ID:308397
 
How am I supposed to know what BYOND can and can't handle? Man, I really wish at times like these that I had a fast-paced action MORPG to reference made by Tom and Lummox JR to really see the limits of this engine.

Do you have any advice on procedures to avoid and such? On how to conserve memory any way possible? If so - please post away!

(This post is hidden from the front page: language) ?... Really?
just profile everything you do and then take out cpu consuming procs one by one. it should be pretty simple.

if your problem involves having lots of mobs over a large map that require while() procs, then have a global proc that tells them to do their loop every tick and have said proc skip mobs that aren't within 10 or so tiles of a player.

err other than that i can't really think of any issues you'd have if you're making a big rpg. pixel movement doesn't seem to cause lag and i run Epic in 30 fps with no issues, so yeah.. just get a good server when you're done if you want to hold like 30+ people.
When you use higher framerates the limit can depend on your graphics hardware. Also, when you're talking about a multiplayer game, the limits depend on each player's internet connection. There's no way to define what the limit is.

It would be nice if BYOND provided some information about what typical performance is. They (the BYOND staff) could create a simple benchmark program and list the maximum framerate for different hardware. Or host it, have a client connect, and measure the average latency. Not only would these numbers be good to have, it'd also show that the BYOND staff is concerned about performance.
Tom has already stated that he's not an expert at the language, so he most likely can't show you the "limits" of the engine. He just makes the software, it's up to the user base to push it to its full potential.
In response to EmpirezTeam (#3)
EmpirezTeam wrote:
Tom has already stated that he's not an expert at the language, so he most likely can't show you the "limits" of the engine. He just makes the software, it's up to the user base to push it to its full potential.

I'm getting tired of pushing it to its limits (however, probably not what you'd call its "full potential").
In response to Forum_account (#2)
I'd like a way to judge ping. I figure getting the pager to track this sort of thing would allow for some useful information collection as well.
In response to EmpirezTeam (#3)
EmpirezTeam wrote:
Tom has already stated that he's not an expert at the language, so he most likely can't show you the "limits" of the engine. He just makes the software, it's up to the user base to push it to its full potential.

Not necessarily. To push it to it's limits you can use a simple program as a benchmark. Just create a bunch of mobs and call walk_rand() for each one, then connect a client and measure the framerate, ping time, and bandwidth usage. It doesn't really matter that it's not an actual game, you just care about the number of objects that are being updated and drawn on the screen.
All the more reason someone besides Tom, who has enough work as it is, should do it.
In response to EmpirezTeam (#7)
EmpirezTeam wrote:
All the more reason someone besides Tom, who has enough work as it is, should do it.

I'm not sure why this task being easy gives more reason for someone else to do it. There are also parts of it that the BYOND staff would have to do (measuring ping times and framerates).

Aside from that, the reason the BYOND staff should do this is because they're the ones who stand to gain from it. It doesn't help me to know that BYOND can draw 2000 sprites at 40 frames per second, but that number might impress game developers who are thinking about using BYOND. Providing these performance measurements would also show that the BYOND staff is concerned about performance. While I do benefit from the BYOND staff being concerned about performance, I don't benefit from them simply showing that they're concerned about performance (but they do).
In response to Forum_account (#8)
People assume just because Tom and Lummox are the ones developing DM, they're expert game developers. Fact is, you could probably make a game better than they could, which is my point: the "pushing" should be done by the user base since they spend more time using the language as opposed to Tom and Lummox who spend more time making the engine.

It doesn't make sense for them to stop what they're doing to make an MORPG when the user base could do a much better job. I'd rather have them keep pumping out updates.
In response to EmpirezTeam (#9)
I think it can be as simple as this:
http://www.byond.com/developer/ADT_CLONE/ ServerStressTestProgram

Maybe a newer version might be nice. The point is, no one here is asking for a huge game from Tom, just a sufficient stress test of the DM language and its features.
In response to EmpirezTeam (#9)
EmpirezTeam wrote:
People assume just because Tom and Lummox are the ones developing DM, they're expert game developers. Fact is, you could probably make a game better than they could, which is my point: the "pushing" should be done by the user base since they spend more time using the language as opposed to Tom and Lummox who spend more time making the engine.

It doesn't make sense for them to stop what they're doing to make an MORPG when the user base could do a much better job. I'd rather have them keep pumping out updates.

I guess I agree with you
In response to Kaiochao (#10)
Kaiochao wrote:
The point is, no one here is asking for a huge game from Tom, just a sufficient stress test of the DM language and its features.

MORPG and stress test are two different things. I'm referring to Orange's OP.
No need to make a MORPG to test what it can "handle". Its limits are the same as every other language (sans 3D support) beyond its processing power limit.
In response to SqueakyReaper (#13)
SqueakyReaper wrote:
Its limits are the same as every other language (sans 3D support)

You mean engine, right? You could make your own 3D renderer in the DM language just as you could in virtually any other programming language...
In response to Tomy (#14)
Tomy wrote:
SqueakyReaper wrote:
Its limits are the same as every other language (sans 3D support)

You mean engine, right? You could make your own 3D renderer in the DM language just as you could in virtually any other programming language...

Why do it here when you have other languages such as C++? o.O
I don't think DM can support model files.
Languages, by default, don't support model files.
I can read and render stuff stored in the .ply format (yes, it's basically a text file, but it still counts as a valid model format, right?)

@Ocean King
Why do it in C++ if you can do it in an assembly language :p
Anyway, I really am working on something like this. It's still far from being optimized or even working as intended.

Oh, and a raytracing engine seems to have been already done in BYOND
http://www.byond.com/games/Xooxer/mobRay
Huh. Alright, you win that point. I still put out there that the current lack of 3D support, either through a user made library or engine-based features, makes it impractical to make a 3D game. But I guess a large application of the raytracing engine would beat that? I dunno.
Page: 1 2