The Star Citizen Thread v5

Status
Thread Closed: Not open for further replies.
But surely, you cannot expect someone to adopt such an obscure, inefficient, and proprietary bit of code as rsync! Be sensible!

e: also, unreliable and definitely not available on any useful platform!

Not to mention, you can't ever get the sourcecode for it. Ever. You *have* to reinvent the wheel ;)
 
This is the same idiot who came up with this nugget of information.

https://clips.twitch.tv/grayheadedgamer/QuaintAardvarkVolcania

He's right, though, SC is nowhere near close to Infinite Warfare, just not in the way he's implying. But I like the notion of "real, physicalised physics". As if SC, with its countless examples of janky dynamics and hopeless collision detection (as demonstrated by the same guy, confusingly), has more correct physics than other games. Chris Roberts would be proud.
 
Found that:
Falling over from gravity change and standing up sometimes has your "facemask" jump away from you which makes it really obvious that its an overlay.

Do I understand it correctly, and edges of the helmet viewed from first person perspective are just like Elite's Remlok, and they're a static piece of 2d graphics? If so, was the third person HUD that was a focus of a discussion some time ago something of a hack, not a great example of fidelity?
 
Last edited:
Found that:


Do I understand it correctly, and edges of the helmet viewed from first person perspective are just like Elite's Remlok, and they're a static piece of 2d graphics? If so, was the third person HUD that was a focus of a discussion some time ago something of a hack, not a great example of fidelity?
Nah, the helmet's definitely an in-world object, otherwise it wouldn't do depth of field (and have so many bloody depth of field bugs). The need for it to be a 2D overlay is essentially a Mac Problem.
 
Well, it's after 4 PM there (https://www.timeanddate.com/time/zones/pst), but perhaps there's still a chance. Still, I don't see them releasing it before 26th, unless their management is way more scummy than I think it is and there will be crunch so close (or even during) Christmas.

Are there any new info from status report?

Judging by information and thread member reporting, it is not ready for prime time. Still they can do tomorrow...or 26th.

Anyways, back to trying getting ED running under Wine.
 
@Zyll Goliath - so simple answer yeah no matter what atm you will have to download all game again. :(

Thx I guess that´s expected from CIG.....they just can´t do such a simple task and if I remember well I think they told us that with 2.5 we get improved launcher???
 
Last edited:
This is not just physics, its super-fidelity fysics, I tells ya. For, a mere touch-interaction will send a heavy object spinning and hurtling into the cosmos, you dont get that in boring real life do you eh?

Sold to the man with the beard :

[video=youtube;lyfw1R86rCI]https://www.youtube.com/watch?v=lyfw1R86rCI[/video]
 
Thx I guess that´s expected from CIG.....they just can´t do such a simple task and if I remember well I think they told us that with 2.5 we get improved launcher???

I am going into rumor/vague memory territory here, but wasn't patch updating problematic due of how CryEngine prepared these 'games'? Something about encryption etc.?

It might be not so easily solvable as we might imagine.
 
I am going into rumor/vague memory territory here, but wasn't patch updating problematic due of how CryEngine prepared these 'games'? Something about encryption etc.?

It might be not so easily solvable as we might imagine.

Other games that are build in CE solved that so it´s can´t be that hard.....These days I been playing Warface so that I can compare it with Star Marine when it comes out...and guess what I had few regular updates for past 3-4 weeks....no issue at all or need to download a whole game all over again....
 
Last edited:
Nah, the helmet's definitely an in-world object, otherwise it wouldn't do depth of field (and have so many bloody depth of field bugs). The need for it to be a 2D overlay is essentially a Mac Problem.
Why's it a mac problem? I'm confused.

Also are you on holiday yet? Hope you booked it early!
 
I am going into rumor/vague memory territory here, but wasn't patch updating problematic due of how CryEngine prepared these 'games'? Something about encryption etc.?

It might be not so easily solvable as we might imagine.

All due respect Eagle it's easily solved, I'm not sure why they haven't done it.
 
Why's it a mac problem? I'm confused.

Also are you on holiday yet? Hope you booked it early!
Yeah, been off all week.
Reason it's a Mac Problem: Depth precision, and more specifically near-clipping distance. Traditionally, there was a rule of thumb that your near clip distance needs to be as far out as possible, as it's the main thing that affects Z-precision. Because of the equation used, if your near clip is 10cm away, you use up 50% of your number range over the next 10cm, 25% over the next 20cm, and so on (this isn't exactly right but with a big draw distance it's close enough).
More recently, some clever person realised that if you use floating point depth, and use 1=near, 0=far instead of the traditional 0=near, 1=far, the increasing precision of floating point close to zero almost perfectly balances the other effect (the precision ends up kind of a wiggly line but it hovers around the ideal value). Depth precision issues solved, basically, with no downsides, and you can move the near clip distance to 1mm or whatever, set far clip to infinity, and no one cares. At that point you can put a helmet over the camera without worrying that a slightly wider monitor will cut holes in it.
Ok, so there's one downside: OpenGL traditionally uses -1=near, +1=far, which puts the extra precision in the middle even if you reverse it. To benefit from this new trick, an extension was added to make it clip at 0 and 1, but Macs are locked to a version of OpenGL that doesn't have that feature. IIRC it might also not have floating point depth buffers either. Pain in the backside really.
 
I am going into rumor/vague memory territory here, but wasn't patch updating problematic due of how CryEngine prepared these 'games'? Something about encryption etc.?

It might be not so easily solvable as we might imagine.

It's not a CE problem — it's a CIG problem. They suffer from the illusion that they must sign their PAK files to prevent… something, and that the patcher cannot do this locally because… reasons. It's actually really trivial to do — they're just zip files with a couple of extra headers that could be trivially manipulated by any of a bajillion different third-party tools or existing libraries if they wanted to.

…and that's assuming you even use PAK files to begin with, which is completely optional, and actually not even a particularly good idea while in testing. CE will happily accept a raw directory tree as a source for its asset files — PAKs are just a convenient way of delivering all of that in a single package on-disc or through a download service once the final version is ready and you don't expect the files to change all that often.

So it's only problematic because CIG is paranoid for all the silliest reasons.
 
Last edited:
It's not a CE problem — it's a CIG problem. They suffer from the illusion that they must sign their PAK files to prevent… something, and that the patcher cannot do this locally because… reasons. It's actually really trivial to do — they're just zip files with a couple of extra headers that could be trivially manipulated by any of a bajillion different third-party tools or existing libraries if they wanted to.

…and that's assuming you even use PAK files to begin with, which is completely optional, and actually not even a particularly good idea while in testing. CE will happily accept a raw directory tree as a source for its asset files — PAKs are just a convenient way of delivering all of that in a single package on-disc or through a download service once the final version is ready and you don't expect the files to change all that often.

So it's only problematic because CIG is paranoid for all the silliest reasons.
I don't know if it would actually be an issue here, but there could be a downside to using a raw directory tree - loading many files that way can end up hellishly slow because all the requests have to go through Windows directory management. I remember I-War invented a "slab" file format that was just all the files uncompressed in a big lump, as a way around that issue.

- - - Updated - - -

I didn't realise they were going for a Mac version Ben?
Hence why the helmet is a 3D model :)
It's a 2D overlay in Elite, which does have a Mac version to support.
 
Status
Thread Closed: Not open for further replies.
Back
Top Bottom