News | Forum | People | FAQ | Links | Search | Register | Log in
Mark V - Release 1.00

* Nehahra support -- better and deeper link
* Mirror support, "mirror_" textures. video
* Quaddicted install via console (i.e. "install travail")
* Full external texture support DP naming convention
* Enhanced dev tools texturepointer video inspector video
* IPv6 support, enhanced server capabilities
* Enhance co-operative play (excels at this!)
* Software renderer version (WinQuake)
* "Find" information command (ex. type "find sky")

Thanks to the beta testers! NightFright, fifth, spy, gunter, pulsar, johnny law, dwere, qmaster, mfx, icaro, kinn, adib, onetruepurple, railmccoy

And thanks to the other developers who actively provided advice or assistance: Spike (!), mh, ericw, metlslime and the sw guys: mankrip and qbism.

/Mac version is not current yet ...; Linux will happen sometime in 2017
First | Previous | Next | Last
@johnny - After Typing That Reply I See ... 
Well .. I'll test those examples tomorrow and do a version 1.98b.

I took a diverse pool of zips with many different configurations, but apparently a different have a diverse enough set in the pool.

Perfect is the goal. So ... perfect it shall be. And perfect is what I demand it must be be ...

Hah, perfect isn't required. :-) I will just be doing a test run through the examples I listed earlier. 
Actually, when the next build is up, I'll go through all of the 50+ installs that the SP starter pack currently supports. Not sure it will happen this weekend because of family stuff, but as soon as possible. 
Alright ... qump (what an incredible start map!) and mapjam6 now unpack ok.

1.98 updated

Only that binary has been updated while I kick tires on a couple things that don't affect the windows operating system. 
Hm, the mouse speed seems dramatically lower in the new build. I've always used a "sensitivity 10" setting, but now I need to set it to 20 to have the same feel (and the slider only goes to 11).

The SDL version seems... weird.... Text and other colors are off, and the sound it tinny. But it has the same freeze-up delay of around 5 seconds when I enable the music, as it loads track04.mp3 (and this was after encoding it in various ways to get that delay down from the original 14 seconds). but if I use an ogg renamed as an MP3, the delay is less than 1 second....

I don't remember all the testing we did in the past (it's all on the oldpage), but I think FTE and Quakespasm had no delay at all for loading MP3s for me....

Question: would the split-screen players be able to connect to an online server? 
1) Mouse now uses DirectInput so sensitivity is a bit different.

According to the research I've done, Windows updates in recent years introduced some behaviors that could cause "frame lagging" with the "ancient way" that Quake did mouse input (which involved recentering the mouse cursor every frame), hence the change to DirectInput which I believe will eliminate the frame lag some users say they experience when using, for instance, a high refresh rate.

Aside from having to tweak the sensitivity.

2) The SDL build uses 44100 as a forced sound frequency because of some sort of SDL2 oddness. Hence the much more crispier sound. I prefer the way that Quake traditional sounds myself but Mark V has a sndspeed cvar that can be set to 11025, 22050 or 44100 and is read on launch. (Has no affect in that SDL2 build, 44100 must be forced).

3) Too bad the mp3 has a delay on your old Win XP computer even for the build that uses a completely different Windows mechanism. I had hoped that on your machine that it would resolve the issue. I tried. I know reformatting that Win XP isn't an option.

4) Split-screen connecting to an online server ...

Well ... I wasn't planning on that.

If I did have it do that, would you be able to test it?

But here is a big question: I am planning to target exclusively the Direct X 9 for that feature. I have to really abuse Windows API to do split screen and SDL2 serves as middleware and isn't going to let me use the abusive Windows API evil that I do for split-screen. 
Ah yes, I remember back in the day I would use the -dinput switch, and that would have a similar effect of requiring me to ramp up my sensitivity. I think this does make my trackpad movement look very slightly smoother too. But you might consider allowing the Sensitivity slider to go up to 22 or so, since this will likely have a similar effect on other people.

Yeah, the XP MP3 issue is weird. On my other (even older) XP laptop, I recall I did't have the issue. But I also remember seeing a post on Quakeone where Dutch said Mark V has "a difficult time playing external MP3" on his WinXP computer.... I don't think he ever gave additional details about that, but I'd very likely assume it was the same issue.

Oh, I would certainly give it a test if split-screen could connect to an online server. Hm. But there are all kinds of issues around that, like if someone wanted to chat on the keyboard, which player's guy would speak? I'm guessing there would only be one "master" player which was attached to the keyboard, and others would be limited to a certain gamepad? 
Adjust slider range to reflect? Excellent idea, gunter! 
Gunter, Re:Split Screen 
I'm guessing there would only be one "master" player which was attached to the keyboard.

That is correct. Only one player can keyboard.

I once considered that I think Spike said RawInput can provide access to 2 keyboards. And perhaps someday I'll consider that route. But I don't feel like coding for such a situation right now.

Multiple keyboard support is probably very low priority. 
Add: Iriyap List + Parubaru 
Revisiting found the comprehensive list that iriyap wrote up that had some ideas/suggestions for improvement (self note it is post #1805 -- I keep losing track of it)

In particular, I have my eye on the few seconds of lag issue on video restart that affects some machines.

I know how this was introduced. I was actually trying to stop a sound jutter, and while fixing it on some machines indirectly caused this to happen on others. 
Install testing is looking pretty good -- I got about halfway through, might be able to finish it off tomorrow or Monday.

A minor feature request for your consideration: an optional argument to the install command to indicate "please put this in its own mod directory" (regardless of whether it could work in id1). Like, if I'm installing terra and I want to make sure it ends up in its own "terra" folder. 
Small interesting thing when installing func_mapjam5: there's a zerstörer.wad inside the zipfile, but it gets extracted as zerstörer.wad (let's see if that ends up showing correctly in the post here).

That's just in the "source" directory though so it doesn't matter functionally. Something to keep an eye out for though I guess. 
I can add an argument to allow you to do that. Maybe even one to force a folder (like a Quoth set). 
Cool. FYI the install tests all look good now. :thumbsupemoji: 
@Baker Arrow Key Bug 
I've experienced this both in Quakedroid and now in 1098. I don't think it's my hardware in this case. Here's an unlisted tube showing the issue. 
Meant To Add 
That a reboot fixed this issue but it's still bizarre. 
I'll certainly look into it.

And if you have anything that I have missed, please let me know.

I've had such a large list of touch-up or in some cases rewrites (like what johnny + I are discussing) that keeping a handle on every single thing has been tough. 
I think you need to unplug your PS4 controller.

PS4 doesn't using the same analog axis mapping as Xbox, it successfully detects your controller and then thinks you are holding a stick up.

/Educated guess.

I do plan to extend Mark V controller support beyond the currently supporting Xbox mappings, Mark V uses native Windows API so I have code this myself (which isn't a problem) but I haven't code such yet (and haven't obtained a PS4 controller yet). 
Shadows "3" 
I can confirm it's working again in the latest build with improved models. Thanks a lot, Baker! ^^ 
haha That's (almost) exactly what it was. I had forgotten I plugged in a PS3 controller to charge via PC the night before. When I read your post I looked down near my PC and there it was still plugged in. Fixed. Thanks! 
@NightFright - ;-)

@dumptruck_ds - Until I get my hands on a non-Xbox controller like a PS4 one, I'm going to force non-Xbox controllers to be ignored (if I can, and I believe I can). In theory, I could code the axis/button remapping without having a controller, but I know from practice that "on the blind" coding usually is a quagmire.

Eventually, I'll get my hands on a PS4 controller ... but I'd like to ask you this.

Is getting a PS4 controller to work with Windows as easy as plugging a USB cord in (it is that simple with the Xbox One Controller S, as an example). 
@iriyap - Your List 
I looked through your list.

First 2 items resolved. #1 already fixed. #2 - sound stutter on vid_restart that occurs on some machines (I've witnessed it first hand on a couple) should be fixed.

I looked at the sky in GLQuake vs. Mark V vs. Quakespasm with gamma 1 and contrast 1 which shouldn't be setting neutral.

The sky does indeed not look quite the same as GLQuake! But ironically appears to look about the same as WinQuake.

So now, I'm not quite sure what to think. WinQuake is kind of the baseline that GLQuake was supposed to look like. Still, maybe an option will happen.

Other things: tga fonts + alpha masking for replacement fonts. Older replacement fonts supported FuhQuake/JoeQuake which has less flexible rules. The newer ones tend to be oriented towards DarkPlaces which has different rules regarding the alpha masking (more proper rules, I would say).

The main issue that has caused me to do nothing is that the replacement fonts for DarkPlaces don't tend to be a 1:1 drop-in replacement but rather really need larger resolutions or they look terrible in resolutions like 640x480.

I had once drafted up a solution been eyeing a solution which is to determine the minimum scaling when loading the replacement font. But the fact I didn't do it bothers me --- I may have run into an issue like HUD replacements and font replacements might require the same base scale under certain circumstances.

WinQuake: A 16-bit renderer would be a massive undertaking. Mankrip lives and breathes the software renderer. My level of interest in the software renderer is very close approproximation to the original WinQuake feature set + modern maps. I would eventually like to get alpha masked textures working (I've made several runs at it). As for HUD scaling, FitzQuake has a ton of canvases and scales all over the place. Mark V's software renderer isn't actually based on WinQuake, especially not the placement and canvasing. A dedicated WinQuake-only engine like Engoo or super8 need not worry about canvasing consistency with a Direct3D/OpenGL version, but one of the main features of Mark V being able to switch between the WinQuake/hardware versions on a whim and get nearly the same experience. Re-writing all that canvasing would take far too much time, especially when the less capable but fairly accurate approximation "stretch mode" is available in video options.

Also looked at bloom. Then realized that at the moment I cannot bloom to work in the Direct3D build but I believe it is fault of something on my end (as opposed to the wrapper). I'll probably look at in the future.

Likewise, multisample isn't supported in the Direct3D version at all and to some extent I'm not 100% that multisample is relevant these days because simply using a higher resolution with a larger scaling factor typically gets approxiamately basically the same results. Multisample with modern very high resolution monitors on 3D renderering is going to be mighty tough to notice, but on older hardware like 640x480 crt modes yeah you can easily tell.

The Direct3D 9 wrapper's level of performance (rendering speed) and stability (AFAIK not even a botched NVIDIA update can mess with it, so it may be totally immune to those kinds of problems). 
Direct 3D supports multisampling but the details are interesting.

The OpenGL philosophy is "the programmer can go nuts and the driver will do it's best to sort out something that works".

The Direct 3D philosophy is "the programmer absolutely positively cannot go nuts, and that sorting out something that works business is on your shoulders, matey".

In a native D3D application this is actually easy. You just enumerate the supported formats and follow the rules.

In an emulation layer that has to translate from the GL spec (and oh lordie if it's pre-3.0 stuff) to the D3D rules the path of least resistance is to just not advertise the extension.

Sure it can be done. But time, patience and sanity are finite resources round my way. 
I have yet to actually game with a PS4 controller on a PC. Or a PS3 for that matter. When I plugged in the PS3 controller via USB drivers were installed but I don't think they were with the PS4. I do know Steam supports PS4 and you can use them with Sony's streaming setup. I have zero experience with Bluetooth on my PC so that may be the ticket. No idea and doubtful.

So using those controllers is NOT plug and play. That makes sense to me as Xbox controllers are native for the most part in Windows and Sony interfacing with PCs has been historically clunky - usually their software is pretty wonky there. 
The following will be available for the installer. Needs run through your series of tests, I'm rather confident it will pass them all.

1. Forced to specific gamedir:

install elek_neh_episode4 nehahra // a Nehahra expansion
install thehand quoth // A Quoth map
install qt_pre02 travail // A Travail expansion


2. Special "self" keyword

install bnt self // Maps only release forced to its own gamedir quake/bnt

@dumptruck - Thanks for info, trying to plan out how long I need the controller and what could go wrong. 
First | Previous | Next | Last
You must be logged in to post in this thread.
Website copyright © 2002-2022 John Fitzgibbons. All posts are copyright their respective authors.