running it in low res screen and disabling the bilinear filtering that was on by default made it playable. It is still slow that said but not much than what i remember playing day of the tentacle on a 486 ![]()
I played the first monkey Island on an 8086. Scrolling would have been so slow that they disabled it
Honestly if Thimbleweed Park runs badly it’s entirely on the devs, my guess is they tried to recreate the SCUMM engine using a modern toolset or something because, aside from the color depth, resolution and the tutorial messages (Ugh), I see no difference with old SCUMM games, which ran mostly fine even on 486.
Maybe they recreated the engine because of licensing issues but it’s clear they didn’t care about optimization. (because of course with PCs as powerful as they are now who would care about optimizing a 2D game when it could eat system resources at will instead?)
Honestly, even the fact that ScummVM supports it via a separate engine instead of using/expanding/adapting their implementation of the SCUMM engine for it makes no sense to me.
The game doesn’t use the scumm engine but a completely different engine.
Scummvm, despite the name, implements various game engines
Games running in residualVM are Not working anymore in SCUMMVM…
Try Escape from monkey Island 3d!
So said I am not a friend of bundling all Kind of engines in one APP
True to BeSNES and SNES Emulator too!
Both worked faster as standalone Emulator s!
Look, I know, I said as much in my post, but it just sounds stupid. It’s clearly a game made for retro-pandering, what was the point in not using the actual original engine? (Other than possible licensing or technical issues)
They could’ve collaborated with ScummVM in making an updated version of the engine or something instead of rushing a proprietary engine that mimics SCUMM without any of the optimizations that engine has.
Also, I’m not sure I like ScummVM implementing various game engines, as Bruno points out some of the engines they “support” are broken.
They recently broke Nuvie’s enhanced UI and made nonsensical edits to its strings for some reason. I don’t see what was the point in absorbing a wide variety of independent projects if they cannot effectively maintain them all.
EDIT: I don’t want to be antagonistic, I’m just wondering why the game turned out to be like it is now. (And why ScummVM devs decided to dig themselves into (IMO) an untenable situation)
MAME / MESS supports a huge number of different sistems, I don’t see anyone complaining
MAME doesn’t usually break stuff from one version to another all willy-nilly.
(They do tend to break old ROMs overtime as they refine the engines they’re emulating but IMO that’s fair game)
It also started out like that (more or less, MESS was incorporated a long time ago), while ScummVM only got to that point fairly recently.
I’m not trying to bash ScummVM devs but it seems to me they can’t manage all the engines they absorbed, or else things wouldn’t be breaking like I’m experiencing.
I see this kind of complains about foss simply unfair. Given I like scummvm very much, it is undeniable it gives thousands and thousands users the only way to play their loved games - hundreds! - on LOTS of different os/platforms. This is HUGE work. And we have it for free.
Does it have bugs? of course it does, but bugs can be fixed. Did anyone filed a bug report about twp being slow? I can’t find any. So why any developers should put any effort in fixing something that’s not pointed out to be broken?
You say Thimbleweed Park is slow? play it on a different platform using the exact same version of scummvm. Is it slow as well? yes/no? are the very same compiler options applied for haiku build? yes/no? Those are just starting points: even non-dev users can help foss.
BTW, scummvm devs have always been handy, as haiku ones.
At the moment, they are introduced in MAME or MESS, games are running often slower than on dedicated engines. It’s probably because first engines developed for an hardware are optimized to be playable and devs are taking shortcuts to speed up rendering. Once, the engine itself is mature enough, the accuracy becomes mandatory for the integration in a bigger emulator. That also requires implementation of features also available for other supported platforms therefore use of shortcuts is no longer possible. It takes generally few years to retrieve some speed but, on the other hand, you get support for few more games. The biggest advantage is that MAME will always work on a newer OS. Otherwise, good luck to make old emulators work on a recent version of windows moreover when they were closed source. Keeping track with several emulators versions in addition of maintaining rom sets is also tiring.
Dude, I don’t even own the game. I’m not blaming ScummVM’s devs over TWP’s performance, that’s on the original authors for not using the actual SCUMM engine.
What I’m blaming on ScummVM is their team direction, the fact they essentially ate up other FOSS projects and now their code could be rotting in there, unmaintained and getting worse with every release. (To be clear, this is just my impression)
I filed bugs, I didn’t get an answer on most of them, one of them got “fixed” but the fix doesn’t work, so why should I bother filing more?
When they did all of that merging I have no doubt their intentions were good, I just don’t like the results.
MAME and MESS are not really about making playable emulators, but about making things as close as possible to the real hardware, and not taking shortcuts for performance. That makes them useful for developping new code to run on the same platforms, and for understanding how the hardware works.
Try to develop code with the help of an emulator that takes shortcuts, and you’ll find out it won’t run the same on real hardware.
That goal is incompatible, or at least conflicting, with squeezing the most performance out of it.
Still sounds more productive than ranting about it on another project’s forum which is guaranteed to help nothing at all, doesn’t it?
You’re underestimating the therapeutic powers of ranting. ![]()
(Also, having a bug not being fixed isn’t that productive of a thing tbh)
I guess they are so used to proprietary development that they overlooked the option to use the original engine.
Don’t know if it is actually untenable. I guess you mean the complexity of maintaining multiple engines in one code?
I’m going to be talking about assumptions because I’m not code-literate enough to see the full picture so take this with a grain of salt.
Compared to other massive projects like this (MAME, RetroArch etc.), which have a main UI which then delegates the rendering for games to mostly-independent sub-engines, to my understanding ScummVM ships a main engine to which all other sub-engines hook onto. This creates a situation where 1 change in the main engine can jeopardize every untested implementation, instead of limiting damage to just 1 engine.
Again, this is what I gathered after looking at a tiny bit of code in the included engines, so I’m not confident enough to state this as a fact.
About the rest, I do fundamentally agree that ranting about a FOSS project is bad form, I do apologize for that. I am a fan of ScummVM myself, I’m just stuck in a position where the engine that could make me replay a lot of my childhood favorites is falling short on that promise for some of them, hence the frustration.
I don’t know if you are serious or trolling, but …
Although the main developer IS the one who created the Scumm engine, it’s a game made by a totally different company vs the one who has the right to use the Scumm Engine.
It’s like asking the Skyrim developers why they did not use Dragon’s Age engine.
Not trolling at all. Using the original engine or some form of it was a possibility they overlooked.
What could possibly have stopped them from partnering with ScummVM and base the game off of a GPL’d version of SCUMM? (Other than “FOSS is piracy hurr-durr”)
Even if they wanted to go with the original SCUMM and paid some form of license to LucasArts, how much would a license for an engine like that cost in this day and age? More or less than creating a new engine for it? I honestly don’t know.
Bethesda is not Bioware. And Dragon Age’s engine doesn’t have a full FOSS replacement ready for it.
Because it’s HIS game and he wanted to have full control over It maybe?
And believe me, having followed Tumbleweed Park’s development since day one: no, they did not overlook that possibility
I’m somewhat glad to hear that (for lack of a better term), thanks for clearing that up.
Full control with barely any optimization, just like an AAA studio would do it, right, right.
IIRC the engine does things that SCUMM cannot do, like actual lighting. While it does resemble the look of old games, it’s engine has a couple of more features that couldn’t be done back then. The original game runs well, anyway; I haven’t yet tried to run it in ScummVM.