Construction of Skybuck's UFO Alert Tool log

Skip to first unread message

Skybuck Flying

Jul 31, 2022, 10:56:46 AMJul 31
Construction of Skybuck's UFO Alert Tool log:

Day 1, 31 july 2022, Scouting if UBER H3 compiles/builds/works and can be used from Delphi, SUCCESS !

Video log of my activities of Day1:

1.1 UBER H3 library works, it was build with VS 2019, CMAKE GUI was first used and later CMAKE command used to make DLL version.

1.2 Delphi's TDUMP was used to inspect the produced H3 dll library and exported functions.

1.3 VIRUS TOTAL was used to confirm the H3 dll is a PE32+ library which means it's 64 bit.

(1.4 Lazarus/Free Pascal's H2PAS and H2PAS Wizard was tried (twice, different versions) to convert from C to PAS but FAILED to produce anything useable.)

1.5 (Nes's) CHET was used to convert C api headers into Pascal headers some further work will be necessary but it already works and is usefull.

1.6 (Nes's) CLang for Delphi was used to build CHET successfully.

1.7 UBER H3 library/api was loaded into Delphi console application and some simple functions tested OK.

(What you could do is try and create some nice sphere with this library to provide some example code of how you did it, this could help me to speed up the construction ! =D)

Bye for now,

R Kym Horsell

Jul 31, 2022, 11:29:23 AMJul 31
Well forgive me if I'm confused but... you can now locate things
in a coordinate system no-one understands?

R Kym Horsell

Jul 31, 2022, 11:52:51 AMJul 31
R Kym Horsell <> wrote:
> Skybuck Flying <> wrote:
> Well forgive me if I'm confused but... you can now locate things
> in a coordinate system no-one understands?

OK. That was maybe a bit brief. :)

I dont know what you've experienced, but I've found people divide
into 2 main camps. Those that have no interest in what you've done and
those that cant understand what you've done.

Taking a hint from S Lem's "Golem 2000" this situation can be overcome
by mastering the art of the introduction.

An introduction proffers a top-down description of you and your current
world of interest, hopefully leaving out all the boring bits people could
guess from the parts to do list.

You start with the most general points and becomre more and more
specific as you go along.

Add definitions for words you feel you have been using differently from
the schnook in the street or other PhD's in astro.

Above all keep you sense of fun and humor intact because the world is
structured in such a way as to require a lot of that as well as
having evolutionary advantages.

Phlatt Smax

Aug 1, 2022, 4:54:56 PMAug 1
You are wasting your time. There are no alien UFOs. The earth is flat
and outer space does not exist. Man never landed on the moon. Satellites
are high-altitude ballons with radio boxes. NASA is the world's largest
consumer of helium. There is no edge of the flat earth. It is a endless
plane in all directions. There are undiscovered continents beyond the
Antarctic ice ring. The earth is not moving. Heliocentrism is a
religious cult. Your mama's so fat, she freebases ham.

I really wish you would stop posting your paranoid nonsense here.


LEVEL : Flat Earth Documentary

The Next LEVEL

DITRH - Deep In The Rabbit Hole

Research 'flat earth' on google.


Phlatt Smax

Skybuck Flying

Aug 3, 2022, 10:53:21 AMAug 3
Day 2, 1 august 2022, Implemented H3CellIndex in Delphi which will be superior to C/C++ thanks to Delphi's property language feature ! ;) =D :P*

Video log of my activities of Day 2:

2.1 UBER H3CellIndex implemented in three different versions: 0.01 (fast), 0.02 (resolution range protected), 0.03 (cells integrated into one property)

2.2 Test program written to test UBER H3CellIndex

2.3 Investigated the ammount of pentagons and hexagons, the latter of which is huge, possibly special rendering library needed to handle it.

2.4 Contacted who has a nice geodesic grid rendering engine, it appears it does not yet use or support UBER H3, filled
an issue/request/idea that maybe it might interested him to implement/integrate it (off/after video/live recording):

Day 3, 3 august 2022, Trying to understand H3 indexing/coordinate systems + Investigated possibility of rendering a sphere/earth with OpenGL

3.0 Trying to understand how H3 indexing works, it has multiple coordinate systems, it could get quite complex, I don't yet fully grasp the coordinate
systems, so this shall require some more studieing and experimenting.

Further clarifications about coordinate systems may be found here:

3.1 Custom OpenGL was tried, by missing 3D perspective matrices.

3.1.1 Experiment with threaded renderer/opengl, surprisingly it worked somewhat.

First a custom opengl renderer was tried, but I more or less discovered that OpenGL does not seem to apply 3D effect automatically.
To create the 3D effect some kind of perspective matrices must be computed/created which is very boring to do, so I decided to investigate GLScene which
has already implemented all of this.

3.1.2 Investigated the possibility of creating custom 3D perspective matrix, but just boring and waste of time for now, maybe in future.

This man explains how to create those 3D perspective matrices as well as other youtube videos/links.
Basically it's a sign to use GLScene and avoid all that work. However GLScene might throw some exceptions in the future and then maybe a custom renderer
could be tried to avoid that those problems, or hack GLScene instead to fix those exceptions, so far GLScene seems promising.
as well as this:

3.2 Switch to GLScene library which has lots of support classes/components for working with 3D scenes and OpenGL.

3.2.1 Downloading GLScene (I downloaded and used the zip file (GLScene_v2.1_VCL_Win32_Win64) from sourceforge)

3.2.3 Installing GLScene

Following the installation guide/pdf allowed me to install GLScene flawlessly, it is not necessary to copy any DLLs into the windows system folder,
I also highly recommend not to do this. These DLLs if used for for example sound and such should later be distributed with the application anyway.
So it's good to be able to test what the effect would be if the DLLs are missing.
The only real step required to install GLScene is to add the source folder to the library path inside Delphi and then open the GLScene.groupproj
and choose compile all, then per BPL/DT (design time) install it into the Delphi environment, only GLScene_DT needs to be installed to get GLSCene/OpenGL working,
the rest is optional for additional functionality.

3.2.4 Compiling/building GLScene

Delphi compiler is lightning fast and compiled GLScene pretty quickly, in about a minute or less. C/C++ would have taken forever ! LOL
GLScene consists out of at least 2.7 million lines of code. The earth's executable size was still below 4 megabyte which for now for me at least is acceptable.
The GR32 RT (runtime) and DT (design time) BPL did not compile due to missing code for GR32 which is a graphics32 library but this code is not necessary.
Only GLScene RT and DT where installed for now.
Installing and compiling/building GR32/Graphics32 might speed up the application even further, but might also introduce some more complexity and potential for bugs,
but could be tried, the source code was somewhat hard to find, but can be found here:

3.2.5 Testing GLScene thanks to quick start guide
Following this very simple quick start guide proved that GLscene can work:

3.2.6 I came across a "Earth" advanced demo which could be perfect for Skybuck's UFO Alert Tool.

While browsing the GLSCene website and it's demos/examples I came across this "Earth" advance demo which is awesome it even has sky light scattering.


GLScene (version GLScene_v2.1_VCL_Win32_Win64) has a very good advanced demo of Earth and it should be just fine for Skybuck's UFO Alert Tool.
And GLScene is necessary to get some "3D" look, otherwise it would be necessary to implement perspective matrices which is very boring for me to do.
The software rendering mode was tried and does work, it is possible to set it. Though the earth advance demo does produce some integer overflow and
range check errors in software mode. The advanced earth demo renders some sky/light scattering thingy, not sure if this is the cause of it.
Most likely the real cause is some kind of texturing issue and mipmapping issue that might not work so well in software rendering mode.
(I report this issue on github and also on the sourceforge page of GLScene, maybe somebody will fix it, but I doubt it for now.)

Most likely it is possible to re-create this earth demo step by step to see how far it will work in software rendering mode.

(Perhaps an option can be added to switch to software rendering mode for the user, however this should be in some ini file/text based so user can change
it in case of problems starting up the application and/or the software should automatically fall back to software rendering mode in case of problems
however GLScene already claims to do that so should be ok. In my case the GPU is half broken and it did not automatically switch to software rendering mode,
which in my case allowed me to at least run the demo to get some kind of impression on this broken GPU.)

Discovering this advanced demo of Earth and the usage of GLScene will most likely seriously reduce my development time for this UFO Alert Tool.
That is the power of Delphi ! =D Quality code as your fingers tips ! ;) =D

Maybe this will be helpfull in the future perhaps for routing network to figure some things out, for now I include it in this log:

Not sure if this will be usefull, maybe:
HexASCII: a file format for cartographical hexagonal rasters

*New* Construction of Skybuck's UFO Alert Tool Playlist:

Skybuck Flying

Aug 6, 2022, 11:06:49 AMAug 6
Day 4, 5 august 2022, Fixed advanced earth demo for software rendering. Massive fails: Tried to download/load better 3D Earth + Google maps

Video log of my activities of Day 4:

4.0 Even though day 4 ended in a massive fail, it was still an important day. This day I discovered how to fix the software rendering for GLScene.

This is especially handy for the advanced demo: Earth.
In folder:

4.1 Fixing the Advanced Earth Demo for software rendering:

4.1.1. Change GLSceneViewer.Buffer.ContextOptions.roSoftwareMode to true.

4.1.2. Change GLSceneViewer.Buffer.AccumBufferBits from 0 to 8, 16, 24 or 32.

4.1.3. Deleting the flare component from the sun component:

GLScene(right click)->Show Scene Editor->LSSun->GLLensFlare1

Right click->delete object

4.1.4. Disable following two lines in GLScene library unit: GLS.Graphics.pas

for slice := 0 to d - 1 do
// Skybuck: disabled, buggy in software mode
// Build2DMipmap(GetLevelAddress(0), LAddresses, fColorFormat, fDataType,
// AFilter, GetWidth, GetHeight);
for level := 1 to fLevelCount - 1 do
Inc(PByte(LAddresses[level - 1]), GetLevelSizeInByte(level) div d);

4.1.5. Disable the following line in fEarthd.pas:

procedure TForm1.GLSceneViewerBeforeRender(Sender: TObject);
// GLLensFlare1.PreRender(Sender as TGLSceneBuffer);

4.2 A bigger/more detailed Earth 3D model was tried. However it sucked, the 3D model itself was just a sphere with triangles the texture maps
where weird. Most 3D Model sites found with google suck and require payment of money.

4.2.1 Conversion from FBX to OBJ was done:

4.2.2 Difficulty loading MTL into GLScene, FAILED, undocumented, unknown.

4.2.3 Difficulty loading Texture maps into GLScene, not even tried, because of 4.2.2

4.3 Google Earth library 1.5.3 for Delphi was tried. However it cannot be used because Delphi's TWebBrowser component relies on Internet Explorer and
Internet Explorer does not support the javascript api of google earth. Very maybe chromium browser could be used by changing some include file.
Howebver chromium is a different browser than chrome from google. Chromium requires some kind of building of files, way too much work for nothing most likely
so this was not tried.

Day 5, 6 august 2022, Loading JSON file with country coordinates/lines and visualized/converted them to X,Y,Z coordinates on globe, quite awesome !

(Forgot to post this link in day 3 log so for those that missed it here it is):
Video log of my activities of Day 3:

Video log of my activities of Day 5:

5.1 This day was an awesome day, it made me so happy I played the starwars theme song as I completed the mission for this day:

Load the country borders of earth, try to visualize them in GLScene and added bonus, try and convert the coordinates from latitude, longitude to X,Y,Z
3d coordinates on a sphere. Big Success ! JSON was parsed using a library I already had installed and used before:
For easy to use, very powerfull, very fast JSON parser !:

5.1.1 JSON data download for earth country lines:

5.1.2 JSON Parser used:

5.1.3 Explanation of Earth X,Y,Z coordinate system:,_Earth-fixed_coordinate_system

5.1.4. Explanation of lat/long to x,y,z conversion formula:

5.1.5 Converted this python code to Delphi to convert Latitude, Longitude coordinates to X,Y,Z:

5.2 Scaled Earth in Y direction (important to visualize lines better).

I also played a bit with earth scaling options, this is kinda cool, now earth can be a bit more realistic by scaling it in the Y direction slightly,
this also helps to render the lines for the countries better.

5.3 Playing with atmosphere.

I also played a bit with the atmospheric rendering/constants to try and make it look a bit better, using real earth radius and such in km:

5.4 High detail of country lines (10m a little bit slow on software rendering)

I also managed to load a more high detail version of the lines to show the country lines better, this does slow down the renderer a little bit, a slightly
less accurate version could also be tried. So far the 10m seems to be the most accurate. A 50m version could be tried to speed it up a little bit.
(Same link as above, mentioned one more time here ;))

5.5.1 Huge Visible Earth Map (failed to load, too big)

I then proceed to try and download and load a VisibleEarth High Resolution Map (43200x21600) assuming it would simply display and load in glscene.

5.5.2 Windows 7 System Crash, and Image/Photo/Graphics Software Failures

Boy was I in for a surprise: Not only did windows 7 completely crash and restart, GLScene/Delphi also fails to load it because TBitmap cannot handle this size,
All other software I tried also failed: paint,, glassimage, gimp and photoviewer. All for different reasons, mostly out of memory reasons or 3D issues.

So unfortunately the day ended in another mass fail, but good progress was made ! ;)
(I didn't see the crash of windows 7 because I bailed to my bed to rest lol, but when I came back the system had rebooted and I checked the system log):
Bugcheck: 0x00000018 (0x0000000000000000, 0xfffffa8005fe0570, 0x0000000000000002, 0xffffffffffffffff)

There were some other peculiar thing in the system log, like the firmware limitting certain processors, possibly system getting to hot.
Main cause of crash seems to be failure of system to respond in time, possibly caused by out of memory situation.
I checked the youtube stream nothing is visible on it concerning the crash lol. It would be diserably to have a seperate system for live streaming
to catch these crashes lol, not that it matters that much, just a funny thing to see/notice, but could also be used to attack windows systems !

Skybuck Flying

Aug 9, 2022, 9:39:38 PMAug 9
Day 6, 9 august 2022, Provinces and Hexagon World

Video log of my activities of Day 6:

1. Provinces was tried/implemented, however there is a backface culling issues with the input data (version 4).
I have found the original source which might have a better version, version 5, issues also reported on multiple sites, however the original source
uses a different format, no JSON is available, but dbf is available dbase database file format which can be handled by Delphi.
Original source of data seems to be:

2. Hexagon world implemented by using the UBER H3 library. It isn't completely perfect yet, but already looks pretty good. Still trying to wrap my head around it.

To solve the problem with the provinces two hexagon worlds could be created. One for zooming, and one which describes the "ground" of earth and it's provinces
and/or countries. This could be done with UBER H3 compression technique where large parts are covered with large hexagons and smaller ones with little ones.
However for routing purposes it would probably be better to main the same size, but for visualization purposes this compression might help.

Possibilities for further research is either zooming functionality where level of detail is adjusted as user zoom in closer. Or a selection technique
where only the selected hexagon to user is pointing towards is zoomed in... which would be a cool/funny technique maybe ! ;)

Now that hexagon grid is working, more time should be spent on the networking and gui design of ufo alert tool and the routing of messages and/or broadcast
messages to alert people of the presence of a ufo, especially if within their region. Plus perhaps also some time on some kind of server based/algorithm
to distribute the initial peers to the main hexagons or so and perhaps further research into "leader" selection per (main) hexagon or so...

R Kym Horsell

Aug 11, 2022, 7:01:47 PMAug 11
Skybuck Flying <> wrote:
> *******************************************************************************************************************************************
> Day 4, 5 august 2022, Fixed advanced earth demo for software rendering. Massive fails: Tried to download/load better 3D Earth + Google maps
> *******************************************************************************************************************************************
> Video log of my activities of Day 4:
> 4.0 Even though day 4 ended in a massive fail, it was still an important day. This day I discovered how to fix the software rendering for GLScene.


I've just spent 10+ days trying all combinations of settings to get
my modem working again after 2 of the local ISP's decided to shut down
their services in the same week. (Naturally, they were the 2 I used
to connect to the net).

Man, I was going nuts. (More nuts than usual).

Skybuck Flying

Aug 12, 2022, 6:53:06 PMAug 12
Hmm sounds like you still on dial-up ? Is there no broadband in your neighbourhood ? :)

Skybuck =D

Bzzz Bzzz Bzzz Bzzz BEEEEEEEPP tutututututututut. =D

R Kym Horsell

Aug 12, 2022, 7:28:21 PMAug 12
Rural Australia. Think annual floods and bushfires and 100 km to the nearest
supermarket. Wirelss broadband. While I wasnt happy the service suddenly
stopped and for a week or more no-one could tell me whether the
service has been switched off or my modem was just broken
turns out they "reorganised bands in preparation for the 3g-4g
upgrade in mid 2024". and didnt imagine anyone was affected


Aug 13, 2022, 7:35:20 AMAug 13
R Kym Horsell wrote on 13/8/22 9:28 am:
Wasn't there a Satelite set up as part of the NBN Switch over?
Shy-thrust or some such!!

No idea of its cost but would have to have been better than Dial-Up!!

R Kym Horsell

Aug 13, 2022, 10:03:52 AMAug 13
Satellites not always the answer. If they are geostationary
there is a pretty big latency time so you cant do many
realtime operations over it. If you're downloading movies
overnight I guess it would be fine. ;)
But students talking to their teacher maybe not so fine.
A real bad application is getting nav data for your car
in realtime. They are supposedly planning to have 5g
control self-driving cars. We don't need an extra 1-2-3 sec
of wait time in that loop. :)

The NBN and 4g/5g have all been wrapped up as "the
same thing" in the public mind. I am sure confused with
some of the literature I get in the mailbox trying to
scare me into buying some service for $60/mo now
I have been getting for $50/year for the past 20y.

All part of the legacy of having a merchant banker
and bunch of accountants design the national internet
and then privatise it to precent any losses appearing on
their books.

NBN at risk of being a `white elephant' unless speed is the focus
The Australian Financial Review, 02 Aug 2022
Telecommunications experts have called on the company to abandon its old
regime under the previous government, as more ...

Skybuck Flying

Aug 18, 2022, 10:59:07 PMAug 18
Day 6.5, 13 august 2022, Studied hexagon routing

(No video log available, too hot and I wanted to concentrate intensely and keep hexagon routing designs private for now until more confident it's good.)

6.5.1 I studied document "An_Optimal_Permutation_Routing_Algorithm_on_Full-D.pdf" in quiet without live streaming, also because it was a hot day.

6.5.2 I believe the algorithm can be extended with an additional axis to go up and down the uber h3 hexagon world resolutions.

6.5.3 I did some initial code sketches for data structures and routines and such.

6.5.4 It would be possible to adjust packet coordinates every time it moves, so that it would seem that each node is the center of the world to prevent/avoid
wrap around problems with big coordinates suddenly going zero, testing will have to prove if this can work and if it works well.

6.5.5 I believe this algorithm described in this document will work and is doable.

Day 7, 15 august 2022, Further plans

Video log of my activities of Day 7:

Now that the previous objectives have more or less been met it's time to formulate new plans.

This day and the days around it it was very very very hot so I kept a low profile, not trying to sweat to much, however these plans came out quite nicely.

I went pretty deep with these plans but that's a good thing, that might make the tool awesome in the end, goals to strive for ! ;)

I will briefly sum up the new plans/goals in this log here:

7.11 Visualizing/triangulatization of the hexagons, connecting the centers to form triangles
7.13 Auto-update feature

Day 8, 18 august 2022, Selecting hexagons, Zoom experiments, Text experiments

Video log of my activities of Day 8:

Still very hot, also sleeping is hurt somewhat by worrieing about aliens being real =D,
so I am taking it easy, plus watching ufo related material is fascinating and I want to stay current/up to date :)

8.1 Biggest achievement was being able to select hexagons this was actually quite easily achieveable thanks to great design of GLScene
it's not perfect yet but needs a bit more work to not confuse the mouse button with rotation and such.

8.2 Zoom experiments have been done, also with earth coordinates in the millions now. Zooming will need some further work too.

8.3 Text experiments performed... will also need some further work, like rotate text so it's nicely aligned with cell.

8.4 A possible problem discovered with Uber H3 library. It seems some APIs might be missing/not implemented during build. Not sure why,
github issue was filed about this. If not solved an entire port of Uber H3 to Delphi might be necessary ! ;)

Skybuck Flying

Aug 20, 2022, 4:43:12 PMAug 20
Day 9, 19 august 2022, Explored the H3Index coordinate system, want the IJK coordinate system.

Video log of my activities of Day 9:

9.1 H3Indexes coordinate system explored/visualized in 3D/Sphere/Globe with TGLFlatText from GLScene.

9.2 Wanted to explore IJK coordinate system but was not yet possible because of 8.4 and 9.3:

9.3 Came to the conclusion that the missing UBER H3 API: _h3ToFaceIjk was a serious shortcoming and would prevent the implementation of hexagon routing.
Thus a solution should be found to get this API working. Different solutions came to mind and the ammount of time needed for each solution was considered and

9.4 Decided to wait a day to see if I or anybody else could come up with a solution and to give this a bit of a rest for today, instead of beginning
a "crazy C to Delphi port" ;) :) to save myself a lot of time on the long run and to benefit from future improvements and versions of the UBER H3 library and
saving myself a lot of days/time.

Day 10, 20 august 2022, Multiple spheres idea tried, UBER H3 DLL fixed, IJK Indexes Explored, 3D hooking described, Server Protocol described

Video log of my activities of Day 10:

10.1 Showed the zooming ideas I had in folder/document: Plans and Designs/Zooming.txt
(To be tried and implemented on a later day)

10.2 Described the server protocol, still more work to do to flesh it out.

10.3 Came to the conclusion that it is very important to get IJK coordinate system working. It will most likely affect the design of the algorithms,
such as the peer assignment algorithm inside the server for assigning/allocating peers to the main hexagons.

10.4 Decided to shift gears and try to fix the UBER H3 DLL/LIBRARY by manipulating the C source codes slightly to try and export the needed _h3ToFaceIjk
function implementation. Did so successfully.

10.5 Later discovered some feedback (from github user: isaacbrodsky) on github how to issue the correct CMake build command to export ALL functions:

Step 2: Configure for DLL release:

Step 3:
cmake --build . --config Release

This fixes the H3.DLL and makes sure all function/api implementations are present, at least it should, at least the _h3ToFaceIjk function implementation was

10.6 Examined and explored and visualized the IJK coordinate system in 3D/Spheres for the hexagons at different resolutions, such as resolution 0, 1, 2 and even 3.

10.7 Explored and visualized the multiple/scaled up spheres, one sphere for each resolution. This could be used to visualize how peers are connected for the routing
protocol (in the future once this routing protocol is implemented). This is what I called "3D Hooking", the peers conceptually hook into this
3D-multi-level/multi-resolution multi-hexagon-grid structure.

10.8 Uploaded Delphi bindings for UBER H3 library to github and issued a pull request to try and link their bindings documentation to (my) Delphi bindings.

Skybuck Flying's Delphi bindings for UBER H3 Library (plus some extra helpfull TH3CellIndex record/unit) currently at my internal folder/unit version 0.05:

UBER H3 bindings for other languages document:

Skybuck Flying's modified UBER H3 bindings for other languages document:

10.9 Signed some UBER H3 CLA agreement to be able to do pull requests/commits on UBER H3 github/master:

10.10 For now came to the conclusion that the decision taken at 9.4 was a good/correct decision. The UBER H3 DLL/library was fixed, at least for now it seems
to be in a good/complete state. However keep in mind it does seem to depend on external DLLs/runtimes, I will sum them up in here just in case, according to
TDUMP, so it might still be usefull to port UBER H3 to Delphi to reduce dependency failures/risks on missing DLLs on end-user systems:
I am not sure about this, but for now, to me, it seems like this is the case:
Dependencies for release version of UBER H3 DLL/Library (for Microsoft Windows operating system):
1. VCRUNTIME140.dll
2. api-ms-win-crt-heap-l1-1-0.dll
3. api-ms-win-crt-math-l1-1-0.dll
4. api-ms-win-crt-stdio-l1-1-0.dll
5. api-ms-win-crt-runtime-l1-1-0.dll
6. KERNEL32.dll

Maybe there is a way to get rid of some of these DLLs to reduce the chance/risk of a dependency failure.

For now the gain in days/time (for not having to port it to Delphi) is nice, but it might bite this project and it's users in the ass later, time will tell, if so
a port could still be attempted. One benefit would be that by then the UBER H3 library has advanced, however it could also mean more code to port. In case
many many new features would be added to UBER H3 library an older version 3 could also be ported, which is the version this project uses for now. However
the ammount of changes to UBER H3 will probably be somewhat limited, probably mostly some API renames, then again time will tell. New powerfull features
would be welcomed by me anyway, the library already has many powerfull features it seems.

10.11 A quick glance/exploration was taken into an alternative geogrid called DGGRID:

10.11.1 However there are a couple of issues with it: First of all is it truely an API library ? Or just an application to generate files ?

10.11.2 Second the library depends/requires GDAL which byself requires many many many many other libraries and this would turn into dependency and make
and build hell, and what cost a lot of time trying to build/make this, and I wasn't even sure if it offers an IJK coordinate system, because that was the reason
why I looked into it as initially the UBER H3 library had some issues with the IJK coordinate system and missing APIs for it, but fortunately that problem/issues
with the UBER H3 library has been solved. A quick glance was taking into the manual of DGGRID it does not seem to be suited for API/Library usage and I could
also not find any reference/documentation about an IJK coordinate system.
So my decision was to stay and stick with UBER H3 library and attempt to fix it, which was done in two seperate ways.
Reply all
Reply to author
0 new messages