The Alien Isolation Linux Testing That Isn't Happening

[raonlinux]

Anyway I'm more interested if that game might support SteamVR, when we finally get that? SteamVR needs some games for launch...

Let be honest Steam Vr didn't release yet, maybe the next year and I doubt it this game will be release on SteamVR, but if is release it in a future just wait better I bet it will cheaper when they release it at less they need to change their code and need do a redux on the game, first we need steam machine when they appear and hit the market steamvr will appear this will depend if are going to linux, until I know steam machine are coming in november.
On linux is just walking on video game, we are not running yet, so be patient, SteamVR or anything a VR still in development they will need to fix a lot of issue they already have as hardware and software, the most important what will be the standard of that tech, right now too many options with pro and con.

[dungeon]

As said already, I don't play computer games, don't have the time.

With hundred articles per month you are already gamer - speedruner

[moilami]

The reason they didn't port the benchmark mode is to avoid performance comparisons with the Windows version of the game. Simple as that.

They already got some heat with Shadow of Mordor performance and don't want customers using Phoronix benchmarks as proof that their port is under optimized. That way they can claim that the client performance issues are a isolated and difficult thing to fix.

Exactly.

[chilinux]

Nice abuse of the refund system there...

Ever thought of ya know , playing it !

There should be no need for any build-in benchmarks in Linux games - the frame rate can be measured by simply overloading certain OpenGL functions in libGL.so

This is the wrong attitude to have if you really care about the game play experience on Linux. Based on this type of attitude, a game company only needs to make a game playable for just the first two hours of gameplay on Linux/SteamOS. They can both check it off as being ported and keep the resulting money from the sales regardless of how well the game plays past the first two hours.

A good built-in benchmark should give an idea of the performance across a wide variety of levels. In a 35 Gigabyte game, the majority of those levels are going to still be locked during the first two hours of gameplay.

To get a feel for how little of the game's overall performance can be experienced and benchmarked in first two hours, consider this:

Steam's statistics show of Alien: Isolation players:

- An average playtime of over 11 hours
- Slightly more than half have completed 3 out of 18 missions
- Slightly more than 40% have completed 4 out of 18 missions

A well written built-in benchmark should tell me in much less than two hours if my hardware or the quality of the port will produce poor FPS on missions beyond mission 4. The idea that I should be able to evaluate the FPS of all 18 missions by just playing during the first 2 hours is absurd.

So I guess the next obvious question is "if you can't evaluate it in less than two hours, why not just play it longer?"

The answer to that is simple: the Steam refund policy doesn't allow anything beyond two hours. If a game port tricks you into playing for at least two hours before running into problems, according to Valve/Steam, you are no longer entitled to a refund.

The next obvious question is "even if you run into problems beyond the refund period, why not just post a negative review to encourage them to fix it?"

The answer is that Steam's method review system doesn't allow reviews about poor Linux ports to impact sales. As such, the review will largely be ignored and the game company will continue to trick people into playing the first 2 hours of a poor port so the policy considers them to no longer be entitled to a refund. The bottom line is they still make their money and put a SteamOS icon next to their game.

In the case of Alien: Isolation, there is already 11777 reviews. Another review will impact the review score by less than 0.01%. So, the current review score is 93% positive and regardless of if another 1 review is positive or negative, it will still be a review score of 93%.

These reviews are then broken into: Helpful, Recent, Positive, Negative, Funny

There is no concept of a #Linux or #SteamOS or any sort of review keyword search at all.

A negative review about a poor Linux port will be 1 of 805 negative reviews where the majority of negative reviews are not related to the Linux port. How many people are manually going through over 800 reviews to see if the quality of the Linux port is a factor?

While Steam has the concept of "Performance Satisfaction by GPU," it does not have any performance satisfaction by OS.

So this brings me right back to Steam's policy that allows just 120 minutes to try to get a feel for the performance of the game while only getting to experience only a small percentage of it. And I think the next obvious question should be "why was the benchmark left out of the port?" I think there is two possible answers:

(1) The company porting the game doesn't want it to be easy to get an overall performance data because it reflects badly on the quality of the port.

- or -

(2) The company porting the game cut so many corners in performing the port of the benchmark system also got cut.

Regardless of which of those two answers is true, I consider them both to be a bad sign. Unlike the rest of the game where the game needs to handle arbitrary interaction with the players, a good benchmarking system is automatic and should be the easiest thing to port. Also, a good benchmarking system can help with the quality assurance in giving feedback on the performance and provide an automated way to look for bugs in rendering the game on different hardware. Lastly, it provides another tool to help technical support by allowing them to request objective data on how the game is performing on an user's system. For example, a benchmark system may show that for hardware X, shader Y has a performance issue which is larger than expected but on shader Z which was thought to be a possible issue actually is performing fine.

So, in general, the gutting of the benchmarking system in the port is not just an issue for Phoronix but should be taken as an issue for all gamers that can't possibly evaluate 35GB of an 18 mission game in less than 2 hours.

If anyone is abusing the refund policy of Steam, it is SEGA and Feral Interactive. They are releasing a poor port and leaving out a critical tool for customers to reasonably determine the quality of the port and a critical tool in getting good technical support. Shame on Feral for pulling this crap and shame on SEGA for letting them release it under the SEGA name.

[bjoswald]

Why is this crap still happening in 2015? Where is the QA department? Did one guy port this over the weekend, by himself, from his home PC? To think that AAA games like this get released with these issues and 60% performance (than Windows) is insulting.

[adakite]

While some people seem to complain about the port quality of Alien:Isolation on Linux, I want to say that I'm really happy of the game. It runs smoothly on my Fedora 21, high FPS with all graphic options setup on the highest level. I have 8GB of Mem, GTX Titan Black (using NV's blob) and a dual socket 6-core Xeon based workstation (with a low CPU freq. though). I realize that this is not a standard PC for gamers, but the Linux port on this beast runs really nicely and on a similar FPS than what I have on Windows 7 (using dual boot on this very machine). My only concern is about VR. While I saw some videos on people playing A:I with an occulus (so I guess under Windows), I wonder if I can do the same with the Linux port.
As for the benchmark mode missing, maybe it was not part of the deal, simply. There's many other ways to check/compare the port with the "native" windows version. As I said, on my computer, I do not see evidence for any bad performance on Linux compared to W7.

[tomtomme]

Well, to be fair, Feral did put out an update for Shadow of Mordor - fixed a problem with AMD Phenom CPUs. In addition, I had a problem with the "WBPlay" integration, and they ran a beta. Not sure if that patch has been pushed out to mainline yet.

yeah, but normal bug-fixing is not what I meant. They certainly are fixing severe bugs, like bugs that stop the game from starting at all.
What I meant is they do not fix performance.

[kaprikawn]

Why is this crap still happening in 2015? Where is the QA department? Did one guy port this over the weekend, by himself, from his home PC? To think that AAA games like this get released with these issues and 60% performance (than Windows) is insulting.

Calm down, you can't pinpoint the issues. It could be that the Windows version is super-optimised (see http://www.nvidia.com/download/drive...px/77837/en-us for example, the Windows version has optimisations in the driver), it could be that the Linux driver is bad. Or it could be the quality of the port.

What's likely is that it's some sort of combination.

[WOLF308]

I'm beginning to regret giving money to this site.

[mike4]

Alien isolation runs great with my GTX770. Also very nice graphics indeed, well no wonder for a Game of the year 2014.