Is Quest for Glory IV even in consideration?

Freddo · Post by **Freddo** » Mon Sep 15, 2014 2:18 pm

balpat wrote:Might I add that in both games - GK1 and QfG4 - the main portions of the graphics are so badly upscaled that I suggest not bothering with hires at all.

Agreed, I prefer the games in lowres.

Napalmtree · Post by **Napalmtree** » Mon Sep 25, 2017 7:17 am

I might add, the windows version uses static black and white mouse pointers as opposed to the sweet animated color dos ones. And you will hit bugs in all versions even with all the unofficial patches. If ever a game needed a port/rewrite it's this one because it's one of the best adventure games ever made.

m_kiewitz · Post by **m_kiewitz** » Fri Sep 29, 2017 8:06 pm

Napalmtree wrote:I might add, the windows version uses static black and white mouse pointers as opposed to the sweet animated color dos ones.

When running it in ScummVM, both versions use upscaled colored mouse cursors.

We are doing the same for Gabriel Knight 1. The DOS hires version originally had really tiny mouse cursors. Those are also upscaled.

And you will hit bugs in all versions even with all the unofficial patches.

We will try to fix them all.
And at least my goal is to even not need any of those unofficial patches anymore. Sometimes those cause additional issues.

If anyone got some time on his/her hands, please play through QfG4 using ScummVM. If possible without any fan made patches. Most timer related issues should not be a problem already.

Still even the slippery slope right at the start is not patched yet, because it's a really complicated issue to patch and I was also busy with other things.

lwc · Post by **lwc** » Sun Dec 17, 2017 11:15 pm

ScummVM v2.0.0 is out and QFG4 is now officially about the only unsupported SCI game, and specifically the only unsupported SCI2 game. Actually, even SCI3 games are supported now. What's so unique about QFG4's engine?

Post by **digitall** » Mon Dec 18, 2017 12:23 am

I am not a SCI engine developer, but a quick code search throws up this commit:
https://github.com/scummvm/scummvm/comm ... 985b9bae73

So it looks like QFG4 and PQ:SWAT are still marked as UNSTABLE since there has been no full playthrough of either game?

tsoliman · Post by **tsoliman** » Mon Dec 18, 2017 12:25 am

My understanding is:
- Lots and lots of script bugs.
- Pathfinding issues due to the differences between original Sierra pathfinding and ScummVM pathfinding.

(I am not an SCI dev)

snover · Post by **snover** » Mon Dec 18, 2017 3:35 am

tsoliman got it exactly right.

I would have really liked to officially support QFG4 in the 2.0 release. Unfortunately, it has a significant number of game-breaking script bugs which are also quite difficult to patch, and because of the game’s non-linearity it is very hard and time-consuming to adequately play test it. The NRS patch set helps ease a lot of this, but there was reluctance to require this patch set, so that was not used as the solution for the script bugs problem.

The differences in pathfinding algorithms are also a larger problem for this game than many of the others, for whatever reason. Being able to walk through the balustrade at the inn near the start of the game made it feel pretty badly broken, and notwithstanding the game-crashing script bugs, it wouldn’t feel appropriate to me to call such broken behaviour in a game “stable”.

I hope in the future that someone who really likes this game will come work on the necessary testing and script patching so it can finally be added to the list of officially supported games!

2Mourty · Post by **2Mourty** » Fri Dec 22, 2017 4:20 pm

2Mourty wrote:Congrats to the developers! This release is awesome and covers almost any game I could want to play from the sierra catalogue. Official support for Gabriel Knight!!!

Now we just need to get support for Quest for Glory 4: Shadows of Darkness. I love that game beyond all reason. I would offer to help, but I'm a professional choir director not a coder.

Again congrats for the awesome release!!

I LOVE this game. 1st computer game I ever bought. I do not know how to the technical part of fixing the broken scripts, but I would be a willing tester for this game. I have played it over 100 times and gone through all of the paths!! Sign me up.

DarkSoul42 · Post by **DarkSoul42** » Sat Dec 23, 2017 5:25 am

First of all, many, many, many thanks to everyone who pulled in all that hard work to even get it to this state.

I just gave the game a quick try.

In 2.0, wraiths do not activate unless you click the hand on them to fight them close-up (which kills you by accelerated health drain as in the original game). (However, this worked in the Dec 9th 2.0 stable build, so it seems someone has caught on ; again, I want to stress my appreciation for people going out of their way to make this work)

The dark one cave slippery path worked for me, though pushing the stone to get back on top has some weird sprite interaction (the hero is covered by the pillar). I also have seen the funny thing occuring with pathfinding in the inn, the hero makes a beeline on thin air to open his door if you click on it from the ground floor. I could actually play quite a lot of puzzles the way I used to back in the day and got as far as 100 points-ish out of 500 on a Paladin character with a propensity to fix puzzles the Wizard's way. This includes going in the Monastery, the Adventurer's Guild and training a bit, Dr. Cranium's puzzles and copy protection, meeting Anna's Ghost...

However, I got a surprise crash when answering the Leshy riddle about the Rusalka. (I don't think it would be appropriate to report a bug at this time ?)

The game actually displayed :

Code: Select all

<Messager>
msgType set to 0 or 579&#58; 0, 0, 0, 1 not found

Before dropping to the debug console :

Code: Select all

ERROR&#58;
parameter 0&#58; 0000&#58;0000 &#40;null, integer&#41;, should be list
parameter 1&#58; 0000&#58;00b5 &#40;integer&#41;, should be null, integer
parameter 2&#58; 0000&#58;0000 &#40;null, integer&#41;, may be any &#40;optional&#41; &#40;more may follow&#41;
&#91;VM&#93; kListEachElementDo&#91;5a&#93;&#58; signature mismatch in method talkerSet&#58;&#58;eachElementDo &#40;room 579, script 64999, localCall ffffffff&#41;!

This to say, also count me in for testing/reporting if needed.

2Mourty · Post by **2Mourty** » Sat Dec 23, 2017 11:08 pm

I was able to reproduce the above error. I tried answering the Leshy riddle before and after visiting the Rusulka for the 1st time. Error happened both times.

I can continue this quick fighter play-through, or should I wait until this error is fixed? A few quest trees are messed up by not getting into further conversations with the Leshy.

DarkSoul42 · Post by **DarkSoul42** » Tue Dec 26, 2017 4:41 pm

Funny pathfinding quirks :
- South, then west from the gate of the town, the hero walks to the center before walking back to the right of the screen
- In the Borgov crypt, if you go through the twisting/curving tunnel that goes to the other side of the screen, the hero remains displayed behind the relief background elements.

The crypt also sometimes yields these messages :

Code: Select all

WARNING&#58; Uninitialized read for parameter 1 from method hero&#58;&#58;changeGait &#40;room 510, script 28, localCall ffffffff&#41;!
WARNING&#58; Uninitialized read for parameter 1 from method hero&#58;&#58;changeGait &#40;room 510, script 28, localCall ffffffff&#41;!
WARNING&#58; Uninitialized read for parameter 2 from method hero&#58;&#58;changeGait &#40;room 510, script 28, localCall ffffffff&#41;!
WARNING&#58; Uninitialized read for parameter 1 from method hero&#58;&#58;changeGait &#40;room 510, script 28, localCall ffffffff&#41;!
WARNING&#58; Uninitialized read for parameter 1 from method hero&#58;&#58;changeGait &#40;room 510, script 28, localCall ffffffff&#41;!

DarkSoul42 · Post by **DarkSoul42** » Tue Dec 26, 2017 4:45 pm

A fight with a Necrotaur will also crash with :

Code: Select all

WARNING&#58; TODO&#58; Horizontal shake not implemented!
Invalid arithmetic operation &#40;bitwise OR - params&#58; 002d&#58;1200 and 0000&#58;0001&#41; from method hurtMyself&#58;&#58;changeState &#40;room 810, script 870, localCall ffffffff&#41;!

2Mourty · Post by **2Mourty** » Tue Dec 26, 2017 6:26 pm

I'll double check the pathfinding issues you mentioned DarkSoul42.

The interesting thing is I got through a battle with a necrotaur just fine without error. I will fight a few more and see if I can reproduce the error.

Thank you for working on this DarkSoul42!! I love this game I can only hope a coder will be able to fix these issues.

DarkSoul42 · Post by **DarkSoul42** » Wed Dec 27, 2017 1:53 am

2Mourty wrote:The interesting thing is I got through a battle with a necrotaur just fine without error. I will fight a few more and see if I can reproduce the error.

I actually wonder if I could finish a battle with a Necrotaur by only casting spells. I crash after swinging my sword at one...

As for revenants, outside of the battle screen, I can easily finish one with spamming flame darts. (I tried this with Wraiths to get Pyotr's sword first thing in the game, but THAT one is too overpowered to think about attempting that without Reversal)

tsoliman · Post by **tsoliman** » Wed Dec 27, 2017 2:00 am

DarkSoul42 wrote: I actually wonder if I could finish a battle with a Necrotaur by only casting spells. I crash after swinging my sword at one...

The vagueness of this statement is why this game is a PITA to test. In any other game it is very specific and "good-enough" for a bug report.

But without a savegame to go along with it, there's at least 3 followup questions (if not more):
- Fighter or Paladin?
- New game or import?
- Which necrotaur?

I am not expecting answers here because if you file a bug report you'll attach a savegame (right?)

I am just making a point about how many variables a given situation has in this game vs say Space Quest