johnparker007

Yeah great work @Laylow with all these Dutch layout releases, much appreciated And I thought I recognised that name, I used to be quite VP-focused many years back!

Have done some fixes/improvements to Indy scrape/manual layout, getting there now Reels are slightly misaligned in this (and also the MPU4 machines I did with MAME), not urgent, but shouldn't be too bad to sort out when I get chance to work on it. Indy still occasionally has a 'Error 7.3', it didn't in this vid, but that'll need fixing MAME-side.

Moved onto looking at JPMs thru MAME rescaled the Vogue cab, it's apparently right if my measurements are to be believed, though I thought it would be slightly taller than the old Barcrest MPU4s... Converted Indy, working with various issues: reels wrong (this was expected as I need to do some work there), alpha display reversed (linked to MAME data layout automation work yet to do), main issue is an emulation-related one, get an error after machine has been running for a bit - this'll need fixing in MAME driver. Also, no input mapping yet, as I need to do work to get this new tech mapping across (so playing on the weird keys with MAME window focused).

Have added support for scraping 7 segment displays from the MAME data layouts (These data layouts are still being created manually for the moment). This machine actually has 'hardwired' decimal points (rather than ROM-controlled), which are always on... not sure this is supported in MAME yet, but I'm holding off on that front for now, as there's so much work to do on MAME source, and I need to keep focused on the simulator

For now the plan is to continue to use MFME for all layout creation, and then convert over (The current converter is super fragile though, can't even move the mouse while it's converting or everything goes wrong). If new techs become available in MAME (that aren't in MFME/are buggy in MFME), MFME can still be used to create the layout, then convert across for MAME to handle the emulation. This was how @A:E and I created the first playable door closed DOND DX many moons ago - created layout in MFME then ran under MAME via MFME2MAME (legacy) converter. Of course it's very much the other way round for a good while yet (MFME is ahead of MAME on all techs, to varying degrees, by tech)...

Done more on the MAME MPU4 stuff - inserting coins works now (as they are just simple buttons for these techs) - more work on the 'dual converter' to build the machine as either MAME or MFME - manually (for the moment) made more MAME data layouts Haven't done Vfd scraping yet, so the second machine in the vid (Crackpot) has blank number displays for now. Also a few characters wrong in the alpha display (my old friend the pound sign characters again) Quick vid:

Made some progress on the refactoring of input code (to support MAME/MFME input sending). Here I have the standard mpu4 button inputs being sent to MAME (via keyboard or mouse input) Doesn't support coins yet, going to need further work to suss out the best way to do that...

I use to sequence back in the day (Noisetracker, protracker, Octamed), was good times If you sail the high seas there's this: Amiga Mods Anthology - 4cds A massive 4 CDROM-Set, 18,000+ files (Protracker, S3M, XM, Amiga Synth...) Many, many, artist represented (with their OWN permission, info-files, pics and previously unreleased mods!) from the origins of the scene (Amiga days) on up to early 1996 featuring most of the best PC-Multichannel musicians. This CD is readable under all major platforms (MS-DOS, Win, Mac-os, UNIX, Amiga-DOS...) Hope that is of use - would recommend most stuff Romeo Knight did, one of my fave scene musicians from back then Edit: also found this collection of torrents, covering a 29gb base collection plus update torrents: http://tracker.modarchive.org/

I believe he's just taking a step back for a while...

I haven't tested this yet, but as @Reg mentioned earlier in the thread, I can scan for MFME's entries in the Windows registry - so ideally the arcade sim program won't need to run as Administrator, I think I can get those at normal privileges. For MFME, the 'turbo start' feature will require a permission I believe - but it will work without that permission (without turbo-starting) if a user chooses to deny permission to the add-on .exe that makes all that work. Not due to the nature of the code, but Windows now requires .exe's to be 'signed'... so if it proves popular, donations could be used to pay for a signing certificate to remove that issue. While we're solidly in pre-alpha territory, I'm looking to keep costs low, so that popup will be an annoyance on the first run of arcade sim running MFME machines. On an alternative strategy to 'unthrottling' for fast startup; For MPU4 under MAME I've had success using 'save states'. If they don't work correctly for other techs (haven't tried) I may be able to fix them up in the MAME source. These mean that even newer techs, that are slow to start up, could be up and running in less than a second All stuff for further down the line...

(Sorry, correction in next post)

As an aside on the MFME/MAME situation: The plan is to have this all seamless, so if someone has MFME installed (as I'd imagine everyone on here has), arcade sim will find it, and the MFME machines will be visible in the library, also arcades including MFME machines that have been shared will be visible. For users who do not have MFME installed, the MFME machines/arcades will be hidden (as they will not work correctly without MFME). I hope this is a good balance for the FME community. I am also thinking about auto-added wall-hung picture 'props', and for standard (non-MFME) arcades that have a few fruit machines; they would contain images directing people to Desert Island Fruits to learn more about FME/get more fruities, perhaps even launch a browser with the URL when selected. I'm looking to strike a balance between 'working perfectly out of the box' for new users, and signposting new fruit machine enthusiasts to this wonderful FME scene. I feel the project had a wobble recently with the MFME/MAME situation, and I hope this clears us a healthy path forward that everyone is happy with... I welcome any further suggestions on this stuff

Thanks bud The multiplayer machines will be supported, but only as MFME machines, as MAME cannot do them yet (see next post for more info).

Another quick tech update vid: I've expanding the Lua scripting... and success So when I press/release Space Bar in arcade sim, it is changing the state of the start button on/off in the MAME instance in the background. All manually hacked together for now, but seems to be working as good as the MFME keypress sending system as a proof-of-concept I believe this should support multiple keys (holding down cancel to slow down skill stops then hitting etc)

A little tech update; on looking at the best way to get input into MAME whilst it's hidden in the background, with Arcade Sim fullscreen in the foreground (the same as how it works with MFME now). Here I'm experimenting with Lua scripting (the other usual techniques for input faking to an unfocussed window don't work well/at all for various reasons). So this quick vid shows me sending pause, then resume commands, while MAME is 'in the background', and the arcade sim has the foreground focus. My hope is that I can expand on this, to set the input switch matrix states as buttons from the arcade sim side are pressed/released

MAME segment alpha display now scraping correctly, including decimal points Next I'll look at both passing keyboard input direct to process (so it can be hidden behind the fullscreen arcade), and then the remapping (I'm playing this in MAME with a really weird set of keys; Ctrl, Alt, Space, Shift, C, 5, 1 etc). I remember having a go at MFME -> MAME automated key remapping before, but I think there was some problem - though I honestly can't remember what it was, it was like 5 years ago at least! Maybe I'll suss it out this time...

There's stuff happening in the VP community around VPE (I think as a spin off from the earlier proof of concept Unit3d). Once that's got to some level of maturity, I'd planned to experiment with it to see how I might integrate it wrt. texture/model streaming etc. Probably a long way off though, but that'll a way to achieve it (as both projects are using Unity game engine already)

Some progress on the new MAME segment display scraper - it's now scraping some (about half) the characters correctly... hopefully be able to iron out the remaining characters and other issues over the coming week

Getting there with the alpha segment display; split out characters/brightness on the MAME data layout, correctly working brightness. Next need to write stuff to decode this to my alpha display 'font' (which should descrambled the characters rendered in arcade sim)...

In the spirit of avoiding repairing the dishwasher and doing other real life responsibilities, I have the first reels working (24 symbol reels) Really should fix the dishwasher now, though I feel the segment alpha display calling me as the last main visual thing...

Meant to be doing real life stuff today, but managed to squeeze in some initial work on MAME lamp scraping/data layouts Some tech info: MAME ultimately won't need data layouts as I have the source, so can expose the raw data (which will reduce 1-2 frames of input lag which is great)... but since MFME works on data layouts, I've decided to do this as an interim approach, in order to get things moving Also I'm playing this via a few keys directly into MAME itself, there's a whole thing to figure out for getting hopefully some automated remapping, so then the keys/clickable buttons will work the same as the MFME ones. DIP switches/jackpot keys/percentage etc are on MAME defaults for now, again this all needs sussing out.

That's a great summary @depdawg and that's pretty much where I am @Mavroz I think that side will be ok, it was discussed recently on here, already Virtual Machines are really good (i.e. performant, whilst they are on the same architecture). Also, with what Apple are doing with x86 -> arm emulation, I think while it may be slightly more fiddly, it will be possible to run MFME 20.1 in the far future, probably indefinitely. After the strong reactions after my recent post about dropping MFME as the core FME engine ...I will be keeping MFME support and my thoughts around this were pretty much what you've written; I think it does make sense to ultimately move to a unifying open source tech such as MAME, but until that is there and at least as good as MFME the best interim solution would be to have MFME as an optional bolt on. As Reg suggested if MFME is detected then the Arcade Sim could add the supported MFME machines for selection during the arcade creation? I guess the downside of not having MFME bundled would be that online arcades that have MFME machines in it wouldn't be accessible to the users that don't have MFME installed, but I'm sure the Sim could easily recognise what arcades or machines are valid for the user. So some kind of autodetecting system, but new users would not see any of the MFME-requiring machines until MAME is at that level (or they install MFME). It could even be on a machine by machine basis, so the Arcade Sim can download fresh binaries of MAME as they are released to its internal directory along with the updated compatibility list from the Arcade Sim server, then slowly enable currently MFME machines to the user, as they become emulated at a good enough level in MAME for use in the simulator.

Thanks bud, much appreciated - the project will be continuing, just shifting focus to MAME for primary FME... Checking the sheet, the project currently has £193.50 in donations, and has incurred £113.58 in costs, full details here: https://docs.google.com/spreadsheets/d/1pVJB5VBvI1eTtB1-7DKGJ67j6RSHk-dBaM9irmZShBI/edit#gid=0

I could perhaps do something along those lines, so it detects they already have MFME installed, so they can manually get those machines working in some kind of local mode, outside of the core MAME system. I'll put the unoptimised sim back together enough (was part way through optimisations, so lots of things disabled), and start working on getting MAME working next - probably just a scraper/data layout at first, similar format to MFME, as that'll be quickest to get up and running. Later, I'll do a custom MAME build to pass video/lamp data (and perhaps audio stream) direct to Arcade Sim. Edit: have sent A:E a message to see, I hope he still stays a member