New Content??

[Tsar Hoikas]

We fixed that bug where OnServerInitComplete doesn't get executed due to delayed loads BTW

[Karkadann]

So in theory I could remove the plane that has the Door Texture on it, and leave the door way.
One thing I did notice about the texture on the inside of library upper floor where the inside of the door should be
the texture looks a bit distorted, almost as if the the door is a separate two sided plane with no uvw map on the inside
any way all this great but again I can produce a lot of different things as far as structure are concerned but Python is still a it of a mystery to me.

[diafero]

You can disable the visual in Python easily, but if you want to change the shape you also need to disable the physics - and that's only possible if the object has its own CoordinateInterface, which unfortunately is often not the case.

[Trekluver]

EDIT: There's a new page on the wiki... http://www.guildofwriters.org/wiki/Logi ... evelopment

Yes, I put that there. While we're on the subject of new content, I'll mention that I've been talking with Deledrius about restarting the StartUp Age redesign (Something he began, and placed aside, a while back). I want to help with that in any way I can, which, since I'm not very good with code, I'm not entirely sure how that will be yet. Hopefully we'll figure out where to start (and where I'll fit in with it) in the near future; in the mean time, maybe he can expound on this a little.

[Deledrius]

So... I've spent most of this week working on converting some content.

I've successfully (to an extent) converted two small simple Ages as test subjects, Tre'bivdil and FehnirHouse. I have loaded them up on my private dirtsand server and can explore them. I can even read the journals in Fehnir, which I've updated to use proper loc files and the global xJournalBookGUIPopup instead of the alternate PotS-FanAge scripts. In doing so, I also made some changes to Plasma which will make it easier in the future for Fan Ages to take advantage of this sort of thing without needing to shove data into places in which it shouldn't be.

So that's the good news.

This wouldn't be Uru without some bad news, so here goes: Due to the conversion from Havok to PhysX (or to anything, really; this is mostly a Havok-induced problem) not all collision is immediately usable. As it has been explained to me, in Havok these surfaces are all double-sided, so most artists paid little or no attention to the surface normals as they had no need to do so. This is unfortunate as it means arbitrary faces are the wrong way 'round once converted, making them not function as intended. Also, being a question of arbitrary geometry as part of the objects, this is impractical to do any automatic conversion on; it's solely an artistic question, and difficult to address non-visually.

What this boils down to is that without assistance from the authors, or a substantial amount of clean-up work against the exported files we have, it will largely be infeasible to convert even these simple Ages for use on a non-Havok-using client.

What we can take away from this, however, is that all age builders should try to be mindful of the surface normals on their physics objects if they ever expect to use them on the MOULa-era Plasma engine, either directly or through conversion.

[janaba]

Wow!!! This is awesome news, thanks, Deledrius, and as of the more unfavourable reference,
I hope that many age creators will read and/or however come to know of your important request ...

[dendwaler]

oh, the bad news seems worse to me, than the good news good.
Can you clarify me a few things?
The direction of the Normals is easy to see inside Blender, that should not be a problem.
But occluder bounds are often set within the logic window.(F4)
The properties are: Triangle mesh, Convex Hull, cone, cylinder sphere and box.
Convex Hull for instance is space " inside " a box , but at the same time the normals of it can pointing " outside " the same box.

I do not understand the relationship , nor how to handle it.

[janaba]

Haha ... In this very moment I was thinking about adding something to my above statement,
and dendwaler's as well as yours again, Deledrius, showed me that I'd be right in adding that

it'd be even better when Cyan eventually let it all go and hand it all over as initially intended
when offering this all publicly as 'Open Source' (I still see the big message 'Cyan makes it official:
"Myst" now in the hands of its fans'), so that those incredible arduous efforts would cease and
could more effectively be applied to making everything new from scratch and with special tools
or facilities or even an open environment/engine (you all know what I mean lol) that is able to
easily integrate fan ages and whatever changes and enhancements etc. like every other game ...

[Deledrius]

oh, the bad news seems worse to me, than the good news good.
Can you clarify me a few things?
The direction of the Normals is easy to see inside Blender, that should not be a problem.
But occluder bounds are often set within the logic window.(F4)
The properties are: Triangle mesh, Convex Hull, cone, cylinder sphere and box.
Convex Hull for instance is space " inside " a box , but at the same time the normals of it can pointing " outside " the same box.

I do not understand the relationship , nor how to handle it.

So far I've only seen this issue with the ProxyBounds physicals, which I assume (someone will correct me I'm sure) is the TriangleMesh option. The regions all seem to be correct, I'm assuming thanks to them being regions and not surfaces; the system makes the only correct guess on those (or perhaps they're written more sanely to begin with, I don't know).

[dendwaler]

But if triangle mesh is chosen for surfaces and the Normal is pointing to the wrong direction, then it is also impossible to see a texture, unless the twoside texturing option is chosen.
Personely i seldom choose twoside texturing and avoid it, only for tree leaves or simular situations where it is absolutely necessary.