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

IT WORKS? IN ONCE?

Test I (and wish me luck)

Or most likely totally fuck it all up

And that should link it all together

And tied it to the boxtext routine

Part of all this done

Let's first make sure everything is compiled

And with this I got the stuff to work on the BoxText voice acting

And now stuff is really linked

Glue Script

NOT THERE YET!

Linked to scripting engine

Channel playing checker

I can't help but to love this song

That does at least link stuff to the database, that does not yet tell much about this works in the game, which will very likely just ignore my hard work... for now

I've put the references to the voice acting in

So now we can get down to business

So far so good.... I now saved some junk and if the same junk comes back I guess I can conclude it works.... right?

It compiles, but does it also work?

Let's see now!

Perhaps that fixes that

Not good... yet!

At least the compiler doesn't through any errors, and all I can do is hope that is a good sign

I've made the needed adaptions.... Not sure if they will hold, but they should

I should see if my BlitzMax version of the editor has this covered, but as the BlitzMax compiler has gone completely cuckoo on me, that's no longer an option... Unfortunately

I see my scenario editor even needs to be adapted for this

This will not yet make John Lovejoy's voice actually speak, as the code for voice acting was deactivated for Dyrt (as I didn't need it at the specific time and could therefore also not properly test it)

And what it comes down to now is first linking the audio files to the scenario database

A last check up confirms that everything is cool now!

I've dummied my debug lines as they will only needlessly log things now... (I can always undummy it should the situation ask for that)

Take XIV

Forgot to compile the fixes

Take XIII

MO-RON!

Still no good!

Take XII

I'll get to test this later!

Idiot!

Still no good?

Take XI

Fixed it!

Ahah wrong key set in my C++ code

And the question is now how this could happen

What I see is that all meta fields have been found, but their values are not send into the scripting engine

Take X

needed to run "barf"

Take IX

That was not the source of all evil here... :(

Linkup issue

Take VIII

DOH!

Take VII

These extra debug lines have to explain this rubbish!

"" is music? That cannot be!

Take VI

Should fixed that idiocy

Did it now anyway

Forgot to run "barf"

Take VI

Linkup data transfer

Take VI

You needed to run the C++ compiler for this fix again, MORON!

Take V

Forgotten link

Take IV

(I forgot to "barf" and that's why it didn't happen first, but it also denotes it would have gone wrong anyway)

Neil Syntax error in link up script

Take III

Perhaps this will fix it (Felix Jr.)

I've added a little debug line in JCR in order to sort this one out!

I think the stringmap reader in JCR6 is not acting the way it should

Take II

Extra debug lines

And now I need to find out WHY!

Fatal Error! Error: Entry reader went past EOF! State: C++

Take I

And then it's time to put it all to da test, eh?

And the Map Loader has been scripted to load music tied to a Kthura Map

No idea yet if it all works... that will take testing, but before I can start on that I need to work out some stiff

And that should link everything together

The Neil part of the routines has been written

It doesn't end here,yet!

And the actual possibility to read-out the Kthura Map metadata

Kthura Meta field list

That wil do!

Script adapted

I see that importing is no longer needed... The file's already there... Good!

I see I didn't make the maps contain the music data by temselves. I will go for that this time, most important reason... Saves me a lot of code, and the times scripts are needed are always in the minority and that is still easy to refer to then.

I've set the title... This can at least allow me to check if metadata can be read, however I see the music field is missing

Before I move on I need to make sure that music loading in combination with loading a map becomes possible. For that I first of all need to import a certain file first which my system is now looking up. Then I also need to adept the map to make sure the music loading can be done. Next, the Kthura main metadata cannot yet be read by Apollo Scripts, and to make that possible I need to adapt Apollo... All in all I got work to do!

IT WORKS!!!!!

I've undummied the WaitMinTicks and now see if the text appear speed is acceptable!

Well That "fixes" the game at least, but the bug causing this remains, so this is why I said "void" in stead of fix

And let's see what happens next

I've forced the system to create a new table in a new class record

I think there's a bug in neil when it comes to tables in a class

That's odd

Now that's odd... The table pointer DOES appear to be different every time....

let's see

That didn't make any difference

Let's see

Okay, all textboxes contain the same data.... Something must have gone wrong when translating the scenario loader to Neil

I need another test to get a few things straight for myself

The scenario file looks the way it should, so clearly something must have put the loader out of order

For some reason all lines of a scenario are always put in the same textbox and also in all textboxes... Oddly enough the lines do not all get generated on screen, so I wonder

Let's see!

Now I wanna see why the number of lines is damn high

A little error elsewhere did at least allow me to find out that the generation routines themselves are good... But the crux lies in that for some reason, all textboxes are claimed to have 103 lines... That should never be the case, so I need to sort out why this is happening

The math formula shouldn't be wrong (if it was, Dyrt would have had the same issue)... It is likely though that some basis numbers to start stuff with is not in order.

It's just the height of the textbox that is failing

And the ember bar is generated accordingly

What does go well is the positioning and howing of the portrait of the person who speaks, though

What I do see though is that the generation of the textbox is completely bogus

And now the loading bar also appears

Okay, where were we?

Something missing in the flip call, and it didn't cause a syntax error?

Okay, the events poller does work (that allows me at least to get out of the game without trouble and risking memory leaks), but the flipper does not, and I wonder why

Let's see

A few obvious mistakes have been fixed, though

I've put minticks on dummy in the script to make sure that cannot be the evil

I already had the feeling events and flipping were NOT properly handled by the script, but only by the core

Infinite loop

...

Wrong method reference?

Well, let's see now!

Case error... Yeah, when making direct links with Lua that is still possible

But doesn't explain why WaitMinTicks doesn't work and is considered a "nil" value

That takes care of that one

Perhaps this should avoid this issue

Ah, analysing what happened (I should have known this), it's clear " is the evil here (as always)

I've seen a lot of crap happening before, but never in my life as a coder have I seen scenario text being dumped as an error message

Test numero... what number are we?

Link up code typo

And sometimes I wish Windows wasn't so damn stubborn about which app to show, when I need another!