Talk:Hitscan attacks hit invisible barriers in large open areas

Is this different from Flawed collision detection?    Ryan W 22:41, 1 Oct 2005 (UTC)

Flawed collision detection usually refers to hitscan attacks sometimes missing their target even when they are supposed to hit it. I don't know if this invisible barrier thing is caused by the same bug though. Janizdreg 17:01, 3 Oct 2005 (UTC)

Somewhat hard to tell, until someone writes the aformentioned article :) I for one would prefer that the articles were written before they are referred to, where possible. -- Jdowland 19:03, 3 Oct 2005 (UTC)

Well, I agree that it oughtn't to get too out of hand.  On the other hand, lists of nonexistent articles have sometimes helped us get a better grip on what kinds of articles we want to have here (without just plunging in and writing hundreds of new pages that we'd just end up deleting).  Also, there are certainly times when working on article A suggests forcefully that article B should exist, but the editor of A doesn't know enough about B to really make a useful contribution.  I myself probably cut and pasted the link Steffen "Rini" Udluft at least five times the other day, but I have no idea where even to begin with an article.  (Indeed, that is why I wish there were more articles about speedrunners — I want to read them!)    Ryan W 20:51, 3 Oct 2005 (UTC)

This has almost nothing to do with flawed collision detection as described in it's article. That is all about thing's bounding box not being detected outside of it's containing block, which often messes up events where things collide with bullets and other things. I'm pretty certian that this is not really the engine's fault, but something wrong with the nodebuilder used to make this level. I have read somewhere that (id?)bsp.exe had some quirks in it that made invisible barriers, slime trails, and other abnormalties.

It looks like there is some glass was attempted using the old engine, what does it show up as on the map? That might help the article better than a demo and a screenshot. B10Reaper 20:50, 16 December 2008 (UTC)

Um, please ignore 3 year old ranting by me if it makes it easier to add content to the wiki.  :D   As I understand it, "invisible barriers" refers to intercepts overflow caused by linedef 0 being very long, while "flawed collision detection" refers to this.  Or am I just repeating what you said?

There is certainly "glass" in a few places in Final Doom (TNT 09 and Plutonia 31 come to mind), which, as we know from COMPET-N, runs fine even with an unmodified doom2.exe.    Ryan W 22:05, 16 December 2008 (UTC)