Doomworld Mega Project 2016 (v2 is on /idgames)

[t3hPoundcake]

This is the first mega project I've taken part in, is there going to be some official feedback thread/discussion thread for this or are these strictly a fun project to get community interaction?

[NuMetalManiak]

TimeOfDeath said:

Getsu Fune: the skullkey bars are close together and block the use actions in classic ports. So I assumed it was played in zdoom.

Try this then.

I noticed that in PrBoom, they don't open from the front side, but if you use them on the north/south sides (you can by the way), they would open. This widens them a bit so it should work when you use them from the front side now.

[Yugiboy85]

t3hPoundcake said:

This is the first mega project I've taken part in, is there going to be some official feedback thread/discussion thread for this or are these strictly a fun project to get community interaction?

I suppose feedback will be provided in this thread but, this is mainly a project for fun heh. I know that at least john suitepee will do a series of streams (on twitch) for this like he always does so I guess there is that feedback to look forward to :D

Then, you have rdwpa who seems to give feedback on the different submissions here quite often.

[Suitepee]

The_Trigger said:

I know that at least john suitepee will do a series of streams (on twitch) for this like he always does so I guess there is that feedback to look forward to :D

I will get around to streaming this project for certain, perhaps after dealing with the majority of my remaining Doom wad requests.... (might play this before tackling Holy Hell as last)

I'll be judging each individual map this time instead of the project's consistency as a whole, which I wish I'd done for the past 3 DMP's.

[t3hPoundcake]

Suitepee said:

I will get around to streaming this project for certain, perhaps after dealing with the majority of my remaining Doom wad requests.... (might play this before tackling Holy Hell as last)

I'll be judging each individual map this time instead of the project's consistency as a whole, which I wish I'd done for the past 3 DMP's.

To have the man, the myth, the legend, John "Heckin" Suitepee give me feedback on a map is my dream. All hail. See on twitch<3

[Grain of Salt]

TimeOfDeath said:

Grain of Salt: is it ok if I make the exit bars pressable to open and add a skytransfer?

Yes. Thanks a lot!

Btw (@ anyone who cares) you can indeed play past MAP33 in prboom-plus. You just need to use -warp, have CWIL patches for those maps, and use -nomusic if you're playing past MAP36. You also need to use -nomusic if you're watching a demo recording of those maps, which is kind of a pain.

[Walter confetti]

You can also play MAP33 to map35 in vanilla / chocolate doom...

[riderr3]

50 authors this is impressive and means there are still people interested and community still growing.

I like the overall idea for HUB but it will be better if HUB have extra portal for playing maps in numerical order (instead this silly mirror).

[Tristan]

I've run v1 through rhinoduck's tnswutil utility.

The following maps are missing P2-4 starts for co-op:
06, 12, 16, 18, 19, 20, 21 (gj me), 23, 24, 31, 32, 34, 35, 36, 37, 41, 45, 47, 49.

On MAP07, linedef 2560 is a W1 teleport. I haven't played the map myself, but if the destination is important and the player dies prematurely, the map will break in co-op - best solution (if required) is to put a line in the destination teleporter that opens up a teleporter pad near line 2560 so other players can get there.

The same goes for the following:
MAP19; the circle of W1 teles with tag 42.
MAP23; the hexagonal sky pit - tag 17.

[Spectre01]

Eris Falling said:

MAP19; the circle of W1 teles with tag 42.

Woah, that's me. That circle warps you to a barrels-o'-fun death exit. I could've added coop starts as I don't think anything would break with multiple players.

[Tristan]

Ah so it does. In that case they can be left as is.

[Scotty]

I'll have to specifically modify my map (23) for co-op i think, as p2-4 will have to teleport into the area from specific starting points to avoid breaking the intention of the map; if this is acceptable, give me a few days and i'll get that sorted along with the linedef problem.

[TimeOfDeath]

Does zdaemon really not just spawn players from the player 1 start if there are no coop starts?

[Tristan]

It does, but that's a pain in the arse. You can't spawn if someone is standing on the spot - sucks if there's a lot of players, and sometimes people spawn and don't move...

[TimeOfDeath]

I thought they fixed that so you can spawn even if a player is standing on it.

[Memfis]

The way it usually works is at first you can't spawn if someone is standing on the spot, but after like 10 seconds or so the system understands that and allows everyone else to spawn anyway. Looks kinda fuuny.

[baja blast rd.]

Do people spawn on others' heads, so it's possible to have a big stack of Doomguys? That would be great!

[Keyboard_Doomer]

I didn't directly check the code but from experience I strongly believe it works like this:

step 1: ZDaemon chooses one of the player starts and tries to spawn a player on it.

step 2: If that place is currently occupied by another player it tries to use the next available player start.

step 3: If all of the player starts are currently occupied you get a message to try again later.

From this follows if there's only one player start placed in the map ZDaemon can only try to use that one. If it stays occupied, no dice. There's no option in ZDaemon that would allow players to walk through each other so if it tried to spawn a player where there already is somebody, they would become stuck into each other.

One option that ZDaemon could use is to check if a space immediately next to an occupied player start is occupied and try to spawn the player there. However, when ZDaemon tries to spawn a player it never checks if a space is occupied by another non-player object (it has no reason to, the object could be placed there intentionally by the mapper) so this would lead to a risk of getting the player stuck inside something or easily telefragging a monster that was, let's say supposed to teleport away immediately after being alerted by the first spawned player. ZDaemon also never checks if the place to spawn a player is far from the coordinates of other player starts (again, no reason to do this, could be intentional) so it's possible if, let's say player 1 start is placed on a small pillar, it would try to spawn another player in a damaging floor next to the pillar.

If you try to spawn while all of the player starts are occupied you get a message that says "please try again in 5 seconds" but this doesn't reflect how ZDaemon checks for available place to spawn, you don't need to wait another 5 seconds. It's like when a website tells you to try to reload in 1 minute when the page doesn't load because all of their serves are busy. It's quite likely the page loads if you try to reload immediately, they're just trying to lower their server load.

rdwpa said:

Do people spawn on others' heads, so it's possible to have a big stack of Doomguys? That would be great!

No, you can't spawn on other player's head. ZDaemon checks the exact coordinates of a player start. If they're occupied, you are spawned on another unoccupied player start (if available) or have to wait.

I believe Memfis' observation is not correct. Players would get stuck into each other if they were spawned in the exactly same spot. It might ostensibly look like the program "figures it out" after 10 seconds but it doesn't make much sense from programming perspective. What in fact most likely happened is that some of the players moved slightly in those 10 seconds and there's now enough available space to spawn another player.


Sorry for the lengthy explanation.

TL;DR: ZDaemon has no way to bypass an occupied player start. To spawn another player either another player start needs to be placed in the map or the occupied player start(s) need to become unoccupied.

[TimeOfDeath]

There is this from the 108-109 changelog:

72. Fixed various problems with coop spawning, including spawning of
players above the number of coop starts in the map, spawning
under bridge things, blocking of spawn points by other players
or monsters, etc. If a spawn is blocked by a player or monster,
they will get telefragged except for the first 5 seconds of
the level. That should give players enough time to move off the
spawn points.

But then later there's this:

9. New server CVAR "coop_telefrags". It's a boolean. It controls
whether player-to-player telefragging is enabled in coop/survival
modes. The CVAR is forced to false for all public, non-passworded
servers.

I wonder if that also disables the player start telefragging?

[Keyboard_Doomer]

What I described applies to survival. On a coop server you can set the behavior to telefrag the player that's blocking a player start which kinda solves the problem but you can't have this behavior on a public survival server.

If a spawn is blocked by a player or monster,
they will get telefragged except for the first 5 seconds of
the level. That should give players enough time to move off the
spawn points.

This applies in coop, not survival. I assume this means if the gametype is set to coop you get the standard survival behavior I described before in the first 5 seconds.

The CVAR is forced to false for all public, non-passworded
servers.

This is true. I think if the gametype is survival it's forced to false even if the server is passworded but never tried so don't quote me on that. Either way, being teamkilled doesn't subtract from the number of lives you have in survival.

TimeOfDeath said:

I wonder if that also disables the player start telefragging?

The way it works is, if set to false you can't telefrag a player if they're blocking a player start, instead you get the behavior I described before - ZDaemon checks for another unoccupied player start, if it doesn't find one you get the message to try again in 5 seconds.

If set to true you telefrag the player that's on the spot you're supposed to spawn at. I think it doesn't even check for unoccupied player starts before that.

The CVAR also applies only for players, if there's a monster standing on the player start, it gets telefragged regardless of the value the CVAR is set to.

[Walter confetti]

Played from MAP01 to MAP10 (and a little of MAP11) with prboom, doesn't found any game-breaking bug expect:
- some little slime trials in MAP02
- I found myself stuck in MAP04 after pressing the yellow key switches (is normal that they activates without any key?), and roaming around the map...
Where is the blue key btw?
- a not really clear path on MAP05 (just something to understand the doors with keys and the exit are...)
- MAP06 have a exit? Or maybe is just me that i made something wrong...

Overall pretty solid and fun mapset so far, congratulations to everyone!

[pcorf]

riderr3 said:

50 authors this is impressive and means there are still people interested and community still growing.

I like the overall idea for HUB but it will be better if HUB have extra portal for playing maps in numerical order (instead this silly mirror).

Could have been 51 but I had no time to make a map.

[NeedHealth]

My computer shit itself about 3 -2 days before new year's eve, otherwise there would have been 52 maps in this set ... That disc what ever error. You other know what I'm talking about if you're having the same problem - I solved it by downloading an slightly older version of DB2.

[Memfis]

Is there anyone that wants to manage the 2017 project? I hope we won't have to wait 8 months again. :)

[scifista42]

I could once again do the job of running a thread and regularly updating its first post with latest versions of current submissions during the entire year. Then, at the end of the year, somebody else could replace me or help me to make the final compilation. How does that sound?

[Memfis]

Sounds good to me. I think you did a fine job last year. I would also like to see custom textures return (with name prefixes).

[Jaws In Space]

Memfis said:

Sounds good to me. I think you did a fine job last year. I would also like to see custom textures return (with name prefixes).

It took Scifista nearly 11 months to release DMP2015. How is that a fine job?

[Ninehills42]

Memfis said:

Sounds good to me. I think you did a fine job last year. I would also like to see custom textures return (with name prefixes).

I dont want costum textures to return :( ive already done part of my map for dmp2017, using only Doom2 textures.

[scifista42]

Memfis said:

I would also like to see custom textures return (with name prefixes).

If I was the one to set DMP2017 rules, this is what they would be:

-Custom textures, flats, sprites, weapons, monsters and other thing types allowed, but they must not replace or modify any stock ones, and their names must start with a unique 3-letter prefix same as your DW forum username does OR they must come out unmodified from some publicly released texture pack.
-Custom skies must NOT replace any RSKY* patches or SKY* textures, but they must be defined as custom textures (with all restrictions from the point above) and applied either via sky transfer line actions in Boom-compatible maps, or via MAPINFO definition in ZDoom-compatible maps.

Ninehills42 said:

I dont want costum textures to return :( ive already done part of my map for dmp2017, using only Doom2 textures.

No matter whether custom textures will be allowed or not, nobody will force you to use custom textures over stock ones.

[TimeOfDeath]

I'm going to sit out the 2017 project, so feel free do whatever you want.