussed that > didn't make it into these notes but it was all I > wrote down (note to > self and others: its good to elect a secretary at > the beginning of the > meeting). > > The attached odt file contains the notes. For those > using pc's where you > are unable to read that format there is a badly > formatted text version > here: > > A few notes from the Zweibrücken meeting 20-Sep-08. > > Present were: > Rein, PA0R > Roberto, IS0GRB > Pär, SM0RWO ("elected" secretary of the > session) > Rolf, DL0IMA > Thomas, DJ4WL > Franz-Josef, DB3CF > Jörg, DL9YCS > > Matters discussed: > > Before taking notes (and being appointed secretary) > Rolf and Thomas gave > a presentation about Intermar and their activities. > They also described > the need for and use of a windows version of the > client. A gui design > idea was presented and the aim of that was to create > a client that would > run on windows and be custom made for sailing > yachts. We had a rather > lenghty discussion about this windows client and > neither yours truly or > Rein were very interested in taking on the job of > forking and building a > client for windows only. Personally, your secretary > here again, I would > rather see a platform independent client and I can > have a look at that > (perhaps run by cygwin on windows). But, the gui > design ideas that Rolf > and Thomas presented were very interesting and I can > see why they would > want a windows version. We should definitely > consider this and see what > can be done. > > Beacon policy > > There are now several servers with active beacons. > If a client should > try to connect during the first five minutes of the > hour then that > client will have a hard time, especially so on 30 > mtrs (10.148) as there > are several servers transmitting beacons. > > The meeting agreed that servers should only beacon > once per frequency > and hour. There may be concerns for servers using > several modes on the > same frequency, no decision was made on that. > > Note from the secretary: > The beacon was added when there was just one server > (sm0rwo) and one > client (pa0r). It was then a good way to let the > client see that the > server was active and what exact frequency it was > using. Now the traffic > intensity is much higher, with aprs beacons and > message handling, with > more servers and clients so the need for the beacon > is less obvious. The > total beacon amount should be in inverse proportion > to the amount of > other traffic on the frequency. > > QSY and common frequency on 30 meters > > Rolf expressed that it would be good for a client to > have a world wide > frequency for pskmail aprs. A sailing yacht crossing > the atlantic was > mentioned, it would be good to have a common > frequency so as not have to > hunt for servers on different frequencies there. > The meeting decided that we should move the current > european common 30 > meter frequency to 10.148 instead of 10.148,25 > (center frequency). Also, > it was decided that the frequency should be used > mainly for aprs/pskaprs > and connected sessions should be handled on other > frequencies. > Connecting on the frequency should be fine however, > the client should > then use the QSY command and move to the servers > traffic channel. Or, if > the server is scanning, the client should connect > when the server is on > the traffic channel. > > Note on Dial frequency and center frequency > > This is just a comment by your secretary, Pär. It > was discussed wether > frequencies should be written as dial (vfo) or > center frequencies. By > center frequency we mean the actual center > frequency, that is dial fq > +/- the center audio fq being used. An example is > 10.147 and USB > displayed on the rig, if we then use 1000 Hz as > "sweet spot"/center fq > then the actual transmission will take place on > 10.148. > If we only write dial fq then two stations using > different offsets > "sweet spot" will not be able to communicate. As > pskmail limits the afc > search range and servers use narrow filters it is > very important to know > where the actual traffic should be handled. That is > why the wiki only > lists the center frequency for each server, they can > then use whatever > sweet spot suits their equipment best (for instance > I use 900 Hz and > receive in cw using bfo offset). > > Development, client maintenance > > Rein declared himself happy with the current status > of the client. He > wanted to free himself of the maintenance workload > and have time to > develop the next version of the on air protocol. He > wanted someone to > step up and take over development, maintenance and > support of the > client. > Pär Crusefalk, sm0rwo, voluntered to handle the > client. (comment: I > don't know if we came to any decision here but I'm > ready to handle this > (sm0rwo)). > > FAQ and WIKI for the client needs to be updated, if > you see questions > repeated over and over on the mailing list then they > and their > respective answers should be added to the FAQ. > > New modes, MFSK-64 and THOR-22 > > THOR22 is currently being tested on the 80 meter > path between Australia > and New Zealand by John, VK2ETA. Patches has been > submitted by John and > those have been worked in to the main development > tree. > Of these perhaps Thor-22 is the most interesting, > its supposedly > (secretary has not tested yet) about as fast as > PSK-63 but very robust > and adapted to the QRN/M on 80 meters. MFSK-64 is 1 > kHz wide and that in > itself is a problem. > > New modem for pskmail > > Franz Josef (DB3CF) has isolated the fldigi mdem and > has a testable > version running with pskmail. > The idea is to fork the modem, make the modem as > lean as possible and > integrate it into the pskmail client and server. The > basic idea is to > migrate only the necessary controls into the GUI, > and set most of them > via a modem config file in ~/.pskmail. This means > that the modem can run > without X and can for instance be run as a daemon on > a server. Franz has > implemented the current posix pskmail-fldigi > interface but he has also > added an IP-based interface to this modem. Here a > discussion === message truncated ===