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!

C# code copied into C++

The first thing that comes to mind now is to simply translate my C# code to C++, but if that is gonna work I dunno.... C# is not as much as C/C++ as you may initially think, so this can be harder than it may initially seem.

I do not recognize a[3][4] as a multi-dimensional array... That's just cheating with an array of an array, and for the blockmap certainly not desirable. a[3,4] that is what I need.... Dim2 does provide that... well almost, as C++ also doesn't support properties like C# does, but at least this will help me along the way.

Dim2 library

The blockmap can be a bit of an issue, since C++ does not support multidimensional arrays

It crossed my mind that before I begin on the actors it may be better to focus on the blockman generation first. The actors need that.

Well a new day has officially begun, but that's for later :P

Repositories up-to-date?

Not to mention that I got now 1 entire page for this day.... Yeah in this stage of the game production that can happen, I suppose

Ah yes!

Well, do I got anything more to update?

Game Updated

Apollo Updated

Running

DAT WERKT :)

AGAIN!

I did this too fast, and ended up with English

Now that I am still at it, I might as well try to see if the game works in Dutch

Fix confirmed

But I gotta test again!

I think I fixed this

When no boxtext picture is there, or when it's non-existent, the game crashes

It was to be expected

Take I (and the only test I hope)

Well all shit aside, let's test if the scenario rolls accordingly

And next Monday I expect my post order and I intend to watch that first... I need to take a break every now and then anyway.

Although on Sunday I've always very little time

I've set the scenario to roll, however I need to begin on the actors soon, as before they are coded I cannot do much anymore on this game

In stead of worrying too much about that cockroach now I think I can better focus on the business at hand

It keeps insisting to destroy BOXTEXT twice

I've Forced MyState to NULL and hopefully that will settle things?

No idea why the destructor for the BOXTEXT state is called twice (causing Lua to crash in the process)

YAY!

And let's put this to da test now!

But hopefully this will work now

And to adapt Kthura

I had to expand my TQSG library a bit for this

Yeah it was still set to do, and with the high crap load of work I have, this one escaped my mind.... STOOOPID!

One should note that the ability to animate was not present in the initial version of Kthura, and this being the very first map ever created in Kthura (not counting my test crap) it is not that unlogical it has wrong data on this department. As BlitzMax did properly cover this up from the start, there was never an issue, and apparently I did cover this also well in C#, but not in C++, I suppose

The mystery appears to be cleared up.... Kthura tries to animate, but apparently it is not properly prepared to reset

But I think I got on track though, but I do need to set a few things in Apollo first before I can debug

Even my wallpaper in Windows disappeared as a result

Well, a total failure

This is getting stranger by the minute

A very irritating debug line has been added, but hopefully it will answer something!

Kthura is programmed to throw an error if something is wrong here, so why didn't it do that?

So it is the layer that does the evil still then?

Take IV

Fuck you!

Take III

More investigation required I suppose

Apparently, there's more to it!

Take II

I think I know what this happened

I KNEW it!

It was all too easy

And it looks like the lack of an audio file that is to blame....

total crash-out?

Although the actors won't appear yet, I've set up a small thing to allow a bit ot switching already... The actors themselves come later

My repositories have all been updated (I think)

And besides I'm tired.... Mind you?

To get the next portion of the game definitely on the move I will have to get the Actor support for Kthura on the run, and that requires more time than I currently have

Well, not that I'm complaining, but I am a bit surprised at the least, yeah