Create an account


Thread Rating:
  • 2 Vote(s) - 3 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Proposal to change the mapping procedure for official Xonotic maps

#1
Brick 
Hi!

In this very long post I'd like to point out some of the major problems I've encountered during mapping for Xonotic, and I'd also like to give some solutions on those problems.
Ok, let me state this as fast, straight, honest, and clear I can: the current mapping procedure for Xonotic is slow, bureaucratic and generally unpleasant to work with, and sometimes it doesn't even give the satisfaction in the end.
Getting a map ingame takes about 4-5 months. Yes, you've read that right. You might ask, why does that take so long?
Let me explain:

First 1-2 months:
Building the map. This means, you have all the required sources, like:
A neat concept, that you have in mind: the general layout, item placement, style etc. of your map. This is the most fun part of the whole procedure, it lets your creativity fly, and if you don't run into (you always do Big Grin) in any sort of bugs of the compiler, the engine, or netradiant, the most easy part too.
However, this is the part where the whole mapping procedure can go wrong, and the problems here didn't found and fixed will escalate during the whole process, and getting worse and worse to fix later.
There are a couple of major problems here from my experience:

1: You overestimate your knowledge, and hence overcomplicate the map. You don't have enough experience, time, and additional knowledge to finish it. Time is the easiest thing to solve. The others, are not so much. Making a good, high quality map requires skills in:
-modelling
-making textures and additional shaders for your map, or fine tune existing ones for your liking
-generating or getting sounds

2: You don't fine tune the basic layout and item placement, and it turns out later in the end, the whole map sucks gameplay wise.

This is the part where you probably keep the map to yourself. It's in alpha stage, not ready for gameplay (you think). False. But it'll get back to this later.

Second part:
Bughunting, detailing up, fine tuning. Given the input from others you get (sometimes ZERO), this can take about a month.
You try to polish up the map as good as possible. Fixing small texture misalignments, detailing up the map, fine tuning lighting, placing sounds, and in the end, optimising for best performance. Official Xonotic maps have a very high quality check on visuals and performance too, and I'm not even sure this is needed in all cases at all. Too bad, the same can't be told for gameplay. To get input in this issue, this is the stage where you do beta releases (if you do), and wait for input from the rest of the community. You even try to contact server admins to put the map on their servers, so people can test it.
Problem: nobody does.

Third part:THE MERGE REQUEST
You finished your map, and you're generally happy how it turned out. Finally, you hit the merge request button. The only thing you need to do, is to wait. And wait. And... wait. This is the most frustrating, and worse part of the whole procedure, waiting on the core team to try it and vote on the map. It can take a good MONTH, while nothing serious happens. It can also happen, that you still work on the map, fixing a few stuff here and there and in the end, the latest stage of the map's branch is a good 10 commits ahead compared to the merge request's. This is very bad and confusing for the Core Team, and the players too.

Fourth part THE MAP IS MERGED!!!
Yes, you've made it. After a couple of months of development, your map is finally in master. The next autobuild contains it. People finally notice it, and try to play... and in the end, it turns out, the map SUXX BIG TIME! Nobody plays it, because it doesn't play well, or have performance issues... and you say: "Fuck this shit, I'm not gonna touch that file again, I've spent a good half year on it, and I'm fed up with it! It even makes me puke just to look at the file in radiant!"

So... this is how mapping works right now for official maps. There are several problems we need to solve, none of them is easy to fix, but not impossible. So, let's go through all the issues, and how we together can fix them.
First of all, I'd like to propose the idea of a mapping team. Ideally, this consists of the mapper, a modeller artist, someone who can do textures, a coder, a server admin, and several playtesters. This in the end will speed up the mapping process, and help to make good quality maps. Let me explain in a more detailed fashion:

A: Planning the map
I think the most problems lie in the very first step, at square one: the idea, and the concept is kept in secret, and only the mapper knows about it.
I tried this aproach numerous times, it doesn't work. I think once the mapper has the idea of the map, that he/she would like to sooner or later get into an official release, they should make a top/down blueprint first, maybe map a proof of concept room of the style (this takes a couple of minutes for an experienced mapper) or anything else, so others can figure out how the map should look like. SHARE THE CONCEPT, share the idea here, on the forum. This leads to the following:
The community gets to know about a new map in the work, and can give feedback, right at the beginning about the style, and gameplay issues. This is very crucial, because this step can eliminate the problem I raised above: making a map that turns out to suck in the end. Also, this helps to form the mapper team too: those, who find enough interest to help the mapping process can offer their help, speeding the whole process up - with proper oragnisation, the more the people who work on the same issue, the faster it'll be done. I know this might be the hardest part, but we already know we have tons of talented people all around here, so why not team up?
Also, this part will lead us to another issue, that was not easy to realise for me first, but is very crucial:

B: What is the purpose of the map?
Let's face it: the Xonotic community is already torn apart into two factions: casual and "pro" gamers. This is inevitable, and it happens to every game, that the more competitive gamers find to be worthy enough to play. This can also be sign of Xonotic getting mature. Like it or not, you can't please the two crowd with the same maps. Let's see the differences here:
What does a casual gamer want from a map?
-great visuals, that is pleasant to even look at. Let's face it, good graphics sell even the most shit games, everybody knows this. Even if those maps doesn't play so well, have you ever walked around Red Planet just for the graphical amusement? Or Lab 610? I did, numerous times, that map is AWESOME. Or Glowplant. Or Lightspeed (ok, the last one might not be the best, but I think I got the visuals there well).
-Tons of weapons, items, health: you want to have FUN. What gives you more fun than tons of powerful, and playful weapons? Hell, imho the most fun weapon in the game right now is the T.A.G. seeker (right after the tuba, ofc.)
-Interesting map layout: hallways, big rooms, maybe TRAPS. Hot Grounds anyone?
-Oh, did I mention VISUALS???

On the other hand, what does the pro player want from a map?
-GOOD, TACTICAL MAP LAYOUT. This is the most important.
-Great visuals? Nah, this is NOT IMPORTANT AT ALL. We already know that most competitive players turn off every disturbing effects, that keep them distracted from fragging. Is there any reason at all in that case, to map and build those awesome details? The cold and unfogiving truth: there is none. It woulnd't be appreciated anyway. What you should however keep in mindm that you have to use textures more tactically, so players get a goot contrast.
-Good, tactical weapon, health and item placement is CRUCIAL, and competitive players don't like the fun weapons at all.

Prime example: q3dm6. It has a "pro" and a normal version. The normal version was shipped with Quake3, and the pro version was later added to the game, by the request of the more competitive players. In Quake3 the two maps were different files too, but thanks to the "gametypefilter" entity key, if the map isn't different in the layout, the fun and pro map can coexist in the same bsp.
From all the above, you can clearly see, you cannot please everybody with the same map. Or rather, you can, but is very hard, or you have to sacrifice a bit from the visuals (speaking of PRO maps). One of the best example is Stormkeep. Let's face it, that map is underdetailed to the rest of the official maps, however, it still doesn't look bad, and plays well. I've heard that Afterslime is the border on eyecandy, but the item placement could be improved for competitive gameplay (but it's good for casual FFA).

What I'm trying to say is that one can't please the causal and the competitive players with the same map, if mapped with the same way.

I propose a change of mind in the sence of mapping here: we have to decide about the purpose of the map from the very beginning, and focus on different goals:
For casual, the map has to be: highly detailed (Xonotic flagship map, Red Planet detail), and FUN to play
For competitive gameplay: details kept at minimum (around Stormkeep level), but more focus on tactical elements of gameplay


Separating these views will lead to more rapid, and focused development of maps, because everybody will know the goals and the audience the map tries to reach. This also means, that developers with different taste can team up more easily.

C: Development
Once we've came to conclusion about the goal of the map, the visual concepts and enough people gathered, then the real work can start. I think the best would be to start a forum thread, where the mapper group can keep in touch, and share ideas, tell requests, ask for help etc. Once the map reaches beta stage, server admins and players can be notified, so they can playtest. This further ties the knots between developers and players, and also gives opportunity for additional suggestions and bugreports from the community.

D: Merge request
At last, the map is finished, hopefully good enough quality for it's purpose, and can be merge requested. I think if we'd followed the procedure described above, the voting phase of the Core Team can go down a lot faster, and more safe.


So to sum up, I have the following suggestions for official map development:

-Any official maps have to go through an incubation period, when players and developers can share and discuss their ideas about the initial concept
-Assembling a team of map developers, who work together during the whole process
-Clearly making a difference from square one between "FUN" and "PRO" maps.

What do you think about this? Given my reasons can convince most of you, I'll create the incubator thread for one of my new maps in work, so we can finish it together.
[Image: 561.png]
"One should strive to achieve; not sit in bitter regret."
Reply

#2
There could be a somehow official map-testing server which only provides "new", "to-be-tested" and "beta-stage" maps.
Perhaps its possible to have a script which automatically syncs the maps from the autobuild server every 24h.

Via sever stats, mappers could see which maps are played the most.
Even better would be some kind of input/text-box (could be shown between end of previous map and beginnin of new match) where people could give maps a rating and provide a quick feedback (like "textures suck", "theres a bug here and there" or "XY can be exploited to foobar the quux").
(00:30:26) jkwood: We have no bugs. We have random features! [IRC Sun, 21 March 2010]
Reply

#3
@ C.Brutail. : I agree with you!
I would join such a group as I'm unable to create models and textures. For my latest map ( You may saw the very little part I already built) I just drew a layout which was very hard because I played Urban Terror for years and the tactical gameplay is quite different. So it would be great if I could hear someones opinion about it or if somebody got a good layout...
I wanted to create a "FUN" and "PRO" map but after I read this it may becomes more a "FUN" map.
Reply

#4
I agree with what you say, C.Brutail.

And I am surely willing to test whatever map, and give CONSTRUCTIVE feedback.

But a server for playtesting doesn't really seam handy. It's all extra work, while testers can just download the map and put in their data directory, and if I test a map I want to look to the map, instead of trying whole the time to not get my ass blown off.
Reply

#5
Very well said C.Brutail! I agree with all the problems you mentioned.

There are also some very good article on this website which I want to share: http://lunaran.com/page.php?id=9

C.Brutail Wrote:I think the best would be to start a forum thread, where the mapper group can keep in touch, and share ideas, tell requests, ask for help etc.
^this
Reply

#6
Where can I open a thread like this?
Reply

#7
Probably under suggestions.
Reply

#8
hmmm, I'm not so sure about this, simply because I have gone through that process. I can do all of the things necessary for a good map, and I can blend gameplay elements with eyecandy in such a way that it pleases everyone.

I do agree that something needs to be changed, and a core "mapping team" would be a good solution, however I think the mappers on that time should still be allowed to work individually insofar as they are able to, and request help only when they need it.


Another thing I want to point out is that xonotic's current selection of maps is not really geared for xonotic gameplay. Maps like red planet and glowplant are fun, and look good, but they seem like they are designed for a slower, more tactical game than xonotic.

At some point, after we have this mapping team set up and a good selection of maps, we should have a vote to pick what maps stay, and what maps are removed from master. This will hopefully give xonotic a much stronger lineup of maps.
Master of mysterious geometries

Imgur Gallery
Reply

#9
Nice post C.Brutail... it certainly gives me a few pointers for my mapping ideas (if they are ever realized)
Damn it, my stats are crap so far lol
Reply

#10
Yeah Xonotic needs, as same as Quake, arenas not real places. Again, read this: http://lunaran.com/page.php?id=192
Reply

#11
(04-11-2012, 08:34 AM)C.Brutail Wrote: Third part:THE MERGE REQUEST
the worst part of the whole procedure! At the moment it's just 'hey, i think my map is good enough for being merged'. If the map looks good and got an acceptable gameplay the map will be added. That's the only thing which matters and the maps have no clear overall consensus. Another problem is that offical maps will always get played on servers no matter how many slots are there. Well, thats normal but the problem is that there are maps like lightspeed, which are simply unplayable with more then 4 people

(04-11-2012, 08:34 AM)C.Brutail Wrote: A: Planning the map
I tried this aproach numerous times, it,they should make a top/down blueprint first, maybe map a proof of concept room of the style (this takes a couple of minutes for an experienced mapper) or anything else, so others can figure out how the map should look like.
The community gets to know about a new map in the work, and can give feedback, right at the beginning about the style, and gameplay issues. This is very crucial, because this step can eliminate the problem.
I cant agree with you with making it public as soon as possible. There are several reasons. The main reason people who give you suggestions can think that it should be only this way, if you don't realize their suggestions/idea (for whatever reason) they'll get pissed off and stop saying anything. On the other hand side there are people who agree to everything, which is also quiet pointless. There are also people who arent reading and consider it as a finished map and it to servers or complain that it's unfinished.
I always ask other players i know from sometime for their opinion and some playtesting. it work quiet well so far.

(04-11-2012, 08:34 AM)C.Brutail Wrote: B: What is the purpose of the map?
Well, that is up to the designer. The problem is that there are maps like red_planet which got every possible mode on this map. Which makes an average map even worse and with some modes unplayable.
With some experience you get a good feeling for the Item- and Weaponplacement and of course playing experiance from xonotic/nexuiz and other similar games.
Mostly dm maps are interesting for "pr0s". Smaller maps for duels and bigger maps for tdm. If it's in your interest then work together with some duel players, but not with too many, they might start discussions like "this map needs a nex! No, it dont!". When you dont care about "pr0s", just add one megahealth and one mega armor, and place all powerful/important items at different corners. Then the map will most likely work in duels and ffa.

(04-11-2012, 08:34 AM)C.Brutail Wrote: D: Merge request
At last, the map is finished, hopefully good enough quality for it's purpose, and can be merge requested. I think if we'd followed the procedure described above, the voting phase of the Core Team can go down a lot faster, and more safe.
Don't know if anyone follows the mapping mergerequest on the tracker, but the voting from the core teams part should be removed. They mostly yes to everything and some dont even respond after months (so everything takes ages.)
<Samual> I am the most unprofessional developer ever
<bluez> halogene, you make awesome music, but you have no clue about ctf.
<Halogene> I didn't know mappers include some mysterious waypoints so members of the BOT clan can navigate a map?
<divVerent> if you don't pay for a premium account, your movement speed is limited to 100qu/s
Reply

#12
(04-11-2012, 08:34 AM)C.Brutail Wrote: What do you think about this? Given my reasons can convince most of you, I'll create the incubator thread for one of my new maps in work, so we can finish it together.

i generally like most of your ideas. And yes, the best way now is to give it a try with some test map.
<Halogene> ok "n1" means "nice one"
<Halogene> "gl" means "good luck"
<Halogene> "fu" means "wow that was wickedly nice that frag"
Frogg on Instagram
Reply

#13
I do post my maps here from the very first stage off development

but not many people comment ... i think they just don't want to try something alpha/half baked ... I also think they may think of me as sort of annoying (do you?) ....


also ... I think that redplanet does fit xonotic very well ... the doors open fast enough for quick bunny hopping action ... and it looks nice... ... it does have low fps though (particle effects )


also also ... glowplant is the prime example of a map where eye candy does not sacrifice performance ... the map looks great ... plays great ... its an old nexuiz classic ... and it is optimized to the death... i don't see why pros would have aything against it
Reply

#14
(04-11-2012, 05:11 PM)hutty Wrote: but not many people comment ... i think they just don't want to try something alpha/half baked ... I also think they may think of me as sort of annoying (do you?) ....
I think people are just too lazy or too busy to help working on your project. At least that applies for me. Blush

(04-11-2012, 05:11 PM)hutty Wrote: also ... I think that redplanet does fit xonotic very well ... the doors open fast enough for quick bunny hopping action ... and it looks nice... ... it does have low fps though (particle effects )
It could be optimised more, especially the outdoor part as quite some people said that they have performance issues there. And the doors may open fast but you often crash against some edge and get stuck. It´s that the layout isn´t made like an arena. Xonotic needs arena maps.

(04-11-2012, 05:11 PM)hutty Wrote: also also ... glowplant is the prime example of a map where eye candy does not sacrifice performance ... the map looks great ... plays great ... its an old nexuiz classic ... and it is optimized to the death... i don't see why pros would have aything against it
Playing duels on it isn´t very funny. Too long hallways, big outdoor part which makes sniping stronger, many elements which are looking good but don´t have a positive effect on gameplay, etc.

In my opinion, Xoylent is the best example where you can see that good looking maps can also be played professional.

EDIT: end of offtopicness here
Reply

#15
Quote:18:12 <+CuBe0wL> http://forums.xonotic.org/showthread.php?tid=2875
18:12 <+CuBe0wL> read this
18:12 <+CuBe0wL> please
18:12 <+CuBe0wL> and reply there
18:12 < Soelen> okey

I dont know what to say about Merge Request since I dont like to merge anything of my work, but I could write down my workflow if I may.

My maps are mostly made with blender, just because I'm too lazy to learn netradiant I guess. After some mapping I release very early versions on a public xonotic homeserver so I can prevent mistakes from the start and implement preferences from the average player. It is important to ask players from all different countries and gametypes to catch a good average response. I also tried once to throw some information about this map, giving people a webaddress to response if they want to

[Image: YS7Ua.jpg]

But I fear noone was interested enough to do so, heh. Anyway,

(04-11-2012, 12:32 PM)Cortez666 Wrote: I always ask other players i know from sometime for their opinion and some playtesting. it work quiet well so far.

I'm not quite sure if Cortez666 is talking about the #xonotic.de, but I get a lot and good response and help from that channel.

So that's it I guess, releaseing very early and getting good response/help from #xonotic.de.

Now people would probably say getting early opinion / response is bad since it wont let you work individual anymore:

(04-11-2012, 11:54 AM)theShadow Wrote: I think the mappers on that time should still be allowed to work individually insofar as they are able to, and request help only when they need it.

Well this may be true, depending on your decisions, but

(04-11-2012, 08:34 AM)C.Brutail Wrote: People finally notice it, and try to play... and in the end, it turns out, the map SUXX BIG TIME! Nobody plays it, because it doesn't play well, or have performance issues...
(04-11-2012, 08:34 AM)C.Brutail Wrote: I've spent a good half year on it
(04-11-2012, 08:34 AM)C.Brutail Wrote: input from others you get (sometimes ZERO)

I dont want to hear exactly that, map sux big time after 6 months of work D:

I do believe in cooperative working/testing/discussions in a team together, and I would join such a group which is all about testing maps.

(04-11-2012, 09:04 AM)matthiaskrgr Wrote: There could be a somehow official map-testing server which only provides "new", "to-be-tested" and "beta-stage" maps.

I hate to say this but he is right. I would love to see such a server which I can easily just join instead downloading maps by hand and throwing them into the xonotic directroy. Also a mutator which you can give maps on the end a rating would come handy.

(04-11-2012, 05:11 PM)hutty Wrote: I also think they may think of me as sort of annoying (do you?) ....
For me you ain't!
[Image: signature.php] [Image: DVlpo.png]
Reply

#16
Please do not derail the discussion. Thanks.
*This was for Maddin and friends.

Also, I don't want to shoot down anybodys work or... hmm... how can I tell this...
This thread is about to be OFFICIAL maps. My previous experiences tell that maps that were not ment to be included in the game in the first place have a very hard time getting in, for numerous reasons.

Also, I forgot to mention (which I thought will be quite obvious, but anyway): mapping for inclusion means using git and the autobuild server, because using those tools, one can be sure that all the work is properly shared, the newly created media is there to try out asap, and using the autobuild requires the SOURCE map file too. And that's one of the requirements of a map added into Xonotic.
[Image: 561.png]
"One should strive to achieve; not sit in bitter regret."
Reply

#17
(04-12-2012, 03:16 AM)C.Brutail Wrote: My previous experiences tell that maps that were not ment to be included in the game in the first place have a very hard time getting in, for numerous reasons.

Well as a mapper you mainly design a map for yourself. I cant agree with you on this point. i didnt made darkzone with the intention to get it included, neither the upgrades on techassault.

(04-12-2012, 03:16 AM)C.Brutail Wrote: mapping for inclusion means using git and the autobuild server,
another overrated point in my opinion. I never used it and never will use it.
<Samual> I am the most unprofessional developer ever
<bluez> halogene, you make awesome music, but you have no clue about ctf.
<Halogene> I didn't know mappers include some mysterious waypoints so members of the BOT clan can navigate a map?
<divVerent> if you don't pay for a premium account, your movement speed is limited to 100qu/s
Reply

#18
Cortez666 Wrote:Well as a mapper you mainly design a map for yourself.
it's your personal preference.For example i don't map for myself, ok i enjoy it, but i do it so others will hopefully have fun playing it. (offtopic sorry)

Cortez666 Wrote:
(04-12-2012, 03:16 AM)C.Brutail Wrote: mapping for inclusion means using git and the autobuild server,
another overrated point in my opinion. I never used it and never will use it.
cortez, but you don't map for inclusion so why should you care anyway. I think it does make sense. Either you learn how to use git or cooperate with someone who is using it already (it's not that difficult, if even i was able to handle it more or less)
<Halogene> ok "n1" means "nice one"
<Halogene> "gl" means "good luck"
<Halogene> "fu" means "wow that was wickedly nice that frag"
Frogg on Instagram
Reply

#19
(04-12-2012, 04:02 PM)kuniu the frogg Wrote:
Cortez666 Wrote:Well as a mapper you mainly design a map for yourself.
it's your personal preference.For example i don't map for myself, ok i enjoy it, but i do it so others will hopefully have fun playing it. (offtopic sorry)

sorry for the wrong expression. with for yourself i mean you make the maps you like.
<Samual> I am the most unprofessional developer ever
<bluez> halogene, you make awesome music, but you have no clue about ctf.
<Halogene> I didn't know mappers include some mysterious waypoints so members of the BOT clan can navigate a map?
<divVerent> if you don't pay for a premium account, your movement speed is limited to 100qu/s
Reply

#20
slightly on/off topic ish here

it seems that many mappers are not following the texture naming rules...

map specific textures need to go in the maps sub texture group thing

I ... being a person who actually enjoys grabbing maps off the froum ... tossing them into the data dir ... and then evaluating them extensively often find my radiant textures list crammed overfull with those maps textures ... making it quite difficult to remember which textures were safe to use...

(note ... my digsyite map does this too ... but that is because I intend for the textures to be a texture set used in multiple maps when finished so thats a special case)


So it seems that the communication between the stickied forum thread on naming standards is not reach the mappers somehow ...

I also found it very difficult to figure out how to add custom textures at all ... cause there is nex to no documentation (and any existing documentation either skips a important step ... or is only a tiny blurb) I had to find most of my info from the irc ... which ... to be honest ... is not a guaranteed source of info

I suggest we make the wiki the center for all infos ...

Then we just point the players to the wiki and they can get all the infos that they need

the bzflag porject has a very very nice and informative wiki bzflag's wiki

we already have a wiki ... time to start using it
Reply

#21
yes brutali!
Reply

#22
Can someone summarize everything here and lay out the information relevant to me more simply? Way too much to read with the time I have right now.
Reply



Possibly Related Threads…
Thread Author Replies Views Last Post
  Official Overkill balance martin-t 38 29,219 10-06-2016, 05:08 PM
Last Post: martin-t
  dpmaster.deathmask.net master server IP change Willis 2 5,835 03-15-2016, 02:23 PM
Last Post: Willis
  Relief Mapping - Performance Improvement - Beta v2 master[mind] 39 48,677 11-11-2014, 03:42 AM
Last Post: Lee_Stricklin
  Xonotic Game Engine, Mapping, Development - General Developer Questions p14r 6 10,938 08-04-2014, 10:24 AM
Last Post: p14r
  [SOLVED] misc_model : Support for specular and relief mapping? Maddin 5 7,112 11-10-2013, 06:34 AM
Last Post: Maddin
Exclamation Join the Xonotic mapping team Mirio 2 12,241 06-14-2013, 04:31 AM
Last Post: Maddin
  Mapping Tutorials hutty 57 49,461 08-31-2012, 11:43 AM
Last Post: aa
  Are maps in Xonotic hinted? Majki 18 17,185 06-15-2012, 10:38 AM
Last Post: Majki
  My mapping ideas - pick one! CuBe0wL 28 32,545 07-07-2011, 06:48 AM
Last Post: CuBe0wL
Question Mapping with Blender? unfa 8 9,335 03-31-2011, 08:26 AM
Last Post: Sepelio

Forum Jump:


Users browsing this thread:
1 Guest(s)

Forum software by © MyBB original theme © iAndrew 2016, remixed by -z-