PDA

View Full Version : Full SDK? Rene!



Laokin
6th Oct 2008, 19:39
I was just wondering if maybe you could comment on an SDK? I believe it's in the best interest to allow DX3 to be FULLY modable unlike IW. If IW was fully modable, I think it would have been 100x better of a game -- the community could of fixed it, would have fixed it. I believe making the game as close to DX 1 is only half of why IW failed, the other half was the community was DOA.

Oh and the other question I wanted to ask, is it possible you could drop a hint or two about which upcoming magazines will be sporting DX3 info?

Thanks in advance.

P.S.
I sure appreciate all the answers you've given the community.

K^2
6th Oct 2008, 21:38
I'm planning to get Tomb Raider Underground upon release and work on taking apart various files. I'm not going to make any serious promises, but I have been successful in the past in breaking the format and writing import/export scripts for 3D models. So I'll be working on that, and with any luck, textures and at least some details of the maps.

Naturally, making full mods would be difficult sans SDK. In particular, most games supporting mods load either a dynamic library or a script to run along with the engine. Some of that can be tinkered with using code injection, but it would be far more limited than if EM chooses to support dynamically loaded content.

René
6th Oct 2008, 21:42
I was just wondering if maybe you could comment on an SDK?

I think it's too early to say. At least, I can't confirm anything right now, sorry.

DXeXodus
7th Oct 2008, 05:27
I believe that a full SDK is really something that should be taken seriously. It can really add a lot to a game. The lifespan of the game increases, the community base will grow even stronger. A SDK release will be a very good move in my opinion.

Igoe
7th Oct 2008, 05:41
If there is one, they'll worry about that AFTER the game is out the door.

If you want an SDK, the best thing to do is BUY the game and convince all your friends to BUY the game.

K^2
7th Oct 2008, 05:49
There are some dangers involved with an SDK, which is why some developers might be very reluctant about releasing them. Especially, if online multiplayer is a feature. Without SDK, the only thing you have to be paranoically careful about is networked data.

Now, imagine that you also have custom maps to worry about. Unlike network data, which tends to be pretty simple, not in the least due to bandwidth limitations, custom levels would be rather complex data sets. Making sure that no buffer overflows occur in that is far more challenging.

And then there are dynamic libraries. That's just asking for vulnerabilities. There are things that can be done, but essentially, you'll be leaving it up to the player to make sure that a mod doesn't compromise their security. And when a few people do get attacked because they downloaded a mod they shouldn't have, guess who gets blamed.

This gets particularly bad with network play. Lets say you connect to a remote server. Lets say that remote server is running a custom map that causes a successful code injection via buffer overflow. Once you are connected, congratulations, your system has just been hijacked. And no, the firewall won't do jack, since you are the one who initiated the connection, and all of the data can be pumped through game's own networking code.

Now, the odds of such vulnerability existing are pretty low, as long as programmers are competent. However, slip-ups happen to everyone. And there are enough people out there who will try putting N+1 nodes into every list that explicitly expects N. So the engine would have to be tested rather extensively for that sort of thing, and that takes time and money.

All of this can still happen, of course, without an SDK. But at least if some people with nothing better to do figure out how to put in custom content, like it has been done with IW, the guys at Eidos can always say its not their fault.

Laokin
7th Oct 2008, 06:26
There are some dangers involved with an SDK, which is why some developers might be very reluctant about releasing them. Especially, if online multiplayer is a feature. Without SDK, the only thing you have to be paranoically careful about is networked data.

Now, imagine that you also have custom maps to worry about. Unlike network data, which tends to be pretty simple, not in the least due to bandwidth limitations, custom levels would be rather complex data sets. Making sure that no buffer overflows occur in that is far more challenging.

And then there are dynamic libraries. That's just asking for vulnerabilities. There are things that can be done, but essentially, you'll be leaving it up to the player to make sure that a mod doesn't compromise their security. And when a few people do get attacked because they downloaded a mod they shouldn't have, guess who gets blamed.

This gets particularly bad with network play. Lets say you connect to a remote server. Lets say that remote server is running a custom map that causes a successful code injection via buffer overflow. Once you are connected, congratulations, your system has just been hijacked. And no, the firewall won't do jack, since you are the one who initiated the connection, and all of the data can be pumped through game's own networking code.

Now, the odds of such vulnerability existing are pretty low, as long as programmers are competent. However, slip-ups happen to everyone. And there are enough people out there who will try putting N+1 nodes into every list that explicitly expects N. So the engine would have to be tested rather extensively for that sort of thing, and that takes time and money.

All of this can still happen, of course, without an SDK. But at least if some people with nothing better to do figure out how to put in custom content, like it has been done with IW, the guys at Eidos can always say its not their fault.

You seem very knowledgeable in this field, so I won't try to impede. I do however see many, many, many, many, many, Dev's take this "huge risk" before though, so here is to hoping Eidos will follow suit. On another note, they haven't mentioned multi-player to my knowledge -- so I don't see this being a big deal. Likely like invisible war, DX is really no place for multi-player unless it's set up somewhat like Dystopia. (Cyber-punk HL2 mod with augmentations, multi-player only.) That being said, mod's would most likely be SP, so really all they allow you to do is edit base stats of the game, art, and camera views.... as well as let you add in sounds and levels. HL2 has an SDK, Quake 1,2,3,4, Doom3, Every Unreal, Gears of War, Crysis.... Most of the games companies expect to live a long time. Even DX 1 had an SDK, leaving the SDK out of 2 was the other big mistake from Warren Spector. Either way I feel this is an important thread that warrants an answer in the future. Show some support for the SDK.....:-)

Oh and thanks a lot Rene for taking the time to drop by and say something.... It's so cool to feel like I'm just a question away from an answer. Make's me trust in EM even more.

K^2
7th Oct 2008, 06:52
Multiplayer could be coop. It would introduce all the same problems.

And yes, there are a lot of developers who take the risk, but there are more of these who don't. As I said, it is manageable. It just costs more to release a secure app with SDK than without, even if all the tools are already there.

jordan_a
7th Oct 2008, 09:04
Whether you need a thread about SDK or multiplayer, please be aware that we already have them.

[HP]
7th Oct 2008, 09:24
I believe that a full SDK is really something that should be taken seriously. It can really add a lot to a game. The lifespan of the game increases, the community base will grow even stronger. A SDK release will be a very good move in my opinion.

Indeed, putting a SDK and Editor on a game, expand the lifespan of the game a lot more! I' d love to see this, being a modder and level designer myself.

Mecranth
7th Oct 2008, 13:40
A utility like this would very much benefit everyone, including the developers. People with the skills would be capable of changing a lot of the issues that many others have expressed their distaste of, which would in return get Eidos Montreal more buyers. It would also dramatically increase the lifespan of the game, a welcome addition even to a lengthy title.

Bluey71
7th Oct 2008, 16:25
I believe that a full SDK is really something that should be taken seriously. It can really add a lot to a game. The lifespan of the game increases, the community base will grow even stronger. A SDK release will be a very good move in my opinion.

Agreed.

The stronger the community the more sales you'll make with DX4.

Lady_Of_The_Vine
7th Oct 2008, 16:28
We do have a discussion thread for SDK and Fan Missions.

http://forums.eidosgames.com/showthread.php?t=75249

JCDentonMale
7th Oct 2008, 17:48
I also think having a SDK will be great. Just as it is great to have the ability to do our own unreal maps/mods.

Also I would like to say : René, you have a private message on the jordan_a website http://unatco.org/, search a little to find it ;)

Laokin
7th Oct 2008, 17:57
We do have a discussion thread for SDK and Fan Missions.

http://forums.eidosgames.com/showthread.php?t=75249

No offense, but you get on my nuts because I missed a thread -- so like you pointed it out to me, I'll point it out to you. We already have a post in this thread containing this information.

"jordan_a Whether you need a thread about SDK or multiplayer, please be aware that we already have them."


So really, I ask, is it that big of a deal? Am I really supposed to post an important question in and old thread that Rene may have never read? I made it a thread for a reason, so the question can get asked in a place everyone is sure to see -- Rene had no problem with it, why do you?

Lady_Of_The_Vine
7th Oct 2008, 18:13
No offense, but you get on my nuts because I missed a thread -- so like you pointed it out to me, I'll point it out to you. We already have a post in this thread containing this information.

"jordan_a Whether you need a thread about SDK or multiplayer, please be aware that we already have them."


So really, I ask, is it that big of a deal? Am I really supposed to post an important question in and old thread that Rene may have never read? I made it a thread for a reason, so the question can get asked in a place everyone is sure to see -- Rene had no problem with it, why do you?

I'm not offended... don't worry. :)
But my post was nothing like a "big deal", as you put it. I thought you might not have known about the other discussions, so I was just letting you know.
That's why I linked to the main thread of ALL discussions, not just the SDK one.
This is done often by other members about other topics, it is meant to offer up further information for new members.
I just posted a link... not a "hey, you idiot, you need to discuss this in here because we already have a thread about it, blah blah" :D

I don't have a problem at all, so why do you imply I do? :rasp:

Laokin
7th Oct 2008, 18:40
It's just that internet is the most visceral battlefield in existence, and I understand how forums work, especially this one. I been here since the day this baby was open I just usually keep to myself until there is actually a point in discussion that refers to the game the board is about. This magazine thing was the first big deal since announcement. I apologize for lumping you into the category of viscous wolves, and thank you for making a welcoming effort.

AaronJ
7th Oct 2008, 18:43
lol...the chances of this are incredibly slim, as much as I want it to happen.

K^2
7th Oct 2008, 19:03
lol...the chances of this are incredibly slim, as much as I want it to happen.
But that's what hackers exist for.

AaronJ
7th Oct 2008, 19:18
But that's what hackers exist for.

Worked for Invisible War, didn't it?

K^2
7th Oct 2008, 19:29
To an extent, at least. Project to get 3D models replaced never got enough momentum.

Bluey71
7th Oct 2008, 19:38
lol...the chances of this are incredibly slim, as much as I want it to happen.

Sadly I think you are right - if Eidos could just look beyond making quick bucks, and nurture a Deus Ex community , fans would be happier (imho) and DX4 sales would be assured.

Lady_Of_The_Vine
7th Oct 2008, 19:47
It's just that internet is the most visceral battlefield in existence, and I understand how forums work, especially this one. I been here since the day this baby was open I just usually keep to myself until there is actually a point in discussion that refers to the game the board is about. This magazine thing was the first big deal since announcement. I apologize for lumping you into the category of viscous wolves, and thank you for making a welcoming effort.

Understood... and no problems. I didn't realise you were an established member because of your low post count, sorry. :)
All is cool. :cool:

Romeo
7th Oct 2008, 19:48
I believe that a full SDK is really something that should be taken seriously. It can really add a lot to a game. The lifespan of the game increases, the community base will grow even stronger. A SDK release will be a very good move in my opinion.
No doubt. I've seen games outlast their initial release by more than a decade with some good mod support (See: Total Annihilation) and still other games get fresh ideas and innovation (See: The Elder Scrolls series and once again, Total Annihilation). If Eidos doesn't, I do hope K^2 does.

There are some dangers involved with an SDK, which is why some developers might be very reluctant about releasing them. Especially, if online multiplayer is a feature. Without SDK, the only thing you have to be paranoically careful about is networked data.

Now, imagine that you also have custom maps to worry about. Unlike network data, which tends to be pretty simple, not in the least due to bandwidth limitations, custom levels would be rather complex data sets. Making sure that no buffer overflows occur in that is far more challenging.

And then there are dynamic libraries. That's just asking for vulnerabilities. There are things that can be done, but essentially, you'll be leaving it up to the player to make sure that a mod doesn't compromise their security. And when a few people do get attacked because they downloaded a mod they shouldn't have, guess who gets blamed.

This gets particularly bad with network play. Lets say you connect to a remote server. Lets say that remote server is running a custom map that causes a successful code injection via buffer overflow. Once you are connected, congratulations, your system has just been hijacked. And no, the firewall won't do jack, since you are the one who initiated the connection, and all of the data can be pumped through game's own networking code.

Now, the odds of such vulnerability existing are pretty low, as long as programmers are competent. However, slip-ups happen to everyone. And there are enough people out there who will try putting N+1 nodes into every list that explicitly expects N. So the engine would have to be tested rather extensively for that sort of thing, and that takes time and money.

All of this can still happen, of course, without an SDK. But at least if some people with nothing better to do figure out how to put in custom content, like it has been done with IW, the guys at Eidos can always say its not their fault.
Yes, you always run a slight risk, and there's a second problem that wasn't mentioned (If people want to share their creations, they need an avenue to do so, which requires quite a bit of bandwidth, and therefore, costs more). If no SDK is released, I would really like to see what you come up with, because the sole reason I still play Total Annihilation, Morrowind and Oblivion is due to their mod-suppot. On a similar note, there's little danger if DX3 is a purely single player (no co-op/competition), I do believe, correct?

No offense, but you get on my nuts because I missed a thread -- so like you pointed it out to me, I'll point it out to you. We already have a post in this thread containing this information.

"jordan_a Whether you need a thread about SDK or multiplayer, please be aware that we already have them."


So really, I ask, is it that big of a deal? Am I really supposed to post an important question in and old thread that Rene may have never read? I made it a thread for a reason, so the question can get asked in a place everyone is sure to see -- Rene had no problem with it, why do you?
Rene might not have noticed, but they're correct. Besides, there's no difference asking your question in that thread, or starting a brand new one, this just clutters up the forums. Secondly, and more importantly, both Jordan and MrsP have been here quite a while, have both contributed great ideas, and are both respected by the moderators, so it would definately be in your best interests not to lip off to them. This thread is now in lockdown.

René
7th Oct 2008, 20:03
René, you have a private message on the jordan_a website http://unatco.org/, search a little to find it ;)

I like cafes. And coffee frapuccinos.