In my RPG viewer tool which I'm now also rewriting in C#, I've now translated the tool that has to generate HTML data to C#. There is still a lot to do though before it can be properly put to action.

And let's test!

Recompiling

Let's see if this gets me more information on the next error

Well that issue's been fixed... Or so it seems....

I think I found out where it goes wrong.... I guess that was the woe of a multi-state engine :-/

Further investigation of the code of that party bar rule the possibility out even more....

Everything about this error is impossible, so if you've eliminated all the impossible, whatever remains, however improbable, must be the truth... However once I've eliminated all the impossible... NOTHING remains so that must mean the truth doesn't exist. If this error was correct, the party status bar would not work, and it works, and the data in it is in order... None of that could be possible that way, so none of that makes sense....

The error thrown is a bit odd... Is further investigation needed?

Okay, brb, guys!

I've also combined this with the work on my RPG Viewer... 2 projects in once, yeah... only I am stupid enough to even TRY that!

Updated

License and versions

Like I said I'll tackle that one later, as I really need a break now!

#34

Before that test can really start I must rebuild NALA, but I've set it all up in a sequence... just wait for it :P

and regardless if new bugs pop up... I'll first take a break... But only if I can confirm the fix below...

An now Take TWENTY-ONE must tell me if there is more stuff that is wrong

Anyway this should be fixed now... (At last)

In stead of (state,call,chunk) it was (statecall,chunk)... Everybody knows Lua knows a nil will be produced and that parameter 3 has become parameter 2, so this was bound to go wrong. This could happen because the glue code in this department was written in pure Lua (as I had the illusion that would be quicker, but I guess I was wrong).

The try catch did now cause the real error message to appear, and also to reveal stuff was sent to the API that was as false as hell!

The Truth has been found... at last!

Take TWENTY!

Another Try{}Catch scope must confirm this now

I think I found the source of the "evil", but I'm not sure... At least it seems that a few suspects I had before have now been proven innocent!

Take NINETEEN

I've done a few more breakdowns, so I can really pinpoint more precisely what could be wrong, and make sure all other stuff is REALLY impossible now

STILL NO GOOD!

Cockroaches! Lovely creatures aren't they?

Take NINETEEN!

Another try{}catch

Take EIGHTEEN!

Compilation done

NALA must be recompiled as a result though

As the interstate communication appears to be to blame I've added some exception catchers in there

Indeed no difference, but it was expected, and at least that is one possible truth less for sure now!

Take SEVENTEEN!

I've investigated the import order in AnyWay.nil, and I've seen that GameVars is done later, than the time module... Very likely this is an impossibility given the fact that then the translation of time.nil would already have failed before I could even select a savegame slot, but I wanna be sure I can rule this one out, completely!

That does indeed seem to be one of the "impossibilities", but more elimiation may be needed here

Take SIXTEEN!

Now what I did see is that I missspelled "os.date", but I deem it unlikely that is the case, but I gotta run another test to rule this out as a possibility

that being said, the only "possibility" appears to be the communication with the time generator

If you've eliminated all the impossible, whatever remains, however impropable, must be the truth!

The breakdown DOES show that the error took place during the base configuration

Take FIFTEEN!

Forgotten ")"

Computers, don't even TRY to understand them!

And now it doesn't happen....

Now THERE is an error that is not possible... WTF is going on here?

Let's GO! Test FORTEEN!

Some debug lines must provide me some more answers!

Mostly I just need to find out which API is being called, as the evil can lie there (not always, though, as I found out errors during #use clauses also cause this error, but that doesn't seem to be likely now given the time this error is given).

In other words "get lost, asshole!"

".NET error in user code"

Take THIRTEEN

Forgotten "$"

Forgotten "Hover."

ALL NON-PRIMARY TEST ERRORS ACCOUNTED FOR!

Take TWELVE

At one place I placed too many "end"s, and at another too little... all resulting into an "unexpected static", but I guess I gotta love NIL for that, as those kinds of mistakes can give headaches to fix

Take ELEVEN!

Forgotten () in a void definition

Time to feed Kitty

Take TEN (and still that what I had to test, has NOT yet been tested)

And that fixes another issue

Generated a new anyway.nil (I forgot about that)

Take NINE!

And fixed an "unexpected static" in the process!

One "end" too many

(and then to think the primary thing didn't even yet come into play)

Take EIGHT!

Same error, different line

Take SEVEN!

It's "sprintf" and not "sprint".... dummy!

All 'field.h' errors accounted for?

is '$' not working well on group constructors?

Now what?

I may soon have to rewrite This tool as it's an important debugging tool, but as I've always had issues getting lzma to work in BlitzMax, a complete rehaul to C# is needed, plus, that was what I was planning to do soon anyway :P

Take Six!

Misformed function name

Take FIVE!

AAARGH!

Take FOUT

").." seen as unknown term? (NIL should never have recognized that as ONE term anyway)

Take THREE!

Incorrect meta-table definition

Take TWO!

Misformed string

Time for another testing round... Take ONE!

Directory data should be saved now (nothing more than that yet... it allows me to test the UI more fully once I fill more stuff in).

that should be fixed now

Unfortunately, the mouse pointer does not appear

That appears to be wroking

And make sure you can see a "Hover" took place too...

I've put in the "Hover" system in the Save Game UI

The next part will be saving itself, but I hope you'll forgive me that that won't be for today

Anyway, the scroll icons come later, as I want them to appear only if scrolling is possibe, and with only the "Create New Slot" bar, not much need for scrolling

Fortunately I don't have to get out of the house today anymore

It's raining here....

Running

Up-to-date

For now it shows what I want to see

Sidebar added

That is news worthy of hitting all the front papers, you know!

I saw what I wanted to see

So I fixed that!

ScrollY did indeed have to be static... Silly me

Yeah, what, did you listen to Eric Idle's advice? Always look on the bright side of life?

I did get an error, but it's no longer a .NET error, so progress was still booked after all

No more crashes.... so maaaaaybe.....

(I mean, praying to Microsoft DOES feel pretty awkward, you know).

(and hopefully I don't have to pray).

Let's test this....

Well, with these kinds of errors ANYTHING is possible so then the sherlock holmes method, to elimitate all the impossible to come to the truth, won't work, you know....

????

Oh wait, I think I found out what's wrong.... (I hope)

Guess.....

Let's see if stuff works properly now .... :-/

I guess Microsoft still operates deep back down, so I guess it's always a possibility anyway

I think I fixed the issue, but this should NEVER have resulted into a .NET user error.... :-/

When creating a box for the side bar it already goes wrong.... Oh odd, since that routine doesn't call .NET

Let's test once more!

It's sad this is what it takes

All steps broken down in CSAY statements

Apparently that was not it....

And let's test, to see what happens!

Running my compiling script... again

From what it looks like, this is because the 'exists' API had to look through a directory that doesn't exist, so it couldn't look up if a certain file was there either. Since this is the most likely error that can pop up in this process, I'll just throw a warning in the console, but let NALA move on....\nThe API will always return 'false' then, since if the directory doesn't exist, the file cannot exist either, eh?

that means "something is wrong and I am not in the mood to tell you what, asshole!"

".NET error in user code"

And antoher test coming up!

NIL callbacks only work if they are declared global... dummy!

I've at least linked the save game management flow to the debug console, so I can at least be sure that I can easily get out of the game!

And what I see is.... NOTHING AT ALL

Yeah, NIL can have a few cons, but they don't outweight the pros, or at least, not this one!

Forgotten declaration

Testing once more!

And another one

Code Typo

And now with a few quick "$" I could fix another error... This time maybe more work than intended, but in the future.....

Rebuilt NALA so the game should now support this too, and hopefully this'll save me some work

jpbubble/NIL-isn-t-Lua#43

Well that that does it for the with stuff.... yay!

For classes this now works for non-static methods.... Static methods (which are for me the most important, actually) are an issue, but I think I got a solution.... But that will take some effort

Right, with is now recursive and in groups this works too... Only classes are a bit of an eyesore, as "$" doesn't want to work in classes

Nope, groups don't work either...

The "with" keyword in NIL does at least work, however classes and groups should also take care of this, but that doesn't appear to happen. At least not in classes, I need to check groups though...

I've set the basic scope starter in NIL to see _G as the 'width' variable if no proper width has been set

I'm first gonna set something up as this has been annoying me for far too long, so I need to set it right!

Forgot to note... I've added a few arrow arrow aliases for quicker access to a few features later in the game

Let's see now

File not found!!!! (404) :P

All direct parser issues accounted for, as far as I can see so far

(And I hope it'll work then, as I'm getting hungry)

Of course, another test iis in order....

Having to modify the engine itself a lot is normal for such a brand new engine seeing its first real action...
As the game progresses, these kinds of issues will get less and less so I can focus on the content more and more.... In other words, nothing has happened that is out of the ordinairy, but annoying it surely is!

It is one that obligates me to recompile NALA.... AGAIN!

Code typo

And Time for another test

Recompiling NALA

I think I found why a certain line which was merely a method call was seen as a delegate definition and even an invalid one too

Take... Wait... I forgot to count... Oh, crap!

So let's do that, and try again!

A few issues with the garbage collector pop up, but likely this is due to not yet re-compiling NALA

Another test

Group mixup

A basic setup for the interface has been done.... Point is due to lack of savegames, this is now hard to test, but at least I can do something... well I suppose...

So this makes the temporary link at least available, and functional

Command set up to activate the save game screen

Link up code to actually enter the SaveGame menu.... (Since no save spots exists now, this code is not yet accesible, duh!)

A link up code for loading the save game manager from the field... This will of course be needed when accessing save spots (although, mostly this load process already took place due to loading a savegame, but since you can also start a New Game, I may never assume this!)

A basic setup has been done to generate a slot to create a new file in.... This is all in preparational form, though so no big results can be expected.... YET!

Set up a checklist in #33, so I won't forget what data a save game file needs.

I've made it possible for the BUBBLE API to load text files in "lines" (Please note, it only takes Unix based text files with only LF based line ends, and files must be in full ASCII).

And so far for now

Both converted to PNG

They are both made in DPaint, so I got them in PCX format for now... This is however easily replaced

Arrow Down

Arrow Up

Well, a neutral tag this time.... An error did pop up, BUT, it could only do so, because no longer in use Kthura maps are INDEED disposed of! COOL!

MOAR Kthura case error fixings

The folder that I intend to use for "dynamic data" (save games, swap files and main config files) is correctly set

Kthura case error

I swear... I should make NIL classes case INsensitive :P

workdir issue

Although on second thought.... i do think it can wait... I've a better idea

I just realized that before I can get properly on the savegames I need to set the global config right. The global config will contain data that will always be in order no matter which savegame has been loaded... This file will allow the NG+ to start, or the True Abyss and what to put in the JukeBox and not to mention some vital data about the savegames themselves as well (not the content of the savegames, but some general data)

Skeleton for the save game interface

SaveSpot Red

SaveSpot Green

SaveSpot Blue

Well, later!

#29

THAT WORKS!!!!

I've set the texture of the actor to update automatically... This should allow Irravonia to turn around in accordance of the direction she faces

I've worked out stuff a bit in the NIL glue code on actors... Hopefully this can prevent memory leaks in the future, but I cannot yet be sure

Added tag OPTIMIZATION

Right now Irravonia won't turn her face depending on the direction she goes... This is not a bug, this is simply because I did not yet put in the code required to make tha happen.

YAY! All issues fixed

I think I found the "evil".... A correction on the x-coord where it had to be the y-coord.... Could make perfect sense in the first go, since Kthura exit spots are (especially in my forest levels) rarely bound to the grid (something I often do for fun) :P And thus correction is needed, and the y coordinate is very low in the start position.....