Author Topic: Re: [APP] - version 3.37.+ ( 18. 3. 2019 )  (Read 9162 times)

Offline balloni55

  • Global Moderator
  • Professor of Locus
  • *****
  • Posts: 2448
  • Thanked: 93 times
    • View Profile
  • Device: Motorola G3 MM 6.01
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #120 on: May 03, 2019, 07:34:47 »
Geocaching
Oki, so if even geocaching master @balloni55 :) does not have anything against it, I'll remove this option to next (Beta) version.
LOL....
thanks hint from c.s.g. long time ago  8) each of my logs start with real time from "offline field log"
{t} > HH:mm
Locus PRO 3.38.3
 

Offline menion

  • Administrator
  • Professor of Locus
  • *****
  • Posts: 11223
  • Thanked: 271 times
    • View Profile
    • http://www.asamm.com
  • Device: SGS7
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #121 on: May 03, 2019, 11:35:06 »
*** new pre-weekend Beta version 3.27.2.10 ***

Huge list of changes after app start.

Only known problem > start on Android 4.x should be probably still there. Fighting with it again and again and ...
Ideas, wishes, problems
Advanced topics, public discussion, sharing of knowledges, testing beta versions: you're here!
 

Offline john_percy

  • Master of Locus
  • ****
  • Posts: 635
  • Thanked: 89 times
    • View Profile
  • Device: MotoG
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #122 on: May 03, 2019, 12:23:30 »
Regarding label layers and so on, Emux's Cruiser GL on android is a great example of what can be done with a separate text layer. Two finger rotate and perspective view keeps the texts horizontal. It even has an optional 3d building view. The rest of the app isn't a patch on Locus however but as an example of map rendering and routing it's excellent.
It uses Mapsforge maps and themes but (I think) a different rendering engine.


Sent from my moto g(6) plus using Tapatalk

 
The following users thanked this post: michaelbechtold

Offline menion

  • Administrator
  • Professor of Locus
  • *****
  • Posts: 11223
  • Thanked: 271 times
    • View Profile
    • http://www.asamm.com
  • Device: SGS7
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #123 on: May 03, 2019, 12:29:08 »
Yes John, it is a different story. Emux is quite hard working on VTM & MapsForge system so it is used as a core map engine in his apps.

I can't imagine how to use it in Locus Map, sorry. Using of rendered tiles from MapsForge library is one thing, using a complete map system as a base for the whole application is something completely different.
Ideas, wishes, problems
Advanced topics, public discussion, sharing of knowledges, testing beta versions: you're here!
 
The following users thanked this post: john_percy

Offline CabrioTourer

  • More than Newbie
  • *
  • Posts: 22
  • Thanked: 1 times
    • View Profile
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #124 on: May 03, 2019, 18:00:45 »
@Menion
To focus on new map core and finalize navigation later is fine.
Just that you are aware of some problems which might not be directly related to navigation logic I want to show you some things upfront.

Simulation in V10 is much slower than V9.
Funny thing: Airplane is slower than car in V10. Airplane means it counts meter by meter. That's slow ;)
I checked recalculation in Simulation in V9 and V10. In V10 it's detect "back on track" but pretty late. V9 it doesnt detect recalculations before back on track isn't improved.

Simulation might not show what happens on the street.
I tried the tour I did on Tuesday with Simulation multiple times. The results are sometimes different. My simulation started more or less same as Tuesday but later it changed and I see something different.
I uploaded the settings, planned route, simulation route and video which shows the issue here. All from V9 because simulation in V10 are too slow.

https://app.box.com/s/xfaxre1nil8n3ljv07az9xtab2nmtpf2

Then a question.
Which parameter influences when second, third etc. recalculation starts ?
The first recalculation is influenced by settings but all other seems to be random.
Sometimes after 300m, sometimes 600m etc.
Can't find any settings to change the time/distance. Tried all settings (I think) in navigation and guidance.
I'm asking because the place where recalculation happens influences the result of recalculation.
Without able to influence recalculation time Simulation might show other things than on street.

Again : it' s fine to focus on new map core first. Just that you are aware of some issues any maybe my sample might be good for testing.
 

Offline john_percy

  • Master of Locus
  • ****
  • Posts: 635
  • Thanked: 89 times
    • View Profile
  • Device: MotoG
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #125 on: May 03, 2019, 18:24:31 »
Increase Map Resolution for vector maps unchanged in the new beta, I think. Reducing zoom level and then magnifying really doesn't work. Texts are enlarged, yes, but other details disappear, depending on the zoom level at which they appear. And roads are no more visible.

Sent from my moto g(6) plus using Tapatalk

 

Offline jonny.blue

  • More than Newbie
  • *
  • Posts: 81
  • Thanked: 4 times
    • View Profile
  • Device: SM-G901F 6.0.1
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #126 on: May 03, 2019, 18:40:33 »
Quote
The first recalculation is influenced by settings but all other seems to be random.
Sometimes after 300m, sometimes 600m etc.

Same to me ... but when you leave Locus and switch back: recalculation starts (most times  :-\)immediately.
« Last Edit: May 03, 2019, 18:51:43 by jonny.blue »
 

Offline menion

  • Administrator
  • Professor of Locus
  • *****
  • Posts: 11223
  • Thanked: 271 times
    • View Profile
    • http://www.asamm.com
  • Device: SGS7
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #127 on: May 03, 2019, 19:10:31 »
@CabrioTourer
slower simulation is possible, not all "activities" have defined average speed. Best is to use base activities like walk, run, bike, car.

Thanks for the very precise bug report. The problem here: you took a really huge shortcut here. To improve performance, app tests only following 100 trackpoints, so seems that whole missed part is more than 100 trackpoints. As a save-button, in navigation menu is "nearest point" button. Anyway this is interesting and probably common use case. I'll think about it, thanks.

Frequency of recalculation: there is a limit to once per 30 seconds. Was there always. Anyway seems, based on @jonny.blue confirmation, there is any new problem. I'll check it.

Map core has priority, but without navigation working at least same as in 3.37.2, I cannot publish new version, so thank you for testing!

@john_percy
it is correct. I spend on this today maybe two hours and it really isn't so easy task. Currently not solved without a clear idea how to easily solve. It just needs some sleep and good tea :).
Ideas, wishes, problems
Advanced topics, public discussion, sharing of knowledges, testing beta versions: you're here!
 

Offline balloni55

  • Global Moderator
  • Professor of Locus
  • *****
  • Posts: 2448
  • Thanked: 93 times
    • View Profile
  • Device: Motorola G3 MM 6.01
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #128 on: May 03, 2019, 19:44:28 »
if i start route planner map center to Prague :-[
in my case blank map is visible, available world map isn´t loaded.
With center button inside routplanner window, map center return to last known position
« Last Edit: May 03, 2019, 20:36:08 by balloni55 »
Locus PRO 3.38.3
 

Offline CabrioTourer

  • More than Newbie
  • *
  • Posts: 22
  • Thanked: 1 times
    • View Profile
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #129 on: May 03, 2019, 20:05:01 »
@Menion
You talk about trackpoints, not sharp/via points, right.
Then it's a pretty short way for a lot of activities. That's for sure a problem.

Solution might be not to choose 100 in a row but every x trackpoint where x depends on the activity (or actual speed?)

If this does not work for activities which needs defined ways (street etc) then selecting trackpoint located on crosses etc might be the solution.

Just an idea.
 

Offline poutnikl

  • Padavan of Locus
  • ***
  • Posts: 439
  • Thanked: 41 times
    • View Profile
    • My Brouter Profiles
  • Device: Xiaomi MI A2 (6-128G) / And9-Pie / LocusMap Pro
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #130 on: May 03, 2019, 20:31:53 »


@Menion
You talk about trackpoints, not sharp/via points, right.

I see you repeatedly use the term "sharp points", possibly because they are displayed as triangles, that are sharper then squared via points.

But AFAIK, Menion has coined for them the term "shaping points", giving to the route a particular shape between start/destination/viapoints., being e.g. a manual correction of the computed route between those points.

Just for completeness, they do not count as points in context of the point priority recalculation,  in contrary to the via points you "must visit".


Odesláno z mého Mi A2 pomocí Tapatalk

https://github.com/poutnikl/Brouter-profiles/wiki
When a user digs profiles up from the GitHub repositories,
they will be hopefully wrapped in some wiki knowledge....  :-)
 
The following users thanked this post: menion

Offline CabrioTourer

  • More than Newbie
  • *
  • Posts: 22
  • Thanked: 1 times
    • View Profile
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #131 on: May 03, 2019, 22:25:27 »
Just for completeness, they do not count as points in context of the point priority recalculation
This is what I thought. I just asked to be sure because 100 trackpoints isn't much for activities like car and even not much for bike in rural area.

Thanks for "shaping point" correction.
 

Offline poutnikl

  • Padavan of Locus
  • ***
  • Posts: 439
  • Thanked: 41 times
    • View Profile
    • My Brouter Profiles
  • Device: Xiaomi MI A2 (6-128G) / And9-Pie / LocusMap Pro
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #132 on: May 03, 2019, 22:57:18 »



This is what I thought. I just asked to be sure because 100 trackpoints isn't much for activities like car and even not much for bike in rural area.

When I analyzed GPX files of multiple bike routes, generated by BRouter, I realized, that as a "rule of thumb", the is in average 1 trackpoint about every 50 m. I suppose less dense for car routes, more dense  for hiking routes.

Trackpoints from recorded tracks will be probably much denser, depending on the recording settings.


Odesláno z mého Mi A2 pomocí Tapatalk

https://github.com/poutnikl/Brouter-profiles/wiki
When a user digs profiles up from the GitHub repositories,
they will be hopefully wrapped in some wiki knowledge....  :-)
 

Offline menion

  • Administrator
  • Professor of Locus
  • *****
  • Posts: 11223
  • Thanked: 271 times
    • View Profile
    • http://www.asamm.com
  • Device: SGS7
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #133 on: May 03, 2019, 23:07:15 »
Well, for the planned route it may be a few kilometers. Anyway, for the manually recorded track used later for navigation, this may reduce to a quite short distance, so agree it is not an ideal solution.

As I think about it, testing on nearest trackpoint may not be so big problem. Because automatic recalculation is trigger max. once per 30 seconds, app should search over whole track without any limit.

On the second side, this quite nicely solves a possible issue with round routes.

EDIT: as Pountik wrote.

Generally, using any exact limit is usually not the best solution. So why 100 trackpoints? Usually it will work, sometimes not, like on your case.
On second side, what if you ride in the middle (on your video it's a first char "a" in name "Leiblachtal"), app snap to the bottom line, but it was only some detour and then you return back close to start of this route?
Just want to show that any change may on cause malfunction of different use-case. So I still see here as backup solution easily accessible button "nearest point".
Ideas, wishes, problems
Advanced topics, public discussion, sharing of knowledges, testing beta versions: you're here!
 

Offline john_percy

  • Master of Locus
  • ****
  • Posts: 635
  • Thanked: 89 times
    • View Profile
  • Device: MotoG
Re: [APP] - version 3.37.+ ( 18. 3. 2019 )
« Reply #134 on: May 03, 2019, 23:32:54 »
Slight change to hillshading is not sufficient IMHO. The screenshots here are of OAM v4 GreatBritainML using Locus internal theme at zooms 16-12. The road is quite visible at high zooms but disappears into the shadows as the zoom gets less.

PS This forum software is trying very hard to prevent me from posting but I will not be beaten.

PPS you may need to download the screenshot on a PC to view it properly.