Waypoints Disappearing

[nohal]

I've just commited a patch making the shared waypoints appear as WPT in any exported gpx file, which should improve the way we interoperate with other software. The representation in navobj.xml does not change - the shared points are rtept with opencpn:shared set to 1. The reason for this is that there is no mechanism to define a relationship between a WPT and RTEPT in the GPX standard but we still need to maintain this relationship for OpenCPN's use.

My thinking is that the way to kill this beast properly would be as follows:
When a route is using a shared waypoint (Use nearby...) keep the current behavior of hiding it with the route, but make the waypoints tab reflect it and make it possible to turn it's display back on independently of the route.
Anybody sees any problem with this solution?

Pavel

[kiltym]

Originally Posted by nohal

My thinking is that the way to kill this beast properly would be as follows:
When a route is using a shared waypoint (Use nearby...) keep the current behavior of hiding it with the route, but make the waypoints tab reflect it and make it possible to turn it's display back on independently of the route.
Anybody sees any problem with this solution?

Pavel

To me, this seems like a good compromise. It allows current users who like it they way it is now to have nothing change.

And people like me, that want to view the waypoints independent of the route, can turn them back on. As long as this is "persistent" from session to session, which I don't know why it would not be, this seems good to me.

Good idea!

[HappySeagull]

.. there's logic to the way it is now , if the incorporated Mark/POI icons also disappeared in the Route Manager when the route is hidden...As it is, marks incorporated as wpts in a route reappear when the route is deleted and that is logical. They have reverted to their status as Marks or POI and have not been lost.

Note that I call them Marks.
I think there's a semantic confusion between the "proper" waypoints that are turns on a route created by the Route Pencil. A "Waypoint" =a point on the way, right? while the (so-called!) "Waypoints" on their own page of the Route Manager) would better be named Marks or POI (Points of Interest). It's been bugging me for awhile.

[kiltym]

Originally Posted by HappySeagull

.. there's logic to the way it is now , if the incorporated Mark/POI icons also disappeared in the Route Manager when the route is hidden...As it is, marks incorporated as wpts in a route reappear when the route is deleted and that is logical. They have reverted to their status as Marks or POI and have not been lost.

Note that I call them Marks.
I think there's a semantic confusion between the "proper" waypoints that are turns on a route created by the Route Pencil. A "Waypoint" =a point on the way, right? while the (so-called!) "Waypoints" on their own page of the Route Manager) would better be named Marks or POI (Points of Interest). It's been bugging me for awhile.

I agree that there is logic, but it may not be the best way to do it.

If I use a Mark in a Route, I feel the Mark should stand independently of that Route, not be consumed by the Route, as it is now, and as you would propose it currently works (If the Marks disappeared from the Route Manager).

As far as semantics on the Route Manager, I would agree that the tab would be better labelled as Marks, but this is a minor (technical) change that people would just have to agree on. IMO, Waypoints are listed when viewing Route Properties.

But, a Mark can be used in a route as a Waypoint in a Route, and in turn it becomes a "shared" <rtept> in the navobj file. This is as close to a "reference" that could be done since GPX does not support this concept it seems.

In practice, I often re-use Marks in multiple routes, and this is why I ran into this "issue". If I create a Mark, I want it to ALWAYS be visible on the charts so I can re-use it again if I create a new route. I am not sure what happens when the same Mark is used in multiple routes inside the nabobj file as I have not tried it yet, but presumably this works OK as I know I have done it before without issues.

I think nohal's idea would work, and not "break" any existing functionality or usage in the way people might be familiar with it working currently. It just gives the user an option of allowing the Mark to continue to be displayed, without the route being displayed.

[HappySeagull]

I agree with you per Mark visibility being separated as a "feature". I was just pinning down the actual inconsistency- that is: the marks being visible in RouteManager when not visible on the chart .

I was just trying Marks in multiple routes...Works fine. A shared Mark remains visible in a visible route and doesn't disappear even if the other route using it is invisible.

[fgd3]

Originally Posted by HappySeagull

.. there's logic to the way it is now , if the incorporated Mark/POI icons also disappeared in the Route Manager when the route is hidden...As it is, marks incorporated as wpts in a route reappear when the route is deleted and that is logical. They have reverted to their status as Marks or POI and have not been lost.

Note that I call them Marks.
I think there's a semantic confusion between the "proper" waypoints that are turns on a route created by the Route Pencil. A "Waypoint" =a point on the way, right? while the (so-called!) "Waypoints" on their own page of the Route Manager) would better be named Marks or POI (Points of Interest). It's been bugging me for awhile.

"Mark" is a generic term that describes a number of objects. Most commonly marks are physical aids to navigation.

Waypoints are named locations.

Route points are locations used in a route. They are usually locations where the route changes direction but they don't need to be. A route point may be a previously defined waypoint or a newly defined location.

When an existing waypoint is incorporated in a route the waypoint data object should be copied into the route. Using the same data object is what causes the problem reported by the original poster.

Fabbian

[Viking Sailor]

It seems to me that when a mark is added to a chart it would be presumptuous of OpenCPN to try and second guess what the user's intentions are for the mark. The user may intend on using the mark in many different ways.

If the mark is later used as part of one or more routes it should be cloned as a route point for each of the routes. Operations (saving, hiding, moving, .etc.) on a route should only effect the route points for that route and have no effect on the "Mark" or any other routes.

[kiltym]

Just wanting to close the loop on this issue.

Is there consensus with nohal's suggestion, or does that not do enough and something more drastic should be done to really separate the concepts of Marks and Marks re-used in Routes......?

I guess I want to make sure a decision has been made, if so, what it is, and that it's on the to-do list somewhere where it will get implemented.

It sounds to me like nohal's solution will provide a quick "fix" for the issue, and not break anything. So I vote for that based on the ability to get it out with a sooner release.

Longer term, from a conceptual standpoint, I think a Mark re-used in a route should be "referenced" from the route, not consumed, but as nohal stated about the limitation of GPX, it is not easily done and assume some longer term thought would have to go into it.

I don't want to step on any toes with this, just wanting to make sure it gets put on the right list somewhere for it do get implemented.

[HappySeagull]

Pavel is on it that's more than enough for me.

....But I like to pick at scabs and maybe the thing would not arise without the "Use NearBy Wpt?" dialogue, because a route can be drawn over a Mark. That's what I could do on the few times it ever will pop up, because my Icons for Marks are pretty distinctive from the bluediamond for Wpts or, as I see now,better named-Rtepts.
... me- user do not know what me are doing and haven't the handle on the GPX format that the developers do. They leap about hammering plugs into holes in user-head...

[nohal]

Originally Posted by kiltym

Just wanting to close the loop on this issue.

Is there consensus with nohal's suggestion, or does that not do enough and something more drastic should be done to really separate the concepts of Marks and Marks re-used in Routes......?

I guess I want to make sure a decision has been made, if so, what it is, and that it's on the to-do list somewhere where it will get implemented.

As nobody seems too offended and I didn't hear any argument making it not work, it's on my TODO and I will implement it as soon as I find time to do it.

It sounds to me like nohal's solution will provide a quick "fix" for the issue, and not break anything. So I vote for that based on the ability to get it out with a sooner release.

Looking at my table it will probably not be for the very next 3.1 beta, but soon.

Longer term, from a conceptual standpoint, I think a Mark re-used in a route should be "referenced" from the route, not consumed, but as nohal stated about the limitation of GPX, it is not easily done and assume some longer term thought would have to go into it.

I don't want to step on any toes with this, just wanting to make sure it gets put on the right list somewhere for it do get implemented.

The WPT (call it mark or whatever) is actually not "consumed" by a route, just our internal usage of GPX format in navobj.xml does not have the wpt element in it as long as the WPT is used in one or more routes to save some space and computations.
For interoperability with other tools we will (already done, will be in the next beta) duplicate the shared mark as both rtept and wpt elements in the GPX exported from route manager. That should hopefully satisfy everybody (It technically can be taken as it implements Fabbian's suggestion to copy the mark to the route).

Pavel

[fgd3]

Originally Posted by nohal

(It technically can be taken as it implements Fabbian's suggestion to copy the mark to the route).

Pavel

I'm not sure I understand your fix. It seems to me it will do nothing to resolve the original poster's complaint because you will continue to use the same data object internally for a waypoint and a route point. So if the user creates a route which uses an existing waypoint (with "Use nearby waypoint") and hides the route the existing waypoint will not be displayed. How will it help to change exported GPX files?

Fabbian

[fgd3]

Originally Posted by nohal

My thinking is that the way to kill this beast properly would be as follows:
When a route is using a shared waypoint (Use nearby...) keep the current behavior of hiding it with the route, but make the waypoints tab reflect it and make it possible to turn it's display back on independently of the route.
Anybody sees any problem with this solution?

Pavel

This looks like a kludge to me. You're making the program logic needlessly complex trying to achieve behavior which would occur naturally if the design were correct in the beginning. If you put a copy of the existing waypoint in the route rather than a reference to it the problem simply wouldn't occur.

If you have an existing waypoint that has been used in a route, why would you want it to vanish when the route is hidden? I can understand that route points which exist only in the route should not be displayed when the route is hidden, but if a route passes through a previously identified location hiding the route should not make that location invisible.

Fabbian

[nohal]

Fabbian...
The change in the exported GPX files has nothing to do with the OP's problem. It's a fix for interoperability with other software. The fix for the display vs. route manager inconsistency still has to be implemented as I wrote in the earlier post.

Pavel

[nohal]

Originally Posted by fgd3

This looks like a kludge to me. You're making the program logic needlessly complex trying to achieve behavior which would occur naturally if the design were correct in the beginning. If you put a copy of the existing waypoint in the route rather than a reference to it the problem simply wouldn't occur.

If you have an existing waypoint that has been used in a route, why would you want it to vanish when the route is hidden? I can understand that route points which exist only in the route should not be displayed when the route is hidden, but if a route passes through a previously identified location hiding the route should not make that location invisible.

Fabbian

Fabbian...
You are as usually absolutely free to implement any part of OpenCPN in a better way and if it really proves to be better, I'm sure it will be accepted. In this case I simply don't agree with you that the current design is broken and to implement what I described earlier I don't need to change it in any way.
I would not even have to change it to implement what you call for, the reason I will not do it is simply that I don't think that's the correct behavior.

Pavel

[fgd3]

Originally Posted by nohal

Fabbian...
The change in the exported GPX files has nothing to do with the OP's problem. It's a fix for interoperability with other software. The fix for the display vs. route manager inconsistency still has to be implemented as I wrote in the earlier post.

Pavel

By "display vs. route manager inconsistency" I think you mean the problem of existing waypoints disappearing when a route that uses them is hidden from view. Is that right? There might be a simpler way to deal with that: tell the users if they don't like that behavior always reply "No" when asked "Use nearby waypoint?". That's what I do.

We've discussed this before, a year or so ago. At the time you seemed to like sharing waypoints because it allowed you to edit multiple routes. I don't see the point of that, but if other folks find it useful that's fine. Maybe all you need to change is the text. Make it say "Share nearby waypoint?"

If a saved waypoint is a fixed navigational feature you don't want it to be dragged around just because a route used it and then was changed. But if your saved waypoint is a race marker and you've planned a number of possible routes I can see that you might want to move that marker for another race where the marker has been placed differently and you want all your route planning to be updated. (I don't race so I'm just guessing why that feature might be useful.)

So it seems the fix may be just a better explanation of why that feature exists.

Fabbian