|
Combined build
#284492 - 05/01/12 01:11 AM
|
|
|
What is the level of interest in a combined mess & mame build?
Should it have everything, or just the working games?
Should it have pinball?
Should it have poker machines?
I've made a combined messui with working games only, no mechanical, no pinball, no poker machines, total 7466 systems & games. I can't release it until u8 appears, but i'd like comments in the meantime.
Thx 
|
|
|
| CiroConsentino |
|
Frontend freak!
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 6211
|
|
Loc: Alien from Terra Prime... and Brazil
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Robbbert]
#284495 - 05/01/12 01:26 AM
|
|
|
|
|
|
|
|
|
| redk9258 |
|
Regular
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 3968
|
|
Loc: Troy, Illinois USA
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Robbbert]
#284497 - 05/01/12 01:36 AM
|
|
|
How hard is it to do? I wouldn't mind doing a TINY MAMEUI build with about 40 arcade games and maybe a couple of game consoles added (Atari 2600, NES, SNES, Gameboy, etc.).
|
|
|
|
|
> not a good idea. I prefer to have MAME separated from MESS.
Those builds will continue to exist... for now.
|
|
|
|
Re: Combined build
[Re: redk9258]
#284502 - 05/01/12 01:48 AM
|
|
|
> How hard is it to do? I wouldn't mind doing a TINY MAMEUI build with about 40 arcade
> games and maybe a couple of game consoles added (Atari 2600, NES, SNES, Gameboy,
> etc.).
I used a script to create the 2 files.
In your case you only have a few systems so it could be done by hand.
|
|
|
|
Re: Combined build
[Re: Robbbert]
#284512 - 05/01/12 03:27 AM
|
|
|
the main problem is it's going to be a bug festival.
|
|
|
| B2K24 |
|
MAME @ 15 kHz Sony Trinitron CRT user
|
|
|
|
|
Reged: 10/25/10
|
|
Posts: 2663
|
|
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: kevenz]
#284526 - 05/01/12 05:45 AM
|
|
|
> the main problem is it's going to be a bug festival.
That and no one will totally agree upon what games should and shouldn't be included. I hope you got a flame suit 
|
|
|
|
Re: Combined build
[Re: Robbbert]
#284591 - 05/01/12 07:20 PM
|
|
|
Very interested. This is something that should continue to be refined and worked with over time, even if there are a few initial teething issues.
|
---
Try checking the MAME manual at http://docs.mamedev.org
|
|
|
Re: Combined build
[Re: B2K24]
#284592 - 05/01/12 08:02 PM
|
|
|
If anyone here starts flaming over someone making a custom tiny build I'll delete them and hand out temp bans.
|
|
|
| Cable |
|
retro gamer
|
|
|
|
|
Reged: 08/30/08
|
|
Posts: 131
|
|
Loc: UK
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Robbbert]
#284606 - 05/01/12 11:53 PM
|
|
|
Its a good idea for something a bit different. To be honest i really do hope mame will always be released seperately from mess but i would have no problem with a joint release. The choice would be welcome
|
|
|
| Naoki |
|
|
|
|
|
|
Reged: 11/10/09
|
|
Posts: 1998
|
|
Loc: United Kingdom
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Robbbert]
#284612 - 05/02/12 12:15 AM
|
|
|
I might be potentially interested too in the concept.. 
|
----
On a quest for Digital 573 and Dancing Stage EuroMix 2
By gods I've found it!
|
|
|
Re: Combined build
[Re: Robbbert]
#284857 - 05/04/12 04:49 AM
|
|
|
I prefer MAME to definitely be separate from MESS. Maybe years down the road once MESS' emulation improves a great deal, I wouldn't mind. But until then, I would like to keep them separate.
But I'm sure some other people would like it. So if you want to make a combination build, you should go for it.
|
|
|
| CptGuapo |
|
Beat'em-ups Lover
|
|
|
|
|
Reged: 03/18/08
|
|
Posts: 342
|
|
Loc: Off to Never Never Land
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Robbbert]
#284859 - 05/04/12 05:22 AM
|
|
|
I prefer separated too. Arcades and alike in a side, console and computers in another one.
|
"Mythology is what grownups believe, folklore is what they tell children and religion is both."
|
|
|
Re: Combined build (yes, and thanks).
[Re: Robbbert]
#284865 - 05/04/12 06:38 AM
|
|
|
> What is the level of interest in a combined mess & mame build?
Thanks for approaching this issue seriously, Robbbert. A combined build is such a good idea for so many reasons, it's kindof ridiculous. The future will be the judge of that though, of course.
I think it's a good idea to include everything (pinball, poker machines etc.) in one build partially because pursuing the alternative is really just delaying the inevitable. It reminds me of the days when gambling machines were added, then considered taboo and removed, only to be re-added later. Recall the logic in the decision to re-add and apply here.
|
|
|
| Sune |
|
Connected
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 5648
|
|
Loc: Lagoa Santa, Brasil
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Robbbert]
#284867 - 05/04/12 07:31 AM
|
|
|
> What is the level of interest in a combined mess & mame build?
>
> Should it have everything, or just the working games?
Everything of course!
I know Haze have been providing a way to make a combined build for a while and that most of the devs seem to be against merging MAME and MESS. Now that the MESS project maintainer is also the MAME maintainer I guess this could change.
I'm not for or against merging, but that's because I don't really understand what the issue is or what the full implications are. So far, personally I don't see a negative side but I guess that's because I am not a developer.
History will show that whatever is decided tends to work out in a way that's best for the project, at least as far as people on the outside can see.
Whatever happens I will continue to do what I do, which is absolutely nothing, besides sitting back, watching cool things happen. Oh yeah I did place a MAME logo sticker on my bass drum.
S
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Sune]
#284899 - 05/04/12 05:03 PM
|
|
|
> I know Haze have been providing a way to make a combined build for a while and that
> most of the devs seem to be against merging MAME and MESS.
My brain hurts. The projects ARE merged, for all intents and purposes. There is one list and one leader. However, merging the projects has nothing to do with shipping a combined build. It's nice to have as a prerequisite, but one doesn't require the other.
Given that a majority of users (and devs) don't want the combined build, and that it would greatly increase bandwidth costs for MW and our other mirrors, it would be dumb to do it.
|
|
|
| AeroCityMayor |
|
Entity formerly known as alien_mame
|
|
|
|
|
Reged: 11/11/03
|
|
Posts: 2438
|
|
Loc: Wakefield, West Yorks, UK.
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: R. Belmont]
#284960 - 05/05/12 12:14 AM
|
|
|
>
> Given that a majority of users (and devs) don't want the combined build, and that it
> would greatly increase bandwidth costs for MW and our other mirrors, it would be dumb
> to do it.
|
Coherance of posts inversely proportional to Foster's consumption!!!
|
|
|
Re: Combined build
[Re: R. Belmont]
#284963 - 05/05/12 12:32 AM
|
|
|
> Given that a majority of users (and devs) don't want the combined build, and that it
> would greatly increase bandwidth costs for MW and our other mirrors, it would be dumb
> to do it.
As far as users go, it's split about 50/50 if you are using this thread as a gauge (which is only about 10 people anyway - hardly representative). And I'm willing to bet not all the folks who are voicing opposition fully understand the benefits of a combined build - when confronted with an unqualified choice, it's human nature to go with whatever you're comfortable with and resist change. I do it all the time.
Just a thought, but I think this thread could benefit from a dev's brief summary of the pros/cons from an objective viewpoint. Perhaps different results would be obtained - and probably a more statistically significant number of them.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Shoegazr]
#285230 - 05/07/12 12:42 AM
|
|
|
Since you guys seem to be... counting (like every MAME and/or MESS user is in here), count me in for merged build.
Actually I'd even want merged build to be the DEFAULT build, but anyway I wouldn't mind if we actually get three builds and don't even advertise the (combined) third. Let us that want it, have it.
Also sorry but I read some not very smart comments, possibly coming from ignorance?
For example:
- What about the question about what to include? Yes MESS allows for software lists but this is not part of the actual emulation (I for one accept TOSEC as the only real list and MESS should just use that). A potential "ROM testing" (like from clrmamepro) involving MESS or combined builds shouldn't be bothered about the software lists, only the emulated machines (for MESS part) and emulated games (for MAME part). Now I understand that in some second stage in the future, a coin-op game based on Megadrive hardware COULD be classified as such and be included in a megadrive list. But we don't have to worry about this yet.
- As a general philosophy, aside from MAME and MESS everything should be emulated that is filtered out by (questionable) moral rules or whatever. Let lists separate things, let it have a nice parental control or something embedded, but the system should support as much as it can.
- More bugs? Sure you know what you are talking about? Actually LESS bugs. The source tree is already more or less one. Why separate the bugs and even the teams? Something resolved (or found) in one side, is quite commonly same for the other side.
- Someone else says he will wait for MESS emulation to progress more first. Erm... do you realize that MESS already exists in MAMEs bones an the opposite? Nobody expects for MESS to replace WinUAE for example, but the 68000 emulation MAME and MESS use is the same (one and only). So what are you talking about?
- Someone else prefers to separate because of an arbitrary difference. Arcade on one side he says. Well many newer arcades are actually well known computers or consoles inside. So what. Is the "insert coin" that bothers you? Why not have a separate MAME for each game classification too? Or a different MESS for 8 bits, for 16 bits an for 32 bits? Come on, all these things can be FILTERED by a nice front-end (already).
---
I think some people don't want the merge because of political reasons. Politics should be OUT of an open source project. Am I wrong?
Other people don't want the merge because they are not really aware of the internal workings of those two (?) projects (that are actually like 95% common already). And those other people that I said above (the politics) don't help these people here actually understand.
I think both obstacles are temporary and the merge inevitable. So get things going already.
As I said, give separate builds for those that want to stick to separate builds for their own reasons, but allow for an OFFICIAL merged build (that if you ask me eventually will take over and we ALL know that).
Thank you all for your time. I didn't intend to insult anybody, sorry if I sounded harsh.
|
---
NLS
|
|
| Sune |
|
Connected
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 5648
|
|
Loc: Lagoa Santa, Brasil
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: NLS]
#285258 - 05/07/12 04:21 AM
|
|
|
> Let lists separate
> things, let it have a nice parental control or something embedded
It's up to you what you want to run in MAME and MESS. It's not the responsibility of anyone else but you. Emulation is one thing, content is another.
If your child discovers the password for the adult channels, it's not the cable company's problem.
> I think some people don't want the merge because of political reasons. Politics
> should be OUT of an open source project.
And so should "parental controls".
S
|
|
|
| Sune |
|
Connected
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 5648
|
|
Loc: Lagoa Santa, Brasil
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: R. Belmont]
#285263 - 05/07/12 04:27 AM
|
|
|
> > I know Haze have been providing a way to make a combined build for a while and that
> > most of the devs seem to be against merging MAME and MESS.
>
> My brain hurts. The projects ARE merged, for all intents and purposes.
I KNOW that ffs. By "merging" I meant the combined build.
Show me how to build MAME from MESS svn on my Mac so I don't have to download diffs and patch with every u release.
S
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Sune]
#285298 - 05/07/12 11:14 AM
|
|
|
We actually agree, you miss my point.
Some parental control could/should be in a front-end.
I don't care for it actually, I am just saying this for all people that want "separations".
Instead of separating things (that are not actually separated), FILTER things.
I would love for MAME to be transposed to... MAME (only this time actually also be MESS in there) and if people want what we call "coin-ops" in some places, to FILTER those through the front-end (or the embedded front-end or even command line).
With today CPUs and RAM (even for... mobile devices), the whole separation which is based on a technicality is moot. In the end, keep providing the unmerged builds if someone so wants. It's like building a specific build with only a subset of engines compiled in it.
In any case everything points to merge eventually.
> > Let lists separate
> > things, let it have a nice parental control or something embedded
>
> It's up to you what you want to run in MAME and MESS. It's not the responsibility of
> anyone else but you. Emulation is one thing, content is another.
>
> If your child discovers the password for the adult channels, it's not the cable
> company's problem.
>
> > I think some people don't want the merge because of political reasons. Politics
> > should be OUT of an open source project.
>
> And so should "parental controls".
>
> S
|
---
NLS
|
|
|
Re: Combined build
[Re: Sune]
#285299 - 05/07/12 11:28 AM
|
|
|
> Show me how to build MAME from MESS svn on my Mac so I don't have to download diffs
> and patch with every u release.
>
> S
get Haze's UME folder containing a ume.mak and ume.lst files including just the MAME and MESS ones.
at every release, just patch the two trees and launch "make TARGET=universal" or ume or whatever Haze liked to call
|
|
|
|
Re: Combined build
[Re: Robbbert]
#285302 - 05/07/12 12:42 PM
|
|
|
I believe you could have the best of both worlds. MESS and MAME can be united, yet become even more separated than before.
Why not have C64 or NES be just like any other hardware platform in MAME, that is ROM images you load in MAME just like Pac-Man and Donkey Kong, with their drivers designed in similar manner as any other PCB in MAME, as "driver", not interwoven within and obfuscating MAME code.
Would not that be more natural continuation of the already existing design in MAME? Why not treat C64 "PCB" just like Pacman PCB? Would not this make for cleaner and simpler code, and make both sides equally happy?
|
|
|
|
Re: Combined build
[Re: LazyCat]
#285303 - 05/07/12 01:31 PM
|
|
|
> Would not that be more natural continuation of the already existing design in MAME?
> Why not treat C64 "PCB" just like Pacman PCB? Would not this make for cleaner and
> simpler code, and make both sides equally happy?
you do know that what you says is *exactly* what MESS already does?
maybe you shall check what you're talking about, before posting 
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: etabeta]
#285306 - 05/07/12 01:49 PM
|
|
|
Indeed but without knowing it probably, he also points what is clear: Those projects are actually one.
|
|
|
|
Re: Combined build
[Re: LazyCat]
#285311 - 05/07/12 02:17 PM
|
|
|
|
|
|
Re: Combined build
[Re: NLS]
#285313 - 05/07/12 02:20 PM
|
|
|
both projects are built over the same core, they share the same mailing list of developers, and the developers prefer to ship two separate executable and sources for end-users which don't compile by themselves.
how is that a problem, when people capable of compiling a combined exe can already do it?
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: etabeta]
#285315 - 05/07/12 02:39 PM
|
|
|
> both projects are built over the same core, they share the same mailing list of
> developers, and the developers prefer to ship two separate executable and sources for
> end-users which don't compile by themselves.
>
> how is that a problem, when people capable of compiling a combined exe can already do
> it?
I am not sure how you talk about the whole team. Maybe you CAN talk about the whole team, I don't know. I think there are different voices though.
The problem is EXACTLY for people NOT capable of compiling a combined exe; which btw traditionally face some discrimination in some open source projects. In those projects (and I am not saying MAME is one of those) it is very common to face an almost racist comment "if you want it like that, do it yourself"... well having been on both sides in different phases of my IT life, I don't like that reply.
The problem also is that the way to build the combined exe is not (YET) official. Why? The best (? AFAIK) effort right now comes from a member that (again AFAIK) doesn't have the sympathy of some of the members of the team (for reasons we frankly do not care). This to an outsider (like me) seems like politics and just politics.
Anyway I think I have made my point (thank you for giving me the space to do so): Combined exe is the way to go. At least let it be an (official) option. I am 100% sure that in the end it is going to be the default.
|
|
|
|
Re: Combined build
[Re: etabeta]
#285319 - 05/07/12 02:53 PM
|
|
|
>> you do know that what you says is *exactly* what MESS already does?
maybe you shall check what you're talking about, before posting
<<
What is it then people who want it "separated" are complaining about?
True, I didn't see much of MESS code, except those terrible "#IFDEF MESS" patches all over MAME source code, that's all I base my assumptions on, glad to stand corrected, but I think that is exactly kind of stuff that should be separated and what people are really complaining about.
>> how is that a problem, when people capable of compiling a combined exe can already do it?
<<
I guess that is the question that needs answering first, what exactly is the problem? I thought "separatists" are complaining about stuff like "#IFDEF MESS" directives mingling with MAME code. If not that, I don't see what else, it's just a matter of compiler options then, as you say, & making it "official" or "more standardized" as NLS says.
|
|
|
|
Re: Combined build
[Re: NLS]
#285321 - 05/07/12 03:12 PM
|
|
|
> I am not sure how you talk about the whole team. Maybe you CAN talk about the whole
> team, I don't know. I think there are different voices though.
I'm just reporting the opinion of the majority of the devs as it emerges from the opinion given on the dev list. no more, no less. since we generally proceed by democratic discussions, that becomes the current position of the team.
> The problem is EXACTLY for people NOT capable of compiling a combined exe; which btw
> traditionally face some discrimination in some open source projects. In those
> projects (and I am not saying MAME is one of those) it is very common to face an
> almost racist comment "if you want it like that, do it yourself"... well having been
> on both sides in different phases of my IT life, I don't like that reply.
>
> The problem also is that the way to build the combined exe is not (YET) official.
> Why? The best (? AFAIK) effort right now comes from a member that (again AFAIK)
> doesn't have the sympathy of some of the members of the team (for reasons we frankly
> do not care). This to an outsider (like me) seems like politics and just politics.
no offense, but reasons have been explained many times and if you still think there is politics involved, maybe you shall check these forums as often as you check Haze's blog.
for one, at the moment MAMETesters is not ready to take care of MESS bugs too, so a merging of the exes would harm our best bug tracking resource; for another, current MESS svn is public, while the MAME one is not and there are bandwidth concerns about making the latter public (closing the former is not an option, of course)...
the situation might change in future, but I think for the moment we can ship separate exes with no harm
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: etabeta]
#285322 - 05/07/12 03:16 PM
|
|
|
OK, allow me to stand by my original views (of an outsider but emulator fan - I don't claim being something different).
BTW I've heard that actually MAME bug tracking system IS moving towards accepting also MESS reports (also because of a recent crash in MESS's bug tracking system?). Is that info wrong? I don't know, just asking.
|
|
|
|
Re: Combined build
[Re: R. Belmont]
#285324 - 05/07/12 03:26 PM
|
|
|
>>
Given that a majority of users (and devs) don't want the combined build, and that it would greatly increase bandwidth costs for MW and our other mirrors, it would be dumb to do it.
<<
If the size is your concern, than main official build should contain no bootlegs, gambling, and maybe even no neo-geo games, just optional at compile time.
Actually, if it was up to me I'd get rid of almost everything... the artwork, bezels, cheats, screenshots, scanlines & HLSLSLHL effects, and anything else that is not ABSOLUTELY necessary just to run the games as the original PCB would. Anything doing more than that I would put in some kind of add-ons, just OPTIONAL to this core MAME.
People working on, say... artwork, will not then be able to screw up already working game drivers. I'd say these things should really be separated from the main core build much more than MESS. MESS should be treated like NEO-GEO.
|
|
|
| mesk |
|
@ the arcade
|
|
|
|
|
Reged: 03/03/11
|
|
Posts: 484
|
|
Loc: Rhode Island
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: LazyCat]
#285327 - 05/07/12 03:52 PM
|
|
|
Remove the Neo Geo games? All the artwork as well?Blasphemy! Glad you are not in charge 
|
|
|
|
Re: Combined build
[Re: NLS]
#285328 - 05/07/12 04:01 PM
|
|
|
> OK, allow me to stand by my original views (of an outsider but emulator fan - I don't
> claim being something different).
>
> BTW I've heard that actually MAME bug tracking system IS moving towards accepting
> also MESS reports (also because of a recent crash in MESS's bug tracking system?). Is
> that info wrong? I don't know, just asking.
it's correct and true. and that's one of the reason why I've said that things might change in the future, just not right now
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Shoegazr]
#285333 - 05/07/12 04:23 PM
|
|
|
> As far as users go, it's split about 50/50 if you are using this thread as a gauge
> (which is only about 10 people anyway - hardly representative).
I do have other inputs, from sites I can't mention here IYKWIMAITYD.
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: etabeta]
#285334 - 05/07/12 04:32 PM
|
|
|
> I'm just reporting the opinion of the majority of the devs as it emerges from the
> opinion given on the dev list. no more, no less. since we generally proceed by
> democratic discussions, that becomes the current position of the team.
Specifically, devs often incur full builds on a several-times-a-day basis. I have a 4.7 GHz i7 with 8 GB of RAM and that still takes a chunk of time that I'd rather be using to, you know, improve the emulation of the systems. (Never mind that I'd have to go to 16GB to comfortably link a combined project).
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: R. Belmont]
#285335 - 05/07/12 04:33 PM
|
|
|
Is this "statistic" private between you two?
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: R. Belmont]
#285337 - 05/07/12 04:35 PM
|
|
|
8GB extra is like 40 euro?
Just saying.
Also, I've been far from development for some years (I am just a glorified user and clueless IT manager right now), but I am not sure why (I am not saying you are wrong at all! Don't misunderstand me) someone has to make FULL builds every time to debug a specific system. If a module is irrelevant, why it has to be compiled in the day-to-day debug builds? AFAIK MAME does have several compile switches right?
Edited by NLS (05/07/12 04:39 PM)
|
|
|
| CptGuapo |
|
Beat'em-ups Lover
|
|
|
|
|
|
Reged: 03/18/08
|
|
Posts: 342
|
|
Loc: Off to Never Never Land
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: mesk]
#285340 - 05/07/12 04:41 PM
|
|
|
Quote:
Remove the Neo Geo games? All the artwork as well?Blasphemy! Glad you are not in charge
LOL! I second that. LazyCat, please, take a cup of coffee, calm down and don't exaggerate. If you insist in that Neo Geo games exclusion idea, you'll suffer... 
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: NLS]
#285342 - 05/07/12 04:42 PM
|
|
|
Look, I'm saying I'm all about choice. We have that right now: if you want all-in-one, you can build all-in-one. It's simply not the default, supported, option. (Supported in the MAMETesters sense; MAMEdev have in fact gone out of our way to make sure the unified build option does work, because we love you and we want you to have the choice).
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: CptGuapo]
#285343 - 05/07/12 04:42 PM
|
|
|
> Remove the Neo Geo games? All the artwork as well?Blasphemy! Glad you are not in
> charge
>
> LOL! I second that. LazyCat, please, take a cup of coffee, calm down and don't
> exaggerate. If you insist in that Neo Geo games exclusion idea, you'll suffer...
LazyCat is almost certainly Abracadabra, if you remember that dude (Haze agrees with this analysis, for those of you who listen to him rather than me ;-) His posts make far more sense through that filter.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: R. Belmont]
#285348 - 05/07/12 04:58 PM
|
|
|
Fair enough, still not a solution for the "common user" (for whom something that is not there in the main site with an .exe extension, doesn't exist at all). It's like you politely told me "well if you like it make it yourself" (and I discussed about that kind of talk above).
I already said it shouldn't be the default (...yet - because in the end I find it inevitable that it WILL be the default, give it a year), fine, but it should be there in the main site, as one of the official builds (and possibly a separate MESS-only build too, but not a separate site and team any more). It's not only that unified would be "better" (in so many levels), it is also that MESS doesn't get any of the "limelight" of MAME project and is a pity - being "embedded" would strengthen it and its "spirit" would live forever through MAME.
Another thing, I am not really sure why the site couldn't handle such a thing (I am saying this because "web load" reason has been discussed above too), since the separate servers used already for the now separate builds could share the load.
Anyway. Thanks again for discussing this whole thing (which I feel was like a taboo some months ago?).
|
|
|
| mesk |
|
@ the arcade
|
|
|
|
|
|
Reged: 03/03/11
|
|
Posts: 484
|
|
Loc: Rhode Island
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: NLS]
#285349 - 05/07/12 05:06 PM
|
|
|
I have seen a few people talk about all these benefits to a combined build,NLS,shoegazer.Interestingly enough nobody has actually said what these benefits would be,just that the amount of benefits is ridiculous.
Personally I do not know enough to really care one way or the other @ the moment.The only thing I have read is that a combined build would increase bandwidth usage and that is never a good thing for a site that runs on donations
Anyways I am curious as to what these benefits are,and just how they would be a advantage to the average end user.To those opposed to a combined build,same question to you. 
|
|
|
| CTOJAH |
|
MAME Addict
|
|
|
|
|
Reged: 07/13/10
|
|
Posts: 980
|
|
Loc: Macedonia,Veles
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Robbbert]
#285351 - 05/07/12 05:07 PM
|
|
|
Putting MAME and MESS in the same basket is against the nature laws; You know : Survival of the fittest - MAME will eat MESS for breakfast, but then again, the very next moment MAME will die by poisoning from MESS rotten ingredients.
So, at the end of Your experiment (mixing oil & water), as a final product we will get 2 corpses instead of 1 health grown up man (MAME) and 1 messy but promising child (MESS).
Obviously I am against this "so called" unity because marriage(unity) without love would not last long.
That is what I think on this subject as a emulation fan, no hard feelings.
P.S. Mechanical games are cancer in MAMEs health body and I hope that relevant people will realize/see that disease before it's too late. Anyway, it's another subject. (but anyhow related to this thread)
|
|
|
|
Re: Combined build
[Re: mesk]
#285353 - 05/07/12 05:08 PM
|
|
|
> Remove the Neo Geo games? All the artwork as well?Blasphemy! Glad you are not in
> charge
I understand your reason, let me explain mine.
I think MAME core should be handled like the development of Linux kernel - modular, or "Object Oriented" if you will, where adding support for artwork would a simple matter like adding network support to some Linux distribution.
One core, many distributions and flavors for everyone's taste. The main point of this design and development scheme is that everyone should be able to easily make (compile) their own distribution (MAME build) by simply enabling or disabling these add-on features at compile time, only I would separate code completely instead of use #IFDEF or anything like that. What you really need in any case is a just a good compiler front-end which should be no harder to use than say MAME .ini file, that's all.
The most import benefit of such modular design is independence of one piece of code from another and self-contained functionality, which makes for many little "Objects" that can be re-designed, re-written and optimized without having impact to other parts of the code. This means that particular people with particular skills would be able to concentrate on specific things without unrelated stuff obscuring the code or standing in a way for better implementation. Smaller independent chunks of code, like game drivers, are much easier to re-write, optimize and maybe most importantly *test* , as you can do it in small steps and independently of everything else. Each module itself, including core MAME, would become easier to improve and changes to any of these separate modules (add-on features) could not possibly have any impact on how actual game drivers or other modules work.
Does that sounds better?
|
|
|
|
Re: Combined build
[Re: LazyCat]
#285354 - 05/07/12 05:12 PM
|
|
|
Guess what? It already is. Drivers can be EASILY removed before compilation. There's no point in separating drivers out into, say, DLL files post-compilation. RAM usage wouldn't change, CPU usage wouldn't change.
|
|
|
|
Re: Combined build
[Re: R. Belmont]
#285359 - 05/07/12 05:23 PM
|
|
|
>>
LazyCat is almost certainly Abracadabra, if you remember that dude (Haze agrees with this analysis, for those of you who listen to him rather than me ;-) His posts make far more sense through that filter.
<<
I said kind of the same thing as you did, we are on the same side here, only I personally would be much more extreme than you, that's all, and I think it's for good reason which I explained above. Do we really disagree?
It's about options and choices, and it should be, the only thing that is missing is just some voting system, so majority can decide. Right?
|
|
|
|
Re: Ninja Assassin MAME devs
[Re: LazyCat]
#285360 - 05/07/12 05:26 PM
|
|
|
A brief reminder that MAME is a not-necessarily-benevolent dictatorship run by a cabal of highly trained ninja assassin arcade game-playing fanatics.
Remember all those stories about how playing videogames will make you violent and murderous? The MAME devs are what you get at the end of that road.
Do. Not. Cross. Them.
|
|
|
|
Re: Combined build
[Re: CTOJAH]
#285361 - 05/07/12 05:29 PM
|
|
|
> P.S. Mechanical games are cancer in MAMEs health body and I hope that relevant people
> will realize/see that disease before it's too late. Anyway, it's another subject.
> (but anyhow related to this thread)
just ignore those sets. nobody put a gun at your head to force you into launching them.
|
|
|
|
Re: Combined build
[Re: Firehawke]
#285363 - 05/07/12 05:37 PM
|
|
|
> Guess what? It already is. Drivers can be EASILY removed before compilation. There's
> no point in separating drivers out into, say, DLL files post-compilation. RAM usage
> wouldn't change, CPU usage wouldn't change.
I was not talking about separating drivers, they are already separated. I was talking about scanlines, artwork, bezels, cheats, save states... anything that is not necessary to reproduce what actual PCB does, it's called "emulation".
HLSLSHL effects, bezels and stuff like that is not what PCBs do, it's what monitors made of glass do and what cabinets made of wood do, that's not "emulation", it's "simulation".
You have to decide is core MAME about documenting electronic s (PCBs) and emulation, or is it about simulation and documenting monitor quirks and cabinet decoration. I think core MAME shold be only about emulation and not about simulation, which is welcome addition, but is sufficient to be optional rather then default for all the reasons I mentioned in the post above.
Also MESS is not separated from MAME completely as it could be, like Pacman driver is. All these things I say could be separated from MAME core just like game drivers are.
Edited by LazyCat (05/07/12 05:45 PM)
|
|
|
|
Re: Combined build
[Re: LazyCat]
#285367 - 05/07/12 05:46 PM
|
|
|
No, emulation is simply more than just the board for arcade machines.
It's as accurately representing the actual game state as much as possible. Thus why things such as MAMEHooker are supported, such as they are, by the MAME devs.
Some games actually REQUIRE the artwork to be functional, as well-- Spy Hunter has five separate lit pieces of artwork detailing car status, for instance.
On top of that, some games may have been designed to take into account glass curvature or other effects of the beam, making HLSL the choice for getting more accurate output.
|
|
|
| CTOJAH |
|
MAME Addict
|
|
|
|
|
|
Reged: 07/13/10
|
|
Posts: 980
|
|
Loc: Macedonia,Veles
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: etabeta]
#285368 - 05/07/12 05:49 PM
|
|
|
>
> just ignore those sets. nobody put a gun at your head to force you into launching
> them.
I can do that, but how would You react if You know (hypothetical) that there is a cancer in Your body ?
1. Ignore it...
...or
2 Remove it.
I would go with 2.
Just my two cents.
|
|
|
|
Re: Combined build
[Re: CTOJAH]
#285370 - 05/07/12 05:55 PM
|
|
|
You take this way too seriously.
|
|
|
| CTOJAH |
|
MAME Addict
|
|
|
|
|
|
Reged: 07/13/10
|
|
Posts: 980
|
|
Loc: Macedonia,Veles
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Firehawke]
#285372 - 05/07/12 06:00 PM
|
|
|
MAME is a serious project !
OK, I will calm down - maybe should not take so radical example, but this is only a discussion about ours beloved emulator and I mentioned before :
CTOJAH said : No hard feelings.
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: LazyCat]
#285375 - 05/07/12 06:07 PM
|
|
|
> I think MAME core should be handled like the development of Linux kernel - modular,
> or "Object Oriented" if you will, where adding support for artwork would a simple
> matter like adding network support to some Linux distribution.
If you look at what MAMEdev's actually shipping, you'll see that we've been converting the entire codebase from fake OO with macros to real OO in real C++ over the last 2 years. MESS is actually farther ahead on this than MAME, but the core is ready in either case. This has enabled a whole slew of good things to happen, including the combined build under discussion and much better separation between chip emulations and drivers.
And just like the Linux kernel, we reserve the right to change the API and ABI any time it makes sense to, lest anyone get any dumb ideas about subjecting users to drivers-as-DLLs.
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: CTOJAH]
#285376 - 05/07/12 06:09 PM
|
|
|
> MAME is a serious project !
No, it really isn't. Or at least I don't take it anywhere near that seriously. As per my devquote, emulating old games is not the cure for cancer, so let's not treat them as equivalent
|
|
|
| DrArcade |
|
AKA Mr. CAST
|
|
|
|
|
Reged: 03/27/12
|
|
Posts: 136
|
|
Loc: Latin America
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: CTOJAH]
#285377 - 05/07/12 06:10 PM
|
|
|
> >
> > just ignore those sets. nobody put a gun at your head to force you into launching
> > them.
>
> I can do that, but how would You react if You know (hypothetical) that there is a
> cancer in Your body ?
> 1. Ignore it...
> ...or
> 2 Remove it.
> I would go with 2.
> Just my two cents.
LOL, Mechanical games in MAME = Cancer. Damn near spit coffee all over my monitor.
Where do I submit documention for RLG's (Real Life Games)? Ring Toss, Horse Shoes, Badminton.. I want to watch the cancer metastasize, in the true sense of preservation. 
Mr. CAST
|
|
|
|
Re: Combined build
[Re: R. Belmont]
#285378 - 05/07/12 06:11 PM
|
|
|
Agreed, but preserving old games is better than preserving old movies in some cases IMSHO.
I.E., depends on the game being preserved vs the movie being preserved.
|
|
|
|
Re: Combined build
[Re: R. Belmont]
#285385 - 05/07/12 07:13 PM
|
|
|
Funny you should mention the drivers-as-DLLs thing. I actually had considered bringing it up as a particularly bad idea, but decided not to mention it for the sake of not giving anyone horrific ideas.
|
|
|
| mesk |
|
@ the arcade
|
|
|
|
|
|
Reged: 03/03/11
|
|
Posts: 484
|
|
Loc: Rhode Island
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: CTOJAH]
#285387 - 05/07/12 07:20 PM
|
|
|
> P.S. Mechanical games are cancer in MAMEs health body and I hope that relevant people
> will realize/see that disease before it's too late. Anyway, it's another subject.
> (but anyhow related to this thread)
Do they really bother you that much? Personally I really dont even notice them,granted I use Hyperspin the majority of the time and those games do not appear in my wheel since they were simply removed from the HS XML,I also use QMC2 I guess I just dont notice them. 
|
|
|
| CTOJAH |
|
MAME Addict
|
|
|
|
|
|
Reged: 07/13/10
|
|
Posts: 980
|
|
Loc: Macedonia,Veles
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: mesk]
#285391 - 05/07/12 07:25 PM
|
|
|
>
> Do they really bother you that much? Personally I really dont even notice
> them,granted I use Hyperspin the majority of the time and those games do not appear
> in my wheel since they were simply removed from the HS XML,I also use QMC2 I guess I
> just dont notice them.
You already answered to Yourself ! (look at the underlined text)
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: CTOJAH]
#285392 - 05/07/12 07:27 PM
|
|
|
Man you are way too harsh to progress. Maybe this is why you like retro?
Cancer, rotting, dead bodies... come on.
You are a very radical "against" person, without realistic arguments though (at least for 2012). No offence.
BTW all these actually make me believe we are actually moving towards that merge. :P
|
|
|
|
Re: Combined build
[Re: mesk]
#285393 - 05/07/12 07:31 PM
|
|
|
I'm neither, but bandwidth is expensive and this site is having financial issues, so I don't see a need to make bad things worse.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: greybeard]
#285394 - 05/07/12 07:37 PM
|
|
|
> I'm neither, but bandwidth is expensive and this site is having financial issues, so
> I don't see a need to make bad things worse.
Guys this is no real argument, come on.
In such a change, MESS site would actually share the bandwidth.
Also nobody forces nobody download from the master servers.
Most modern sites that give out some important piece of software have alternative download sources.
There could be use of file sharing services. Most (if not all) Linux distros etc. come out as torrents too.
This thing about bandwidth is a non issue.
(because there is no real issue to be found maybe? :P)
|
|
|
|
Re: Combined build
[Re: Firehawke]
#285395 - 05/07/12 07:39 PM
|
|
|
>>
It's as accurately representing the actual game state as much as possible. Thus why things such as MAMEHooker are supported, such as they are, by the MAME devs.
Some games actually REQUIRE the artwork to be functional, as well-- Spy Hunter has five separate lit pieces of artwork detailing car status, for instance.
<<
Ok, when you say "game state" you mean whole cabinet, I mean just PCB. It's because I wanna play it in a cabinet and you wanna play it on a PC.
You make a good point and yes it's nice to have it all done "digitally", it's just that I believe you should preserve "wood" by restoring wood, and restore PCB by replacing it with PC, so that PC needs to do only what original PCB did.
In any case I am not suggesting to remove these features, just separate their source code from the main core build, make it more like some "plug-in", move it from run-time to be compile-time only options in order to make it easier for the development and is overall better design for such huge & public project, I believe.
|
|
|
|
Re: Combined build
[Re: LazyCat]
#285397 - 05/07/12 07:44 PM
|
|
|
Too many cabinet types out there, and MAME is a project to preserve what it can. Believe me, if MAME didn't do artwork, a lot of it would have been lost by now.
The entire MAME concept is to make it preservable on pretty much any platform that might happen in the future. That means computers, tablets, even consoles if you're so inclined to port it (SDLMAME on PS3 Linux). Removing that functionality to modules does nothing more than to increase the complexity of the codebase for zero real gain.
You, as a user, would save probably 10k off the binary.
MAMEdev would gain about three extra bleeding ulcers maintaining it as a module when it's part of what they consider the core *meaning* of the project.
|
|
|
|
Re: Combined build
[Re: NLS]
#285398 - 05/07/12 07:46 PM
|
|
|
The compilation time IS an issue, but could be worked around in SOME cases by using tiny builds-- still, when you're ready to put out a final release, that might require several trial runs with minor tweaks as errors crop up in the final compilation of everything together. (Not to mention the RAM usage..)
While I believe that EVENTUALLY they'll probably end up in one binary together, that's likely to only happen when the impact to actually compiling the thing is less to the developers.
|
|
|
|
Re: Combined build
[Re: R. Belmont]
#285399 - 05/07/12 07:47 PM
|
|
|
>> If you look at what MAMEdev's actually shipping, you'll see that we've been converting the entire codebase from fake OO with macros to real OO in real C++ over the last 2 years. MESS is actually farther ahead on this than MAME, but the core is ready in either case. This has enabled a whole slew of good things to happen, including the combined build under discussion and much better separation between chip emulations and drivers.
<<
I see Aaron's re-write of drawgfx, it's finally free of that self recursing noodle soup, but looking a bit more into the whole re-write thing I see some kind of pointless effort to make MAME "modern" (for some future OS, as Derrick says). You know, having 17 bits precision stored in 32 bits based on funny assumption it might become "standard", one day. Kind of stuff I'd call "unnecessary bloat".
So, anyway, are you saying all the #IFDEFs are going to be gone? MESS actually separated to its own files rather than mingling within MAME core? If so, I applaud that and I'd be glad to help, just shine the sign up in the sky.
By the way, I've sent you link to my Android MAME build, you have to try "Crazy Scoreboard" and "Crazy Tilt"... super-cool, if I may say so.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Firehawke]
#285400 - 05/07/12 08:00 PM
|
|
|
> The compilation time IS an issue, but could be worked around in SOME cases by using
> tiny builds-- still, when you're ready to put out a final release, that might require
> several trial runs with minor tweaks as errors crop up in the final compilation of
> everything together. (Not to mention the RAM usage..)
>
> While I believe that EVENTUALLY they'll probably end up in one binary together,
> that's likely to only happen when the impact to actually compiling the thing is less
> to the developers.
Good point. For me to fight I mean.
So you say that the actual compile time of the final build must be... an issue that is strong enough to actually count against the universal binary?
Erm... sorry I don't see it.
PLUS, I suspect that building MAME + building MESS is actually longer than building universal. Because some do that. Remember the head of the MAME team is the head of MESS team right now... I suspect, it would be faster for him too...
|
|
|
|
Re: Combined build
[Re: NLS]
#285402 - 05/07/12 08:09 PM
|
|
|
Well, RB insinuates pretty directly in http://www.mameworld.info/ubbthreads/sho...part=1&vc=1 that the actual development compiles take significantly longer -- at least longer enough that it's cutting into actual work time without getting a major upgrade.
Remember, this project is on a volunteer basis. Things that are going to make the codemonkeys require newer hardware for a non-paid project are likely to be shelved. Especially in this economy, demanding things that are going to be serious inconveniences to the devs aren't going to be taken with less than an entire bucket of salt.
That'd probably be the biggest reason right now. It's simply too much of a pain at the moment. Nobody ever said it's going to be like this forever, mind. There may be a complete change of attitude on it in another three years.
Or putting it down to the absolute simplest equation:
*THIS* is a bad time to force such a changeover.
While I actually SUPPORT the idea of flat-out merging the two as a single release, I believe that it's a long term goal at best and not really a great idea for the moment.
|
|
|
|
Re: Combined build
[Re: R. Belmont]
#285407 - 05/07/12 09:15 PM
|
|
|
> > As far as users go, it's split about 50/50 if you are using this thread as a gauge
> > (which is only about 10 people anyway - hardly representative).
>
> I do have other inputs, from sites I can't mention here IYKWIMAITYD.
I actually don't care if the binaries will be merged or not, but I see obvious benefits in merging the source tree.
So I think I can be classified as pro merging (if you guys are measuring the public opinion).
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: LazyCat]
#285414 - 05/07/12 10:06 PM
|
|
|
Apparently you haven't seen what we've been doing at all. #ifdef MESS hasn't existed for well over a year now. There is one single object-oriented core. You really think a unified build was possible with #ifdefs?
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Ramirez]
#285415 - 05/07/12 10:09 PM
|
|
|
> I actually don't care if the binaries will be merged or not, but I see obvious
> benefits in merging the source tree.
In case I haven't been clear enough about this: the source tree is being merged. That isn't even a debate.
|
|
|
| R. Belmont |
|
Cuckoo for IGAvania
|
|
|
|
|
|
Reged: 09/21/03
|
|
Posts: 9711
|
|
Loc: ECV-197 The Orville
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: NLS]
#285416 - 05/07/12 10:12 PM
|
|
|
> So you say that the actual compile time of the final build must be... an issue that
> is strong enough to actually count against the universal binary?
Yes. One of the reasons nobody initially wanted the job after Kale resigned is because it takes many hours to do the builds for release.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Firehawke]
#285418 - 05/07/12 10:19 PM
|
|
|
> Or putting it down to the absolute simplest equation:
>
> *THIS* is a bad time to force such a changeover.
>
> While I actually SUPPORT the idea of flat-out merging the two as a single release, I
> believe that it's a long term goal at best and not really a great idea for the
> moment.
I on the other hand believe we have already come a long way to finally arrive at this goal.
Think of the state of things last year (in the two teams). Think of the available cheap hardware right now (I even said about the 40 euro an additional 8GB for the 8GB Belmont's machine - AND I AM GREEK!).
I think we are there and if we are not, we are not three years away. Six months at most.
(quote me when I fail)
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: R. Belmont]
#285420 - 05/07/12 10:26 PM
|
|
|
> > So you say that the actual compile time of the final build must be... an issue that
> > is strong enough to actually count against the universal binary?
>
> Yes. One of the reasons nobody initially wanted the job after Kale resigned is
> because it takes many hours to do the builds for release.
Of course you are the expert.
So how many hours is the build as is now?
How many hours is MESS to compile on same machine? (from same tree, so clean whatever is related to MAME)
How many hours a combined build?
Asking so that I will know.
Also asking you as the expert.
How many times a COMPLETE build is needed between releases?
Why?
Why this cannot be substituted by "sub-builds" (with many or most modules that are untouched in the release, disabled) and only do the final build once?
(and I believe that is what you guys do anyway...?)
Finally asking you as the member of the team.
What would drastically improve compilation time? Stronger CPU? A second CPU? More RAM?
RAM is a non-issue nowadays. I have three machines and 20GB RAM total in my home and I am a middle class GREEK (you know what is happening here aren't you).
In any case asking the member of the team...
If someone is responsible for the builds (or even more than one person), how hard is it to raise a fund to pay for the extra hardware needed? We are not talking for four figures here (ok maybe close). I think it is something that IS doable (IF it is so much an issue). I've seen fund raising here before for PCBs (EXPENSIVE), for domain registrations and hosting etc. So... is it REALLY an issue?
Thank you for any feedback on the above.
Edited by NLS (05/07/12 10:34 PM)
|
|
|
|
Re: Combined build
[Re: R. Belmont]
#285422 - 05/07/12 10:45 PM
|
|
|
> > I actually don't care if the binaries will be merged or not, but I see obvious
> > benefits in merging the source tree.
>
> In case I haven't been clear enough about this: the source tree is being merged. That
> isn't even a debate.
Don't worry, you was clear enough. That was me saying that I'm exited about it.
|
|
|
|
Re: Combined build
[Re: R. Belmont]
#285437 - 05/07/12 11:19 PM
|
|
|
>> Apparently you haven't seen what we've been doing at all. #ifdef MESS hasn't existed for well over a year now. There is one single object-oriented core. You really think a unified build was possible with #ifdefs?
<<
Ay, caramba! That's what I've been doing in my build for the last year too. What a terrible synchronization, my mistake.
I guess I missed out quite a bit then... errr, what "unified build" means? In summary what's the biggest change?
>> In case I haven't been clear enough about this: the source tree is being merged. That isn't even a debate.
<<
I think people are talking about two different things while thinking to be talking about the same thing.
a.) separation/merge of source code
b.) separation/merge of the binaries
Someone should summarize what each of those really mean and what kind of difference it makes to end-users and developers.
For example, I don't care about binaries at all, even simple console script (.BAT) can be made to make it all very easy to compile and build whatever binary someone might wish, with neo-geo, maybe just some individual games, with bootlegs, with MESS, mechanical, gambling, mahjong... or without.
I do care about source code separation/merge. I don't wanna see any Artwork, Pacman or MESS related stuff in core input, core timing or core video handling module.
In other words, I want to see MESS source tree merged, but I also want MESS code in it's own separate files. Are we on the same frequency here?
|
|
|
|
Re: Combined build
[Re: Robbbert]
#285441 - 05/07/12 11:24 PM
|
|
|
I think mech games should be left out as I always looked at them as a nusance.
|
|
|
|
Re: Combined build
[Re: jumpmaniac81]
#285448 - 05/07/12 11:37 PM
|
|
|
> I think mech games should be left out as I always looked at them as a nusance.
Left out from what? Source code, binary?
The point is to make compiler "front-end", which would be easy to use, as easy is to configure MAME at run-time, possibly simpler.
Linux kernel does it with a simple console script where all you have to do is just check or uncheck some check-boxes and press [compile] button. Don't wanna mech, neo-geo or MESS, uncheck the box! Simple as that.
Then you do not care about what is in the source tree, you simply make your own build with or without any feature you want or do not want. The end, eh?
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: LazyCat]
#285454 - 05/08/12 12:07 AM
|
|
|
> I think people are talking about two different things while thinking to be talking about
> the same thing.
> a.) separation/merge of source code
> b.) separation/merge of the binaries
I am talking about #b - because #a is already "almost there" AFAIK.
And since mechanical, mahjong, gambling etc. were mentioned too, I support that the official release builds have EVERYTHING on. MAME is not an ethics school (like the old ban of gambling machines), it is EMULATION software. It is a matter of others to do a possible separation...
The simplest separation is any front-end and even stock build UI and console switches. No need for separate builds, just make on/off mode switches or the functionality ALL front-ends already have, folder filters (version filters, cpu type filters, adult filters, whatever filters... so add more filters "mame", "mess", "mechanical", "mahjong").
More "specific to a task" separations a matter of 3rd party compilers. For example someone might like mahjong ONLY... Make a mahjongmame from the source and release it, who cares. Or... a mess only build. :P Or an old-school-mame-only build. :P
I don't get all this separation of the core project personally.
|
|
|
|
Re: Combined build
[Re: NLS]
#285466 - 05/08/12 12:55 AM
|
|
|
>> And since mechanical, mahjong, gambling etc. were mentioned too, I support that the official release builds have EVERYTHING on. MAME is not an ethics school (like the old ban of gambling machines), it is EMULATION software.
<<
You seem unaware of how easy it already is to compile MAME, and how much more simpler yet with more options can be made. That's what changes everything.
MAME should not even release any binaries. What are you making issue about is just a bunch of options set as default, or not. Just a bunch of check-boxes that needs to be checked or unchecked. I'd set them all to unchecked by default, so that "official build", or better to say "default build" you can make with "official compiler" is the smallest size of all.
It should be about "official compiler" and "official source code", not about "official binary". Official binary you get by using official compiler and official source code, no matter what game drivers or extra features you exclude.
|
|
|
|
Re: Combined build
[Re: NLS]
#285485 - 05/08/12 03:25 AM
|
|
|
It's entirely up to the comfort zone of the MAME development team is what I've really been getting at. They will choose to move at the pace they're comfortable with.
Honestly, you come across as somewhat badgering on this particular point. While I agree that a full build is probably a logical step forward, it's OBVIOUS to me that this is entirely outside the comfort zone of the team right now. Even more so with RB's admission that the team actually balked at picking a leader over the compilation headache.
That's more than a minor issue, that's the kind of issue that could actually *stop the MAME project cold* at this juncture.
|
|
|
| B2K24 |
|
MAME @ 15 kHz Sony Trinitron CRT user
|
|
|
|
|
|
Reged: 10/25/10
|
|
Posts: 2663
|
|
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: NLS]
#285516 - 05/08/12 07:28 AM
|
|
|
I have a great Combined build that I'm using right now!
It consists of 2 self compiled official binaries paired up with QMC2 Front End and I make use of the MAME/MESS switcher.
User is happy and MAMEDev/MESSDev is happy
|
|
|
|
Re: Combined build
[Re: NLS]
#285519 - 05/08/12 08:09 AM
|
|
|
how many question are you still going to ask?
> Why this cannot be substituted by "sub-builds" (with many or most modules that are
> untouched in the release, disabled) and only do the final build once?
> (and I believe that is what you guys do anyway...?)
because if you don't build the whole tree, you could break some dependence (e.g. of a different driver from the one you're working on, which shares some components) and doing so you could stop some other devs' work. and checking over and over which components are interconnected takes sometimes as much time as compiling the complete tree from scratch.
verifying that everything compiles fine before a commit is mandatory (things already break this way from time to time, due to e.g a single .h file missing to be updated, without adding more entropy)
also, most of the complains were about building time of the maintainer before a release: you do realize that before every release the maintainer compiles baseline, debug, SDL and (when applicable) MSVC builds to be sure that everything compiles fine and only after that packaging a release? add the times you have to fix a compile and start from scratch, and maybe you will start to understand the magnitude of time we are talking about
> If someone is responsible for the builds (or even more than one person), how hard is
> it to raise a fund to pay for the extra hardware needed?
because this is not a job. taking care of releases should not be an obligation due to people having paid for that, or it would get very close to obligations towards paying clients at work
|
|
|
|
Re: Combined build
[Re: etabeta]
#285526 - 05/08/12 09:25 AM
|
|
|
For end-user, compiling GNU C/C++ based projects is the same thing as installing software on Windows, just instead of pressing "install" button you type "make", and that's how user "installs" software. Easy.
MAME should not release any official binaries, just make sure it's easy for the end user to compile their custom build where they can exclude all those games and features they do not want.
Why make yourself a problem with releasing any binaries when you can leave the problem to end-user, which will not be a problem at all if you make it simple enough, it would be very universal solution.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: LazyCat]
#285530 - 05/08/12 10:07 AM
|
|
|
> You seem unaware of how easy it already is to compile MAME, and how much more simpler yet
> with more options can be made. That's what changes everything.
> MAME should not even release any binaries. What are you making issue about is just a bunch
> of options set as default, or not. Just a bunch of check-boxes that needs to be checked or
> unchecked. I'd set them all to unchecked by default, so that "official build", or better
> to say "default build" you can make with "official compiler" is the smallest size of all.
> It should be about "official compiler" and "official source code", not about "official
> binary". Official binary you get by using official compiler and official source code, no
> matter what game drivers or extra features you exclude.
Ah! We are now (back) to "do it yourself". So you are honestly saying that every "plain user" should setup his own compilation environment to build their own breed executable? Oh come on.
BTW, I would love for all builders out there that it is indeed a simple procedure. Seems they contradict you though, saying about weird dependencies (people that are not active in the development may not know about), huge build time etc.
Hmmm... Let all open source projects stop making executables for users whadayathink? Let executables be mystical products only for the initiated. :P
Jokes aside, I don't think your point is very valid (at least as a general policy). I hope devs will not take it seriously.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: Firehawke]
#285531 - 05/08/12 10:09 AM
|
|
|
> It's entirely up to the comfort zone of the MAME development team is what I've really
> been getting at. They will choose to move at the pace they're comfortable with.
>
> Honestly, you come across as somewhat badgering on this particular point. While I
> agree that a full build is probably a logical step forward, it's OBVIOUS to me that
> this is entirely outside the comfort zone of the team right now. Even more so with
> RB's admission that the team actually balked at picking a leader over the compilation
> headache.
>
> That's more than a minor issue, that's the kind of issue that could actually *stop
> the MAME project cold* at this juncture.
That on the other hand is a valid point (although I wouldn't go that far as to believe that a project like MAME could ever stop cold). Since team devs are not comfortable with this yet, I cannot say much more. I guess I'll wait until they are comfortable.
(which I don't think will be in three years)
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: etabeta]
#285532 - 05/08/12 10:15 AM
|
|
|
> how many question are you still going to ask?
Is there a set limit? I think you are giving out useful info.
> > Why this cannot be substituted by "sub-builds" (with many or most modules that are
> > untouched in the release, disabled) and only do the final build once?
> > (and I believe that is what you guys do anyway...?)
>
> because if you don't build the whole tree, you could break some dependence (e.g. of a
> different driver from the one you're working on, which shares some components) and
> doing so you could stop some other devs' work. and checking over and over which
> components are interconnected takes sometimes as much time as compiling the complete
> tree from scratch.
Clear. Also means that there is need to be some more restructuring to make it more oo and modular? (which AFAIK MAME takes pride that IS like that already)
> also, most of the complains were about building time of the maintainer before a
> release: you do realize that before every release the maintainer compiles baseline,
> debug, SDL and (when applicable) MSVC builds to be sure that everything compiles fine
> and only after that packaging a release? add the times you have to fix a compile and
> start from scratch, and maybe you will start to understand the magnitude of time we
> are talking about
Also clear, still I am curious to read real numbers on a typical machine.
> > If someone is responsible for the builds (or even more than one person), how hard
> is
> > it to raise a fund to pay for the extra hardware needed?
>
> because this is not a job. taking care of releases should not be an obligation due to
> people having paid for that, or it would get very close to obligations towards paying
> clients at work
Of course this is not a job. I see what you are saying. Funding some better hardware I don't think crosses the line though in obligating someone (that already IS doing THIS part).
As a general comment personally I think this thread is becoming a great read on the whole subject of merging. Maybe something to revisit sooner or later. (hopefully sooner)
|
|
|
|
Re: Combined build
[Re: NLS]
#285535 - 05/08/12 11:16 AM
|
|
|
>>
Ah! We are now (back) to "do it yourself". So you are honestly saying that every "plain user" should setup his own compilation environment to build their own breed executable? Oh come on.
<<
I know just the fact you need to type something in the console makes you not wanna do it, that's why I propose there be "compiler front-end" so it all looks exactly like some standard software installation on Windows. You simply check or un-check the options you want and click the start button.
Would you not prefer to have that freedom of choice to make your own build with whatever features you wish at any time you want? It would be no more complicated than installing say Microsoft Word, it would only last longer, I guess, but in turn you could have it optimized for your particular processor. Does that sound better?
>> Seems they contradict you though, saying about weird dependencies (people that are not active in the development may not know about), huge build time etc.
<<
All I know is that at the end there surely must not be any problems with compilation, otherwise the build is simply not ready for the release. For *END-USER* compilation must always work if using official compiler and official source code, even across various operating systems. That's not a problem, that's what MAME is already good at, just needs some more user-friendliness, that's all.
>> Hmmm... Let all open source projects stop making executables for users whadayathink?
<<
Yeah, that's exactly what most of them actually do, especially ones that strive to be cross-platform portable since that's the most universal type of software distribution there can be. Perhaps lately the trend has become to release binaries too, but that's not for the benefit of the user, it's for their laziness, and in return they get less than optimal build for their particular system and needs.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: LazyCat]
#285536 - 05/08/12 11:34 AM
|
|
|
Well interesting twist of the original thread.
Actually if I call myself a plain user (no problem to do that), I would be one of the less uncomfortable (or more comfortable, to eliminated the double negative) to do compiles myself. Check my registration date here. I was a dev (not for MAME) in my early days and after all I started when computers were 8 bit and one of my vacation hobbies (ok just during evening sleep of the rest of the family - I am not THAT a geek) was to write BASIC ON PAPER (no portables back then). Interestingly enough most didn't even need debugging back when typed on the computer. Anyway... point is that even for me, what you suggest is not a realistic solution.
Most plain users even with a friendly GUI will not even touch something that needs re-compile for every update. 99.5% of them will prefer the drop-in-place exe. Imagine if all software was like that! For people coming from Linux of course this way is more common (BUT REMEMBER BUILD ENVIRONMENT IS ALREADY PART OF THE OS SETUP) still even there, it is a dying trend. Even if it becomes as easy as installing Office (you see I make it easier than installing Windows that you say), I don't think we could convince anyone reinstall Office with the rate MAME builds come out. It is unrealistic.
Personally I would prefer a 10% or 20% slow down of betas, if official merged build was there. (everybody talks for himself of course)
Only way I would "stomach" a user-re-compile for every update would be if it was a transparent part of a very smart installer. (which btw should also support UNC paths as my whole emulation is on a huge share :P) I think this is a big project BY ITSELF. (much simpler than just providing universal-all-enabled exe I think)
|
|
|
|
Re: Combined build
[Re: NLS]
#285540 - 05/08/12 01:59 PM
|
|
|
It's all the same to me, the only practical issue here seem to be just bandwidth then. Binaries-shminaries, I thought this was about source code.
|
|
|
|
Re: Combined build
[Re: LazyCat]
#285545 - 05/08/12 04:12 PM
|
|
|
> It's all the same to me, the only practical issue here seem to be just bandwidth
> then. Binaries-shminaries, I thought this was about source code.
All I see here is you're not clearly thinking because:
a) Binaries don't bother you, you compile your own builds.
b) Because of the previous point, whatever mamedevs do with the distribution can't harm or benefit you in any way.
c) You're suggesting to remove set off by default features on a PC emulator intended to work on a PC. Because we know every user using MAME has a PC stored inside a cabinet with an outdated CRT monitor (that was sarcasm).
d) Your elitist thinking would lead to limit MAME to only Linux users owning arcade cabs and NEOGEO PCBs so you don't miss them. AFAIK that's called a custom build and I'm glad you're not in charge too. :P
e) How could be reducing bandwidth distributing the source AND the compiler environment so everyone can make their own custom builds? Let alone wait up to 3 hours (depending on your own machine) to install something smaller than Office. Most users have Windows machines and no compile environments, for every 100 PCs only 5 or less have Linux. You do the math, that's what this trend to release binaries is for.
|
|
|
|
Re: Combined build
[Re: BIOS-D]
#285553 - 05/08/12 05:26 PM
|
|
|
>> a) Binaries don't bother you, you compile your own builds.
<<
From your tone I'm guessing this is a bad thing, can you explain what is your objection really about?
>> b) Because of the previous point, whatever mamedevs do with the distribution can't harm or benefit you in any way.
<<
That's the power of the source code. I don't understand your objection again, what's bad about any of it?
>> c) You're suggesting to remove set off by default features on a PC emulator intended to work on a PC. Because we know every user using MAME has a PC stored inside a cabinet with an outdated CRT monitor (that was sarcasm).
Because of the first point, this too does not bother me. However, if you wanna pretend the purpose of MAME is to accurately document (emulate) arcade games and not for people to play them on their PCs, then perhaps you should not make it obvious you are in fact going extra mile to make it so.
>> d) Your elitist thinking would lead to limit MAME to only Linux users owning arcade cabs and NEOGEO PCBs so you don't miss them. AFAIK that's called a custom build and I'm glad you're not in charge too. :P
Ah, you just feel like arguing. I'm not that radical and there is more reason to my position than you want to give a credit for. In any case I think there should be voting system and majority should decide, even though majority can easily be wrong, but "Vox populi, Vox Dei".
>> e) How could be reducing bandwidth distributing the source AND the compiler environment so everyone can make their own custom builds? Let alone wait up to 3 hours (depending on your own machine) to install something smaller than Office. Most users have Windows machines and no compile environments, for every 100 PCs only 5 or less have Linux. You do the math, that's what this trend to release binaries is for.
<<
There are pros and cons to everything, it's just a matter of individual priorities. Source code you have to distribute in any case and combined with build environment it makes for download that is not much bigger than the binary, if bigger at all.
It's not a problem for me to find 3 hours to spare and let computer do the job while I sleep or go out to buy some drugs, and on drugs time flies. I think your problem could be that you are not using enough drugs.
|
|
|
|
Re: Combined build
[Re: LazyCat]
#285567 - 05/08/12 07:00 PM
|
|
|
> It's not a problem for me to find 3 hours to spare and let computer do the
> job while I sleep or go out to buy some drugs, and on drugs time flies. I
> think your problem could be that you are not using enough drugs.
I'm sorry if my comment sounded rude, that was not the purpose. But the way you argue and counter argue makes it almost obvious your only purpose is to argue, and you'll be doing everything your way doesn't matter what.
If you'll be doing everything as you like, then there's not purpose to argue how MAME developers do their own stuff, right? Maybe you really need to not go into drugs. I do and I think very clear.
Remember when I said less than 5 PCs for every 100 use Linux? I forgot to add those 5 PCs have different Linux distributions fragmenting the user base a hell a lot more. See how successful the way Linux thinks it is (not saying Linux is bad though).
If MAME adds not only arcade PCBs, but anything arcade related it's more than obvious their purpose is to document hardware than playing with them. So removing everything you don't like, as you wish and that doesn't work will make it less obvious for people the emulator is intended to play them on their PCs? That's a wishful thinking.
|
|
|
| NLS |
|
Computing Nostalgia Adict
|
|
|
|
|
|
Reged: 10/28/03
|
|
Posts: 76
|
|
Loc: GREECE
|
|
|
|
Send PM
|
|
|
Re: Combined build
[Re: LazyCat]
#285568 - 05/08/12 07:00 PM
|
|
|
Guys I think you moved away from the thread's topic.
The topic is about "combined build" (so it cannot even mean the source, as the source is already more or less one).
Also I think with 10-20 users participating in the thread, a vote would be stupid. Hell even if all EmuChat voted, it wouldn't mean anything. AFAIK thousands use MAME (and maybe 1/10th of them also use MESS - also I hope 1/2 of them KNOW MESS at least).
As a member of the team (?) pointed out, there are some practical issues for this to become reality yet ...that honestly I don't see and I asked more "educated in the subject" people and weirdly enough agree with me, but anyway, the thing is, the devs are the devs and until someone that is a MAME team dev steps up and says "I will handle the unified build and produce it for all major updates", then for us is not much to do. Yelling doesn't help either (I usually tend to be pretty vocal).
I am that close to make a bet that we will start having an official unified build within the next six months, but knowing that there are people that actually push against it (for reasons I am not sure are valid), I will not make that bet.
I'd love more expert opinion from "my" side (I know and you know that at least one voice is not interested to be heard here), but in any case I've done what I can.
(which is not much)
|
|
|
|
Re: Combined build
[Re: BIOS-D]
#285581 - 05/08/12 09:06 PM
|
|
|
>> I'm sorry if my comment sounded rude, that was not the purpose. But the way you argue and counter argue makes it almost obvious your only purpose is to argue, and you'll be doing everything your way doesn't matter what.
<<
I misunderstood for the first half what the discussion is about, but I was not that much off topic and I still hold the reasons behind my proposals are valid. Of course I have to try defend and justify what I proposed if challenged, that's where the logic will either fail or persist, and so far I'm still convinced to be right - according to my own views, priorities and goals.
It's easy to disagree, I prefer to find agreement, but we can not agree no matter if both our logic is valid if our priorities and goals are different. That's where majority vote should make the cut so we always have more happy than angry people, and then all angry people have to do is just take some more drugs. Drugs always solve everything.
>> If you'll be doing everything as you like, then there's not purpose to argue how MAME developers do their own stuff, right?
<<
Right. I wanted to present and explain my opinion, not really to press the issue any more than to make sure I was understood properly, which is when it turned out it is in fact me who didn't understand what others are talking about. What a day!
>> Remember when I said less than 5 PCs for every 100 use Linux? I forgot to add those 5 PCs have different Linux distributions fragmenting the user base a hell a lot more. See how successful the way Linux thinks it is (not saying Linux is bad though).
<<
Linux is irrelevant, the installation file can be "setup.exe", Windows binary file that you run, select your options and press install button, and then the setup unzips the source code and compiler tools from itself and begins compilation. Only fresh compilation lasts long, but if you plan to compile every MAME release then you should keep your source tree, apply only diff patches and the compilation will be done in few seconds.
C'mon, we have computers, we can use them to make complicated things simple for us. Drugs too.
- - - - -
As for the actual topic, I do not see a single benefit or even a reasonable motive to have one big MESS+MAME binary, I only see longer loading times and increased memory usage.
|
|
|
Index 
Threaded 
Combined build
Reply
