Hendricks266, on 01 March 2014 - 02:14 AM, said:
Maphacks should fix the sprite angle, never the model.
Existing misorientations should not be preserved in the name of convenience to the maphack-makers. It would be best to bite the bullet. I also believe that the the list of wrongly-facing models in hrp_todo.txt should eventually all be corrected and their maphacks updated.
Well, ...
should, eventually. I think this isn't the time
yet to bite that bullet. For a decade or more no one has reviewed or complained about those models. And there should be more important things on everyone's to-do list. (Like reaching the BIG maphack milestone, actually implementing the maphack DEF tokens
![:blink:](https://forums.duke4.net/public/style_emoticons/default/smile.gif)
(sorry), etc.
Hendricks266, on 01 March 2014 - 02:14 AM, said:
If it would be any assistance, it would be no trouble for me to write tools that:
1. rotate MD3s
2. identify all sprite IDs of a given picnum in a map
Yeah, I'd really appreciate that. I once tried to rotate a model in Milkshape trial, but the written model was somewhat messed up, and I didn't pursue it.
Hendricks266, on 01 March 2014 - 02:14 AM, said:
With the data from #2 it should be no trouble to generate per-map grep/awk/let/sed bash script magic to update all the maphacks in the SVN, provided the respective map files.
It's not completely trivial. Many models have not only been rotated, but also moved (like intersecting groups of barrels, or brooms to actually exactly lean against the wall). And those models which appeared to be [edit:]
rotated correctly in the first place and therefore do not appear in the MHKs, would then look definitely off. (Could be solved by an automaphack-block appended to the manual MHK, ofc.)
Hendricks266, on 01 March 2014 - 02:14 AM, said:
I see that 99% of the maphacks have tilenums commented already, so this task should be even easier.
Yes, I've designed the UserMapHacks with (semi-)automatic patching in mind. Unfortunatley, the one percent (well, more like ten, actually) are original game's MHKs which still await my full review (redo from scratch and then merge with repository version).
Conclusion: I agree on the principle and I'm willing to work on that, but not before we have:
- EDuke32 supporting the maphack DEF tokens
- a subsequent
full HRP release containing fully reviewed Atomic MHKs (and hopefully the majority from my "big milestone to-do list")
- supporting tools (like #1, #2)
Mark., on 01 March 2014 - 05:35 AM, said:
I use that program and I agree, its free quick and easy to align the models. But that program has a habit of changing the internal "frame" name of the model to a default of none00 and will screw up the defs.
That's no big deal.
eduke32.log will complain, and frame names can easily be identified by opening the MD3 in a text editor.