[mono20]

You cannot cross solid (blocking) floor aligned sprites crouching even when they are exactly on floor height since version... dunno. When you switch to 3rd person mode (F7) while crouching it looks like the player sinks into the ground but I don't know if there's a connection.

TIA
mono20

Attached thumbnail(s)

• 

[Danukem]

That's bad! I just checked this out myself, in E1L1 (see attached)

Here's what happens on that sprite bridge in E1L1: If you crouch, you get stopped at the border between the sector floor and the start of the sprite bridge. However, if you stand up and then walk onto the bridge, you can then crouch down and crawl across it normally. F7 does show you sinking into it, but only by a few pixels (not like in the OP).

Attached thumbnail(s)

• 

[Hendricks266]

I recall we changed something relating to crouched movement relating to conveyor belt (water current) SEs. Perhaps that's related. Any chance someone could bisect which synthesis build caused the issue?

[LeoD]

Hendricks266, on 07 December 2015 - 08:05 AM, said:

Any chance someone could bisect which synthesis build caused the issue?

r4873

[Hendricks266]

LeoD, on 07 December 2015 - 09:30 AM, said:

r4873

Interesting, that was a one-line change (excluding something unrelated about sounds).

[TerminX]

LeoD, would you also mind bisecting which build first exhibits the player sprite sinking into the ground when crouched? It only happens in Polymost. Much thanks if you figure it out, it really saves a lot of my limited free time.

[LeoD]

TerminX, on 07 December 2015 - 11:47 AM, said:

LeoD, would you also mind bisecting which build first exhibits the player sprite sinking into the ground when crouched? It only happens in Polymost. Much thanks if you figure it out, it really saves a lot of my limited free time.

r5071 - OK
r5072-r5080 - compile error
r5081 - Polymost crouch error

[Helixhorned]

mono20, on 07 December 2015 - 07:03 AM, said:

You cannot cross solid (blocking) floor aligned sprites crouching even when they are exactly on floor height since version... dunno.

DOS Duke3D.

Trooper Dan, on 07 December 2015 - 07:45 AM, said:

I just checked this out myself, in E1L1

Comparing DOS Duke3D and a recent version (r5466, self-built):

          | sect. floor -> 1st bridge seg. | 1st -> 2nd segment | 1st -> sect. floor
----------+--------------------------------+--------------------+---------
 DOS Duke | blocks                         | blocks             | proceeds
 EDuke32  | blocks                         | proceeds           | proceeds

So we're less buggy than DOS Duke in this regard.

Hendricks266, on 07 December 2015 - 09:33 AM, said:

Interesting, that was a one-line change

IIRC the bug was that clipping which should occur did not, suggesting that my fix is proper.

IMO, this bug is in the category "could probably be fixed, but would need consideration/testing of whether the fix introduces new issues".

[Mblackwell]

But clipping through a sector floor shouldn't happen to vanilla assets that didn't clip originally.

[Danukem]

Helixhorned, on 19 December 2015 - 03:21 AM, said:

DOS Duke3D.

Wait are you telling us that for all of these years it was impossible even in EDuke32 to crawl across the bridge in E1L1 and no one noticed?

[Lunick]

There's like, almost no reason to crouch on that bridge and move at the same time though.

[mono20]

Helixhorned, on 19 December 2015 - 03:21 AM, said:

DOS Duke3D.

Ah, OK. I'm glad we got that bug ...er feature back

[Helixhorned]

Trooper Dan, on 19 December 2015 - 07:29 PM, said:

Wait are you telling us that for all of these years it was impossible even in EDuke32 to crawl across the bridge in E1L1 and no one noticed?

I'm just summarizing the result of some quick experimentation, nothing more, nothing less.

By the way, here's the post containing more information about the bug that was fixed in r4873. Revision 3207 (where it was introduced) is from November 2012, so the missing clipping was present about two years.

[mono20]

Helixhorned, on 19 December 2015 - 03:21 AM, said:

IMO, this bug is in the category "could probably be fixed, but would need consideration/testing of whether the fix introduces new issues".

We've been talking about the old conveyor belt behavior here. Wouldn't that be the same category? Just asking.

Merry Christmas y'all
mono20

[Helixhorned]

mono20, on 25 December 2015 - 02:13 AM, said:

We've been talking about the old conveyor belt behavior here. Wouldn't that be the same category? Just asking.

Not really. For that one, we know which change introduced it.