 News | Forum | People | FAQ | Links | Search | Register |
Log in |
|
|
From what I know (unless I mix up things) Quake demos only include the entities in the PVS so a feature like that would require a whole different protocol.
As the demo progresses the engine is updating entities with new locations, states (moving, alive, dead etc) and when the map is frozen I just want to fly around and see where all the entities are and what state they are in. Even the current PVS would be better than nothing.
Qrack does this today and is the only engine I know of that allows this. I don't have any experience personally using the feature recently, I'm just pointing you to what can satisfy this desire today.
[I still have no time to work on Mark V and don't see this changing in the known future Daz's recent CG video revealed another Mark5 bug: playing back a demo of sock's map, many models were displayed pitch black, sometimes flashing depending on the view angle. I haven't tested it, but I could imagine the clipping and lighting issue also ccurs on other over-limit maps.
Hey no worries on your time, I just wanted to post here some ideas if you come back to MarkV development in the future.
quakeworld servers can record Multi-View Demos that support that (cl_demospeed 0 to freeze them or whatever it is, attack to toggle free-cam, jump to switch players).
With FTE, you can start a map and issue the 'easyrecord' command and it'll just start recording some randomly named mvd demo. (sv_voip_record and it'll also include any voice chat too, for extra playback privacy invasions *cough*). Something to add to a wish list
* Excluding entities from being recorded in a demo. For example detecting a new key on an entity "nodemo" "1". * Excluding entities from casting shadows (r_shadows). For example detecting a new key on an entity "noshadow" "1" * R_shadows always casts shadows from 360 angle, being able to change this via worldspawn. For example a new key "r_shadowangle" "90" More stuff
* Adding external texture support for model skins (24bit TGA) with use of the alpha channel so that the skin can faded gradually. The external skin could be specified with an extra key on the entity. Like for example "ext_skin" "textures/hiknight1.tga" Drawing alpha-blended models correctly requires sorting every alpha object (if convex) or triangle (of a concave object) in the scene and drawing them back to front. I don't know of any engine that does this (perhaps darkplaces.) But fitzquake certainly doesn't.
People probably haven't complained about the cheap method used in fitzquake and other engines, probably because most mappers use the feature sparingly, and only on simple convex brush-models, and not on many models in the same scene so that you see them overlapping incorrectly. I only mention this because as a perfectionist like me, you will probably notice this problem once you start relying heavily on the feature, on many objects in the same scene, and on objects with convex, organic shapes. The safest way to use alpha given this, is to only put it on convex shapes and make it hard for the player to have a line of sight that passes through multiple layers of alpha objects. I remember now...last time I checked darkplaces doesn't handle this within a model - if you have multiple layers of fur in the model you do get graphical glitches occur where they overlap.
even if you do depth-sort, large objects can still cut through and overlap in screenspace even if the actual brushes don't (or for instance a transparent player standing waist-high in water).
mostly a ext_skin with nice long strings would be far too painful. just use Would a QC controllable sound mixer be a good feature? Adding reverd echo and other dsp stuff. I remember good ol duke3d had a few option with sector effectors. Setting an effect to match a rooms size would be cool.
don't care too much about that, although it would be interesting to use every once in a while.
what I want is moving sound sources and doppler effect! also, while i'm dreaming, sound that travels through leafs/portals would be totally bad ass with volume attenuated by distance traveled instead of just radius.
halflife had various triggers that can apply various eax environment settings.
the forwards-thinking audio api to use is presumably openal, which would give environmental effects as well as doppler effects. but hey, if you want to write your own reverb library and plug it in to a quake engine, be my guest. :P incidentilly, a couple of engines already have rate modifiers, just not editably so... doppler would theoretically be easy enough to add, but as quake traditionally never moved sounds (and with invisible entities generating sounds and thus there not always being velocity information) I'm not sure how much it would break. Greetings everybody,
a few hours ago I decided to install DirectQ engine for Quake 1 and find it to be brilliant. The only (MAJOR) problem with it so far is the fact that "sizedown" command won't work. It's properly binded to "-" but it has no effect. Writing the command in console does nothing too. "Sizeup" works just fine. I run a 1920*1200 resolution on a 24 inches monitor and the fact that I can't "sizedown" the window makes the engine a bit...useless in the end. Can anyone please aid me on this ? Thank you in advance for your help. viewsize cvar. set it to 30 for a tiny view (not always supported nowadays), 100 for default sbar+ibar, 110 for sbar-only, and 120 for not even an sbar.
note that not all engines support a value less than 100. I've no idea what directq is meant to support, but if in doubt specify the viewsize cvar value directly and see what you get. That in the next 3 weeks I'll release a fix-up Mark V final release. I only have effectively one day off a week and not much free time in that one day off, really --- but I worked on Mark V in a 115 different mini-revisions and I know the first one of these with "the issue".
The bad news is I so radically changed the code base that even one "mini-revision" is a heap of changes. But the good news is that if I spend 30 minutes a day for a week or 2, I should be able to repair it. If I'm not way too tired to do so each day I don't have answers, I only have questions. Sometimes I think that Metslime, MH, Spike, the Quakespasm guys (szo, stevenaus, SleepWalker
Which is a pipe dream --- except I don't actually believe that. Such a thing almost seems like it was always meant to happen ? [End Stereotypical Baker post ;-) Hey I gotta be me Lately I've been spending most spare moments of thought on how to get an update out with virtually no "true" free time.
I'm rather determined to get an update out --- I'm just trying to figure out "the how". Some of the great stuff in "Final Mark V" is irresistibly fun --- and I'm still quite irritated about the collision bug. I hope you are doing well, SleepWalkR. Admittedly, my finite time I haven't had much time to read up on your massive contribution to Quake mapping in TrenchBroom. :( Read me: http://quake-1.com/docs/utils/fitzquake_mark_5.txt
Download: http://quake-1.com/docs/utils/fitzquake_mark_v_final.zip Weight: 448k (the engine) Speed: Yes Features: Some Code: Better Map Editor Used: Notepad.exe Is Bug Free: Should be (cross fingers) QuakeC Version: Depends on gamedir Recap of features: Faster, auto-completes about everything including key names, brightness only affects window, skill level recorded in demo, optional multisample capability. Borrowed some of Quakespasm's finesse and tried to largely sync with entirety of Quakespasm's codebase. See read me. External Short version: Light, fast, quick, better. You won't likely notice any of the changes because they were implemented very softly and subtly, in FitzQuake style. Last version of Mark V.
|
|
| This thread has been closed by a moderator. |
|
| Website copyright © 2002-2024 John Fitzgibbons. All posts are copyright their respective authors. |