If we could have one more feild that you put in what device you had and the software could then be altered so non approved devices could not be used for jelly beans and PBS everyone would be happy . I'm sure we could easily crowd fund the cost if that's an issue. I think we all agree the motion and locosys products are the most accurate and all the serious GPS sailors should use them. I just hate seeing people drop out, in the end we all just want to windsurf and see how we went against are mates at the end of the day..

Well said Simon, I agree with you, i can tell you when my GT31 stops working i am out of GPTSC, and that is not good.

All you have to do is persuade Nebs, and a majority of the advisory group.
But Nebs is about a year or so behind, with other mods to the website.

Select to expand quote

bc said..

Boston! said..

Kazza said..

Bradman said..
Send it in Kazza. I sent mine in with their return form and they sent me a refurbed one which was like new. Cost of postage to Taiwan.

Hmmm, maybe I will then. I sort of didn't like to trust the refurbed one either. Can't believe that a watch GW60 that cost $260 or $360 (can't remember) only lasted 2.5 years and others have the same problem with them.
So what's the problem with Garmin and GPSTC, surely an app can be developed? Garmin should become a sponsor for GPSTC, think of all the Garmins that people would buy........

I've got three failed GW60's, Kazza. (I had to borrow one for Green Island!) I've given up trying to get them repaired. They were never fit for purpose. I use a Fenix 5 now and post to KA72 only. The GPSTC honchos have made it quite clear, on these forums, that they will not accept Garmin data, even for the most unremarkable sessions. It's just not going to happen.

I know of at least a dozen sailors who have dropped out of gpstc because their watches have failed it seems to me that for the sake of accuracy that is way over the top we have lost the original idea of team challenge being an inclusive fun activitiy.
From website
philosophy of the Challenge:Inclusiveness: That everyone in the windsurfing community be included from fun and social sailors, to the serious including Pro's. Participation: That it is important to participate, to be involved, to develop skills and improve Personal Best Times. Be involved all season, not just summer. Fun: That most importantly we have fun, and however you get it, from stacking it on a 15 knot run, to doing 40+ knots, or just having a beer/bourbon/wine with your friends after a session. Building relationships: That we connect people with the same passion, meet new people, create connections intra, inter-state, international and generally just build relationships and grow our community

I've never read the philosophy of GPSTC before! A good laugh! It seems things may have drifted slightly since those freewheeling early days. The terrific comp. it is, in its current state, couldn't be less inclusive! (only one worthwhile GPS available which you cant get your hands on for months) Could we, ("the small group of individuals"), possibly start something similar using KA72, minus the device "compliancy" and associated egg headedness?

Select to expand quote

vosadrian said..

sailquik said..

The subject has been explained at length by GPSTC and argued over by the same small group of individuals who continue to choose not get it I'm afraid.

I've seen no evidence of anybody not getting your explanation. We just don't agree with it and like to point out that it seems at odds with the above quoted philosophy of the GPSTC (perhaps you need to update that?).

Just because you have explained it previously does not mean that we will stop presenting better options that would keep more GPSTC users happy such as the two tiered system mentioned above. In fact if you could take the time to point out any problems with a two tiered system, that would be appreciated.

Well clearly you don't get the logic and purpose of it, and choose to ignore the reasons for it.

There is an apt saying: "I can explain it to you, but I cant understand it for you."

But, such an option as you propose already exists and a few members do use it: Post your results in the comments section with you comments and your story and any other palaver you wish to indulge in. That's it! You are in and participating! It won't matter if your 'results' are accurate or pure fantasy. They won't contiminate the database or effect the competition, but you can interact as much or as little as you like. Easy peasy.

Over to you.

Select to expand quote

decrepit said..
All you have to do is persuade Nebs, and a majority of the advisory group.
But Nebs is about a year or so behind, with other mods to the website.

Thanks for that, .so who is on the advisory group?

Select to expand quote

sailquik said..

Well clearly you don't get the logic and purpose of it, and choose to ignore the reasons for it.

There is an apt saying: "I can explain it to you, but I cant understand it for you."

But, such an option as you propose already exists and a few members do use it: Post your results in the comments section with you comments and your story and any other palaver you wish to indulge in. That's it! You are in and participating! It won't matter if your 'results' are accurate or pure fantasy. They won't contiminate the database or effect the competition, but you can interact as much or as little as you like. Easy peasy.

Over to you.

Yeah, working a couple of decades as an electrical engineer in RF devices including GPS devices, I tend to struggle a bit with logic.......

Perhaps the other option is that your logic is flawed?? As pointed out above... the philosophy of GPSTC is inclusiveness, but you can't be included without a device that you can't actually get??!! Something amiss there!! Your logic provides an ideal solution for a small subset of users. You fail to see the logic that a large group of people just want to do this non-competitively for fun and social.

And the option of putting it in comments is not a real option. I am sure I am not the only one who would prefer to look for other options rather than use that option. That option does not allow tracking of my own results (kms, PRs etc.).

Here is my proposed solution (again). When you put in a new session, you can select for each parameter from doppler/track points/unknown. If you have an unapproved device, you select unknown. A few lines of code changed in the score processing code to ignore data that is "unknown". Problem solved!! Everybody happy!!

Select to expand quote

vosadrian said..

sailquik said..

Well clearly you don't get the logic and purpose of it, and choose to ignore the reasons for it.

There is an apt saying: "I can explain it to you, but I cant understand it for you."

But, such an option as you propose already exists and a few members do use it: Post your results in the comments section with you comments and your story and any other palaver you wish to indulge in. That's it! You are in and participating! It won't matter if your 'results' are accurate or pure fantasy. They won't contiminate the database or effect the competition, but you can interact as much or as little as you like. Easy peasy.

Over to you.

Yeah, working a couple of decades as an electrical engineer in RF devices including GPS devices, I tend to struggle a bit with logic.......

Perhaps the other option is that your logic is flawed?? As pointed out above... the philosophy of GPSTC is inclusiveness, but you can't be included without a device that you can't actually get??!! Something amiss there!! Your logic provides an ideal solution for a small subset of users. You fail to see the logic that a large group of people just want to do this non-competitively for fun and social.

And the option of putting it in comments is not a real option. I am sure I am not the only one who would prefer to look for other options rather than use that option. That option does not allow tracking of my own results (kms, PRs etc.).

Here is my proposed solution (again). When you put in a new session, you can select for each parameter from doppler/track points/unknown. If you have an unapproved device, you select unknown. A few lines of code changed in the score processing code to ignore data that is "unknown". Problem solved!! Everybody happy!!

Logic? Yeah, nah...Whatever.....

So you sound like the ideal person to do the extensive experimental GPS testing and maths gymnastics, and to then convince Garmin to include those calculations and data in their devices. Go fot it.

Select to expand quote

F2 said..
Well said Simon, I agree with you, i can tell you when my GT31 stops working i am out of GPTSC, and that is not good.

Why would you even say that if your GT31 is still working, why would you leave GPSTC and not just buy a new GPS when it becomes available and why if you leave GPSTC that it would not be good, not good for who ? I wouldn't like to see you go but hey life's way too short, do what makes you happy.

Select to expand quote

bc said.. Thanks for that, .so who is on the advisory group?

I've written to them, to see if I can answer that.

We've got two conflicting issues here: accurate results and inclusiveness. I think the current situation can be improved, and will outline what would be needed in a separate post. This one is just about accurate results.

1. Accurate results:
This is a must for a competition that (a) uses hundreds of results every month, and (b) is scored automatically, without an expert examining every jelly bean track.
What is confusing is that many non-approved devices are very accurate most of the time. But all GPS devices will sometimes report inflated results. This is true for every single GPS device I have ever looked at, including the GT-31, GW-60, the Motion, and the Suuonto watch.
Software uses a variety of data to automatically identify "spike results". This includes the number of satellites, acceleration, and error estimates (called SDoP or sAcc). The worst spike occur when crashing, waterstarting, and swimming. Anyone who learned foiling with lots of crashes and slow waterstarts can produce tons of examples, I do it every session.
Non-approved devices do not provide the necessary data to automatically identify spikes. That's the problem. The Suuonto files I looked at do not even report the number of satellites, which is essential for swimming spikes. Furthermore, the currently approved devices (GW-60, Motion, and u-blox based prototypes) use higher data rates, which make it much less likely that spikes that slip through the filters affect the result. One wrong number in 1-Hz data will be 50% of a 2-second result; for 5 Hz data, this drops to 10%, and it drops to 5% for 10 Hz data.

I have written GPS analysis software that anyone can use for free. I have quite a bit of experience in algorithm development, and have demonstrated that speed results can be calculated a lot faster than other software does. I could imagine some new software filters that would identify some artifacts in data from non-approved devices, but I cannot think of algorithms that would identify "fake top speeds" without satellite and accuracy data at a level that is adequate for a competition like GPSTC.

Inclusiveness:
The request to allow at least "non-counting" posting from non-approved devices comes up again and again. Several former GPSTC members have stated that they rather dropped out of the TC than spend another $250+ on a device that many see as flawed.
Let me outline what would be needed to implement a two-tier system, where (almost) any device can post, but only approved devices would count towards rankings (and PBs).

1. Web site changes:
The GPSTC web site software would need to be modified so that:
a. For future sessions, only sessions for approved devices are used for rankings and PBs, but other sessions can be posted and shown.
The "Doppler / Trackpoint / Unknown" choice should be changed to something like "Approved Device / Other device".
b. For posting from ka72 and software like GPS Speedreader, two new posting pages should be made, one for approved devices and one for other devices. When posting from ka72 or Speedreader, the software selects which of these pages is used for posting, based on the GPS file. Users cannot change the type from "Other device" to "Approved device" on such postings.

2. Analysis software changes:
Software used for direct posting (that is ka72.com and GPS Speedreader) needs to determine whether or not data are from approved devices. This can reasonably well be approximated by checking the following:
- Original file format. Anything other than .SBN. .SBP, .UBX, and .AOA is unapproved.
- Presence of valid satellite and accuracy data. This would sort out any files that were converted to .SBP (etc.) using GPSResults, GPSBabel, etc.
- Recording rate. Anything slower than 1 Hz is automatically "Other device".

I could easily deal with the necessary changes in GPS Speedreader, but ka72.com is the more important site for posting. It would only be reasonable to pay Dylan for the efforts to make these changes, but raising the money should not be a problem. When he could find the time to do implement such changes could be a bigger issue. The same thing applies to the GPSTC web site changes.

There would be some effort and cost involved, but it seems doable. The end result would be a more inclusive GPSTC, with lower hurdles to attract new members. I, for one, would sometimes love to see the speed results from posters like JJ who currently do not post speeds when foiling. All this could be done without compromising on accuracy.

I value accuracy, and the convenience of a wristwatch
I like the gw-60 very much
I would buy another if mine fails.
I don't have a problem with spending 300 on a new watch every couple of years
That's a bargain for what it offers me.
Post your dodgy garmin recordings to ka72 and keep GPS challenge genuine and accurate
Windsurfers spent 5-10k on go fast gear and then complain about a few hundred dollars on watch

Select to expand quote

tonyk said..
I value accuracy, and the convenience of a wristwatch
I like the gw-60 very much
I would buy another if mine fails.
I don't have a problem with spending 300 on a new watch every couple of years
That's a bargain for what it offers me.
Post your dodgy garmin recordings to ka72 and keep GPS challenge genuine and accurate
Windsurfers spent 5-10k on go fast gear and then complain about a few hundred dollars on watch

I agree Tony, we are talking the cost of a good fin...??

Select to expand quote

sailquik said..

Logic? Yeah, nah...Whatever.....

So you sound like the ideal person to do the extensive experimental GPS testing and maths gymnastics, and to then convince Garmin to include those calculations and data in their devices. Go fot it.

Why would garmin alter their product for the GPSTC..... so they can sell another 50 out of the millions they sell without doing any mods? The only companies that would customise for GPSTC are companies so small that selling another 50 units is of value to them. So you end up with small companies that are not experienced in making volume products for many consumers... and they fall apart/leak/have no modern connectivity features!

The issue you fail to address in any way is that what % of current and future users actually care about that accuracy. And the option discussed above which allows those who care to continue with an approved device, and those that don't to upload data that cannot be used competitively to skew team results. I suspect the reason you don't address this is because you realise it would be open the flood gates and many GPSTC users would use unapproved devices and less would care whether their results count.

Select to expand quote

tonyk said..
I value accuracy, and the convenience of a wristwatch
I like the gw-60 very much
I would buy another if mine fails.
I don't have a problem with spending 300 on a new watch every couple of years
That's a bargain for what it offers me.
Post your dodgy garmin recordings to ka72 and keep GPS challenge genuine and accurate
Windsurfers spent 5-10k on go fast gear and then complain about a few hundred dollars on watch

So Tony, since you are about accuracy above all else, I am interested in your thoughts on how the two tier system mentioned above that allows non competitive GPSTC users to post results in a way that does not effect competition in any way (no team scores, no jelly beans etc.). Would it cause you any problems at all if other users posted in this way?

Select to expand quote

boardsurfr said..
We've got two conflicting issues here: accurate results and inclusiveness. I think the current situation can be improved, and will outline what would be needed in a separate post. This one is just about accurate results.

1. Accurate results:
This is a must for a competition that (a) uses hundreds of results every month, and (b) is scored automatically, without an expert examining every jelly bean track.
What is confusing is that many non-approved devices are very accurate most of the time. But all GPS devices will sometimes report inflated results. This is true for every single GPS device I have ever looked at, including the GT-31, GW-60, the Motion, and the Suuonto watch.
Software uses a variety of data to automatically identify "spike results". This includes the number of satellites, acceleration, and error estimates (called SDoP or sAcc). The worst spike occur when crashing, waterstarting, and swimming. Anyone who learned foiling with lots of crashes and slow waterstarts can produce tons of examples, I do it every session.
Non-approved devices do not provide the necessary data to automatically identify spikes. That's the problem. The Suuonto files I looked at do not even report the number of satellites, which is essential for swimming spikes. Furthermore, the currently approved devices (GW-60, Motion, and u-blox based prototypes) use higher data rates, which make it much less likely that spikes that slip through the filters affect the result. One wrong number in 1-Hz data will be 50% of a 2-second result; for 5 Hz data, this drops to 10%, and it drops to 5% for 10 Hz data.

I have written GPS analysis software that anyone can use for free. I have quite a bit of experience in algorithm development, and have demonstrated that speed results can be calculated a lot faster than other software does. I could imagine some new software filters that would identify some artifacts in data from non-approved devices, but I cannot think of algorithms that would identify "fake top speeds" without satellite and accuracy data at a level that is adequate for a competition like GPSTC.

boardsurfr,

I agree with 99% of what you have presented. I do think that your proposed changes are more than I was thinking (surely we can trust people to manually set the field that indicates whether the device is approved, and then only minor mods are required to GPSTC and nothing else).

On an aside, I do a lot of algorithm development also. I am really perplexed as to how someone crashing could not be easily identified from a speed plot and nothing else. Erroneous GPS data will result in acceleration/decelerations that are just not possible on a windsurfer and the speed would go close to zero in close time proximity to the high speed spike. A simple algorithm that identifies implausible speed data and then masks out a period before and after would surely catch 99% of crashes. It could be developed further to look at position. We could identify a function that has the minimum radius for the current velocity that a highly skilled sailor can achieve and if the position plot is outside of plausibility we invoke a similar masking. Likely this would catch most errors automatically and at least flag them for a human to decide upon.

My point being that with some smart heads working on it, I believe that data errors could be detected in most cases without the requirement for accuracy data.

Select to expand quote

vosadrian said..

So Tony, since you are about accuracy above all else, I am interested in your thoughts on how the two tier system mentioned above that allows non competitive GPSTC users to post results in a way that does not effect competition in any way (no team scores, no jelly beans etc.). Would it cause you any problems at all if other users posted in this way?

You misrepresent what Tony said. He did not say at all that he is "about accuracy above all else,"

At the risk of repeating for those 'who will not see': You can already post your results in the GPSTC as you wish to by posting your numbers in the Comments section only, as others do.

I dont see the complainers doing that, so the question is: Are they really that interested in posting non teamchallenge counting results?

Select to expand quote

vosadrian said..boardsurfr,

My point being that with some smart heads working on it, I believe that data errors could be detected in most cases without the requirement for accuracy data.

OK, so there you go. You dont really want a two tier system.

Well Adrian, you are obviously 'smarter' than all the other people who have worked on this.
It's time to write up your paper on exactly how that can be achieved.

Select to expand quote

Mr Keen said..

tonyk said..
I value accuracy, and the convenience of a wristwatch
I like the gw-60 very much
I would buy another if mine fails.
I don't have a problem with spending 300 on a new watch every couple of years
That's a bargain for what it offers me.
Post your dodgy garmin recordings to ka72 and keep GPS challenge genuine and accurate
Windsurfers spent 5-10k on go fast gear and then complain about a few hundred dollars on watch

I agree Tony, we are talking the cost of a good fin...??

$370 Jason. Did not think you have bought a fin that cost that much yet
If the one shop that sells them in Australia gets sick of the issues with selling them you will not be covered by Australian consumer law covering the GW60.

Select to expand quote

Cocky2 said..

Mr Keen said..

tonyk said..
I value accuracy, and the convenience of a wristwatch
I like the gw-60 very much
I would buy another if mine fails.
I don't have a problem with spending 300 on a new watch every couple of years
That's a bargain for what it offers me.
Post your dodgy garmin recordings to ka72 and keep GPS challenge genuine and accurate
Windsurfers spent 5-10k on go fast gear and then complain about a few hundred dollars on watch

I agree Tony, we are talking the cost of a good fin...??

$370 Jason. Do not think you have bought a fin that cost that much yet
if the one shop that sells them in Australia gets sick of the issues with selling them you will no Australian consumer law covering the GW60.

Oh alright my mistake 1.5 fins....

Select to expand quote

tonyk said..
I value accuracy, and the convenience of a wristwatch
I like the gw-60 very much
I would buy another if mine fails.
I don't have a problem with spending 300 on a new watch every couple of years
That's a bargain for what it offers me.
Post your dodgy garmin recordings to ka72 and keep GPS challenge genuine and accurate
Windsurfers spent 5-10k on go fast gear and then complain about a few hundred dollars on watch

Hi Tony,
The cost of GW60 is $370.
If I could get 2 years out of one I would be happy. As you have only sailed for less than 5 000 km with one you have not used an approved device very long on water. I am on my third one already and never use buttons and wash in fresh water after every use.
Also previous to this I have purchased 4 Canmores, GT31 2 x GW52 .
During 25 000 km I have had no issues with my Garmin which is accurate for all disciplines except for the occasional 2 sec higher reading.

Vosadrian, yes acceleration filters are well used but boardsurfr, specifically referred to low speed crashes while foiling.
And for 5hz and 10hz data the acceleration filters need to be set higher to allow for the quicker rise time of the higher frequency data, from memory 10hz is set at 16m/s2. So high speed crashes will be caught but low speed ones might not be.

There's been a great deal of time spent testing the optimum way to set all filters available. You may do a little better but don't count on it.

Select to expand quote

vosadrian said.. >>> Would it cause you any problems at all if other users posted in this way?

Personally I wouldn't like the idea of competing against dodgy data. You never know if that score that's just a bit better than yours is real or not.
So it would need to be a completely separate comp, I don't think we could convince Nebs of going to the trouble of writing that.

KA72 is your best bet

I'm a bit late to this thread BUT didn't we do this already!!!!!!!!!!!!.
BTW lm on the advisory committee.
No error file , no use for GPSTC.

I like someone's suggestion about repurposing the data type field. Would save having to change the database structure.

Doppler = approved device
Unknown/track point = non approved

does anyone know if the gpstc website already excludes non doppler results from team calcs (ie "jelly beans")?

KA72 knows what type of device you are uploading from by the file format/details. Could the gpstc data type field be automatically populated from ka72 based on file type?

Select to expand quote

kato said..
I'm a bit late to this thread BUT didn't we do this already!!!!!!!!!!!!.
BTW lm on the advisory committee.
No error file , no use for GPSTC.

This statement is false as GPSTC contains data from;
Canmore has no satellite or error data.
All the data from GT 11 has no error data.
All early data from Garmin is "Trackpoint " not Doppler .

Select to expand quote

Cocky2 said..

This statement is false as GPSTC contains data from;
Canmore has no satellite or error data.
All the data from GT 11 has no error data.
All early data from Garmin is "Trackpoint " not Doppler .

Rubbish!
You are being deliberately misleading. What is your game Ian

Of course we do this now. (the canmore is a whole different situation, introduce when the GPSTC was in an untenable bind. The short answer is that is is being phased out by the end of the year). It was not always like that, and back then led to some serious and frustrating problems. Before we had the GT-11 we did use trackpoints, and then, with the GT-11, we used Doppler without error data (already a huge improvement). But, in those days there had to be a hell of a lot of manual checking and validation, everyone had to have their own analysis software, team captains and team experts were called on to check data routinely, and still some clangers caused problems. With the introduction of the GT-31, almost all of those issues went away overnight, so for the last 8-10 years problems are rare, and those are usually picked up quite quickly.

Are you advocating for allowing less reliably accurate and unverifiable devices? Do you see that as a desirable situation?

Select to expand quote

sailquik said..

Cocky2 said..

This statement is false as GPSTC contains data from;
Canmore has no satellite or error data.
All the data from GT 11 has no error data.
All early data from Garmin is "Trackpoint " not Doppler .

Rubbish!
You are being deliberately misleading. What is your game Ian

Of course we do this now. (the canmore is a whole different situation, introduce when the GPSTC was in an untenable bind. The short answer is that is is being phased out by the end of the year). It was not always like that, and back then led to some serious and frustrating problems. Before we had the GT-11 we did use trackpoints, and then, with the GT-11, we used Doppler without error data (already a huge improvement). But, in those days there had to be a hell of a lot of manual checking and validation, everyone had to have their own analysis software, team captains and team experts were called on to check data routinely, and still some clangers caused problems. With the introduction of the GT-31, almost all of those issues went away overnight, so for the last 8-10 years problems are rare, and those are usually picked up quite quickly.

Are you advocating for allowing less reliably accurate and unverifiable devices? Do you see that as a desirable situation?

As usual you will not acknowledge that you approved the Canmore which is no more accurate than Garmin Doppler especially when using Glonass.

All legacy data which does not meet current data standards is not accurate as you have stated many times .

Select to expand quote

Cocky2 said..

As usual you will not acknowledge that you approved the Canmore which is no more accurate than Garmin Doppler especially when using Glonass.

All legacy data which does not meet current data standards is not accurate as you have stated many times .

WRONG! I do acknowledge that and so stated in the post above that you are replying to. Did you actually read it

(although I would point out that whe have good evidence that the Canmore data was mostly better than some other non approved devices)

Of course the legacy data does not meet the standards of the current posts. But as pointed out, it was subject to a whole lot more scrutiny then, and we have confidence that it's accuracy was as good as we could do with the then existing technology. What is your point