Cruisers Forum
 


Reply
 
Thread Tools Search this Thread Rate Thread Display Modes
Old 21-05-2017, 01:29   #1
Registered User

Join Date: Oct 2014
Location: Netherlands
Boat: Halmatic 30
Posts: 1,101
SignalK development ?

SignalK looked very promising at first sight. OpenCPN was a supporter of this new development.

SignalK seems to work well. A SignalK server was developed by Digital Yacht.

But now I don't see any pratical use of SignalK. No information about SignalK in combination with OpenCPN. Is it still work in progress ?

Why seems the further development of SignalK on halt or rather very slow ?
verkerkbr is offline   Reply With Quote
Old 22-05-2017, 21:11   #2
Registered User

Join Date: Aug 2015
Location: Sozopol
Boat: Riva 48
Posts: 1,375
Re: SignalK development ?

SignalK is unlikely to survive without a major hardware vendor behind it. It just will not be supported.

IMHO, we need to stick to established, open protocols such as serial and Ethernet. NMEA2000 I thoroughly dislike (too expensive unless you use automotive parts), Seatalk was great but low bandwidth nowadays and Ethernet cables are difficult to work with for minor connections. Try to stick to computer connections as much as possible.
Pizzazz is offline   Reply With Quote
Old 24-05-2017, 11:39   #3
Registered User

Join Date: Aug 2009
Location: oriental
Boat: crowther trimaran 33
Posts: 4,414
Re: SignalK development ?

I have developed a free autopilot that uses signalk as well as nmea0183

my last passage it consumed total of 2 to 3 watts depending on wave running downwind wing and wing in 10 to 15 knots wind steering half the average error of my monitor with vane
seandepagnier is offline   Reply With Quote
Old 24-05-2017, 13:17   #4
cruiser

Join Date: Nov 2007
Location: Probably in an anchorage or a boatyard..
Boat: Ebbtide 33' steel cutter
Posts: 5,030
Re: SignalK development ?

Quote:
Originally Posted by Pizzazz View Post
SignalK is unlikely to survive without a major hardware vendor behind it. It just will not be supported.

IMHO, we need to stick to established, open protocols such as serial and Ethernet. NMEA2000 I thoroughly dislike (too expensive unless you use automotive parts), Seatalk was great but low bandwidth nowadays and Ethernet cables are difficult to work with for minor connections. Try to stick to computer connections as much as possible.
SignalK is certainly very much alive and well in one place - the openplotter image running on the raspberry pi where it's at the core. And very useful it is too, on a platform focusing on sensors being able to give each sensor it's own place in a data structure far beyond anything nmea 0183 or 2k can offer is a godsend, then view or access the data on any smartphone/tablet over wifi/ethernet.

Seems a bit late for "unlikely to survive".. It's there already, opensource and available for anyone, with an active team supporting and continuing to develop.

There are a handful of offerings from the commercial world signalk enabled and it would seem like shooting themselves in the foot for MFD manufacturers not to implement signalk viewing capabilities over wifi even if religated to sensor type non vital data viewing with the important gps type data hardwired.
conachair is offline   Reply With Quote
Old 24-05-2017, 15:16   #5
Registered User
 
travellerw's Avatar

Join Date: Jan 2014
Location: Martinique
Boat: Fortuna Island Spirit 40
Posts: 2,298
Re: SignalK development ?

Quote:
Originally Posted by Pizzazz View Post
SignalK is unlikely to survive without a major hardware vendor behind it. It just will not be supported.

IMHO, we need to stick to established, open protocols such as serial and Ethernet. NMEA2000 I thoroughly dislike (too expensive unless you use automotive parts), Seatalk was great but low bandwidth nowadays and Ethernet cables are difficult to work with for minor connections. Try to stick to computer connections as much as possible.
You lack a basic understanding of the OSI model. NMEA2000 protocol defines itself on the multiple levels of the OSI model.Originally intended to run on CANBUS, it has since been expanded for other mediums.

The physical medium these connections run on is just one layer of the OSI model. E.G. Ethernet can run on CAT5, CAT3, Coax, wireless links, hell even IBM "Type-1". Likewise, CANBUS can run on multipe physical medium, including CAT5.

The fact is.. NMEA2000 is a good protocol. Its not completely open as that could possibly affect the stability (rogue devices causing havok). Just like HTML it shares a certification structure. Believe me this is a good thing on a boat where devices causing crashes could be a big problem. This is a spot where SignalK falls down (devices don't need certification).

Now as to the choice of physical medium.. DeviceNET (that is standard for the wires and connectors you think is NMEA2000) is a FANTASTIC medium for boat. It robust, expandable, sealed and best of all TESTED. DeviceNET had been used not only in automobiles, but factories, military and mining for YEARS. Fact is, its one of the best possible solutions for a boat.

Now I can hear you saying.. No no, everything should be like my computer with ethernet cables (physically CAT3,CAT4, CAT5 or CAT6). Well let me tell you, there is a reason that CAT cables are a terrible choice on a boat. CAT cables were never intended to be used in an environment like a boat. The connectors were not designed to be robust, sealed or even corrosion resistant. The cables themselves were not designed for movement or flexing (at least no where near the amount on a boat). Frankly CAT cables suck on a boat. In a year and a half of cruising we have had to replace multiple CAT5 cables and devices as corrosion killed them. We had one POE injector completely short out. Number of DeviceNET cables replaced, ZERO!


Alright.. Back to the original post.. I really want to see SignalK succeed.. However I will not use it on my boat until its supported and controlled by and industry body. Unless, of course, I design, build and install my own devices, then SignalK is excellent, however I would rather be cruising than faffing with that stuff!
travellerw is offline   Reply With Quote
Old 24-05-2017, 18:51   #6
Registered User

Join Date: Feb 2017
Posts: 60
Re: SignalK development ?

There's actually a lot going on with SignalK. Version 1 of the spec is coming soon! We are constantly improving the open source server implementations and Digital Yacht is also releasing regular updates.

Really cool plugins for node server are out or being developed. They do things like anchor alarm, autopilot control reporting to MarineTraffic.com and more.

Last November I released an app that only displays info from SignalK sources:
https://appsto.re/us/CQZKeb.i
Sales have been better than I expected!
sbender is offline   Reply With Quote
Old 24-05-2017, 18:54   #7
Registered User

Join Date: Feb 2017
Posts: 60
Re: SignalK development ?

Quote:
Originally Posted by travellerw View Post
You lack a basic understanding of the OSI model. NMEA2000 protocol defines itself on the multiple levels of the OSI model.Originally intended to run on CANBUS, it has since been expanded for other mediums.



The physical medium these connections run on is just one layer of the OSI model. E.G. Ethernet can run on CAT5, CAT3, Coax, wireless links, hell even IBM "Type-1". Likewise, CANBUS can run on multipe physical medium, including CAT5.



The fact is.. NMEA2000 is a good protocol. Its not completely open as that could possibly affect the stability (rogue devices causing havok). Just like HTML it shares a certification structure. Believe me this is a good thing on a boat where devices causing crashes could be a big problem. This is a spot where SignalK falls down (devices don't need certification).



Now as to the choice of physical medium.. DeviceNET (that is standard for the wires and connectors you think is NMEA2000) is a FANTASTIC medium for boat. It robust, expandable, sealed and best of all TESTED. DeviceNET had been used not only in automobiles, but factories, military and mining for YEARS. Fact is, its one of the best possible solutions for a boat.



Now I can hear you saying.. No no, everything should be like my computer with ethernet cables (physically CAT3,CAT4, CAT5 or CAT6). Well let me tell you, there is a reason that CAT cables are a terrible choice on a boat. CAT cables were never intended to be used in an environment like a boat. The connectors were not designed to be robust, sealed or even corrosion resistant. The cables themselves were not designed for movement or flexing (at least no where near the amount on a boat). Frankly CAT cables suck on a boat. In a year and a half of cruising we have had to replace multiple CAT5 cables and devices as corrosion killed them. We had one POE injector completely short out. Number of DeviceNET cables replaced, ZERO!





Alright.. Back to the original post.. I really want to see SignalK succeed.. However I will not use it on my boat until its supported and controlled by and industry body. Unless, of course, I design, build and install my own devices, then SignalK is excellent, however I would rather be cruising than faffing with that stuff!


I don't expect that SignalK will ever be controlled by an industry body. This is an open protocol. We do not want to end up where NMEA is with a closed protocol and huge costs to produce hardware. There's a great group of people developing the spec and taking input from the community. I hope it stays that way.
sbender is offline   Reply With Quote
Old 24-05-2017, 19:04   #8
Registered User

Join Date: Feb 2017
Posts: 60
Re: SignalK development ?

Quote:
Originally Posted by Pizzazz View Post
SignalK is unlikely to survive without a major hardware vendor behind it. It just will not be supported.

IMHO, we need to stick to established, open protocols such as serial and Ethernet. NMEA2000 I thoroughly dislike (too expensive unless you use automotive parts), Seatalk was great but low bandwidth nowadays and Ethernet cables are difficult to work with for minor connections. Try to stick to computer connections as much as possible.
SignalK works just fine over serial and Ethernet. NMEA2000 is not an open protocol and is difficult to write software for.

SignalK does not replace NMEA2000, at least not on my boat. What it does is provide a nice standard mapping from different protocols like N2K and NMEA0183 and allows people like me to write software that uses data from the boat without having to know the gritty details about these protocols.
sbender is offline   Reply With Quote
Old 25-05-2017, 03:50   #9
cruiser

Join Date: Nov 2007
Location: Probably in an anchorage or a boatyard..
Boat: Ebbtide 33' steel cutter
Posts: 5,030
Re: SignalK development ?

Quote:
Originally Posted by Pizzazz View Post
IMHO, we need to stick to established, open protocols such as serial and Ethernet.
Which are just two of the available methods for transmitting signalk.

Quote:
Signal K data is transmitted as a JSON string. JSON is native to the Internet, and to browsers, so transmission is supported on any medium that can access the Internet. That includes mobile data, WiFi, Ethernet, USB, Bluetooth, and pretty much anything coming. But importantly we are just sending a string of characters – so we can send data over serial cables (RS232/442/etc.), Onewire, CAN, in fact almost anything.
Signal K » Overview
conachair is offline   Reply With Quote
Old 25-05-2017, 03:50   #10
cruiser

Join Date: Nov 2007
Location: Probably in an anchorage or a boatyard..
Boat: Ebbtide 33' steel cutter
Posts: 5,030
Re: SignalK development ?

Quote:
Originally Posted by sbender View Post
I don't expect that SignalK will ever be controlled by an industry body. This is an open protocol. We do not want to end up where NMEA is with a closed protocol and huge costs to produce hardware. There's a great group of people developing the spec and taking input from the community. I hope it stays that way.
conachair is offline   Reply With Quote
Old 25-05-2017, 05:57   #11
Registered User

Join Date: Aug 2009
Location: between the devil and the deep blue sea
Boat: a sailing boat
Posts: 20,420
Re: SignalK development ?

Quote:
Originally Posted by verkerkbr View Post
SignalK looked very promising at first sight. OpenCPN was a supporter of this new development.

SignalK seems to work well. A SignalK server was developed by Digital Yacht.

But now I don't see any pratical use of SignalK. No information about SignalK in combination with OpenCPN. Is it still work in progress ?

Why seems the further development of SignalK on halt or rather very slow ?
Please look up my posts earlier this year in Signal K CF thread.

I am not sure why K is being developed. We are having enough ride with SeaTalk(-s), 183, and n2K and pop comes the weasel and people appear who claim K is the way to (but I ask to where).

Imho K should be abandoned and N2K should be deployed.

Implementation of N2K in OpenCPN would be a huge bonus, imho.

Implementation of bastardised 183 would be a bonus too. (see below)

And if we want alternatives, then why not simple text based data exchange. SPEED=4kn implies speed of 4 knots. If we want Signal K people to wrap the whole concept in white papers that are difficult to fathom for a regularly educated adult (who speaks 4 languages and writes code in 3) then let them be.

There sure is reason and space for alternative protocols just the way Signal K is being promoted is completely futile.

This is just my opinion and due to change over time and as I learn new things. But it stands today.

Cheers,
barnakiel
barnakiel is offline   Reply With Quote
Old 25-05-2017, 06:25   #12
Registered User

Join Date: Feb 2017
Posts: 60
Re: SignalK development ?

Quote:
Originally Posted by barnakiel View Post
Please look up my posts earlier this year in Signal K CF thread.

I am not sure why K is being developed. We are having enough ride with SeaTalk(-s), 183, and n2K and pop comes the weasel and people appear who claim K is the way to (but I ask to where).

Imho K should be abandoned and N2K should be deployed.

Implementation of N2K in OpenCPN would be a huge bonus, imho.

Implementation of bastardised 183 would be a bonus too. (see below)

And if we want alternatives, then why not simple text based data exchange. SPEED=4kn implies speed of 4 knots. If we want Signal K people to wrap the whole concept in white papers that are difficult to fathom for a regularly educated adult (who speaks 4 languages and writes code in 3) then let them be.

There sure is reason and space for alternative protocols just the way Signal K is being promoted is completely futile.

This is just my opinion and due to change over time and as I learn new things. But it stands today.

Cheers,
barnakiel
So you really think that everyone that writes marine related software should be coding to N2K? Have you seen the details of that protocol? Do you understand that it is closed? You have to join the NMEA group, which costs lots of money, just to get the details of the standard N2K messages.

It makes zero sense to do this. SignalK is standards based (json, web services, etc.) Very easy to code to and lots of tools available on most platforms.

I'm able to write an app using one standard protocol that gets data from N2K, 0183, web, custom sensors, etc.

These are only a few reasons why SignalK is being developed.
sbender is offline   Reply With Quote
Old 25-05-2017, 07:11   #13
Registered User

Join Date: Aug 2009
Location: between the devil and the deep blue sea
Boat: a sailing boat
Posts: 20,420
Re: SignalK development ?

Quote:
Originally Posted by Pizzazz View Post

SignalK is unlikely to survive without a major hardware vendor behind it. It just will not be supported.

IMHO, we need to stick to established, open protocols such as serial and Ethernet. NMEA2000 I thoroughly dislike (too expensive unless you use automotive parts), Seatalk was great but low bandwidth nowadays and Ethernet cables are difficult to work with for minor connections. Try to stick to computer connections as much as possible.
I see it otherwise: What we do need is Open Source hardware. We want to kick the manufacturer out of our business. We want to be able to build our plotters and our instruments just like Open Source software initiative gave us OpenCPN, qtVlm, Linux, OpenWRT.

The blocks are there. We just need to build and share.

The rich will always buy the most recent and glossy B&G plotter. But this world is more than a shop.

I cannot see any alternatives for N2K. We want many instruments and sensors and all of them talking at once. So this will be something CAN Bus based, even if not N2K.

We could do fine with bastardized n183. Why use lengthy complex sentences where simpler short sentences can do the same job?

Meanwhile, we want a ... " 'translate anything' n183, ST, N2K, Open Source (both open hardware and open software) multiplexer" one able to do serial, BT and wifi/TCP. This would be a great start.

Kplex on OpenWRT bridged with Arduino based sensors is an example. But we need a better protocol for it. I think add N2K and we are nearly home.

Just my opinion.

b.
barnakiel is offline   Reply With Quote
Old 25-05-2017, 07:26   #14
Registered User

Join Date: Feb 2017
Posts: 60
Re: SignalK development ?

Quote:
Originally Posted by barnakiel View Post
I see it otherwise: What we do need is Open Source hardware. We want to kick the manufacturer out of our business. We want to be able to build our plotters and our instruments just like Open Source software initiative gave us OpenCPN, qtVlm, Linux, OpenWRT.

The blocks are there. We just need to build and share.

The rich will always buy the most recent and glossy B&G plotter. But this world is more than a shop.

I cannot see any alternatives for N2K. We want many instruments and sensors and all of them talking at once. So this will be something CAN Bus based, even if not N2K.

We could do fine with bastardized n183. Why use lengthy complex sentences where simpler short sentences can do the same job?

Meanwhile, we want a ... " 'translate anything' n183, ST, N2K, Open Source (both open hardware and open software) multiplexer" one able to do serial, BT and wifi/TCP. This would be a great start.
.
You talk about wanting to build OpenSource, but then promote N2K as the protocol, it makes no sense. N2K is closed and proprietary! I think you must be missing something here...

We have your "translate anything" now. The signalk-node-server does this. On my boat it "multiplexes" N2K, multiple 0183 sources, serial data from a Victron MPPT charger, and wifi temp and humidity sensors. All running on an inexpensive raspberrypi.

Others are building sensors using Adrino and other platforms that speak native SignalK for things like yaw/pitch/roll and compasses.

And then any application that supports SignalK can display data from the boat without having parse antiquated binary protocols that were designed to save bytes.
sbender is offline   Reply With Quote
Old 25-05-2017, 07:58   #15
cruiser

Join Date: Nov 2007
Location: Probably in an anchorage or a boatyard..
Boat: Ebbtide 33' steel cutter
Posts: 5,030
Re: SignalK development ?

Quote:
Originally Posted by barnakiel View Post
Meanwhile, we want a ... " 'translate anything' n183, ST, N2K, Open Source (both open hardware and open software) multiplexer" one able to do serial, BT and wifi/TCP. This would be a great start..
It's so close to being here already - openplotter on raspberry Pi. Just need to crack the seatalk 9-bit malarkey and job done.
Can you make new N2K pgns? If not then how do you transfer data from sensors not included in the closed expensive standard.
conachair is offline   Reply With Quote
Reply

Tags
men

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
New Marina Development in China GordMay Pacific & South China Sea 4 29-09-2009 04:33
Nautical Development 39 (Morgan 39?) riptide Monohull Sailboats 1 22-07-2009 11:53
News: interesting development craft - high speed landing craft Amgine Multihull Sailboats 0 03-11-2008 11:30
Turks and Caicos Development Petition Canibul Atlantic & the Caribbean 5 24-04-2008 18:15

Advertise Here


All times are GMT -7. The time now is 01:37.


Google+
Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Social Knowledge Networks
Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2024, vBulletin Solutions, Inc.

ShowCase vBulletin Plugins by Drive Thru Online, Inc.