I am required to do rebuild the game itself for that, though

I need to find out if the animator is what causes things to turn invisible

I've put in an extra precaution that should make sure all Kthura Maps are unloaded (this because Lua is on the rampage here and I don't want more trouble than I already got).

Take XXXV

An extra debug line should confirm if indeed most of the map was killed (for no apparent reason)

Now the question is, of course, why this all happened

CWRITE> ____________XXX___XXX_________________X
CWRITE> ___________XXX_____XXX________________X
CWRITE> __________XXX_______XXX_______________X
CWRITE> XXX______XX__________XXX___XXXX______XX
CWRITE> ______________________________________X
CWRITE> ______________________________________X
CWRITE> ______________________________________X
CWRITE> ______________________________________X
CWRITE> ______________________________________X
CWRITE> ______________________________________X
CWRITE> XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

Half of the blockmap is gone, which may indicate half of the map is gone as well

At least the actors appear, but now we do have some other bones to pick... They are shown far too small, and their environment has gone.... I'd like to know why that happened.

Take XXXIV

Syntax Error

Take XXXIII

Kind

The destructor should be obsolete now

Constantly All objects are being destroyed, I see

Now I do have a few ideas to make the destructor obsolete, but still I need to check things out!

Now the destructor is called while it shouldn't be (causing a crash in the process).

Take XXXII

!= is not the same as ==, Jeroen!

It KEEPS insisting not not working!

Take XXXI

Whatever causes this bug, it's persistent!

Fucked it all up AGAIN!

I am not definitely pissed off!

Even now that the actor could contain nothing but the correct data I get garbage

Take XXX

I hope

I think I solved the case! (or is that wishful thinking?)

Elementairy, my dear Watson!

This can only mean that the parent layer is somehow not properly set in the Actor record, which should have happened... The question is, of course, why...

Aha... Now the log makes more sense.... The layer for dominance remapping is declared a null pointer....

Take XXIX

I do need to re-test as a few things in the IDE went wrong

Ah... C++ placed the error marker on the wrong spot (what else is new?)

The location where the crash is noted (a debug line just containing a "cout" and a constant string, is what surprises me the most , though.

Apparently.... no work at all! I guess fate is against me after all!

I can at least confirm that normal objects appear to be working under the new code

Well We're still counting on... test XXVIII

I needed a break before the great test would commence

And now it even deems the compilation succesful? Is it really done?

Hmm, it has stopped giving me errors in the Kthura library and now wans the adaptions for the API

That's only half the work, though!

I think the core part of the operation is done....

DUMBLE-ROLLED!

:-P

With C++ desperately trying to get EVERYTHING in a bad manner, there is only one thing left for me to do... I'm gonna fuck up the Kthura library ENTIRELY!

That experimental line was exactly where C++ didn't like me, at all

As expected

Take XXVII

I've tried something but I expect not much from it...

Take XXVI

MOAR stuff

Things appear to go wrong when the actors are being called.... It was to be expected, yet, I wonder what the fuck is going on here!

Take XXV

I've added a debug line in the iterator... Hopefully it answers a few questions

Is the iterator still broken?

The data in the debugging screen doesn't make sense at all.... What the fuck is going on here?

Returning a string is reason to crash?

I see now that the system does follow a different traceback route though, so that is a bit odd

I've ruled out possibilities for memory conflicts now, so NOTHING and I repeat NOTHING could have caused anything here.... Yet it still happens.... and the computer will have to clear up why!

With no more reason to crash, it still crashes

Take XXIV

That was take XXIII btw

Will backspace is a handy key, but still

Why did an identifier get cut in two over two lines?

The blockmap builder has NOT been done. It would a) be too much hassle and b) Actors were ignored by the blockmapbuilder even before this update because c) including actors in the blockmap builder is in 99% of the cases beyond stupidity to do anyway.

Labelmap builder updated

Tagmap builder updated

Dominance regenerator has been updated

But first a natural break

This will NOT yet affect the mappers, that will be the next step

The Kill All has also been set to clear both lists

And the killer will already respond to the correct vector in order to remove actors if they should be.... This should happen automatically

I've put the actors in their own vectors

And while I was in bed I realized why I've been stupid with the object list... After all, and I should have realized that, C++ is primitive, and so a memory conflict in using the way I used vectors was only bound to happen.... Since the listing system in both C# and BlitzMax only store the pointers and have a completely automated memory system around that I never realized this... C++ is, of course,a different story....

Take XXII

ALthough as an intelligent person, I don't understand how this could happen

I've forced the actor spawner to rebuild the ID map though

On Crystal the system crashes on being unable to find the object ID, on Wendicka this didn't bother the system

Take XXI

Internal fixups

Take XX

I hope I got it now

The actor IDs remain 0

texture ""?

Take XIX

This should prevent it

The bug that happened now should not be possible, but did happen

Take XVIII Confirmed

Well I fixed it!

I guess that's Microsoft for ya!

VisualStudio complained about bugs in strings, while the fault was made in my vector usage....

The error gets picked up in the xstring library on the tag of Wendicka's actor... Why this is happening is something I cannot tell, as the traceback only limits itself to the function where the fatality happens, but it doesn't tell me which function called the fatal function.

Take XVII

Inside visual studio then and hopefully that produces some useful insights

Stuff gets even stranger now as the game now crashes without any notice

Take XVII

Compile

Part II done

Part I done

I see my API has a few advancements that now work against me, so I need to experiment on a bit

Take XVI

Adapted API to this experiment

I think my entire approach may be faulty and if I'm right I guess I may be lucky I didn't get a segmentation fault

XV

Can't be it, but here goes nothing

I wonder what happens if.....

Take XIV

Why the tags do not change I cannot tell though

I've at least fixed the ID...

a few more things

A few issues do arise here.... First of all not the correct object ID(), which should happen... But I guess that's only natural, and second NO TAG!

Take XIII

Done

Forgot to compile and that debug line was in the C++ code

Take XII

Ultimate debug line and let's see if this provides some answers!

The forcing didn't help and the created output by the tag list confirms the actor is not listed in the tagmap

Take XI

Extra debug line

I've forced the system to always rebuild the tagmap whenever actors are being created, regardless of underlying configuration settings

So far the logs confirm the entire claim is nothing more but bullshit

Take X

Let's see

Extra debug line

Take IX

Field name != Object tag

That at least covers the issue of having no actor with the specific tag

Take VIII

But I think I got it now

Nope!

Take VII

?

No this has NOTHING to do with corona... I am just traumatized about things that happened over the entire course of my life, and that is getting onto me.

My stress is really getting the best of me

Take VI

Link up error

Take V

no "return"

Take IV

Syntax Error

Take III

Forgotten declaration?

Take II

Fixed Syntax Error?

Take I

Does it work?

Actor texturing

Stress got the better of me... I need a time-out!

I still need to texture them out, though

Actor spawning

It does appear I forgot a few things

Well, I guess that appears to be the case... Of course, there's still le moment supreme

Take I

Now before I get to the actual testing of the actors I wanna test if the link script compiles now. If the title screen appears without question I guess it does.

Compilation gives no errors... Is that a good thing?

Script part taken care of!

Not there yet!
I need to make sure the Kthura classes are able to work this out properly.

WalkTo and MoveTo linked to scripter

Spawner linked into scripter

API functions have been written... Not yet been linked to the scripting engine....

That finishes the hardcore stuff in Kthura, Apollo cannot yet handle all this as the APIs that allow Kthura with the scrips have not yet been written, at least not where actors in particular are concerned

Code compiles, and that's all I can test now, so far: Geen bericht is goed bericht!

Driver updated

And thus that will be the next step, and also a crucial one as I cannot test anything without that driver being updated

Now the actors only work from core, the draw routine will call the driver to actually show the actor and that's also where the road ends for now, as the driver has not yet been updated for that.

Running

And I'm writing in the wrong devlog I see

Empty base script

HOLD THE SHOW! I forgot some mortally important things!

Let's see of the Hall of Heroes works so far!

I've linked this map to the world map

The ones who you should meet in the Hall of Heroes will NOT yet appear!

Basis Map Hall of Heroes

Time was also not friendly enough to do my Dyrt duties, although I think I can do a few tiny things

However the time has come to a point that I'm not really game to test if it works...

The Actor Class has been set up

I will now have to work out the actor class... Now actors are merely an extension to objects, however actors have to be more versatile, and this versatily brings that they gimme more data to cope with...

At least Apollo compiles, and up to this point no news is good news...
I will have to see later if it actually works

You've got to keep in mind though that the pathfinder can only be put to the test once the actors work, and the actors have always been a nasty thing in Kthura. They will require some workout before I can put this all to the test

It's all done now

I've created the Dijkstra driver for Kthura

I've added a Kthura point class... The module I'll use path finding will have to use this class to make sure all output regardless of pathfinding module used will aways output the data in the same format so Kthura can understand it

Now the next course of action will be to attack Kthura to my Dijkstra routine, now in order to make Kthura even more modular than ever before I'm gonna make Dijkstra a driver for Kthura. Now I will likely stick with my Dijkstra routine forever... (I don't have much need for anything else), I can this way make Kthura operate without Dijkstra when no pathfinder is needed (saves RAM) and if Kthura ever gets adopated by other people, they can then use their own pathfinding routines if they desire, so all in all, it's the better run and not that hard to code in C++ anyway (it was not even hard to code in C#, but I was back then too much in a hurry).

The blockmap output shows me what I wanted to see

Take II

That's not good!

Take I

Well, what will this all produce?

My Debug script allows me into this data now

NOT there yet, thougg!

And now the link is complete

Neil code

Part of the link

integrated

I've set up an API allowing the Apollo Scripter to readout the BlockMap for debugging purposes

took a shower

Minor issues

Something was not right in the BlockMap Checker

The blockmap builder has been done, but is not yet fully operational

No this will NOT compile C++ != C# after all, so I need to adapt things as a result!