开云体育

Hello everyone! I've been running the EchoLink and IRLP newsline server for more years that I can remember. I retired last and now that I'm living on my savings and social security I need to minimize my monthly costs... The server consists of two instances of thelinkbox plus a couple of scripts and configuration files running on a virtual private server that I rent for $16/month. The person that takes it over needs to be knowledgeable about thelinkbox and Linux. Someone that already has a cloud server with extra bandwidth would be ideal. I doubt much bandwidth is needed. The person should also have a good relationship with the Echolink and IRLP admins. Please DM me if you are interested in taking over the responsibility for the service. Thanks & 73's Skip WB6YMH

You're welcome. I didn't do much, but looks like it was a minor configuration error that confused tlb. Commenting out the offending line did the trick. For everyone else following, don't use a hostname for SF_Bind2IP. This should only ever set to an IP address, and even then, only if you know you need this setting. Most people don't need to use it. Most common usage scenarios for Bind2IP or SF_Bind2IP are when you're running multiple instances of tlb or tbd on the same physical host that has multiple IP addresses, or where you want to be certain what address tlb reports to the servers or the remote peers it connects to. And it seems using a hostname confuses tlb internally. It binds to 0.0.0.0 (all IPs), but doesn't seem to properly pass traffic on the Speak Freely (IRLP) ports. I consider this a bug in tlb, because the configuration parser should either reject this as an invalid IP(v4) or perform a DNS lookup and obtain the actual IP. The first option (rejecting anything that's not an IPv4 address) is the safest. In this instance, connections could happen on a couple of IP addresses, so you actually don't want to set SF_Bind2IP here, hence the reason I commented it out, rather than setting it to a valid IP address. On 13/7/24 2:58 am, Enrique XE1E via groups.io wrote: -- 73 de Tony VK3JED/VK3IRL http://vkradio.com

Hi guys, M17 has I believe a CAN number and uses the codec2 1600 voice codec. FreeDV can transmit/receive mode 1600 but in the modem, as it stands, doesn't have any 'extra' data other than the voice that we could use to signify to a receiving system, put it as a valid signal to such as TLB. Do we have any intersted people here who could encourage the FreeDV developers to think about this. Or, since M17 uses an old Codec2 library which includes a modem, could somebody modify or change the modem to include in the transmission, this CAN number? Just thinking --------------------------------------------------- Alan VK2ZIW Before the Big Bang, God, Sela. OpenWebMail 2.53, nothing in the cloud.

Hi All! Has anyone solved the problem, TLB with an active radio port as an EchoLink conference, with ID *CONFERENCE*? How does the TLB identify itself on the EchoLink server? And how can I order him to identify himself as TBD? I tried changing, configure.in: -- AM_INIT_AUTOMAKE(thelinkbox,0.56) ++ AM_INIT_AUTOMAKE(thebridge,1.09) But it didn't help. I still get "Not logged in".

Hello and Happy New Year! I am working on a project to integrate AllStar (chan_tlb) with an existing tlb installation that is being used as an IRLP experimental reflector. I have searched around and found various posts and believe that I have both sides set up correctly, however I am unable to connect the allstar side to tlb. I don't see anything in the allstar logs, nor the tlb.log on the tlb side, so it makes me think the connection attempt isn't being made, hoping someone can shed a little light and see if maybe I missed something... I have both tlb and allstar running on a single VM, running debian 10 tlb.conf from the allstar side and tlb.conf from the tlb side are attached to this post.. Thanks! Jim, K6JWN

Hi, The developer team for codec2 (FreeDV) in the latest iteration decided not to include a Destination byte or even a nibble. 2022-2023 Thus, networking cannot be done !!!! Back nearly ten years ago, when the Codec2 mode 700D came about, I saw the possibility of networking. As did another programmer. But FCC rules: No repeaters on HF., the US developer team, having taken over from the Australian protocol designer, said in email to me "not doing it". So, I'm asking for help in programming. Everything is open source and available on github. The library is in C and the app C++. In the 700D and later modes "free text" data has had FEC added to improve it's reliability for the Reporting feature, to http://qso.freedv.org. We, at this point, cannot implement codec2 repeaters on 10m and up as there's no way to determine if the received signal wants to be repeated or not. (As CTCSS tones do for VHF FM) So again, can I ask for programmer help? I do run a "parrot" repeater. It "looks' for ++ in the free text. Though can't see ++ in this new FEC corrected text. (frequency and mode as shown on qso.freedv.org) 0x73 Alan VK2ZIW --------------------------------------------------- Alan VK2ZIW Before the Big Bang, God, Sela. OpenWebMail 2.53, nothing in the cloud.

Due to some Virmach issues I moved my cloud host to a new provider/server and in the midst of that Echolink contacted me about running in multiple places. all that's ok now but they mentioned the version of Echolink I'm running is old: " the version number appears to indicate a very old version of theBridge." Can the LinkBox upgrade this without breaking anything?

Reason for selling them: Moving to Australia Icom IC-7800 ... $3,800 ICOM IC 9100: ... $ 2,600 IC-756 Pro III .... $815 Icom PW1: ... $1300 ICOM 7610: ... $2,105 Elecraft K3 w/ P-3 ..... $1,650 Flex 6300 : ... $2,200 Flex 6600 .... $3,700 Yaesu FTdx 5000 .... $2,500 Alpha 87A: .... $2,050 Yaesu FTDX9000d: ...$3950 Yaesu FTdx-101D: ...$2405 Yaesu FTDX 5000,Yaesu FTDX-3000 Alpha 9500: ... $3750 ACOM 2000A: ... $3250 Acom 1200 Amp: ... $2990 AMERITRON AL 1500: ...$2,300 Kenwood TS-2000LE, KENWOOD TS-990S: ... $4,100 Callsign: K5PHL Name: Philip C Robinson Email me: k5phl.73@... Phone: 308 409 0084 73

Greetings, I've cloned the repo and built TLB on a RPi 4 running 64-bit Raspbian Bullseye. I did the patches for bLogCmd to get it to compile. I got an echolink connection going (it's at W6GKD-L). I created a port0.conf file and set the audio device. I am using a StarTech USB sound card and it shows up at /dev/dsp3. I can run a "speaker-test" to confirm it can output sounds to an external speaker. I don't have a /dev/parport0 so presumably I need to use something like /dev/gpiochip0, which is currently streaming endless IOCTL error messages so probably not EXACTLY that. I need to know how that part works, I need to know what happens with the hardware side. I've been looking at IRLP hardware and an ICS-controllers PI-REPEATER-2X so I am not totally ignorant. I am just doing a test set up right now on my bench. The goal is to link 6 repeaters together. Baby steps. Does someone out there have any experience setting up a RPi or can you reference a recent web page? 73s -- Brian H Wilson W6GKD Astoria Oregon tel:707-827-0001 -- Brian Wilson W6GKD Astoria Oregon

I'm a huge fan of notifications and I've got my Allstar setup to Telegram (or Email) me when a station connects or disconnects from the system because you can call a custom script for it from the connection event. Example: https://github.com/Russell-KV4S/ASL_scripts I would love to have a way to get the same kind of notifications for The Link Box when a station connects and disconnects. If it already supports this and I just need to set it up, please let me know how! Thanks, and magnificent work all around on the project. I really enjoy the software and system it works great and is exceptionally reliable! 73, Russell, KV4S

Hi Tony, Volodymyr and all, When I read, from the standard: For the third generation of Ettus devices, a new type transport protocol was designed. It reduces the complexity of the original standard and uses a fixed-length 64-Bit header for everything except the timestamp. Because it is "compressed" into a 64-bit header, it was dubbed CHDR. Where does this fit in with first, both the source and destination addresses (callsigns, DMR IDs) and the requirement at the Radio end to reduce the data rate below 5000 bits per second? We could have: 100 radios connected to a VKF/UHF repeater. That repeater could be connected to 10,000 other repeaters. And those repeaters to another 100 radios. And, of course, those radio users may well want to belong to different talk groups. So, 32bits for just one, either source or destination, I see as problematic. In my HF interest area, we have to limit the data rate to around 1000 bits/sec. So, there's very very little data space for callsigns and destination addresses. I guess, what I'm asking for such as "thelinkbox" is a transport for both the "high data-rate codec" and the "low data-rate codec" data such that, when that data arrives at the RF equipped node, it could go out as, plain audio or DMR/P25/D-Star/Fusion data or Codec2 data. Leaving the final "codec data to voice" to, the last leg, the user radio. Right now, just for me, I'd like to see a link protocol such that we could implement full duplex HF repeaters because they require separate Tx and Rx sites. Reason: The "parrot" (store and re-transmit) 23cm repeater here in Woolongong shows, it's problematic. The time waiting for a reply. 0x73 Alan VK2ZIW On Wed, 18 May 2022 07:52:39 +1000, Tony Langdon wrote > No it doesn't support USRP at all. That's something I'd like to see as well. > > On 18/5/22 1:55 am, Volodymyr wrote: > Hello! TLB, does it fully support the USRP protocol? And metadata too? Best regards, Volodymyr > > -- 73 de Tony VK3JED/VK3IRL http://vkradio.com --------------------------------------------------- Alan VK2ZIW Before the Big Bang, God, Sela. OpenWebMail 2.53, nothing in the cloud.

I am seeing an odd issue using Skips method of playing audio files followed by a script to disconnect for ARRLNEWS. e.g. 'tonegen -c "script scripts/disconnect.sh"' "file audio/news/arrlan.wav" The disconnect.sh script is never called with the recent AAN-2022-05-13 news file. I think this is the first time this issue has occurred. I suspect it is related to one of my recent tlb modifications for COS/PORT TX/RX indicator or audio silence detection and pause. I've only attempted limited testing and debugging so far. The disconnect.sh script seemed to be called as expected with last month's audio file. There seems to be a few seconds of audio left in a buffer that gets replayed periodically several minutes after the audio file completes. Perhaps one of my changes prevents an audio buffer from being fully flushed, preventing execution of the disconnect script. I don't know yet. I wanted to give notice of this issue in case anyone is using my modified version of tlb. I have temporarily implemented a secondary disconnect script via cron to ensure nodes are disconnected after the news file completes on the *ArRLNEWS*/0099 server. I'll debug further as time permits. David M. WD5M

Hello! ZLB, can block incoming connections by country? By callsign prefixes? If, for example, I want to block all callsigns from China, how can I do that? Best regards, Volodymyr

This COS event issue is starting to look like, maybe, a bug associated with audio samples left hanging and returned from the VoipNode::GetRxAudio routine. That routine is where the COS logging and event call occurs. I've been able to show the audio samples left over after unkeying in debug logging. I am mostly seeing the false COS symptom when I connect three or more nodes, IRLP, Allstar TLB, Windows SpeakFreely or EchoLink. It doesn't seem to follow any particular node type or mix. However, with only two nodes connected, I haven't seen the hanging audio samples and COS symptoms occur. This reminds me of the memory leak issue from last year. But these hanging audio samples are cleared when the nodes disconnect. Maybe this is related to the memory leak? David M. WD5M

I've known there are differences in results of the ".users" command when "linking" connections versus using the tlb ConfEnable conferencing mode. After a bit of experimenting, I'm closer to understanding why. Some use results from the "users" command to display tlb node status information on web pages. I wanted to know why the results are different between some linked nodes. I use a tlbEvent script with the "link" command to enable transcoding between CODECs for connecting nodes. If you connect to thelinkbox using GSM on EchoLink and ADPCM on IRLP/SpeakFreely and uLaw on Allstar/tlb or IRLP/Speakfreely, thelinkbox will transcode the audio if you use they "link" command to connect them together. The ConfEnable setting is normally disabled in this configuration. When nodes connect using the "link" command, the results from the "users" command will display the connections, but some attributes like "T" (talking) and "last tx" with the "-t" switch are not displayed for non GSM (non EchoLink) connections. This is because those attributes are being determined in the older tbd EchoLink conferencing part of the code, which only processes that information for EchoLink or GSM connections. If non EchoLink system connections are made using the EchoLink CODEC (GSM), then those missing attributes appear in "users" command results for those nodes too. I can use the SpeakFreely client to connect using GSM, ADPCM or uLaw. I only see the "T" (talking) or "last tx" users attributes when connected using GSM. Now, can I modify the code to provide the same "users" results for non EchoLink/GSM CODEC linked connections? David M. WD5M

In case someone may find this useful, here is the method I use to download the theLinkbox source from github as a compressed tarball file. These commands pull directly from the current version in github. Watch out for wrapping text in email. These commands are a single line that you can copy and paste. This curl command will download from Skip's source repository. curl -L https://api.github.com/repos/skiphansen/thelinkbox/tarball/ >skiphansen-thelinkbox-master.tgz For my (wd5m) modified branch you can use this one.curl -L https://api.github.com/repos/wd5m/thelinkbox/tarball/ >wd5m-thelinkbox-branch.tgz For my dev branch where I may be testing more recent modifications.curl -L https://api.github.com/repos/wd5M/thelinkbox/tarball/dev >wd5m-thelinkbox-dev.tgz When you extract the contents of these tarball files they will extract to a directory name similar to the following.skiphansen-thelinkbox-63f0dc5 wd5m-thelinkbox-46b6f2a wd5m-thelinkbox-8afef82 The string of numbers and letters on the end of the name will change with each github repository update of that specific repository. More details about this method can be found at <https://docs.github.com/en/rest/reference/repos#download-a-repository-archive-tar> There may also be thelinkbox tarball files located elsewhere in the cloud or in github folders that are prebuilt from variations of these repositories. David McAnally WD5M

I happened to notice an odd log entry on the ARRLNEWS server. A node identifying as "stn7298" did not disconnect normally and may have reconnected as a result. Thinking this might be a bug I need to address, I looked a bit further. Seems IRLP node stn7298 does not exist. The method I use to disconnect IRLP nodes from the server failed since this was not a real IRLP node. It appears this may have been someone using a SpeakFreely client that had set stn7298 as their name identifier. As far as I know, the old SpeakFreely client does not receive a "disconnect" notification from the remote node when it disconnects. The SpeakFreely client will attempt to reconnect when it finally detects the remote audio network connection is no longer active. This may be why that node reconnected soon after the first disconnect. This action could be seen as a feature in SpeakFreely, as it will automatically reconnect from temporary network outages. The IRLP implementation of SpeakFreely adds a TCP network port that listens for extra information from the remote node, including the disconnect notification. David M. WD5M

Skip (and others), Please review my changes from last year to fix a compile error that Alan VK2ZIW raised. You may want to apply the patch to your tlb source. At the time I think Alan reported a problem that affected compiling on the Raspberry PI. This problem also seems to affect Debian 11 on VPS and other computer systems. Debian 10 still seems happy with the unpatched version, and I confirmed it will compile with the patched version too. On Debian 11, compiling the current version of tlb from Skip's github returns something like the following error. /usr/bin/ld: ../src/libtbd.a(conference.o):/usr/local/src/skiphansen-thelinkbox-63f0dc5/src/conference.c:889: multiple definition of `bLogCmd'; nodes.o:/usr/local/src/skiphansen-thelinkbox-63f0dc5/linkbox/nodes.cpp:352: first defined here collect2: error: ld returned 1 exit status make[2]: *** [Makefile:453: tlb] Error 1 make[2]: Leaving directory '/usr/local/src/skiphansen-thelinkbox-63f0dc5/linkbox' make[1]: *** [Makefile:477: all-recursive] Error 1 make[1]: Leaving directory '/usr/local/src/skiphansen-thelinkbox-63f0dc5' make: *** [Makefile:397: all] Error 2 I made the following changes to my version of tlb last year. Corrected int bLogCmd to be extern int bLogCmd and placed in src/conference.h as it is shared between nodes.cpp and conference.c now. Changes linkbox/nodes.cpp, src/conference.c and src/conference.h. This should resolve issues with some compilers. Specifically: In linkbox/nodes.cpp remove the line "int bLogCmd;" In src/conference.c change the line "int bLogCmd;" to "int bLogCmd = TRUE;" In src/conference.h about line 373 add the line "extern int bLogCmd;" I think you can view my modified files with highlighted changes at the following links. linkbox/nodes.cpp <https://github.com/wd5m/thelinkbox/commit/fdb7bbb9feab04c09c792ba91460d2dccc40edac> src/conference.h and src/conference.c <https://github.com/wd5m/thelinkbox/commit/243ba5b14df9c2b993fab9371f728ab7773c6916> David M WD5M