48 réponses à ce sujet

[SlottsMachine]

Aw, that game isn't bad at all.

 

 

Gimme an example.

 

What? Enemies climbing ladders when under fire instead of finding cover isn't bad? 

[Fidite Nemini]

Right people, back on track please. This topic isn't about stupid game mechanics or AI, it's not about GPU marketing fails and neither is it about GPUs spporting certain APIs or not (and to end that particular question, yes, the R9 280X will support DX12).

[SlottsMachine]

Okay.

[Endurium]

Insider info! And not a surprise at all. One example of poor programming discipline comes to mind in The Longest Journey; they used a shortcut in DirectX which worked at the time, but causes UI overlays to not render when forcing AA on modern systems. Dev admitted as much, but wasn't willing to do anything about it.

 

I used to do this a bit with "undocumented cpu instructions" back in the day because it was fun and part of discovery, but anymore I stick to the official instruction set and make sure I'm not changing bits marked "reserved" in any API.

 

Edit: I haven't written a game since the 80s; mostly write modding tools nowadays, when I find the time/inspiration.

[Obadiah]

Explains the rebugging in DA2 nVidia graphics.

[Kaiser Arian XVII]

I don't recall encountering that. But whatever.

 

As for the topic, that doesn't surprise me. Just try to build something so complex and interdependent as a game and then fix everything about it. You can't.

 

I was just playing Inquisition yesterday. I got to a certain spot: It's the broken-down castle in-between your initial camp and the camp near lake lucius or whatever. Solas makes a comment about wanting to dream there, and there's a rift there initially.

 

There's a broken bridge there leading into the castle (though you go into the castle from the side because it's broken), and right next to the bridge is a little patch of ground where there's a plant--Elfroot I think, maybe Embrium.

 

However, when I jumped to this patch of ground from the bridge, I noticed something. There's a hut off in the distance, again in-between the initial camp and the lake camp (but closer to the lake camp)--if you were to go play the game and stand on the bridge, you'd see it easily. However, when you jump onto this patch of ground, the hut  and the surrounding assets disappear.

 

They straight up disappear. The only thing that's on the now-blank ground in that area is a door suspended in space.

 

How did that happen? What could possibly have caused the hut to disappear? The only thing I can think of is seriously, seriously messed up occlusion culling. But it's an immediately noticeable bug. And yet, It's just one of those things that simply happens.

 

You fell into the black fade lol

 

More frequent bug was dialogs between companions (E.g. Solas and Varric) being repeated several time, to the degree that you would facepalm.

[Fast Jimmy]

Right people, back on track please. This topic isn't about stupid game mechanics or AI, it's not about GPU marketing fails and neither is it about GPUs spporting certain APIs or not (and to end that particular question, yes, the R9 280X will support DX12).

 

If anything, I'd say it is an indictment of the multiplatform model. 

 

Or, at least, the model where a game must release on all platforms AT ONCE. If a develop committed to a staggered release (say, maybe, three months between each platform), this could provide laser focus on both their pre-release QA as well as their patching efforts. It keeps more of the team engaged over a longer period of time, but also allows for a smaller (or at least less crunched) team size that could otherwise lay dormant or be occupied with sub-optimal tasks.

 

 

Of course, we know the reason why companies do not embrace this model - marketing. They can get much more bang for their buck with one $100 million dollar advertising campaign on one release than they could three $40 million dollar marketing campaigns over six to twelve months, despite them spending more money overall. Not to mention the perception of "playing favorites" with certain platforms... which still happens in terms of polish and QA, but just in a less overt manner.

[Fast Jimmy]

??? How is multiplatform release causing "broken" games?

 

Attention is divided in three different directions, all at once. A QA team is given the task of assessing the game across all platforms, simultaneously. 

 

Is it easier to write one ten page paper on a single subject, or three separate three page papers on totally different subjects? 

[Fast Jimmy]

That doesn't sound like programming issues, that sounds like design.

 

The overwhelming majority of the assessment is not specific to platforms.

 

I disagree. A quest that may play fine on one platform can bug out with show-stopping crashes on another. A UI that works great for one system may be terribly optimized for another. A control scheme that is easy to use for one controller may not handle well at all on another. 

 

For Microsoft platforms, you need to worry heavily about the online side of your content, as they are fairly restrictive on what downloaded content can consist of (even DLC and patches). For Sony, their certification process is cumbersome, creating lots of delays if all the i's aren't dotted or t's crossed. And PC's struggle in hardware optimization, which requires a multitude of testing across a wide variety of system layouts to capture performance and bug issues.

 

All of the above require different approaches and different focus during your testing and development phase. Trying to be all things to all people is the road to ruin in business and that's what so many developers try to do when they develop for every platform simultaneously. Better to stagger platform releases and ensure each one receives all the attention possible than dividing your efforts three (or more, I suppose, if you throw in something like the Wii) different directions.

[Fast Jimmy]

Like half of what you posted is design and not programming...

 

I guess I'd have to see exactly how programming changes with each system. I can understand it for the 360 and PS3, because they had fairly unique architecture, but outside of optimizing performance for individual chips, I don't see how porting for the "next gen" consoles is a programming issue.

 

How would there be bugs on three different systems that use the same architecture, if they all have the same version of the game?

 

Well, I'm not saying it is programming over design... but that distinction is rather moot. Although your focus on hardward architecture to the omission of operating system nuances is a bit simplistic - next gen consoles are basically PCs, but that doesn't mean they have anywhere near the same software subprocesses going on, which mean the chance for bugs between platforms is definitely possible. 

 

But even if it wasn't... the fact that it is design-based doesn't make it any less real of an issue. Nor any more easily fixed. The Save Import issue is a design issue, not a programming one... it doesn't mean that its inherent limitations are going to be overcome any time soon.

[Fast Jimmy]

You may have the same architecture but don't forget you are running on an operating system. Your software may not support the runtime thay your OS provides

 

How did it take me this long to realize you were Turing/LindsayLohan/the-million-and-one-aliases you have had over the years?

[Inquisitor Recon]

He's right, and it certainly doesn't help when Nvidia gives the driver devs more mess to deal with.

 

*cough* GTX 970 *cough*

[Fast Jimmy]

Fair enough, though I'm skeptical as to the actual affect that would end up having on the game (outside of performance issues).

 

I wasn't saying it was not an issue, merely that this thread is about bugs, not "DA I has sh*tty PC controls."

 

Bugs can be just as tied to design shortcomings as programming ones. It isn't inherent complexity that causes the problem... it is that the complexity of the task is amplified by three with every release. 

 

To use DA:I... it had close to four years in development, with multiple long, intentional delays to improve content. It also release on five different platforms. And people still cynically say it is in beta-test because of some issues remaining. 

 

I think, from both a technical and a design perspective, DA:I would have been a better game had it release on only PC and next gen consoles. I think having to develop and test for five different systems caused some of the shortcomings, as well as some of the delays in getting those issues fixed now, to the project. 

[Kaiser Arian XVII]

Lmao I hid well jimmy

 

and stroke swiftly!

[Fast Jimmy]

That's not a fair comparison at all, because the differences between the old and new consoles are huge, not least of them being memory, particularly in consideration of the ginormous (LOL, the BF thinks that's a word) size of the levels and the high res textures, which create a large and necessary divide between old and new consoles, in addition to processing power (though less so) and GPU power, all of this yet being separate from the additional issue of architecture, of the which the previous generations were unique, or rather unique from PCs, and so required large amounts of optimization, entirely unrelated to the actual game itself, and entirely related to the design of the machine, and in conclusion one might ask how the development of a next-gen-and-PC-only DA I would be affected by dropping either of the consoles, the PC understandably not considerable because of the necessary testing of various hardware, entirely unrelated to the game itself, leaving only two consoles entirely alike to one another and the PC in architecture, and with a considerably improved memory, at least in regards to previous consoles, that levels the playing field a great deal and all of which, in my opinion, decrease the amount of difference between each platform to a very small degree, and thus make questions of programming, if not irrelevant, then vastly outweighed by design decisions.

I freely admit I wrote that ^ JUST so I could make a coherent run-on sentence (230 words! That's Faulkner quality there ). What do you think?

Though I meant it all and it is indeed my argument.

Nah.

[Riven326]

If anything, I'd say it is an indictment of the multiplatform model. 

 

Or, at least, the model where a game must release on all platforms AT ONCE. If a develop committed to a staggered release (say, maybe, three months between each platform), this could provide laser focus on both their pre-release QA as well as their patching efforts. It keeps more of the team engaged over a longer period of time, but also allows for a smaller (or at least less crunched) team size that could otherwise lay dormant or be occupied with sub-optimal tasks.

 

 

Of course, we know the reason why companies do not embrace this model - marketing. They can get much more bang for their buck with one $100 million dollar advertising campaign on one release than they could three $40 million dollar marketing campaigns over six to twelve months, despite them spending more money overall. Not to mention the perception of "playing favorites" with certain platforms... which still happens in terms of polish and QA, but just in a less overt manner.

Can you think of any examples from this generation that were not directly related to hardware differences?