AT LAST one fight without any crashes

Would I really be free of all bullshit now?

Take MMDCXXXI

Well, there are some of those days I suppose

And when I thought it was finally over another bug must come in

I really hope that is the last fucking error I gotta deal with, but I guess I'll have to deal with it.

Take MMDCXXX

Well I recovered that variable, but it remains odd.

A variable disappeared, but how did that compile anyway. Lua should have crashed in the compile phase already (Lua itself... not Scyndi).

Take MMDCXXIX

Database junk

Take MMDCXXVIII

I think I found the motherfucker

So it ignores the front line completely? What the freaking fuck?

Take MMDCXXVII

The no targets bug did not yet give much clarity in the logs. I hope the next test does as I added some extra log instructions

Take MMDCXXVI

And did I not properly save and export the data for a normal attack for Scyndi?

Missing Rosetta tag

Switching should end the turn of the character who leaves the front row

Take MMDCXXV

And that DOES justify another take

Illegal function call

Color bug that does not justify another take

Take MMDCXXIV

Whine!

Take MMDCXXIII

No targetable players for a foe. That is not possible to happen, so a debug line must tell me how the impossible becomes possible

Take MMDCXXII

Crash on less than 6 characters

Take MMDCXXI

Illegal function call

Take MMDCXX

Established

Switching was not linked to the PIMI yet

Take MMDCXIX

Missing #use

Take MMDCXVIII

Two more of those

Take MMDCXVII

Scyndi's declaration of locals bug

Oh yeah, and switching in combat should also work now

Take MMDCXVI

Scyndi's normal attack has been set up

But first a little break

And now it actually get into switching during combat

Switch string for localization

Take MMDCXV

All party members turn skip

Take MMDCXIV

AI should not target back row

Missing link to Scyndi's PIMI

That should be taken care of now

The system did clearly NOT understand to leave the back row alone

Take MMDCXIII

Mini portrait issue

Take MMDCXII

And now the true testing can begin

Some treasures added

A few finishing touches before this place sees its TRUE test

Random encounters configured

Candy

Several abilities for Candy

Sugarland

Maybe I'll be back later to put in the encounters for Sugarland

Updated

Updated

All bugs accounted for (or so it would seem).

Take MMDCXI

API issue on Moving check

Take MMDCX

Link up issue, causing Scyndi to be unable to move over chasms (which was precisely what she was supposed to be able to do)

Property returning wrong value

Take MMDCIX

Now I *THINK* I might have solved this, but the next take will take

A nonsensical error pops up (no way a NIL could EVER come out), and the line number is not shown, making it impossible to properly debug this

Take MMDCVIII

Scroll issue fixed

But I'm not there, yet!

I see two function calls that were not right, so that has been taken care of

That's odd.... No data in the log at all... That really shows something is REALLY wrong

Take MMDCVII

And extra entry to the log should tell me more about this

Apparently something is blocking passage for Scyndi. I need to find out what.

Take MMDCVI

Recompiling SCI

Links it all up

Glue code

Moving API

Take MMDCV

Bad identifier

Take MMDCIV

Code copy conflict

Take MMDCIII

Well, I certainly HOPE this wil work the way it should

The hooking ability should work now

Some workout to player movement

Checkup for free way from Scyndi to spot

I really need a break now!

So far, so good!

Take MMDCII

I placed a little marker. This as the pole will always be outside of the view range. A bit of a woe due to SugarLand originally being desinged for a game where there was more on screen

Take MMDCI

Debug errors

Take MMDC

Back to the actual run-time tests

The translation shows what I wanted to see. Good!

I do still want to check the translation, though!

And it seems to work now and all compiler errors are accounted for in the process.

Take MMDXCIX

Anyway, it's fixed now

In C++ switch cases required "break" after every case. The irony that I would forget that while doing THIS particular thing.

Take MMDXCVIII

One syntax error the warning system popped up when compiling the Scyndi compiler

Take MMDXCVII

Break should now be translated

ERROR> Unknown instruction kind (40) (Internal error. Please report!) in line #68 (Script/Flow/FlowField/PFA_Scyndi.scyndi)

That is the error message I wanted to see. Good!

Take MMDXCVI

Keyword "BREAK" now recognized by parser, but not yet fully set up by the translator.

"Break" added as instruction kind

So far, so good. Next step

Take MMDXCV

C string fault (merging stuff that should not be merged. A comma was required here)

Take MMDXCIV

Tests only to make sure Scyndi picks up stuff correctly

This will not yet make Scyndi support break yet, but most likely throw an error for a keyword it can't handle

Keyword list updated

I see I didn't yet put in break support in Scyndi yet. Let's first take care of that, as I need it!

Take MMDXCIII

I've put in some debug log lines since this all happens under the hood (for now).

A pole can now be accepted if the coordinates are right. Checking for a direct connection possibility without any blocks will slowdown the game, so it will only do that if this check up which is rather fast, is succesful. When no poles have been found at all, the step will therefore be rejected to save time.

So far, so good

Take MMDXCII

Will it work now?

Fix ignored

Take MMDXCI

Once again, if no run-time errors pop up, I must assume everything's alright

At least no compiler errors

Loading the shit completely fucked up

IDIOT!

Take MMDXC

Except for not seeing crashes all I can test is the compilation. If loading a map doesn't crash the game I must assume everything is so far alright

The maploader should load the hookspots now

Pass zones set up of Sugarland

And NOW stuff appears to be working the way it should (at last)

Take MMDLXXXIX

Luckily I got my Kthura hack tool

Take MMDLXXXVIII

A devastating bug fixed in MapAction. Even leads to corrupted data.

Take MMDLXXXVII

The poles in SugarLand were already tagged, due to them being part of Dyrt. That doesn't really matter, though, the way MapAction works makes it alright, it reacts to texture and not to tags, and it will log the poles into the poles data the game needs accordingly.

Tagging of untagged poles didn't work the way it should

Take MMDLXXXVI

MapAction should now support the poles for Scyndi to hook herself to

As I also need to script Scyndi's hook on ability (tutorial about that will be in the sky tower), I will focus on that first, so I will not be hindered by random encounters. The boss will also come later.

I've set my personal save to SugarLands now

Stuff works

Take MMDLXXXV

NOW the generator does what I wanted to see

Again

Tag

I don't know why, but the tagging was undone somehow and that was the cause of these failures.

Take MMDLXXXIV for real this time!

I'm not sure, but here goes nothing

Again!

Hold the show, the compass generator missed something. Maybe a save and generation order conflicted.

Take MMDLXXXIV

The reason for doing this first is due to this being something I'd easily forget, and although the removal of the guard seems cosmetic, it is actually more important than you might thing, as it would create a plot hole.

Compass data updated

Aziëlla's guard. As you won't go to Aeria during the events of this game, he has no more purpose here.

Chain to the Beach and back

I may work out Sugarland first before I work out Independence's random NPC. This to make sure that Scyndi (the character, not the programming language used for this game) is completely correctly configured. It won't be long before Aziëlla joins the party, and Rachel will join too. This can allow me more time to have at least Scyndi in order. New characters are always a chore to work out in RPG games, so I learned over the years I create RPG games.

This should all do for now.

Bio requiring 256 FP for Myrah? That's not even possible.

Take MMDLXXXIII

All of Scyndi's abilities accounted for

Take MMDLXXXII

Restore

Scyndi's IDLE pose

Well, transferring complete. One more thing left to do, which will in Scyndi's case not be too hard.

Updated

The next step will be to make sure all of Scyndi's combat graphics are in place.

Now that will not be too hard as most can just be imported from Dyrt. Only one may have to be added as that was not supported in Dyrt and that was an actual combat stance in IDLE form.

Well that works, and was also an important step to take before I could put Scyndi into battle

Take MMDLXXXI

Small change on the marking system as I feel otherwise the altscreen can cause ugly effects

No switch took place

Take MMDLXXX

Ah, found the little bugger

Now this was only cosmetic, but a true issue is now on the way as bool and int type mismatch and I really need to sort out why it thinks that.

Chosen marking effect was not well

Take MMDLXXIX

Mistake

Take MMDLXXVIII

A cosmetic bug, yet a rather ugly one

Cosmetic When hovering one character the next would turn light blue

Which was pretty stupid

Initial chosen not set

Take MMDLXXVII

Switching should work now (I hope)

Take MMDLXXVI