DSDA Lua script #4464
DSDA Lua script #4464
Conversation
|
Misc domain may have to wait until I gather feedback from Doom people. |
YoshiRulz
added
Request: Feature/Enhancement
kalimag
force-pushed
the
pr/dsda-lua-module
branch
from
96725d2 to
5cc4472
Compare
cleanup
kalimag
force-pushed
the
pr/dsda-lua-module
branch
from
7a7f2aa to
56267ec
Compare
|
I think center zoom is done. Not 100% perfect but combined with mouse drag feels sufficient. |
Assets/Lua/Doom/dsda-data.lua
Outdated
| dsda.LINE_SIZE = 256 -- sizeof(line_t) is 232, but we padded it for niceness | ||
| dsda.MOBJ_SIZE = 512 -- sizeof(mobj_t) is 464, but we padded it for niceness | ||
| dsda.PLAYER_SIZE = 1024 -- sizeof(player_t) is 729, but we padded it for niceness | ||
| dsda.SECTOR_SIZE = 512 -- sizeof(sector_t) is 344, but we padded it for niceness |
There was a problem hiding this comment.
Maybe a Sizes table for these too?
| local read_u8 = memory.read_u8 | ||
| local read_u24 = memory.read_u24_le | ||
| local read_u32 = memory.read_u32_le | ||
| local readfloat = memory.readfloat |
There was a problem hiding this comment.
Similar shortcuts exist in the main script, looks like it makes sense to declare them once and then use everywhere?
They're still available through `line.padded_size` etc.
|
|
||
| if Init then init_mobj_bounds() end | ||
|
|
||
| -- re-init cache after state load, rewind, etc. |
There was a problem hiding this comment.
That isn't raised on rewind
Although I guess it's possible to load a state with the same framecount as the current state, so doing both might be good
Assets/Lua/Doom/dsda-data.lua
Outdated
| assert(dsda.line.size == 232, "line.size does not match sizeof(line_t)") | ||
| assert(dsda.mobj.size == 464, "mobj.size does not match sizeof(mobj_t)") | ||
| assert(dsda.player.size == 792, "player.size does not match sizeof(player_t)") | ||
| assert(dsda.sector.size == 344, "sector.size does not match sizeof(sector_t)") |
There was a problem hiding this comment.
Having these as consts at the top is better because we can be descriptive with their names and reuse them if needed. Especially sensible since we have 2 types of sizes: padded and internal.
There was a problem hiding this comment.
All of them can be accessed like dsda.line.size and dsda.line.padded_size, which seems descriptive enough. The asserts are just there to make sure the script calculated them correctly
But sure, I'll put them them back into tables
Assets/Lua/Doom/dsda-data.lua
Outdated
| dsda.LINE_SIZE = 256 -- sizeof(line_t) is 232, but we padded it for niceness | ||
| dsda.MOBJ_SIZE = 512 -- sizeof(mobj_t) is 464, but we padded it for niceness | ||
| dsda.PLAYER_SIZE = 1024 -- sizeof(player_t) is 729, but we padded it for niceness | ||
| dsda.SECTOR_SIZE = 512 -- sizeof(sector_t) is 344, but we padded it for niceness |
There was a problem hiding this comment.
Having these as consts at the top is better because we can be descriptive with their names and reuse them if needed. Especially sensible since we have 2 types of sizes: padded and internal.
|
Should the script pause during TAStudio seek? |
|
Scripts only pause by default if you seek with turbo, and there's an option to not pause them even during turbo. Outside of turbo we don't normally make them pause, because sometimes it's needed to keep running, and switching that on and off in the script itself would be annoying. Turbo-seek checkbox is faster to use and more organized. |
3 participants
(continued from #4458)
TODO:
Ideally you can write code like
for mobj in pairs(dsda.mobjs) do print(mobj.type) endand all the memory reading is abstracted awayObjectscache inthings-lines.lua(doesn't work with spawned objects)Seems to be a big performance cost on fancy Hexen maps
If time permits:
gui.textworksCheck if completed: