Platform Discussion Thread

+

Which API do you think CP 2077 will use?


  • Total voters
    135
It's the only thing that's cost effective.
You really think training a few dozen in-house programmers in multi-platform programming and the extra time that will be required for any/all programming once it's implemented is more cost/time-efficient then an OS port?
If that were true then every game maker in the world would be doing so because it would increase their profit margin.
 
You really think training a few dozen in-house programmers in multi-platform programming and the extra time that will be required for any/all programming once it's implemented is more cost/time-efficient then an OS port?

Yes, I already explained above why. Because it accumulates expertise in house and allows them to develop the engine properly. That's what Epic, Oxide and other professional developers do who develop their games with their own engines, and who are interested in their engines in the long term.

Anything else will produce poor quality half baked "port" which they can't even support themselves because they have no clue how to. It's not just poor taste, it's not moving their engine forward. Investing in their own engine with in-house expertise is cost effective. But of course if you are coming from mentality of "develop and throw out into the garbage", then sure, everything should be done by temporary subcontractors. It would be cheaper. Quality will be lower as well, and it will have one time use.

So, do you think CDPR are the former, or the later?
 
Last edited:
I think they have a small, but rapidly becoming larger, game development company using a propitiatory engine. I'm sure they have plans for the future that certainly will include some amount of in-house multi-platform support.
Do I think they're going to go the the lengths you propose while doing so.
No.

There are innumerable ideal visions of how things should work. Unfortunately far too often those visions aren't tempered with a cold calculated look at reality and how things do work, and companies that could work wonders, if their visions could possibly be realized, go under. And I don't think CDPR is going to fall into that trap.

Dig around long enough and anyone can find a case or two to "prove" their pet theory. But in the big picture one or two cases are just that ... one or two cases ... they don't validate the theory. Now if it can be applied in-mass, that's validation!
 
Last edited:
As I said, those who are successful - accumulate expertise in-house. It's old and proven wisdom. Outsourcing never pays off. It looks like cost saving, but it's not. And it's about long term / short term approaches as well.
 
Last edited:
You should use AMD Vulcan for Cyberpunk 2077.
In this case every gamer even with casual PC without having a 700$ graphics card inside plus gamers who hate every Windows after Windows 7 can get 9999 frames with photorealistic CGI-graphics and everything maxed out, and no gamer is forced to upgrade to Windows 10 for the newest lighting and texture filter effects in DX12.

Yes I am missing lots of lighting and fire effects and other DX12 effects in all my games, and I think "hmmmm all the games look a lot different in all the Youtube videos than on my computer even with every graphics setting maxed out" but I am NOT and NEVER upgrading to Windows 10 - truth is I made an UPGRADE from Windows 10 to Windows 7 after buying this computer because I almost puked after seeing W10 running when switching it on for the first time, and NOTHING worked from graphics chip to mobile Internetstick and USB-3.0 Slots - and no drivers avaiable....

After the upgrade to Windows 7 everything worked fine :)
 
Can't you run some kind of virtual machine throught Linux to run the game with windows?
(I don't know much about Linux, just aking)
 
Maelcom404;n6974810 said:
Can't you run some kind of virtual machine throught Linux to run the game with windows?
(I don't know much about Linux, just aking)

You theoretically can. But I'm not interested in using Windows for that purpose. Not just because I'd need to pay for it and money would go to MS, but because Windows itself is something I oppose for several reasons. And also because it wouldn't show a demand for the game to come out on Linux.
 
Last edited:
RepHope;n7036430 said:
Microsoft helped with marketing. But they didn't not port W3 to Linux because MS told them not to. They didn't port to Linux because it isn't worth the time or money. I recall reading a developer talking about how they ported one of their games to Linux due to fans petitioning for it, only to find that the actual amount of people who bought and played it using Linux was waaaaaaay smaller than people had claimed.

Face it mate, you're not getting any future CDPR games on Linux. It's not happening. I'm going to go try to find that interview, been a while.

Edit: Found it https://www.reddit.com/r/linux/comme...nough/c89sfto/
CDPR is a "big dog" now. They're not going to go through the hassle of a Linux port for $20k worth of sales. Since the "Year of Linux" was a bust, I just don't see it happening mate. Sorry.

That's an old rant from John Carmack which was debunked by companies which actually release Linux games, so it has nothing to do with CDPR. Not only it was wrong way back then, it's irrelevant today, since situation changed from 3 years ago. I.e. Linux gaming only grew since then. Whether CDPR are big dogs or not, has nothing to do with that rant. If you are pointing to big publishers like EA and etc. then they are a poor example. They don't support Linux not because there isn't a market for Linux games, but because they are legacy publishers. They don't care about reaching more users and making profit, if they can make sufficient profit excluding those users. They can, so they do it. Smaller studios for which the risk of releasing for Linux should be higher (if according to you the market isn't big enough), are releasing games for Linux much more often. So that disproves your (and Carmack's) argument that they don't release it because of lack of demand. Demand is here.

Marcin Iwiński himself said in the past (when Linux market was actually smaller than today, and even before they started supporting Linux on GOG):

If Steam will deliver a constant Linux environment, call it SteamOS or anything like that, we would love to have our game there, because the more people play our games, the better for us.

That is available today. The reason they failed with TW3 wasn't because there was no market or anything the like. They were planning on doing it until very recently. The reason was rather poor planning and a ton of technical issues they hit along the way, after they realized that it's a non trivial effort to do it post factum. I.e. they should be able to learn from those failures and avoid them with CP2077, if their voiced intentions about Linux support are genuine.

I'd agree with you however if CDPR became "big dogs" in the sense of backwards thinking legacy publisher mindset. If that's the case, I'll stop paying attention to them, same as to EA and others of their ilk. But until at least recently, CDPR publicly claimed that they don't want to adopt that mindset (see their presentations to shareholders). I guess we'll have to see, whether they'll be true to their word with CP2077 or not. No CP2077 for Linux - no more CDPR for me.
 
Last edited:
From Wikipedia:
" In 2010, in the first Humble Bundle sales, Linux accounted for 18%.[SUP][42][/SUP] The Steam client tracks usage and reported in March 2014 that less than 1.2% of users are using some form of Linux as their platforms primary operating system,[SUP][43][/SUP] in May 2015 it was below 1%.[SUP][44][/SUP]
The Unity game engine gathers user statistics and showed in March 2016 0.4% Linux users"

Yeah, not a big platform. Less than 1% of a million sales is less than 10,000 sales. That's less than US$600,000. 10 million sales, that's less than $6 million gross. Worth it? Depends on the effort required and support offered. Most businesses wouldn't go to too much effort for less than 1% of their market base - smarter to spend the money supporting/enhancing the 99% of the rest of the base.

Doesn't look too good for CP2077 on Linux.
 
Sardukhar: Steam numbers don't show sales potential, it was pointed out many times, so those numbers aren't useful (as you can see HB numbers are very different). Steam doesn't publish sales statistics which you could bring as an input for analysis. Unity numbers are also pretty useless, because they don't differentiate them for cross platform and non cross platform games. Or to put it simply, a lot of Unity games come out for Windows only, so those numbers are too skewed.

If you want to analyze sales potential on real case scenarios, you'd need to take some multi-platform game, and then figure out usage per platform. Preferably one that was released for all platforms at once, and not Windows first and Linux 3-4 years later. There isn't much of such stats floating around, and from ones that do, Linux can be higher than 2% (HB is one of such examples).

So again, statistics are a tricky subject, since you can twist numbers in a lot of ways, if you don't pay attention to the context. GOG for example could publish usage per platform (since sales per platform doesn't really make sense for DRM-free games, since you buy all versions at once). They for instance can analyze downloads per OS. Unfortunately they don't publish anything of the sort. The bottom line, so far I didn't see any comprehensive proof which shows that Linux sales potential is too low to release Linux games. And in practice more and more companies do it today, which shows that the market is good enough and growing.

But going back to the main subject. CDPR never said they had issues with Linux market size. They released TW2 for Linux when it was even smaller. And they expressed interest in releasing TW3 back then as well. So the size isn't the issue for them, and I wouldn't worry about CP2077 from the market size perspective.

Lock-in into MS mentality however is a different matter and can be a practical barrier which management should be willing to overcome if they care. I.e. if all their developers are too DX oriented, releasing anything for Linux would be a challenge, and would end up in outsourcing, which in their experience can be rather risky (remember their past outsourcing failures).
 
Last edited:
Gilrond-i-Virdan;n7037380 said:
Sardukhar: Steam numbers don't show sales potential, it was pointed out many times, so those numbers aren't useful (as you can see HB numbers are very different). They released TW2 for Linux when it was even smaller. And they expressed interest in releasing TW3 back then as well. So the size isn't the issue for them, and I wouldn't worry about CP2077 from the market size perspective.

We disagree - HB market is much smaller than Steam. The Steam numbers do have value and, yes, W2 was released on Linux and W3 wasn't. Sure it might be later or sure it might be a technical issue but you know what?

It's quite likely the reason is what it appears to be: the numbers don't justify the effort. Sometimes a stone lion is just a stone lion.

The size and investment/reward ratio of the Linux market are absolutely an issue for CDPR - they are professionals. It would be the height of arrogance to disregard or ignore those issues when considering a port. They port to platforms with a big user base. Linux does not have a large user base for gaming and Linux users can and do run Windows on their boxes.

Here's a thing: is there any argument or data that would convince you Linux gaming isn't worth the investment or that the Linux gaming market -is- less than 1%? Any kind of data you wouldn't wriggle around or try to re-interpret? Because if not, it's not the argument or the data that is unclear.

You're pretty close to the subject - so heavily into it that you refuse to even run Windows on a Linux box - and that makes you pretty prejudiced.

I'm fine with a Linux release, but, yeah, Linux gaming is a tiny tiny market segment. See that clearly and accept it and all the games -not- made for Linux suddenly make sense.

 
Sardukhar;n7037480 said:
Here's a thing: is there any argument or data that would convince you Linux gaming isn't worth the investment or that the Linux gaming market -is- less than 1%? Any kind of data you wouldn't wriggle around or try to re-interpret?

There isn't such data and you know it. The most you can argue is, that releasing games for Windows only is a safe bet (massive user base), and releasing games for Linux is a lower certainty of success. That's true, no one argues with that. But as I said, despite some avoiding Linux gaming by sticking to what they are used to, the increasing number of companies release games for Linux today (recent Tyranny by Obsidian is an example). Don't tell me they aren't professionals and don't analyze what they are doing. I.e. they clearly consider it worth the effort.

Sardukhar;n7037480 said:
TW2 was released on Linux and W3 wasn't. Sure it might be later or sure it might be a technical issue but you know what?
It's quite likely the reason is what it appears to be: the numbers don't justify the effort.

Regarding CDPR it's hard to argue that they consider market size a problem in this case. As I said, they released TW2 for Linux when it was even smaller, and then could use TW2 Linux usage to analyze future sales potential. Clearly they saw those as worthy, since they said they were working on TW3 Linux release quite a long time after TW2 for Linux came out. So according to my understanding, they just were caught up in technical difficulties which borked the effort.

The most likely scenario was like this. When TW3 was released, they decided to repeat the pattern they used with TW2. I.e. since they didn't work on the Linux version from the start, and since it was a throwaway engine, instead of investing into in-house engine rewrite they turned to same VirtualProgramming, or some other porting folks for help. They said they'll evaluate it. They probably gave them an estimate and started working even, since that was going on for around a year or so. Along the way they hit some major technical issues, and ended up being over budget. So CDPR decided to halt the effort. It's not that they weren't interested, it just became too difficult because they didn't spend needed effort in advance and got tangled in lock-in too much. If you'd argue that it wasn't the reason, then you'd need to call their claims about working on the Linux version a lie.

If we assume the above is correct (which I assume is, I don't see a reason to assume they were lying), then they have a good opportunity to avoid such pitfalls with CP2077, i.e. they can make the cost of the Linux version reasonable, rather than unbearable by doing right things at the right time and not when it's too late already.

Sardukhar;n7037480 said:
Linux does not have a large user base for gaming and Linux users can and do run Windows on their boxes.

That's incorrect. Same argument can be applied to consoles and by using it you could claim there is no point for CDPR to release console versions since why wouldn't console users run Windows somewhere else as well? But there are many who don't because they prefer consoles. Same with Linux. One of the major reasons to use Linux for gaming is avoiding Windows, so that logic above is pretty faulty.
 
Last edited:
Gilrond-i-Virdan;n7037510 said:
There isn't such data and you know it.

Gilrond. Come on. How could I know there was no such data? I just showed you some pretty indicative data and you ignored it and talked about "sales potential". But even so, how could I know there was no such data showing that Linux is less than 1% of the gaming market? I mean, there might be or might not be. My issue is, you would never, ever believe it, no matter the evidence.

Gilrond-i-Virdan;n7037510 said:
Regarding CDPR it's hard to argue that they consider market size a problem in this case. As I said, they released TW2 for Linux when it was even smaller, .

Yes and didn't release W3 on Linux. Pretty clear they didn't think the W2 release was worth it, post facto. If it had been, you'd have played a W3 release on Linux already. They may, because they are really nice, release it for Linux eventually. However, it was rejected for effort when they needed all the money to cover W3 costs and I can tell you, that says everything to a businessman.

It's too bad but that's how it works. It's really too bad you're going to cut off your nose to spite your face and insist on only playing W3 on Linux. Buy it for both, if it ever comes out. Same for CP2077.

Not buying it doesn't really convince CDPR of the market in Linux. Buying it gives them more money for less fiscally sound choices than the usual Windows, XBox and Sony crap.

Not that I think PC and consoles are crap but, yes, I agree that Windows in many ways sucks. And sucks hard. And lets not talk about the limitations and DRM issues of consoles. Because Ugh.


Edit: also, your consoles argument? No, Gilrond, just no. It's not the same argument. That you say that says a lot about your disconnect with console users and why CDPR has time for the console market and not Linux or Mac OS.

People use consoles because they like sitting on their couch with their controller and not sitting at a desk. Also because console UI is massively simpler? Flavourful? than the PC process. Also consoles are kind of cool looking. Oh! And you can pick a team, people like that.

The Linux and Windows at-the-desk gaming experience is pretty similar. Mouse, keyboard, monitor, speakers, etc.

And that's not even talking about the price-point ratio. Consoles are waaaaaaay cheaper and your giant TV that you use to watch hockey works just fine with them.



 
Last edited:
Sardukhar: Again, you missed the point. May be a timeline will help to clear things up.

1. TW2 for Linux came out in May 2014.
2. SteamOS ad about TW3 appeared in June 2014.
3. Second SteamOS about TW3 appeared in March 2015.
4. CPDR said they are working on the Linux version of TW3 (same March 2015).
5. TW3 came out for Windows in May 2015.
6. CDPR again said they are working on the Linux version of TW3 (August 2015).
7. CDPR said they are "looking" at the Linux version of TW3 (March 2016).
8. CDPR said they are "exploring" the Linux version of TW3 (June 2016).

Let's analyze. There was a year+ between #1 and #3-#6. So they clearly didn't see the market as unworthy in 2015. In the period between #6 and #7 things turned sour, and working changed to "looking" and "exploring", which indicated some problems. I can only assume they hit technical issues, since if according to you they have a gripe about market size in general (let's say based on poor sales of TW2), #3, #4, #6 wouldn't have happened to begin with.
 
Last edited:
Gilrond-i-Virdan;n7037720 said:
Sardukhar: Again, you missed the point. .


Yeah, not really. The point is, CDPR decided that the investment into Linux wasn't worth it. All your timeline does is put that into clearer detail. Either because of technical issues ( which they could solve if there was more money there), because W2 on Linux just didn't sell what they'd hoped or because, flat out, further study showed the time wasn't worth the reward.

There is really no way to wriggle off this hook: there is no Witcher 3 for Linux because it's not worth it to CDPR. And that means not great chances of Linux for CPunk, UNLESS CDPR is feeling really nice or something changes in their original evaluations.

But if time and/or resources are tight, Linux will get chopped.

As for your "they clearly didn't see the market as unworthy in 2015" that's clear only to you, because you want to believe it. It's much more likely that they were keeping Linux as an option and waiting on cost/reward numbers. It's much more likely that was what happened because there is no Witcher 3 for Linux at release or a year after.

You can be working on a Linux version by making three calls to your software guys and asking them about Linux, you know that, right? That doesn't -mean- anything, but it does look good PR wise. Same as a SteamOS ad.
 
Even if those were just "three calls", they still said they were working on it, which means they considered it worthy at that point. Otherwise it would have been "we are looking at the possibility" etc. Their wording was very precise, same as it clearly changed later. I.e. market size back then (assuming TW2 input) was considered OK. But as you said, pay off depends on expenses. If it would take for VP or whoever too long to do the work, the risk becomes higher. As I said, the issue was the post factum work and their whole DX11 dependency mess. So you can't project their pitfall there to CP2077, because they are making the engine from scratch. They can avoid that pitfall and excessive cost by proper design from the start.
 
Top Bottom