David Haywood's Homepage
MAME work and other stuff
May 5, 2012 Haze Categories: General News. 113 Comments on Attack of the Street Fighter Clones

No, not clones as in alternate versions of the game, but clones as in a million clones of all the Street Fighter 2 characters attacking you for hours.

Maybe the idea of Street Fighter characters in a side-scroller isn’t a terrible one, but after the 1,000,000th time you’ve killed a palette swapped version of E.Honda it kinda drags on a bit, especially when the game offers absolutely nothing but hours of repetition.

There are good beat ’em ups, and there are bad beat ’em ups, and then there are beat ’em ups which are physically painful to play because they’re so badly designed. This game, known as “Jue Zhan Tian Huang” is definitely the latter. The entire game consists of either 4 enemies on screen at a time, all being clones of Street Fighter 2 characters, with seemingly random amounts of health, and an occasional boss fight against cyborg type enemies and a token SF2 character who will respawn until the boss is dead. There are no weapons, there appear to be 2 types of destructible objects throughout the entire game. The background are as generic as you can get, with an unhealthy liking for lift levels, it’s buggy, feels half finished in places, and will have you screaming ‘stop, stop right now’ by level 4, but luckily for you there are 8 levels broken into even more stages.

The game is also full of foreground elements which will obscure your view, some of the train levels mess with your eyes in ways which make you feel like they’re going to fall out by rapidly scrolling the foreground and background leaving you with what can only be described as a grid like pattern through which to view the action.

The hardware is Megadrive / Genesis based (the same board as Puckman Pockimon) and had the game been based off the Streets of Rage code it might have been good, but it looks like this is an original creation, or at least if there is any Streets of Rage code in there it’s so well hidden behind poor design you’re not going to notice it.

Here are some screenshots, including the ending, or at least I hope it’s the ending. There are protection checks throughout the game so I had to play until the end to make sure things were working. There’s a sound bank needs hooking up too, but the samples are so poor I’m not sure what’s right and wrong.

This could possibly displace Dragons Lair as the worst ever arcade game I’ve played …. I’m not making a video because that would involve playing it again.

Urgh... Urgh...

Urgh... Urgh...

Urgh... Urgh...

Urgh... Urgh...

Urgh... Urgh...

Urgh... Urgh...

Urgh... Urgh...

Urgh... Urgh...

Thanks (I think) to Yohji, Mr. CAST, B. Stahl, Smitdogg, The Dumping Union


You can follow any responses to this entry through the RSS 2.0 feed.

well it can’t suck as much as Guardians of the Hood ;-)

Or can it?

at least GOTH was somewhat ambitious, but ultimately failed…

Judging from the screenshots, the protagonists are KOF characters and the antagonists are SF2 characters.


Yeah, the 4 main characters are from KOF

I do wonder if it’s a hack of a hack tho, it has some typical ‘hack’ features like being able to change your character at any point with start 1, and there are some protection checks which never seem to be hit

I can’t pinpoint exactly where all the characters come from.

To it’s credit the characters will jump around with their signature moves, but that doesn’t really excuse the tedium of the whole affair.

“It’s too danger!”, LOL.
Here’s a business idea: move to Asia and offer translation services.

There are times when I really wonder… I mean, the game does not even have to be as bizarre as this one.
You get a decent gameplay, and when you finally reach the ending, those five or six short sentences are written as if the original text was machine translated back and forth ten times.

Well when even SNK is best known for ‘Engrish’ translations despite having a large budget it’s hardly surprising that no-budget titles like this are badly translated.

I’m more amazed there was an ending, I was expecting it to either loop or just present me with a ‘congratulations, game over’ screen.

Sorry for the irrelevant comment but well you had comments closed.

I think many people want to hear your take on the new MAME (and MESS?) team leadership.
Will this affect the merge positively or what?

Apparently Megumi Rescue was released on Sega System ‘E’, more info at smsmpower

I don’t think it’s going to make a major difference.

Public opinion on a merge seems to be about 50/50, depending on site / audience, I actually find that impressive considering that the majority of benefits are technical / organizational ones.

Obviously with Micko in charge of *both* MAME / MESS now some things might end up becoming closer purely because it makes life easier for him. I have a feeling by the next major release people will be working from a single source tree, but binaries will still be separate distributions.

Mametesters appears to be going through an update to allow them to take MESS bugs which has mainly come about due to the MESS server crash a couple of weeks back where they lost pretty much their entire bug database and more (hopefully the Mametesters one is actually backed up properly!)

I have asked if the UME target can be included in the official tree, even if those binaries aren’t offered on the official site. We’ll have to see about that one.

Simply put, there isn’t really anything else I can do, or say. If people really want it then they’re going to have to make a point of it on the various MW threads and such or otherwise more neutral ground.

The main opposition still seems to come from people who know it will make it more difficult to have an entirely green light in clrmame due to the size of a complete merged romset if you include softlists..

The problem is my philosophy is simple ’emulate, don’t discriminate’ but in reality everybody wants to discriminate against something. Be it adult games, gambling games, post-90s games, non-english games, or console games there will be a category of games / hardware every single person has an individual belief is ‘less worthy’ than the others and thus not worth including / emulating.

The net result of this is that getting everybody to agree on an everything solution is near impossible, which is why it’s essential to view things from a technical point of view, in which case it makes perfect sense. The game I’ve just emulated here runs on Megadrive hardware, all the code needed to run /any/ Megadrive game is in MAME, why is it turned off? Why are we shipping a ‘demo’ project?

Nice spot w/regards the System E news, I would have missed it otherwise.

Game is probably well outside the price point of anything the DU could afford tho, the remaining System E boards seem to be rather rare and fetch a lot (which is amusing, people say E stood for Economy, ie cheap shit)

Sixtoe should be notified about it at least tho.

Haze, there is one hack i try to find out.
But Smitdogg can´t find it for now, because is so many rare to get this.
Do you know the game call King Of Football 1995 from BMC?
Is the same version, but the background is full black, only have Ryu and Ken to play.
They made some fireballs instead of balls, and they play poker by using hadoukens down.
Player 1 is only Ryu and Player 2 Ken.
The game is Ryu Vs Ken. If you find some more informations, you can get them.
But if i find first i will just send to you.
At least this new hack is very good. ;)


The game doesn’t appear to be for sale. It looks like he bought it from Bill Burger last year for a reasonable price. We missed the auction. I can’t find them all. It now appears to be in a Japanese collector’s hands. If you look at his other videos he has more undumped stuff, probably isn’t mame friendly but someone who speaks Japanese could ask.

Personally I’m in favor of the merge, but I really don’t think that public opinion should be taken into account, people are too selfish to know what’s better for the project. Whether there should be democracy, it’s only among the developers, not among the the general public, most of whom have never looked at the source tree.

Oh wait I recognize that YT name. It’s a rather confusing…shop? Anyway did you see a link to it for sale?

Who are you addressing? if you’re talking about the System E thing you’d probably be best registering over at SMSPower to continue any conversations about it? Maybe those guys are interested in it with it being SMS based…

I was addressing you about the System E game. It’s in the hands of a Japan shop so I don’t think the SMS Power people can help but I did email Charles about it since he has some friends at SMSP. You said something about the price point so it sounded like you saw it for sale?

Rare Sega games are expensive, very expensive. I don’t actually have figures, but you’ve seen the amounts games like Bullet have gone for, and this is even less common.

You’ll probably be paying house prices if a Last Survivor shows up.

and yeah, for the merge, it’s good if people have their say at least.

I can’t agree with you on all points, Tosec for example was never interested in documenting REAL dumps for consoles, which is where MESS stepped up to the plate, and likewise I think the very strict inclusion requirements in some of the others actually put more at risk than they preseve. The idea of the MESS lists is closer to that of the MAME ones, document the dumps which exist, even if they’re not perfect, being able to run something is better than not being able to run it at all. Quality can improve over time.

I see Robbbert has put his ‘combined’ build out, but it’s really just another case of discrimination, in that case against anything flagged as not working, mechanical (including working mechanical games?!) and anything marked as imperfect graphics, even if sometimes the flags are just used system-wide when 95% of games in a driver are fine.

Maybe you should make a new post about the merge subject (so we don’t mess this)?

I agree with Robbbert’s “discriminating” build, but its a step. I actually proposed (in my naivety) you two cooperate.

TOSEC does have its issues I agree. It is a huge project though, so I wouldn’t “just ignore it for something else” (same discussion with Mike from C64/Amiga Forever). I would prefer FIXING IT. I am fore “synthesizing” things, not braking them down.

Well the existing lists are obviously based on things which are out there, although work would be needed to sync things up, and naming schemes are never going to agree between projects.

CD Images and prototypes are a bit area where existing lists are a let down, quite often with something like a prototype you might have one chance to rip it, from a cd-r disc, and it’s *never* going to meet the standards requirements to be in an official list. That doesn’t mean such images should be thrown away tho, which is what has happened in the past, I couldn’t find some Saturn Clockwork Knight proto thing anywhere for example.

Likewise some protos get recompiled / built from source code, the ‘pure’ lists ignore those, but they’re still very valuable resources and thus god to include in the MESS lists.

At the end of the day a decent system of flagging dumps, and their quality, and if they’re sourced from tosec images etc. will be enough for people to filter what they specifically care about. Documenting bad dumps when you already have good ones is stupid, but documenting acceptable dumps when no other is available is *essential*

Also I see the compile time argument is being thrown up again… I’m not sure it makes sense.

With Micko in charge all he needs to do is say that MAME + MESS releases of whatever version are the same sync point (which in a shared tree is what will happen anyway really)

From that point if he builds a combined build he can split out MAME/MESS binaries with 0 extra compile time, which is actually faster than building MAME, clearing out your folders, then building MESS because you’re only building the shared components (cpu cores etc.) ONCE instead of TWICE

I really don’t find the arguments others put forward to be convincing at all, it actually helps with distribution…

So why don’t you participate? Not allowed?

Hi, this is cool read. It makes a lot of sense, it’s nice to see other peoples thoughts. Keep up the great work.

I have no real desire to re-register at MW after previous treatment there.

The posts pretty much confirm what I’ve said above tho, those most vocally against it seem to think MAME should emulate their favourite games, and their favourite games only (the whole Mechanical games are a cancer line of thinking shows a severe misunderstanding of how MAME works), and the technical arguments are flimsy at best, and in some cases already outdated (MT is geared up for MESS reports right now)

You’ve got LazyCat posting nonsense based on his work on outdated Mame4All builds (I don’t think you’ll find any #if MESS code in the current sources) although no surprise because he’s just 720 controls troll from before.

You’ve got others saying compile times are the issue (things like the C++ conversion and newer tools seem to have a bigger negative impact on compile time than any of this, the MESS code isn’t that huge as you tend to find a ‘heavier use of fewer platforms’ trend compared to arcades) That also doesn’t explain why the systems for which the code is already fully included can’t just be turned on.

There are absolutely no real technical blockers on this, the target could be included with no issues at all, builds function with no known additional issues etc.

It just seems a bit bonkers, the only real reasons are historical, and most of them are going away anyway.

We will see what happens after the next release, I do plan on putting the merged builds up then under the ‘UME’ name regardless. I’m hoping the target makes it, because there is little reason it shouldn’t. The only extra workload is on me if I have to compile + distribute the build (when really it could be done at the same time)

To Jue Zhan Tian Huang,the game runs on puckpkmn hardware,so they can use the same sound chip,does puckpkmn can help hook up it?

> You’ve got others saying compile times are the issue

you do realize that before every release the maintainer has to compile baseline, debug, SDL and (when applicable) MSVC builds to be sure that everything compiles fine and only after that packaging a release?

> The only extra workload is on me if I have to compile + distribute the build (when really it could be done at the same time)

be our guest and take care of that. let’s see how many releases you keep in sync before getting bored and deciding that you have better things to do…

Etabeta here, interesting. :)
And I was just going to suggest if I should point MW thread to this thread. :)
(which I am not sure how well it would be taken by mods – you know… talking about non-existing politics now…).

Anyway I hope it all goes well and WILL go well if some (otherwise quite fine) brains get “unstuck”.

Strangely enough I know what a maintainer has to compile…..

It’s amusing, I was saying compile times got bad when we moved to c++, people were saying blah, no issue, top end machines can do it in 3 minutes.. now suddenly compile times are an issue. This is the thing with Mamedev, you flip your reasons to suit your needs all the time.

Given how many u updates Kale ended up putting out on mamedev.org which didn’t compile (resulting in relaunched builds, ‘fix compile’ checkins the very next update) It’s fairly obvious u updates only got light testing anyway (expected) and a full build is once every 2-3 months anyway, which is when you’d really need to do such a stressful compile (and the only time mamedev.org hosts binaries anyway)

Also if it wasn’t abundently clear in my previous post, sycning the release schedule results in LESS work for Micko, LESS compile times, and and easier schedule. If he has to do MAME and MESS releases anyway then the universal build is *FREE*, or more, the other builds are *FREE* if you have a universal one.

Asking me to do it results in MORE work, because it doesn’t get done as part of the normal process therefore I am having to build everything from scratch, yet I am willing to do it?


(which incidently is why your previous tirade against the GIT is stupid too, quite often after a u release the only way to get a build which worked for YOUR compile setup was to delve into the GIT and pull the latest source with post u fixes for linux/msvc/oldergcc applied, it’s been a lifesaver for many people)

Also please note that I *will* debunk nonsense arguments where I see them, because I believe in people understanding the full and clear picture

However, the _ONLY_ thing I’m asking for is for the couple of files which make up the UME target to be included once the SVNs / sources are merged (which is already underway)

I’m not personally asking Mamedev to put up the binaries, or even test the UME builds to make sure they compile / work I’ve said I’ll do that and submit any needed patches.

I do think it would be *easier* for it to be done officially, but I don’t care much about that, just about being able to say that whatever I’m posting is being built from unmodified sources.

I think it’s hilarious that someone thinks Haze has no idea what is required of the project’s coordinator even though they were around when he was the coordinator.

Haze has done great things for the project just like many others have but (correct me if I’m wrong because I just woke up and I’m still half-asleep) he was responsible for changing policy so that “u”s were released instead of waiting months to release a stable Beta and also for making sure whatever is known about a system is added as it’s discovered instead of letting people continue to “…to work on stuff monolithically for years at a time and submit it only when it was done.”

I’m not saying Haze is the only one that has done good things for the project but for some reason he seems to get picked on the most and thought I’d express some appreciation for his work, for what it’s worth. Before you call me a fanboy, I’d like to say I despise all of the gambling and Korean pr0n games he was responsible for adding. :)

Seriously though, I wish all everyone involved in the project could get along better. The politics don’t help at all. Some of the people involved need to have their egos deflated a bit but we know that won’t happen or they’ll threaten to take their ball and go home.

You are correct. One of the problems during Nicola’s later days was that everything was kept quiet, and on his HDD for months on end, you didn’t know what you had to sync to, what would be included, and you could find yourself having to resubmit things a year after the original submission.

Not his fault, it was a growing project, attracting a lot of attention with a lot of people having stuff to submit. Prior to the ‘u’ system I was maintaining what were pretty much my own internal ‘u’ builds and posting them to Mamedev, basically combined patches of everything which had been submitted since the last major release. I also took responsibility for syncing those after a release so that things which were missed did get resubmitted automatically.

So yes, I then later (given the chance to be in charge) introduced the u builds for real (and turned on all testdrivers to show it was a development focused project)

I guess it’s not THAT different to the process here. I’m maintaining these UME builds (although at this point they don’t really need maintaining, it ‘just works’) to show that the idea is a sound one, although in this case if it is to be made official like the ‘u’ builds were it will be up to somebody else, not me.

We’re all going to hate something, I hate the Korean porn games too, and I emulated half of them. You can’t always do things and make decisions based on what you like, but instead simply what makes sense from a hardware point of view. MAME is an ideal platform to emulate Korean Porn games… beyond that no arguments really matter.

AMEN brother.

BTW seems MW forum has a tradition in name calling. I just got that from Smitdogg. Interesting statistic, I registered on MW forum 40… days after him.

(maybe why I have 32 posts only is hinted somewhere here)

and as anybody trying to debate this on MW has found it’s turned from using valid points to create a constructive debate to personal insults and threats of thread locking to invalidate the views of people posting.

I don’t think it touches any person with common sense. ;)
The posts are there to read. I don’t care if they lock the thread, many points are already displayed.

No, but it’s rather sad.

If/when it happens I’m sure the very same people will be championing it.

I’m sure it’s given some people a bit of insight into the real problems anyway. I will just continue to sit here and think it’s all rather silly, especially for cases where all the code is already there or an absolute minimal amount extra is required to unlock a massive amount of functionality.

such politics also don’t hold back other projects like FBA, where they’re perfectly happy to turn on the Genesis, PCE and NeoGeo home drivers where they’ve emulated the arcade systems based on the same hardware.

MAME is becoming a project shackled by it’s own legacy while instead other projects are now showing the way forward.

MW thread about this wasn’t showing up a minute ago (just the bits about robbberts build), although yeah it’s there again now, dunno what happened there.
*edit* my fault I swapped machines, I have news board bookmarked here, and presumably emuchat on the other pc, removed confusing post saying it had gone as to not clog things up.

Interesting… seems from the thread, that there is no real reason NOT to make universal exe. Of course if I write that there (after all those “arguments”), they’ll render me blind and/or stupid. So I write it here (where again they can easily read it – and some will).

Irrelevant, can’t your blog come to 21 century (or even late 20th) and translate time stamps to my own time-zone? It’s like erm… basic and I am like 300% sure wordpress can do it.

Timezones etc. I have no idea, I’m not a web developer, I just update the site, the server guys take care of the reset. Not got anything fancy installed tho and the installation is up to date.

and universal exe, of course there are no real reasons not to make it and there haven’t been for a long time now, wouldn’t seem so silly otherwise.

pitch the reverse idea to mamedev (to split MAME into 80s, 90s and 00s builds) and they’ll fight tooth and nail to tell you why having a single build is best.

> Also please note that I *will* debunk nonsense arguments where I see them,

unless they’re yours… in that case you keep shouting until you convince your supporters that they are perfectly valid arguments.

also, when you were coordinator there was no official support for msvc nor sdl, so you had to test only half of the compiles. but I’m stating the obvious I think.

that said, I found hilarious that you still want so much to dictate times and modes of the merging when you could simply release your own universal build and be the Messiah for all your fans until we sort the remaining inertia that still slows down the creation of a merged build. who’s trying to bring up politics again and again when no politics is involved?

where is the popcorn smilley when you need one…

again, I’ve asked for one very easy thing, which shouldn’t be an issue for anybody.

you’re ignoring what I said too, things like u builds aren’t tested against those criteria, there are often compile fixes after them for the very cases you talk about so you’re talking about a couple of times each year when full releases are done. A full release takes work, that is to be expected.

nobody has put any valid arguments up against the things I’ve said, the threads descended into name calling once people were unable to do so.

it surprises me less every day that the team is so inefficient at getting things done when such small things cause such huge issues.

Personally, I’d rather he be the MAMEiah than the MESSiah. :)

sorry for the long post!

taking into accout that what etabeta says is true(i’m not a mame or mess dev, so just take my opinion as an outsider prospective), it seems the back-end structure of the mame test/bug management system is not ready to be used for mame and mess together yet..
so maybe a steady universal public build release is not good idea “at the moment”.

the crew should think about a good way to “welcome” the new merged userbase without complicate things up(bug management system, release schedule etc), first.

in addition.
i can understand that compiling the whole thing for each u release maybe a fuss(compiling mame and mess after the universal is blazing fast tho, as it’s just a metter of linking already compiled stuff),
but a once-in-a-while universal release would be still good for now, tho, good advertising, think about it.

releases apart, i think that the UME target files should be added to the main tree like.. NOW :D.
the mame/mess code is ready and ume files are harmless and it’s useful to devs willing to start working on both sides of the “wall”, with all the benefits Haze already mentioned before.
…and stubborn devs may try the new build and find it a good thing… :D

Well the MESS bug database was destroyed when their server crashed without backups, and if you look at Mametesters they’re starting to add code to support the MESS bugs there already.

Honestly I don’t know why any of this is a big issue, compared to the task of developing drivers in the first place being it’s a trivial amount of work even in the least ideal of conditions. I guess it’s another thing which doesn’t give me high hopes when it comes to the most important thing, actually improving the emulation.

for the record, I don’t remember any u* release in the past year not building on Windows when released. broken compile on Mac or Unix or PPC environments tends to happen whenever Arbee has no time to check before a release, but windows (both baseline and SDL) tends to all work

there was one intermediate release with broken -lx output, but I don’t recall from the top of my head any which did not compile.

There were some that didn’t build on Windows

worse, Kale silently replaced the packages each time.

worse still, if you did patch with one of the silently withdrawn patches then subsequent ones wouldn’t patch properly either even if the issue didn’t affect you in the first place.

That’s why Roman etc. had problems building sometimes, and ended up reporting them on MW (and even emailing me about them) because the initial source patch that had been put up was broken and untested. downloading said patch from mamedev.org in the first few hours it was up left him with a broken compile.

Cross platform (non-default target) testing is an ongoing process, as you point out, and depends on users compiling to catch any bugs during the normal course of development. It tends to be a good idea to ensure everything builds there BEFORE a major release, but again afaik that doesn’t *always* happen, your only guarantee has ever been that the standard Windows build will compile with the standard Windows tools.

With unification of the build tools between linux / windows so that at least the same GCC errors get reported less general compile issues should be introduced to cause linux problems anyway, the only time they become likely are during massive core changes, which shouldn’t happen close to a final regardless.

I feel stupid now. The whole 100+ post thread about the unified build, started by Robert, now is again pointed by Robert talks about an UNOFFICIAL merged build (much like yours).

I thought this was something that would lead in serious OFFICIAL consideration about a merge. :(

(maybe it still is… I hope)

Yes Robberts project was his own, not one I was too keen on tbh because it strips away the development focus from MAME/MESS at a time when both projects are in dire need of new developers.

His build is more akin to the type of thing you see running in illegal Mame cabs, working games only etc.

One of the more subtle benefits of having things like gambling games, and lots of non-working things in a build is that it makes life a bit harder for people wanting to build such things. As Robbert found out, simply stripping things by flags and status kills off an awful lot of stuff which works just fine for all practical purposes.

I guess his intention was that the build was a ‘best of MAME/MESS’ but in reality you can’t do that. You have systems marked as ‘working’ when it’s debatable they work at all, and systems marked as ‘not working’ because the author would rather be cautious (SegaCD for example). This applies to both MAME and MESS. In the end all you end up with is what feels like an incomplete compile.

Make sure we get a UNI build until this eventually becomes official man.

You (probably) saw that I pushed hard for QMC2 full support of your binary in their release after the next.

etabeta said:
“May 9, 2012 at 05:33

>Also please note that I *will* debunk nonsense arguments where I see them,

unless they’re yours… in that case you keep shouting until you convince your supporters that they are perfectly valid arguments.

also, when you were coordinator there was no official support for msvc nor sdl, so you had to test only half of the compiles. but I’m stating the obvious I think.

that said, I found hilarious that you still want so much to dictate times and modes of the merging when you could simply release your own universal build and be the Messiah for all your fans until we sort the remaining inertia that still slows down the creation of a merged build. who’s trying to bring up politics again and again when no politics is involved?”

Jeezus, why don’t you shut the fuck up? I keep scouring the changelogs, and still, your name isn’t there. What do you do anyway? What have you done, other than run your yap? When you say “until we sort the remaining inertia” surely you aren’t including yourself in “we”? Because you don’t appear to do a damn thing. Please go back to spell checking roms, or whatever low level function into which you’ve managed to insinuate yourself, and allow the adults to continue their discussion.

Everything you say turns out wrong. But apparently you are immune to shame and embarrassment, because you keep running off at the mouth.

Haze convinces readers with logical statements of fact. It’s pretty damn hard to shout on the internet last I checked, and Haze has never posted using all caps, so you’re wrong again.

Haze dictates nothing. He is the messiah of nothing. He intermittently writes this little blog, and does some coding. You do neither.

Haze, in his own blog, allows open discussion, including dissent. The forum where you apparently reside does neither.

Haze in this blog provides interesting details concerning MAME/MESS development. You do not.

Haze has a clear grasp of the specifics concerning the history of mame development. He was there and participated in it. Your understanding of mame history is murky, and usually wrong. I don’t think you have the age or maturity to have been there.

Haze has contributed to the preservation and understanding of thousands of games. THOUSANDS. You have not. And yet, you pretend to have a better understanding of how everything works. Now THAT is fucking hilarious.

You have no standing to even post here. But keep reading because I’m sure there will be more things for you to find “hilarious”.

Haze has a clear vision for the future of MAME/MESS development. You have NO vision for the future of MAME/MESS development.

Haze has simply come up with a good idea. It’s beyond good, it’s brilliant. It’s logical. It’s overdue. People want it.

Your idea? You have none. Oh wait, your idea is to keep things the broken way they are now. That is a bad idea. It is illogical. It is past due to be discarded. And users do not want it.

Your ill-founded, unsubstantiated, and repeatedly and thoroughly debunked assertions to the contrary, UME makes things easier for mame devs, not more difficult.

People like UME. It makes sense on every level. But you are averse to change. It serves your inflated ego somehow to maintain the status quo, and so you rightly feel threatened. Because UME is coming and there’s nothing you can do to stop it. You are the one shouting, not Haze, only problem for you being that no one is listening, because it is so readily apparent that you are a blithering idiot with nothing positive to contribute.

You represent the past. The past way of doing things is what allowed an entire MESS history of changes to vanish. How is this possible? There’s only one way that even makes sense: if it was you, etabeta, in charge of MESS backups. Then I could understand how they disappeared. Because that would require an expert incompetent.

Haters will hate, and you, chump, are a hater. You are so jealous of Haze for some reason, that the stink lines come off of you in waves…are you a love struck girl perhaps? Did Haze once spurn your affections? If so, may I suggest advertising on Craigslist. You will surely find your love there.

Ummm, that’s good vitriol!

Yikes! I still want my popcorn smilley!

Could you scale back the personal attacks please.

Eta does contribute code to MAME (search Fabio Priuli / fabiopri ) and does even have his own blog.

He doesn’t really work on that much *emulation* code however, as in emulating new systems and the like, which is why you might not see his name on things you care about.

He just fixed some playchoice 10 regressions in MAME by um.. merging it with the MESS implementation….

“He just fixed some playchoice 10 regressions in MAME by um.. merging it with the MESS implementation….” LOL

Etabeta isn’t all bad. He did give me a driver I wasn’t supposed to have a few years ago. :) This was back when there were still testdrivers and a certain ROM distribution site was still new.

I’m happy to learn what etabeta does. In fact I just read some of his posts elsewhere on this same subject, and his tone was much more considered and moderate, which makes it all the more confusing why he comes here with such a different attitude.

I’ll scale back my personal attacks, but they are always in response, I have never started them. Consider enabling rolleyes smileys here, then I can ridicule without writing a word. LOL

I found his blog. He says
“Because there are people which refuse to be part of the discussion lists and prefer to whine and rant in their blogs.”

Gee, let me guess who he’s talking about. You put a real burr in his saddle Haze old man.

BTW, what he’s practicing is called Psychological Projection. In layman’s terms, when you point a finger, three more are pointing back at you.

nice to see how people twist statements without knowing the full of it.

for one, my blog post was replying to an old post by Haze which is gone with his old website and was objectively aggressive.

for another, the code I imported from MESS for the PC10 bug was written by myself 2 years ago to emulate more accurately MMC3 mapper than the original implementation, which had been borrowed from MAME in early 00s. since anyway the MAME code worked fine with the older PPU code, I saw no reason to update it until now.
of course, you could have just checked in the mess svn log before doing any sarcastic comment, but apparently argumentations based on facts are an habit that is disappearing.

for the rest, I don’t care much about credits. the only things really worthwhile I’ve done so far, imho, are implementing proper c64 tape support in MESS (with a small, but very useful, help from Robiza), and giveing some help to judge with the softlists code.

my other big project (a CPU core for a minor system) went to la-la-land with an HD failure and since I never got anything to show, and I’m too busy at work to attempt anything about it again, there is not much to be said

feel free to keep insulting me and twisting my words, I don’t really care… people capable to go beyond the current I-shout-louder-than-you-all attitude are probably also capable to make their own opinion based on the bare *facts* and don’t need my (or anyone’s) guidance

I said the PC10 emulation was fixed by importing code from MESS. I find this to be quite rather ironic because

a) You seem to hate the idea of the merge

b) If the projects were one in the first place (as in never split) then there would have never been 2 copies of the code, ever and this long standing regression would have never existed, ever.

It would appear I’m not alone in seeing that as rather amusing.

you don’t seem to understand how this work, which I find amusing since you claimed you had taken a look at NES at some point (but maybe only the documentation or the video chip, who knows… if the latter just ignore my amusement)

if we were having merged code already, you would need (currently) to have all code for all ~210 mappers we support in MESS compiled also in MAME (with a way larger nintendo.a block of code to be linked to the main emu, and probably a larger exe, which brings back to the waste of bandwidth)

that’s the only reason why machine/nes_pcb.c is not in MAME (remember, I’ve never been against merging the cores, only against the shortsighted way you attempted to do it) and why we currently have ~5 different implementations of the MMC3 code

the latter progress on slot devices gives me hope to implement this in a clean way, but I’d prefer to avoid having 230 devices, one for each mapper, like Curt is doing with c64 cart types, when in my opinion only the carts with additional chips (e.g. Jpn Konami games with built-in additional sound hardware) really require it

anyway, nice being back to discuss emulation for once ;)

well yes, if the projects had been one then the code for all the NES mappers would be there, would have been reused where appropriate, and not been duplicated / rewritten but rather fixed, and the PC10 emulation would have just been built on the NES support directly.

not seeing the issue, at all, that scenario would have been far preferable.

the program would have been bigger, but also 10x more functional, and with less bugs due to vastly more test cases.

in a few months mapper code should be shared and all of this moot ;)

I hope in a few months EVERYTHING is shared.

Also since I cannot get in the technicalities you devs know, I can only say one thing. Talking about bandwidth issues in 2012 >NULL.

Need I say about the much better bandwidth availability?
Need I say about the much lower prices for more bandwidth?
Need I say about FREE file uploading services?
Need I say about FREE project services that also serve executables? (like sourceforge)
Need I say about torrent distribution? (already the favorite method of most -huge- linux distros?)

I mean find something serious, not bandwidth issues.
(…or the documentation!)
(…or how the users… will get confused!!!)

it’s not my duty to judge other people concerns (e.g. bandwidth).
and I have extensively explained my personal reasons.

are they not serious enough for you?
well, then be my guest and start offering your magnificent combined build like haze is doing. as long as you give out the source (so that the MAME license is respected), you are absolutely entitled to do it.

like for people which wants the removal of mechanical games (or fruit machines or mahjong games, depending on the years), I don’t see why are you waiting for us to do something you can easily do by yourself ;)

I think we should focus on the positive things here.
Etabeta is not against merging the cores and the release of a combined exe and is
sorting out the remaining inertia that still slows down the creation of a merged build.

I’m also glad to see NLS talking about the much lower prices for more bandwidth and have
good hopes he will donate to running the MAME World mirrors or set up his own mirror.

I am not the kind of person that either advertises what he may have done or will do. One thing is sure, I didn’t drop from another dimension yesterday.

Keep irony to people you know, gentlemen.

On another note,
since comments were closed for the previous blog post about different slot machine generations I have to comment here instead.

First of all, nice summary of the different generations! As you wrote it is a very UK / Euro based list.
I don’t know if you are interested in other systems or not but I just wanted to mention Williams Dotmation.
This site has alot of good information about the Dotmation slots here

What I find interesting about the Dotmation slots is they use a DMD for more than just being a slot machine. Atleast seven of the games have an interactive bonus round and thirteen of the games have a non-interactive bonus round. This sort of makes them a bit closer to regular arcade games than the usual slot machine.

Just to clarify in case there are any misconceptions…. MAMEWORLD is in no way an official MAME forum. Their space/hosting costs and issues have nothing to do with MAMEDEV or have any effect on any of these merging discussions, hosting cost or otherwise. Everything MAMEDEV does is done for free, many developers giving time, resources and yes.. money, to keep the project alive. MAMEWORLD is simply a forum which houses some boards related to MAME which also happens to have a lot of Developers as members of that forum.
The same could be said for bannister.org forums where R. Belmont’s SDLMAME support forum as well as the MESS.org forum resides.

MAMEDEV does not want control of it’s project to be in uncontrolled hands. It is really that simple. There has been history of people looking to circumvent MAME by attempting to either trademark or ‘steal’ the project right from under the creator, Nicola Salmoria and it’s hundreds of contributors over there years. Times would have to be very dire to see MAME source update hosts on free download sites, much less any binaries. Options such as Sourceforce are incompatible to MAME’s unique licensing requirements for its source and would not be allowed.

To sum up all the recent discussion.. merging of source trees is expected to happen, but don’t expect combined binaries in any official capacity. Each project is unique and while each one shares the same core code and discussion list currently, each one also deserves it’s own identity and being given the ability to stand on it’s own. Who knows what project will stand longer.. MAME or MESS? Things haven’t even begun to get interesting yet! Just, as been said in the past, just be patient and wait to be steamrolled by cool things.

Yeah as I mentioned there are others, although I’ve never seen the ones you show around here, and it doesn’t look like much of a ‘US gambling emulation’ scene has ever emerged online, so resources are rare..

.. or maybe not


These were marked ‘WMS Russia’ but I think they’re actually Williams games, maybe Russian market versions, maybe not? The titles match up with some of those on that page. I can’t speak for how complete any of the dumps are tho.

Just to clarify in case there are any misconceptions…. MAMEWORLD is in no way an official MAME forum. Their space/hosting costs and issues have nothing to do with MAMEDEV or have any effect on any of these merging discussions, hosting costs or otherwise. Everything MAMEDEV does is done for free, many developers giving time, resources and yes.. money, to keep the project alive. MAMEWORLD is simply a forum which houses some boards related to MAME which also happens to have a lot of Developers as members of that forum.
The same could be said for bannister.org forums where R. Belmont’s SDLMAME support forum as well as the MESS forum resides.

MAMEDEV does not want control of it’s project to be in uncontrolled hands. It is really that simple. There has been history of people looking to circumvent MAME by attempting to either trademark or ‘steal’ the project right from under the creator, Nicola Salmoria and it’s hundreds of contributors over there years. Times would have to be very dire to see MAME source update hosts on free download sites, much less any binaries. Options such as Sourceforce are incompatible to MAME’s unique licensing requirements for its source and would not be allowed.

To sum up all the recent discussion.. merging of source trees is expected to happen, but don’t expect combined binaries in any official capacity. Each project is unique and while each one shares the same core code and discussion list currently, each one also deserves it’s own identity and being given the ability to stand on it’s own. Who knows what project will stand longer.. MAME or MESS? Things haven’t even begun to get interesting yet! Just, as been said in the past, just be patient and wait to be steamrolled by cool things.

mbit said:
“I’m also glad to see NLS talking about the much lower prices for more bandwidth and have
good hopes he will donate to running the MAME World mirrors or set up his own mirror.”
Why should NLS or anyone donate money to mameworld? Please explain why MW simply does not use Mediafire, Rapidshare, or torrents? Those are FREE, and as reliable or, in the case of torrents, more reliable than MW.

Given how interest in MAME is at an all time low anyway I have trouble believing bandwidth is a genuine issue, and if it is I suspect most of it is being used by non-legitimate downloaders. You’ve got something like 5-10x less interest than 6 years ago, bandwidth is cheaper than 6 years ago, and while the project has grown they’d surely cancel each other out by now.

Also in a combined project you suddenly have the resources the MESS guys were using as an alternate mirror, and if you strip the artwork from the MESS package and distribute it separately (which IMHO is *essential* from a legal POV anyway and why MAME doesn’t include scanned artwork etc.) you’ve shrunk the distribution of that significantly anyway.

What has dropped is overall interest in MAME / emulation, which means MW has less sponsors, and less money coming in, that however can only be addressed in one of two ways, either increase interest in emulation, or reconsider some of the things you’re hosting in the first place, and as suggested start load balancing downloads with public services. Doing nothing won’t help, MW are already having to periodically beg for donations to keep the site alive, and as less and less people download / become interested in MAME because it isn’t doing anything new they care about you’re also going to find less and less people who are interested in donating, and eventually the plug will get pulled anyway. Is emulating consoles alone going to change that? probably not, it’s too little too late, and getting less significant and later all the time. Is *anything* going to change that? Maybe emulating 3D platforms at a decent framerate, pushing boundaries the way MAME used to, but even then it seems anything MAME does now gets a sarcastic clap because it’s taken us 10/15 years to emulate a board. The current public are for the most unappreciative and take emulation for granted. I’m not sure if that’s because they saw the rate of progress when MAME was a pioneering piece of software and current rates don’t compare, or if they missed that period and all the blood, swear and tears which went into it and only see the current devs doing almost nothing but breaking, re-fixing and generally just maintaining the work of their predecessors and just assume emulation is something easy.

Ultimately external site issues shouldn’t be holding back the project tho, the writing is on the wall for MW anyway, as it is for MAME without some major innovation and maybe re-purposing as a free professional virtualization & inline component emulation solution or similar in addition to the current stuff. The original audience no longer exist, MAME did what they wanted years ago and cannot remain relevant without moving beyond that. If all somebody wants to do is run Pacman, or Street Fighter 2 they have absolutely no reason to pay attention to development at all and no reason to support MW, or the project as a whole.

The whole truth? I don’t expect to see it, MW has a very anti-merge, pro-arcade agenda, it’s taken until the recent forum reorganization to even move from a MAME Chat to ‘EmuChat’ forum to reflect current trends and interests.

As a final note, I do also find it amusing also that when MAME did once upon a time support a ‘console’ game (the CPS Changer version of Street Fighter Zero) everybody was happy, and there was uproar when it was ripped out for ‘not being an arcade game’. Supporting it made perfect sense then, and it still makes perfect sense now, it’s just CPS1 hardware with no coin input. Again other emus see this clear as day, their solution was to just leave it there, doing no harm. MAME’s solution was to ship off to MESS, have the entire driver copy+pasted just to support 1 game, double the amount of code to maintain in the process, and leave it subject to code rot while the main driver was improved, it’s an inherently needless, and harmful policy, which is my main issue with it, the way NES and PC10 use different code despite being the same thing is just another example of that. At least now the CPS-Changer driver has been remerged even if it’s still turned off in MAME (common sense prevails in part)

I know some would say I’m trolling over this, but I don’t think I am. Ultimately maybe none of my ideas *will* help, but simply maintaining the status quo isn’t helping either and I’d rather see the project go down with a fight than a whimper.

What an interesting and thought provoking thread and as a ‘normal’ MAME user (since version 0.1) I thought I would pipe up and comment on what was discussed here. Please bear in mind I am not a developer or technical in any way although I have compiled from source a couple of times :D

Merging MESS/MAME: I think the issue I see is one of ‘contamination’, making a universal binary will slipstream development and remove redundancy, but it will (forgive the pun) make a mess of MAME. If I understand what is being discussed here, the binary will be a MESS/MAME hybrid that will basically do both at runtime. How about this… (forgive me if it is not possible)

I have seen somewhere in the source a ‘target=686′ option. Would it not be possible to merge the source and have a target=’mame’, ‘mess’ or ‘both’ option or is this not possible? It would be the best of both worlds…

Why would it mess MAME? If you ask (some) devs MESS is actually “cleaner” than MAME (and some of the clean up job benefited MAME too).

I think what you ask is already possible. The thing is that the official (?) team won’t press that magic button that makes “both” yet. I say give all three builds for a while (MAME/MESS/UNIVERSAL)… you will see that eventually only the UNIVERSAL will be used (and downloaded). In fact this is what some people are afraid.

BTW good to know some fellow “old” user that knows MAME before it was big (I am one of those too).

You could do mame/mess/both (and in fact that’s what the UME patch I’m currently trying to get put in does)

The combined build (UME) isn’t really a ‘mess’ tho, it just offers full functionality at runtime, if you don’t want to run console / computer stuff then you don’t have to, if you do, it’s there and waiting already, ideal if you need to test / compare something.

At this point having consoles in the same binary is no more messy than having 80s,90s and 00 games in the same binary, or fighting games and puzzle games in the same binary, or anything else. MAME is a hardware emulator, and the hardware is much the same; Pacman still runs as it ever did, but if you also want to fire up a NES game without the annoying PC10 menu then (in a combined binary) you can without having to recompile a thing and it will use the exact same code as it would if executing PC10.

Thanks for the clarification. As a non-contributing long time MAME user I fail to see the resistance to this merging idea. The fact that you proposed the concept of options also lessens the credibility of the objections to the merge.

I would also assume (please correct me if I am wrong) that any patches/changes are will be tested on any system that uses that code, the same as it probably is now, but now the MESS side will be tested too.. is this correct? Of course I understand there are specific drivers that are geared to specific systems this does not apply.

I didn’t mean to suggest the code was messy, I was referring to the organisation of my HD, so in this regard as I know and understand it is not a concern, please disregard.

@NLS Good to see other long term users too. :D

Yes, patches will be tested on anything which shares code (within reason, obviously if you’re making a change to the z80 you can’t test EVERY possible game using a z80 even now, that’s what Mametesters reports are for)

Actually patches under a combined-by-default system would end up better tested across projects as you’d have more test cases available at your fingertips without having to recompile, and if you were making large changes to components you’d also be given a greater level of awareness as to what else is using that component in both projects, as opposed to possibly working on something in MAME or MESS and not realising the other project had systems depending on it.

Even when it comes to organization of ROMs at the end of the day ROMs are ROMs, for system roms you’re not going to find MESS increasing the size of your MAME folder much even if you merge it (*everything* is 336.80MB, a drop in the ocean, some MAME updates add twice that in one swoop)

For Software Lists (which MESS uses heavily) you have things clearly sorted into folder by system anyway, and if you don’t care about the software for a certain system you can simply delete the list and never have to think about it.

Another plus for the merge would be the avoidance of redundancy. Currently I guess you could have two coders implementing similar code fixes for the same issue in two emulators.

May I ask. What is the argument against merging? (I mean argument in the reason sense, not disagreement sense). If you don’t want to go into it here or if there is a link to a post, would you be so kind to share it. From what I have read there is a lot of resistance.

MAME has changed so much since the original mission statement, it cannot be around that.

The valid arguments against it are.. um ..

I’m not sure..

MAME is an arcade emulator?

Really several reasons have been thrown up, most of them which can easily be discredited or are seriously exagerating things. The effort need to organize it is one, but honestly if it was done overnight the dev teams would accept it and get on with it, most of the work to merge the teams and the SVNs is already done at this point. There seem to be interpersonal issues between teams, others saying it would create user confusion, but again people would pick things up, and if anything support tools and tutorials would improve. Stability has been put forward, but really is another non-issue, I’ve been using such builds for ages with no extra stability problems, they’re as reliable as any other build and only prone to the same issues. I’m still yet to be presented with a convincing technical argument to say why it would be a bad idea.

Non-technical reasons such as bandwidth have been cited, but again it’s not really convincing, the project is going to grow, it has been growing for years, had this been done years ago you wouldn’t think anything of it at all, and (if you exclude the artwork) it’s only a small bump anyway, plus as I’ve said you have a whole extra server where MESS once was as a mirror.

From a public point of view the opposition seems to be from people who don’t think consoles are worthy, but it tends to be the same subset of people who don’t think various other things (mahjong etc.) are worthy either, and generally such views just show a personal bias towards game types rather than an objective one which would be creating the best quality of emulation. Also some people who insist on having every ROM are unhappy because of the potentially gargantum size of the Mess software lists, but again, if you don’t want them you CAN ignore them… beyond that the public aren’t likely to notice any real difference apart from some additional sets for the systems they can filter out in their frontend if they don’t care about.

Again from a public point of view some believe it will take the focus away from arcade emulation in MAME, by giving no incentive to improve arcade emulation because you can always rely on MESS / console changes to bulk up the changelogs, but again this holds little weight, devs will still work on what they’re interested in and it can only benefit later arcade systems where the hardware was mostly the same as home systems anyway because it was cheaper (there are Arcades based on every major console from the 16-bit era up)

Legal reasons? MESS has a stronger legal ground to exist than MAME anyway, it has the potential to develop into a commercial grade virtualization suite with many real world applications, or develop into an educational tool; I once had a lecturer asking if MAME could replace an aging VB application they use to teach assembly by acting as a real-time editor, debugger, memory / register viewer etc. Obviously in it’s current state that isn’t possible, but it’s another possible application of such a project further down the line which could improve userbase and exposure greatly if it ends up being used as a learning tool.

Conflicts of interest? There are MAMEdevs who have worked on commercial console games, although such personal issues should not come into play, they’re a hindrance to the project, and many of those who have worked on commercial games are MESS devs anyway, so I don’t buy that one, we’re talking about old techs anyway.

As you can see, I’m drawing blanks for the real reason, and legacy ones such as ‘MAME is an Arcade Emulator’ don’t really hold weight.

Note, even now you’ve got people not even knowing MESS exists, and missing out on a huge amount of what the drivers in MAME can do as a result
That isn’t helping anybody, an emulator which people don’t know about is an emulator which isn’t used, is an emulator which doesn’t get tested properly, and ultimately is an emulator which doesn’t improve.

I think the main reason a combined binary will never be the default is because I’m the one suggesting it. Honestly, as much as that sounds like a silly reason there are people on the team who hate the very suggestion because its one I’ve made, and they see it as granting me some kind of control over the direction of the project when they’d probably rather see me dead.

Going back to the game in the post, the background in the second screenshot is from “The Punisher”. Some of the other backgrounds look familiar, but I can’t place them. The ending reminds me of some KOF ending (’99?), but I’m not so sure (anything post-98 I haven’t played that much).

Yeah, wouldn’t surprise me if all the backgrounds were ripped from somewhere, although at the same time they’re VERY generic.

Thank you for your honest and candid opinion on the matter.

and for anybody wondering why I don’t have any interest on re-registering at MW, just head over there now, the childish playground bullying of Paratech (greybeard) on half the things he posts has resumed.

*That* is doing as much harm to emulation as anything else, if people go there expecting to see a civil site where emulation and real issues can be discussed, but just see that while controversial, but still on-topic threads get locked. No self-respecting developer would want to be seen dead there.

What? MAMEWorld just sucks?

Well it is one of the sites where some people are… more equal than others.
Also a site where some things are plain and simply taboo.

Finally I don’t like that things that manage to break the taboo (but are still against some people’s that are “more equal than others” agenda), initially get attacked with some arguments, but if they fail, it simply turns to other means, name calling etc.

Emulation is much larger than MAMEWorld though, so I guess things will eventually self correct and take their course.

Haze it would help if your blog was not just a blog, but a forum. Or even a complete portal. Or join forces with others that agree and use their resources. Because you know if in the end nobody is with you (or us)… maybe indeed we are wrong.

> Thanks for the clarification. As a non-contributing long time MAME user I fail
> to see the resistance to this merging idea. The fact that you proposed the
> concept of options also lessens the credibility of the objections to the merge.

you might want to notice that the option is already there, if you compile by yourself the executable from the source code, and the objections are only about distributing officially an exe combining the two projects together

Unless somebody has actually checked in the UME target the option IS NOT THERE.

Otherwise I could say the Raiden 2 driver works fine, as long as you fix the protection yourself…

Anyway, you can go back to hopelessly arguing that MESS is somehow better than any of the standalones too if you want. Personally I find the combined build handy for convenience, not because the MESS emulation is better than the standalones at present. The fact that you’re trying to pass the NES driver off as ‘as good as standalones’ when it falls over on some very common games which have been emulated 15+ years elsewhere says a lot.

Given the convenience of the console drivers I’ll use them in MAME / UME, if the software I need to run works fine, great, a lot does and in most cases MESS is *good enough* and it saves me some searching. If I’m having to search out another emulator anyway because the convenience isn’t there then MESS is unlikely to be my first choice, ever. I suspect I am not alone in this.

huh!?! where did I stated that NES in MESS is as good as Nestopia or any other standalone emu? the simple fact that Mapper 5 cannot be supported with current PPU code, or that half of Mapper 64 games rely on analogous features that don’t work well with the current PPU, put MESS one step below compared to the other emus (even if t’s now only one step below instead of 3 or 4 as it was a few years ago…)
and I never made a secret of these deficiencies, nor I suggested MESS over standalones for NES…

on MW I have just stated that Akumajou Densetsu should work except for the missing additional sound, and I have stated that probably the guy complaining did not even checked the emulation he’s talking about (we are talking about a guy which complained he cannot find anywhere MESS forums and got offended when pointed to a link clearly available on the MESS homepage…), neither of which sounds remotely what you’re saying…

I’m really puzzled, but feel free to go back to hopelessly arguing that I’ve said the opposite: a lot of people will just trust what you say, even if no post of mine says what you seem to pretend…

and finally

> Unless somebody has actually checked in the UME target the option IS NOT THERE.

you know what? quite often during your tirades, you end up writing very silly things… maybe you should calm down and reflect, from time to time…

have you realized that at this stage there exists no repository which hosts both trees in sync? so where the new target shall be checked in?
currently, you need anyway to download the two trees from different places and put them together with some manual tweak. what would be the point to host the complete target in either place, when there is no insurance it would build?
the files for building it are not hoarded, and people can get them from you blog. so what is the big deal?

once the merging of the svn repositories is complete, we can re-discuss about it: you know I’m not at all against having the target supported, or I would have never pointed you to Micko’s simplified approach and you would still be there to suggest to manually fiddle with lst files…

I’m saying that the people on MW are arguing that, maybe one of the devs (ie you) should actually step in and say that no, none of the systems in MESS *are* as good as the standalones rather than let the misinformation be spread, and making the guy claiming they’re not as good look like a troll (maybe he is, but he is right)

PCE / PCE-CD for example in MESS have ugly scaling artifacts when doing horizontal scroll on many games due to the video code, problems with some CDDA tracks starting in the wrong case, and some of the more demanding games just not working.

NES, as pointed out doesn’t support that many mappers, and appears to have various off-by-one errors.

Genesis code, by my own admission needs serious work too, it did run things *at the time* Kega couldn’t run, but I think Kega does better now, although Kega is probably full of hacks at least for 32x.

The likes of c64 people are going to judge on game compatibility, not support for obscure HDD expansions.

The non-mess site is hilarious in claiming that the Jaguar emulation should be best in field, but is just slow, it’s atrocious in every way because it was designed to run some games which need nothing more than a framebuffer.

I hate the spread of misinformation, and people are trolling the guy who is telling the truth? come on…. it’s just typical I’m afraid.

The emulations are *good enough* for the majority of cases, but nothing more, nothing worth going out my way to set up MESS as a separate emulator.

The same could really be said of MAME tho, the only time it’s the best emulator is when it’s the only emulator, or the others haven’t been updated for many, many years. Note, none of this means I hate MESS (I think it has the potential to be very, very good) but it is better placed as a bonus feature of MAME, where it’s convenient and can aid in delivering a better overall emulation, rather than pretending to be an emulator capable of standing it’s own ground otherwise nobody is really going to use it.

I think the rampant fanboyism either way annoys me most, especially when people are being trolled for stating the truth. MAME/MESS aren’t all that great, they’ve been left behind. Other emulators do have (out of the box):

better features for debugging (yes, rewind can be a debugging feature, systems being able to have their own custom debug displays for vdps and the like is a debug feature, a proper interface for layer enables is a debug feature)
easier interfaces (the old MAME cheat search was easy, and thus people could do more with it, the interface didn’t come back and while the new system is more powerful nobody can use it, see how dead the MAME cheat boards are) same for tape handling and the like which is just clumsy in mess due to lack of graphical interface.
better compatibility
better speed
less artificial constraints (not caring if things are home / arcade)

and that’s without even going into territory which would be unsuitable for a pure emulator like MAME/MESS (artificial gfx enhancements etc. plugin systems) so there is easily room for improvement in both without entering ‘silly’ territory.

the real, and only real advantage of MAME (or UME) is convenience, both user and developer. Everything is right there, in an active project, regardless of platform, it’s a one stop solution which WILL work for the majority of cases, is easy to compile, provides a common codebase and will save you some time looking elsewhere if you don’t care about all the other bells and whistles. Beyond that, it’s no longer 2004, MAME is no longer king of the hill, and MESS never has been.

the UME target could quite easily have been checked into MESS some time ago, maybe it isn’t the latest current MAME code all the time, but there hasn’t really been a moment when it doesn’t build (that’s been my development environment for the last 2 months or so)…. What you wrote indicated that it had already been checked in because you said people have that option, they don’t, at least not unless they dig through my pages for the target.

I’m spreading no misinformations at all and I had already stepped up at the beginning to say which systems MESS is really good at:


and I repeat: for systems like classic Macintosh, or Atari Lynx and for A2600, MESS is as good as any standalone (and quite often it is better)
for the c64, MESS code is much better than you claim (you know we use, with permission, the VIC emulation from frodo? this means that except for some games with timing issues, the result is rock solid)
are these too exotic for you? well MESS does a definitely a good job for the SMS too, thanks to recent fixes by Enik & judge.

and elsewhere I’ve seen that the MESS SID isn’t very good, which is one thing people associate with C64 more than anything else….

simple fact is tho in that thread you’d got somebody being ridiculed for saying MESS isn’t as good when he’s exactly 100% hands down correct and absolutely nobody is stepping in to say that the regulars are wrong.

by all means keep allowing such to be spread, but people will just end up thinking the team are liars, rather than keeping their expectations realistic when trying the emulation.

is MESS good enough for many, many cases? Yes. Is MESS better than any of the best standalones? No.

That extends on many levels, not just compatibility, but many of the other things I’ve listed in the previous post.

So, UME target is now in MESS SVN.
Let’s go back to coding. ;)

and so you see where all Haze’s conspiracy theories go…

what conspiracy theories, or are we back to the name calling and character assassinations?

I’m not the one locking say legitimate discussions on if Mamedev has any real developers left, while letting people be ripped apart and ridiculed in every post they make…

etabeta when you say MESS emulation on Mac, 2600 and C64 emulation is better than their respective stand-alone emulators goes from unlikely (Mac – Basilisk 2), to funny (2600 – Stella project), to insulting (C64 – VICE… esp. for this I have so much to say following VICE and Frodo from their ORIGINAL incarnation, that in fact I won’t bother saying anything).

At least you didn’t say that MESS is close to WinUAE in Amiga emulation (it is not even close to original outdated UAE).

Personally I love MAME, since I haven’t followed much other coin-op emulators anyway and after all MAME is MAME, it what showed the way. I also love MESS but indeed lives like the weirdo cousin in the shadow of the big boy. Merging would give it fresh air.

In general (and I am with Haze in this) we cannot ignore the fact that WE ARE in 2012 and MAME has to really start looking 21st century (we are well in the second decade of it).

For Mac, I repeat it: MESS emulates the hardware to a level that Basilisk can barely dream of. check this thread

I particularly like this post ;)

for a2600, I repeat what I wrote on MW: unless you are terribly interested in latest homebrew multicarts, MESS is basically at the same level as Stella. Actually some improvements in Stella a few years ago were based on the work done in MESS by judge

for c64, where did I say that MESS is better than VICE? I never claimed that (even if Arbee pointed out that the CPM expansion cart is emulated in MESS only).
OTOH, I have said many times, and I will keep saying it, that any user that has not tried c64 in the past year, should give it another try, because:
– c64 emulation in current MESS is better than what people usually expects,
– the overall emulation is pretty solid (with SID issues, admittedly, but nobody said the opposite)
– most games with issues are affected by subtle timing problems (e.g. disk games failing to load due to fastload tricks), not by flaws in the design
Given the c64 driver has been substantially broken between 2001 and 2006, with non-functional tapes, disks and video emulation, and given that a lot of users got their disappointment towards MESS exactly after testing the c64, I think it is important to stress that now the emulation is pretty good, even if not at the level of VICE

People are going to care more about baseline compatibility, performance and easy of use than running an internet browser. Maybe MAC is pretty good, maybe I’d know more if I’d had it in my working tree long-term to dabble with a bit?

I just think some people with want to oversell both projects, or don’t realise just how far the competition has come in some cases rather than trying to sell them / it based on the actual strengths, ie convenience, and being ‘good enough’ in most cases the former I feel ia b eing harmed by having the project split up, the latter people aren’t noticing due to the former.

As I’ve said, UME in the tree will do me fine for my own personal use, although people should really be considering it as the default target eventually because it just makes more sense than the other builds, it plays to the strengths of the projects.

Keep compiling UME for us man!

Anyway, yes, syncing to current adds in the UME target from the official MESS tree, that will save some time, thanks.

I was synced to 15139 for the work I was doing there, target was added in 15179.

It only went in earlier today, so the statement is now correct, users can compile the UME build from an official MESS tree with no additional code from me. (simply do a make TARGET=ume when compiling)

Thanks Micko, wouldn’t have minded an ‘its in’ email, but that’s an aside :-)

I think the point is that with UME in the MESS tree with the eventual SVN merging, all this discussion will be pretty much a moot point as people will have the full choice to choose any presentation of binary they wish. Isn’t that what everyone wanted in the first place?

Pretty much, as long as people remain open minded about it and aren’t going to treat it as some Taboo build / subject.

Official binaries of all of them would be nice, but as I’ve said, I’m willing to provide the UME packages myself.

It will be interesting to see what the default target in a shared SVN is too, technically UME would make most sense (doesn’t matter if you want to work on MAME or MESS stuff that way, you don’t have to change anything) but I guess it depends more on what people consider the primary project?

I’ve mentioned before – regarding the long term primary project – time will tell which project has more appeal long term. While a good number of people may assume MAME will always win out I would not want to put down large amounts of money on it outlasting the MESS project. Even with MAME’s increased target range in recent years, MESS targets basically everything and anything else that is CPU driven. Many people don’t know, for example, that MESS emulates scientific calculators and chess board emulations with artwork, mainframes, workstations and numerous “personal organizers” and other consumer appliances .. all making valid targets for MESS and giving the project lots of content and future promise.

Well yes, MESS has a more secure long-term future, I’ve said that before, it’s another reason I think it would make more sense as a single project, Arcade games are really only another subset of the type of thing MESS does anyway.

MAME will die due to limited scope eventually, the only real debate is over if you want to keep the established MAME name, or hope that MESS gets some brand recognition. I wouldn’t bet on the latter, which is why I’ve been proposing merging things into MAME before it loses all the remaining momentum it does have.

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.