When chemical grenade being activated on any unsimulated turf (for example a space turf), smokeFlow() starts infinite while loop, overloading the server.
With the only one unsimulated turf in pending list while() loop can't finish itself and for() loop can't start.
This fix makes chemical grenades being able to explode in space without server going in infinite loop.
Also fixed issue with smoke being able to pass through glass walls with help of c_airblock() proc. That's the best solution I could come up with.
Zone check code was useless because every single turf besides walls (which allready checked) and unsimulated turfs have it, making algorithm believe that turfs behind the glass wall are being reachable by chemsmoke.
I don't know if it was intended not to spread smoke in airless or unsimulated areas, but a bunch of airless simulated floor tiles in space is good enough for this algorithm to start spreading smoke around.
P.S. this algorithm is good for large, open areas and horrible in narrow maint tunnels. It is creating only a single cloud of smoke for 50/50/50 recipie in 1 tile wide tunnel.
New turf proc: assume_gas(). Maps to air.adjust_gas_temp().
Lots of optimizations to processing, fire, lighting, HasEntered() and
more.
Zones now process all fire data and existance in one go, fire objects
only handle spreading.
Most code has been ported straight so some of it mightn't be ideally
structured for the new gas_mixtures.
Signed-off-by: Mloc-Argent <colmohici@gmail.com>
recently i fixed a couple of issues with /vg/'s alien weeds and nodes,
and i see some applying to Bay12 too.
one of those issues was the order of checks in weeds/Life().
it should cancel the weed's search for tiles to expand to without starting
the direction loop first, saving quite a few proc calls.
another issue is alien weeds having the default OBJ_LAYER layer 3 which makes them grow
over all kinds of items and objects which makes sense i guess fluffwise but
is really annoying ingame for both crew and aliens.
( and this coming from the host of Alium Deathtrap 13 ;) )
nodes stay at layer 3 to appear properly over things like AI-holopads.
a really tiny improvement is setting linked_node after the space-turf check in weeds/New().
link to the /vg/ commit : 4dcb434f72
These are new posters for more postery goodness. Credits for the sprites
go to Nienhaus.
For the sprites of the poster go to http://puu.sh/5QlUx.dmi
2 posters are not included: 6th and 7th due to reasons.
- Renamed the variable for the smoke density mechanic to density. Also clamped it to a minimum of 1 so it cant multiply reagents.
- Added some checks so the chemsmoke effect doesn't run if it's in nullspace.
- Removed an unnecessary distance calculation.
- Added a null turf check to prevent runtimes if a really big smoke grenade goes off at the edge of the map.
- Added a check so the reagent copy code doesn't run if there's no reagents to run it on.
- Complete rewrite of the chemsmoke spawning code (everything that happens after the reaction to create the smoke).
- Moved it into it's own file since it's a little longer than the old one code-wise.
- Tweaked the smoke sprites; gave chemsmoke directions for more randomization, and tweaked the timing of their animations so they are smooth now.
Most of the code is documented, so you should be able to follow what's going on just reading through it. In short though:
- Chemsmoke now affects a circular area with a variable range depending on the total amount of chemsmoke created (no more 1-of-each-ingredient smoke grenades - sorry).
- The carried reagent volume in the cloud balances itself according to how big the affected area is (larger area, less carried reagent per turf).
- The carried reagent has more volume the closer you are to the centre of the cloud.
- The visual smoke effect scales with how big the affected area is.
Fixes#3986
- Foam reagent inheritance fixed:
The trans_to() proc in Chemistry-Holder.dm was calling handle_reactions() every time it would add and remove reagents with add_reagent() and remove_reagent(). I've added a safety flag to both of these procs (and copy_to()) so that trans_to will not call handle_reactions() until it has added all the reagents to the target container. This allows foam to reliably take on the properties of other leftover reagents.
- Fixed a bug in effect_system.dm that wasn't applying reagents from foam to the environment.
- Tweaked smoke to apply reagent effects after it has stopped moving
This was causing huge reagent spam at the center of the smoke cloud as it was spawning, since they were being applied every time the cloud of smoke moved. Also changed it to apply the effects 2-3 times at longer intervals (2 seconds).
- Smoke also only effects tiles in a 3x3 grid now, rather than 5x5.
- Summary:
Chem smoke does slightly less damage (if it contains damaging reagents)
Chem smoke proc calls reduced by 60-70% (significant lag reduction)
Foam works properly again
Foam proc calls reduced by 70-80%
