[Suggestion] Open:MP and MTA:SA

[Jusonex]

Hello,



before I start, let me very quickly introduce myself. I am a member of the MTA Team and actively developing MTA. This post is as well approved by the other team members.



We?ve been silently observing what?s currently happening with SA:MP and you for the past few days. We decided to wait a bit till the turmoil in the SA:MP scene cooled down a bit - that?s why we haven?t said anything earlier.



In our opinion, the current situation is a great chance for MTA:SA and SA:MP to come closer. We do understand that you?re not happy with the current situation, but we believe that starting another multiplayer project from nearly ground up seems counterproductive for an ancient game like GTA:SA. We love San Andreas - just as you do - although player counts are stable, we can?t deny it?s difficult to attract new players, especially considering the change in demographic.



Instead, we?d like to encourage you to think about joining the MTA community. Talking about technical details, we think that implementing an SA:MP compatibility layer for MTA is totally possible and an optimal long-term solution. On the other hand it will also get you faster to an initial release (which we assume is one of your greatest problems).

To mention some stats: During the past three years, we released four new versions in addition to lots of rolling-release fixes and anticheat updates whereas the last official SA:MP release is more than four years ago (according to sa-mp.com).



Apart from a few exceptions, MTA is technically already capable of ?emulating? all features SA:MP has and we?re sure - with your help as experienced developers - we?ll be able to address the remaining exceptions together as well.



Furthermore, we already have a resource called amx (https://github.com/multitheftauto/mtasa-resources/tree/master/%5Bgamemodes%5D/%5Bamx%5D) that runs a Pawn virtual machine inside MTA. Unfortunately, it?s outdated since a couple of years and probably needs refactoring, but the general approach works.



In the end, we?re sure that it will be a win-win situation for both sides and is a great opportunity for both communities: Besides more workforce and fresh ideas, we?d benefit from better gameplay diversity for our players and you?d delight in an influx of new players (e.g. ?die-hard? MTA fans) and lots of new scripting features.



If you have doubts about any kind of possible issues, we?d be happy to discuss them with you.



Kind regards,

- on behalf of the MTA Team

[Gravityfalls]

I expect to hear "no" from the team.

Makes no sense because this will help MTA, and not open.mp. What we're here for actually.

[BloodMaster]

I am in no way representing open.mp community in this post, but here's what I think.



If we, players,?wanted to play/develop MTA, we would.



SA-MP is a mod that's supposed to keep the credibility and looks of San Andreas, while MTA is a custom engine you can mod, and that's the main difference why many more players are better appealed by SA:MP (we hope open.mp in future).?Check out this thread:?https://www.burgershot.gg/showthread.php?tid=401

[adri1]

what?

[javelin]

That's some decent fucking marketing, way to go MTA team, way to go!

[ohmios]

As a player my main issues with MTA are:



1) MTA client is famous for being a pain in the ass?to make?run on linux with wine, to be fair it seems easy now using the latest wine, but I can see that being broken again with more MTA updates.

While samp client on linux just works? and hopefully open.mp client will too when it gets released many moons ahead.



2)?One?MTA?server didn't let me join, the error message from the server literally said it was because I was running wine.?



3) AFAIK?MTA server doesn't support ARM while open.mp will.



4) XML is fucking everywhere in MTA, fuck XML.

[iReal Worlds]

Yeah, MTA is cool and all, but... Lua... Srsly?

[BloodMaster]

Yeah, MTA is cool and all, but... Lua... Srsly?

Well, we're stuck with pawn from 2006... So yeah

[ohmios]

Yeah, MTA is cool and all, but... Lua... Srsly?

Also this, Lua is far from being my favorite

Well, we're stuck with pawn from 2006... So yeah

Isn't Pawn older than that?

[michael@belgium]

Join MTA? Yeah no, imagine everyone switching to lua...



Open.mp might have more languages than pawn so thats a huge benefit

[Fearful]

LUA is far superior language than PAWN. I don't know what are you talking about. MTA:SA has become something that open-mp will avoid, a custom game engine that you can script, most of the MTA servers do not feel like GTA:SA anymore.

[Gforcez]

My problem with MTA is: "Downloading 3,1GB before you can join the server.. current speed: 50kb/s". I have a 250 down connection.. Sure, great I love waiting ages before joining a server and then leaving again because it's not my thing! This happens with multiple servers. Especially the Roleplay ones that have everything modded.

[BloodMaster]

Well, we're stuck with pawn from 2006... So yeah

Isn't Pawn older than that?

Well, it is, but believe it or not, Pawn is getting updated to this date. But SA-MP is using the version from 2006. (maybe even from 2005. considering the mod first went public that year)
You can see the development here:?https://www.compuphase.com/pawn/pawnhistory.htm

MTA:SA has become something that open-mp will avoid, a custom game engine that you can script, most of the MTA servers do not feel like GTA:SA anymore.

THANK YOU!

[Deity]

Interesting

[JustMichael]

We will comment officially, when we have fully discussed this.

This is by no way a small offer, and it will take some time to address all the information that has been
mentioned in this thread.

So don't take this as us saying "no" or drawing out the time. We will reply in due time.

[Jusonex]

At first, thanks for all the feedback!

We will comment officially, when we have fully discussed this.

Sounds good. Take all the time you need. We'd be delighted to chat about any questions and concerns you may be having!



In the meantime, I'd like to respond to a couple of things that were mentioned in this thread.

SA-MP is a mod that's supposed to keep the credibility and looks of San Andreas, while MTA is a custom engine you can mod, and that's the main difference why many more players are better appealed by SA:MP

I read this a couple of times before, so it seems like a pretty important aspect. It's true that MTA got something like a "custom engine you can mod". I personally don't think it's a restriction though. In the end, nobody forces anyone to create something that looks totally different from SA - it's the server's decision. Also, I think you inevitably end up with something like this if you add more and more features.



Just a rough and spontaneous idea to address this issue: Perhaps it would help to introduce a server tagging system that allows for better filtering (e.g. "SA credibility").

1) MTA client is famous for being a pain in the ass?to make?run on linux with wine, to be fair it seems easy now using the latest wine, but I can see that being broken again with more MTA updates.

While samp client on linux just works? and hopefully open.mp client will too when it gets released many moons ahead.

2)?One?MTA?server didn't let me join, the error message from the server literally said it was because I was running wine.?

3) AFAIK?MTA server doesn't support ARM while open.mp will.

(1) MTA depends on a few third-party libraries that are known to cause problems on Wine (e.g. CEF). We invested some time a few months ago to make these components optional. This should have improved Wine support. Also, I know of a couple of people who actively use MTA on Linux.

(2) It's true though that some servers block Wine. The reason is that essential anticheat components do not work on Wine and therefore simplify cheating a lot. So we decided to go for a trade-off that allows servers to block Wine. At the same time, we added scripting features that can be used for per-player whitelisting.

(3) Technically, it's possible, but the demand for it was pretty low in the past.

Lua... Srsly?

In my humble opinion, Lua is a very nice multi-paradigm language that supports imperative, object-orientated and even some functional programming approaches. Unlike other languages, it was also implemented with focus on embedding in other applications - that's a big advantage, especially when it comes to securely using it for clientside scripting.

Apart from that, let me mention an interesting project that was created by the MTA community recently: https://mta-slipe.com/. It essentially transpiles C# to Lua.

imagine everyone switching to lua...

That's not what I was suggesting. We're talking about a compatibility layer that would allow you to run your SA:MP gamemodes (written in Pawn) without modifications directly on MTA. Like open.mp wants to implement additional languages, we'd add Pawn as an additional language to MTA (incl. SA:MP's scripting API).

[Deity]

At first, thanks for all the feedback!

That's not what I was suggesting. We're talking about a compatibility layer that would allow you to run your SA:MP gamemodes (written in Pawn) without modifications directly on MTA. Like open.mp wants to implement additional languages, we'd add Pawn as an additional language to MTA (incl. SA:MP's scripting API).

If you can manage?this i'll move my?server instantly to MTA, without any doubt. Jesus how good that would be.. Too good to be true however :(

[BloodMaster]

Like open.mp wants to implement additional languages, we'd add Pawn as an additional language to MTA (incl. SA:MP's scripting API).

The question is then, why didn't you do it already?

Why do you need open.mp's developers??

Considering how much you've done with MTA, adding compatibility with PAWN seems like a week worth of coding.

I read this a couple of times before, so it seems like a pretty important aspect. It's true that MTA got something like a "custom engine you can mod". I personally don't think it's a restriction though. In the end, nobody forces anyone to create something that looks totally different from SA - it's the server's decision. Also, I think you inevitably end up with something like this if you add more and more features.



Just a rough and spontaneous idea to address this issue: Perhaps it would help to introduce a server tagging system that allows for better filtering (e.g. "SA credibility").

And what would really stop any server from adding the same tag?



I am really jealous of your client-side scripts, but that's about it. I wouldn't like my SA:MP server to compete with other MTA servers.

Possible solution could be adding a whole another master list where you would have "Multi theft auto" and "San Andreas Multiplayer" servers where San Andreas Multiplayer servers would be limited on what can be done, but are able to switch to Multi theft auto whenever, thus, removing the limitations. I mean name "San Andreas Multiplayer"?is by no means copyrighted, so yeah?

I think you inevitably end up with something like this if you add more and more features.

It's been 14 years since SA:MP's release date... 14 years! And I could make huge lists of what could be done for SA:MP while still keeping the game's credibility.

But it's true that I think SA:MP is really close to being a finished product.



I spent good years on MTA's race-dm servers (DDC), but never wanted to develop my own server.

[ohmios]

Just a rough and spontaneous idea to address this issue: Perhaps it would help to introduce a server tagging system that allows for better filtering (e.g. "SA credibility").

Or a way for the player to know how big is the download before having to enter to the server, even better, that the player can sort players by how much he would have to download.?

We invested some time a few months ago to make these components optional. This should have improved Wine support.

That could explain why it was way easier and more stable?with the latest version of MTA and Wine, Linux love is always welcome!

(2) It's true though that some servers block Wine. The reason is that essential anticheat components do not work on Wine and therefore simplify cheating a lot. So we decided to go for a trade-off that allows servers to block Wine. At the same time, we added scripting features that can be used for per-player whitelisting.

That happened to me about 2 years ago, hopefully it has been sorted out now.

[Jusonex]

The question is then, why didn't you do it already?

Why do you need open.mp's developers??

Considering how much you've done with MTA, adding compatibility with PAWN seems like a week worth of coding.

It's definitely more than a week of coding. Adding the PAWN virtual machine isn't the problem. The problem is that many SA:MP functions cannot be mapped directly to MTA as they behave differently. So you have to essentially translate all SA:MP function calls into a set of Lua instructions and transform parameters as needed.



An alternative approach we've been brainstorming about is to implement a transpiler that actually "converts" PAWN code to Lua, similar to how https://mta-slipe.com transpiles C# to Lua. The feasibility of that approach depends a bit on the usage of showstoppers like the #emit statement (as it would require an additional bytecode to Lua transpiler), but the big advantage is that you can start immediately with adding new features.

One of our developers has recently written a blog post about that matter: https://qaisjp.com/2019/04/24/exp-transl...o-lua.html



We are in no way SA:MP or PAWN experts, so there are surely pitfalls we don't know about, but open.mp devs do. Also, it's about manpower.

And what would really stop any server from adding the same tag?

[...] Possible solution could be adding a whole another master list where you would have "Multi theft auto" and "San Andreas Multiplayer" servers where San Andreas Multiplayer servers would be limited on what can be done, but are able to switch to Multi theft auto whenever, thus, removing the limitations.

Nothing would stop them from adding the tag unless it's manually managed by MTA moderatiors. I'm personally open for splitting the server list into categories, but I don't think they should be artificially limited as there'd be a pretty big gap between "SA:MP" and MTA servers then that is difficult to overcome.

Or a way for the player to know how big is the download before having to enter to the server, even better, that the player can sort players by how much he would have to download.?

I'm happy to tell you that we've already reacted to issues mentioned in this thread:

• https://github.com/multitheftauto/mtasa-blue/issues/893
  
  
• https://github.com/multitheftauto/mtasa-blue/issues/896 (to get rid of the download performance issues)
  
  

[MyU]

Before I say anything: Whatever I say is my personal opinion and not really related to OpenMP or the team.



The thing is as some already noticed MTA and SA:MP are two different things, really different things.

While MTA has been innovating SA:MP has been staying "true to the old spirit"; by a lack of updates and ignored feature requests.



I for myself been playing MTA and SA:MP (MTA 2-3 years ago active; and 1-2 months after the OOP LUA addition) and after some time it got too heavy for me I've been using dated hardware back then (but way over the GTA:SA specs ofc) and had many problems regarding CEF after that even though I'm sure they are resolved by now.



While SA:MP comes with close as none dependencies (archive file, main samp.dll, launcher and 1-2 files) MTA:SA comes with all the CEF files, CGUI files, locale files and so on (yes I know SA:MP statically links some stuff too).



Merging both simply can't work in my opinion, they're too different and both have different goals from the public view somehow SA:MP remembers me of the old mods you drop into your game folder - and they work! MTA:SA is more of a "polished" and full fledged multiplayer modification for a mod and well people have preferences.



Simply said:

SA:MP -> lightweight and easy to use but limited in more then a few aspects.

MTA:SA -> heavy but feature rich and sometimes a lil complicated oh.. and a bunch more stable than SA:MP (people use MTA:SA's crashfixes to resolve problems this tells much)



But yes I don't think people would like that we have two different communitys and long term goals I guess and competiton and selection is healthy.

[dkluin]

I think it's better to merely work together but keep both projects seperate. Both have seperate goals and visions on how to do things, I doubt a merge would work

[ohmios]

Or a way for the player to know how big is the download before having to enter to the server, even better, that the player can sort players by how much he would have to download.?

I'm happy to tell you that we've already reacted to issues mentioned in this thread:

• https://github.com/multitheftauto/mtasa-blue/issues/893
  
  
• https://github.com/multitheftauto/mtasa-blue/issues/896 (to get rid of the download performance issues)
  
  

"Show initial download size in server browser" That's perfect! Looking forward to see this implemented. It will be great to just sort the server list by download size and try some servers that need less than 30mb.



Now just need to replace XML or add the option to use something else like?YAML or?JSON or plain text files or anything else.

[nik]

I never got into MTA, it doesnt even feel like something related to San Andreas

also the MTA community is way more different than the one in SAMP, the latter having a superior one

[TheRealBot]

What about CLEOs?

[Gforcez]

What about CLEOs?

what about them?

[ronfino]

what about no

[Jusonex]

The thing is as some already noticed MTA and SA:MP are two different things, really different things.

I think we're starting at the wrong point. We should rather look for similarities than for differences, if we wanted this to work. And in case of specific differences, I'm sure it's possible to find a compromise.



Beyond that, to be honest I don't agree with your theory. In my estimation, the goals of both projects are very similar. Just the implementation differs as a result of a different basis.

Some servers on MTA don't look like GTA:SA, that's true. But the reason is not that they can't, they (some and only some!) just don't want to. From a technical perspective, there's nothing that prevents you from keeping the "true spirit" of GTA:SA. In case of SA:MP, you don't have the choice.



Apart from that, I'm honestly a bit wondering what your goals are for open.mp. If you don't want to continue SA:MP's "lack of updates and ignored feature requests" (as you said it), you'll inevitably end up with diverging from GTA:SA's true spirit at some point (and that's perfectly fine in my opinion).

I for myself been playing MTA [...] and after some time it got too heavy for me I've been using dated hardware back then (but way over the GTA:SA specs ofc)

Bad performance is in almost all cases related to poorly scripted gamemodes (specifically because of bad clientside scripts). "Unfortunately", the sentence "from great power comes great responsibility" applies here.

While SA:MP comes with close as none dependencies (archive file, main samp.dll, launcher and 1-2 files) MTA:SA comes with all the CEF files, CGUI files, locale files and so on (yes I know SA:MP statically links some stuff too).

I wonder if this is really a problem. Most applications consist of more than 10 files nowadays and the regular user doesn't care anyway from my experience (but I principally understand you here and as a developer and perhaps German too I sometimes share that paranoia ;) ).

Now just need to replace XML or add the option to use something else like?YAML or?JSON or plain text files or anything else.

There's only the meta.xml (resource "information" file) that requires you to use XML. For everything else, you can choose whatever you want. We for example have the toJSON and fromJSON function (https://wiki.multitheftauto.com/wiki/FromJSON). I'm not a fan of XML either, but I don't think it's worth adding support for another file format for the meta file as it'd make things more complicated than they need to be (and cause fragmentation).

[Sasino97]

I think the greatest advantage would be the number of players that already play MTA, and the second one is the very advanced level of the features it has. Downsides: their forum is ugly, sorry.
Still I love the very idea of open.mp and the reasons behind it, so I think they should grow as separate projects.
Whatever happens, I will stay here as long as I can port my existing SA-MP code.

XML is fucking everywhere in MTA, fuck XML.

You are right, thousands of CreateObject lines in your game mode that slow down compilation time are certainly much better :D!

[Gravityfalls]

What are you even talking about? The new community compiler compiles so fast - why you have to worry about your thousands of lines of code with CreateObject?

[Chase]

I like MTA, I would have played it more if someone made it compatible with fastman92's limit adjuster to be able to add vehicles, weapons, etc without replacing anything. I know many people here prefer to not make GTA SA a custom engine you can mod but I personally believe that decision should be left to individual servers.

[Mark2]

I'm all for that offer. SA:MP has a large established community, MTA has something to offer this community, MTA gives developers almost unlimited opportunities, greater stability and a minimum number of bugs, and the fact that it is difficult - an excuse for the lazy. SA:MP just "shot faster" and took the initial piece of a large community, while MTA worked hard and gave server developers something that developers and players of SA:MP could not dream of. Choosing so many players is SA:MP - it's not because it's easy to use, just SA:MP was at the right time in the right place =)

[MyU]

@Jusonex: Some arguments may seemed a bit off but yea I was pretty tired writing this.



Well; Simply said in my honest opinion MTA:SA is fine.. it's just not completely for me if I want to roleplay or dm; the extension of custom fancy modern content just ruins the "old-school" feelings for me, and yes sadly SA-MP did the same with the RL releases but that's still pretty limited compared to MTA:SA.

I'm personally fine extending upon GTA:SA functions without needing custom made content.



And to make something I said b4 clear:

MTA:SA is more vanilla GTA:SA as SA-MP considering SA-MP disabled and failed to sync things like: (hydraulic / horn sync, disabled blur, disabled haze, disabled vegetation water plants etc and many other small things).

BUT it provides content creators with a bunch of functions and ways (CEF, Dx drawing etc) or basically said power over GTA:SA and utilities to make it look they want. And that's one thing many people I know dislike about MTA:SA (excluding abnormal "community caused" download sizes which makes quick play for MTA:SA on bad net nearly impossible).



SA:MP is your... well basic crappy but completely custom content hating competitor (ignoring the DL releases in that case) which gave a good mix and a good home for both parties. and well... Open MP focuses on SA:MP - your "competetion" or well let's say lame competitor when it comes to technical components.



So that's why it won't probably fit.

And I would contribute to MTA:SA whenever I could - but the community of developers is so big and communicative which fixes bugs and adds new things faster then I could recon them not playing MTA:SA or being involved in the community.

Take this as an compliment (:

[Reachless]

MTA would have been populated soo much earlier by SA:MP players if certain things were done. For example, as a beginner in Lua, I have found implementing clientside damage incredibly difficult and, when I asked for support on your forums, someone from your staff literally told me to "not implement clientside damage because the sync is just right in MTA". Hell nah, I've set the most effective rates possible in the MTA server settings and more than 50% of the bullet hits were not counted.



You should pay more attention to the support request topics in your forum, because SA:MP server owners (like me, with hundreds of players willing to migrate to MTA) always ask for help but nobody bats an eye over there.

[Sasino97]

AddPlayerClass(1, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(1, 125421, 124214, 214421, 4210421, 241013205, 21401)But, what happened in the end? Do we have any news about this topic?by the open.mp project maintainers?

What are you even talking about? The new community compiler compiles so fast - why you have to worry about your thousands of lines of code with CreateObject?

But why mix the source code with resources??It's just awful.

I don't know you, but in my gamemodes I always load player classes, objects, vehicles etc. from the scriptfiles instead of uglifying?my code with?

Code:

AddPlayerClass(1, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(2, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(3, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(4, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(5, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(6, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(7, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(8, 125421, 124214, 214421, 4210421, 241013205, 21401)

AddPlayerClass(9, 125421, 124214, 214421, 4210421, 241013205, 21401)

I don't use XML, but I like the idea of importing resources from files, and I think that the open.mp API should provide basic loaders for objects, vehicles, skins, textdraws (and all UI),?and other few things.

[IllidanS4]

For years I have been waiting for something that could unify the playerbases, but I don't think centering on MTA is a good course of actions. Running scripts on MTA with a SA-MP compatibility layer would eventually make developers switch to the features of MTA and limit people with the SA-MP client. Opening servers to MTA players is a good idea, but regardless of its features (I have nothing against full modding support, Lua or XML), the focus should be on SA-MP servers and players.



Just my opinion.

[Jimmy]

Close MTA and help making open.mp great! open.mp can get backwards compatible with SAMP and MTA both... It should have functionality of both MTA and SAMP. Whats better than that?