This map is as close to perfect for me as dod_solitude was in 1.3. WHAT SERVERS FILL THIS MAP UP AND DON'T LIE!!! I would love to play this map with a full server, ultranew_b did a professional job with it and it deserves the play time!!

Post server ip's only please, If I like the server I will visit your site on my own :) .


Thanks in advance,

--Ozz

Frames slow to a crawl for everyone with more than 18 people and it's not something ultra is willing to fix since long view distances are part of the map.

And again, the map could still be optimized for a b4 or final version, without taking away any of the gameplay features. I once read here that new_b loses interest in a map real fast, so he maps it all very fast or it never gets done. Don't know how consistent he is with that saying though :)

Either way, pier is a lot better, but it's also way too big for my own liking and the number of players we now see on servers. Now more than ever is the time of small custom maps.

I don't know. Maybe I am just used to playing with low frame rates (I never get above 45) but even with my old vid card ( ATI 9700 pro) I have never really noticed this drop everyone keeps talking about.

It is a shame he won't optimize it more; if dod_pier can be optimized as well as it was he probably could have done a couple more things to sora.

I still love it either way.

Wow ! People still play this?


I'm busy with other mapping projects. If a resident expert on optimizing wants to help out, i'm willing to let them try and give them full credit of course (as long as nothing is changed visually). I would like it balanced as well, so the allies don't have such an advantage.

Otherwise, it cuts into my new map dev time. Therefore making a final release unlikely (from me).

When I released it, I recommended 18ppl then, there was a reason for that. Deciever, If you don't like it dont play it, simple. Complaining does not inspire me to fix it.

:)

Who's complaining? I was simply stating fact. Don't get me wrong, I love playing the map, just as long as there is less than 18 people heh =P. If someone could help him optimize it, it would be great ^_^.

Complaining would be:

"OMFG WTF. stupid un playable map! lazee mapper cna't even optimize FFS!!!1111"

Dave, can you send me the vmf by PM? I'd like to have a look at it (as I remember you protected it from vmex right?)

Originally posted by ultranew_b
When I released it, I recommended 18ppl then, there was a reason for that. Deciever, If you don't like it dont play it, simple. Complaining does not inspire me to fix it.

so servers that have player limits over 18 should just totally avoid this map? you have a fan-base for sora. you should honour and appreciate that. update the map with your best effort to please their comments, as long as they arent absurd

I hope someone takes this up because I really respect this map and would love to see it perform to its fullest potential. There's something about the atmosphere in that church that just blows me away.

Originally posted by Neutrino
so servers that have player limits over 18 should just totally avoid this map? you have a fan-base for sora. you should honour and appreciate that. Neutrino makes a very good point here. Some mappers would kill to have the following and buzz your maps create. That is harder to acheive than you may think ;)

Map is not protected and I remember taking a look at optimisation. Basically the bottom of the map is not sealed so make most hint not working and some building could use areaportals. There is definitly a possiblity to optimise it further from what I remember looking at it. I also looked at the tower and the actual place a sniper can snipe could be left open and still close off other roof with skybox not the one with player access on both ends. Still a challenge to keep the tower view since you see the whole map but only a few place where you can snipe. But definitly could be improve.

I agree with neutrino that the fan love this map and to make it possible to play on bigger server would please many person.

I am currently helping to optimise Ripster remake of dod_solitude so I have my hand full. But take a look at it Furyo I am sure you will find it optimisable.

Any work you guys do to make this one final is greatly appreciated.

Thanks,

--Ozz

Yes, good luck !!

The layout is going to make it very difficult to get a lot better. The long view distances are deadly.

The only major changes needed aside from optimization are in the allied spawn area, maybe removing the quick access route to the tank. I think that would balance the map greatly. (ie. the axis and allies would reach their closest objective at the same time)

A lot of little bugs were reported in the release thread, those would get addressed as well.

If someone can speed it up, kudos to ya. I will fix the smaller issues relating to gameplay (exploits and small graphical stuff.)


:)

I got the vmf from Dave directly, that'll be better than the decompile with its possible errors. I'll take a look at it tonight

Ok first look taken....wow, just wow. There are lots of possibilities on this map for FPS gains, just as I thought.

Here's a quick peak:

http://perso.wanadoo.fr/froughol/907th/dod_sora.JPG

That's a quick shot taken in Hammer, once I unchecked most of the filters, to have only the world geometry show up.

The problems are the following:

- overlapping world brushes
- poor use of clipping tool on the arches, which should A) be made of individual brushes turned into func_detail and B) be snapped on the grid (it doesn't show, but they aren't). Not to mention the angle at which both brushes meet is the fastest way to increased vis leafs totals.

Some other things that showed up at first glance was the exaggerated skybox size (could use a lot of tightening up) and the overwhelming use of func_detail roofs as soon as they are in a /\ shape. While that may seem good, it also allows VIS to render what's beyond the building the player immediately faces, and that's bad on the performance.

I'll take my time with this one, I have dod_rive that's coming closer and closer to a b2 playtest. But there is definitely room for improvement.

PS: This isn't a blast at ultranew_b, whose map designs count amongst the very best in the community, but an attempt to show other mappers what NOT to do.

I just worked on it some, here is what "my" (as in it may very well not be perfect, experts please chime in) version looks like:

http://perso.wanadoo.fr/froughol/907th/dod_sora_furyo.jpg

In red, the world brushes, after some changes (see pic 1 in previous post).

In greenish yellow (numbers 1 and 2), the func_detail brushes, for the arches. Please note the exact same angle was kept from the original version. There are no intersections, only brushes joining at the very best position, both on the grid and sharing similar vertex positions (notice I said vertex positions, not vertices) which is better for vvis and vbsp.

Number 3 is kinda hard to see, but since the right wall cut the left one, I made this extra edge another func_detail. That way, vvis only interprets the two brushes where they meet. The extra little bit of optimization would be to make those two brushes meet in a 45° angle (both brushes having that 45° value), which would mean only one exterior face would have to be rendered, as opposed to two right now (the number of brushes in Source has no bearing whatsoever in a map, it's the number of rendered faces at once that counts more (hence the use of nodrawed faces))

In this particular case, since I used the nodraw texture, the point above is mute

Number 4 is to show that the bottom brushes here are textured with "brickwall002" with the scale of .32/.32, which I left as is for now. The correct scale for textures is .25 usually, with very rare exceptions at .5. Never .32 in any case. This stretches the texture quite a bit, which doesn't necessarily show on small faces (this brush is 40 units high), but that encourages face splitting (how the engine handles texture rendering on faces). The more faces the more processing power needs to be taken for that task, leaving less for the rest of the map.

This texture issue wouldn't make an impact if that's a lone case. But mappers should keep it in mind regardless.

It's official, I give up. The map's brush work isn't quite as bad as a 1.3 map one would have run through a decompiler, but there still is so much work rebuilding it from the ground would save time... From carved brushes to useless displacements, from textures to hundreds of faces rendered for no reason, from all roofs turned to func_detail to the skybox being a single huge box....

If Ultranew_b wants to give it to someone else, then maybe you'll see a better version come out one day. I know I just don't have the time for this type of job (and not necessarily the will either)

I knew Source is a good engine, but damn I gotta hand it out to the engine on this one, it's quite exceptional to achieve such framerates with this geometry...

Thanks for looking at it. But, I already knew it was a ton of work to fix it. I was still rather relatively new to hammer when I released this map. Anyway, I still play it all the time and people seem to like it. If I did anything to it at all, it would be balance it out to remove the allied advantage. It's still to this day in the top 14 custom maps as well as pier, so i'm happy.

From day one I reccomended 18 ppl.

I'm working on a few other maps that run perfectly fine with 32 ppl. I do learn from my errors. :)


@ theozzmancometh : thanks for bringing the map back into a semi-spotlight. I play on full servers regularily, I don't unfortunatly have them in my favs. I use the filter in the game browser and usually find a game quick.


:)

Well I started the thread because I love the map (aesthetically it's gorgeous) and hardly EVER get to get a full game in on it. If I knew anything about Hammer besides how to open it up for use with DODS I would tackle the project, but no way will it be my virgin attempt at hammer because I would undoubtedly destroy it.


Looking forward to your future releases Ultranew_b.

I see sora is all about carved brushes. Yeoutch!

Where Dash?

No, I always vertex edit (wayyyyyyyy faster and also avoids texture scaling),albiet brushwork is sloppy in a couple areas in the map as Furyo pointed out. Carving is perfectly fine using rectangles.

It is a beta map, keep that in mind. I made it within 3 weeks, so it was rushed.

Well the church is one area where you used the carve tool for non 90° angle (or so it seems). The church as it is, is about as Vis unfriendly as it could be. It's a good thing there isn't much geometry behind it or the map would have been in a world of pain.

Strictly speaking there is absolutely no visblocking on this map. From point a, the map is always fully rendered, or close to it. No wonder I get 15 FPS tops.

As for the top 14 custom maps, as much as I really enjoy the opportunity given by Valve, the only two maps that are worth mentioning always sit on top (that's the good news) but the amount of crap that follows is just a sign that the community is lacking. An example would be Omaha. The sole reason it's still up there is there isn't a good beach map yet. The day we see one, this one will go down the drain. At least that's the way I see it. 20K minutes a week isn't a success, it's a sign there isn't anything left to play, it's the bottom of the list.

its a nice map , looks fantastic, plays well & a few of us like to play it , but when we put it on it does unfortunately empty the server, main reasons as above FPS and probably also the download time / size

be great if it could be rebuilt with optimized fps , we'd certainly have it running regularly

Originally posted by Furyo
As for the top 14 custom maps, as much as I really enjoy the opportunity given by Valve, the only two maps that are worth mentioning always sit on top (that's the good news) but the amount of crap that follows is just a sign that the community is lacking.

Yikes, what's with the dissing dude!??

Thanks for calling my maps crap. Your map is on that list too. I personally don't play the top 2 maps ever. Personal tastes.

..and no i didn't use the carve tool ffs. I'm verrry sorry i posted the vmf.

Furyo, what goes around comes around !

:)

The custom map stats are a useful concept and should be a good way to introduce "good" custom maps to the general community.
However, some of the maps have the player minute counts because they run on Mapname 24/7 and/or on 32 man servers. That will definitely pad the stats.
It would be very helpful if the more enthusiastic server admin would update their mapcycles with the latest released versions of maps . (For example, the final version of dod_tiger blows dod_tiger_source out of the water.) Current and quality custom maps don't necessaily get fair representation and for the server admins who don't frequent the DOD Forums, the War Report, as an official Valve resource, is a place they look.

and Furyo:
I repectfully disagree that "all" the maps ranked under anvil and salerno are bad.
Churchyard, battlebridge, chemille( now coire_rc2) and harbor are solid maps for gameplay and are currently being tweaked for better performance.
Dijon is not only beautiful but plays surprisingly well in league play. Sora has been discussed to death. It was a great first effort .

You brought up some very good points Rosie !! The custom map stat page also is invaluable to the mapper for determining the balance of a map, that is, maps that have been out for awhile. Stats on brand new maps would not be useful, as people are not yet acquainted with the level yet. Sora has always favored Allies because of the easy access to the tank. By simply removing one path to the tank the map will, be balanced, I feel.

I'm sorry if I came across wrong, I didn't mean to say that all the maps that followed the first two were crap, but that the amount of crap among those was high.

My main point is this, and that's why I'm calling some maps "crap" is I firmly believe that some maps would get a lot more time if they had taken more time to make. That includes Sora, Dijon (which both suffer from huge download times which players aren't going to take time for) and others that either don't have the level of details or optimization.

So yeah, when I see dod_orange_fight_arena or Omaha get more play time than maps that are a lot better, I can only conclude that, although mappers were willing to push certain things to the players, the players didn't let them. Huge download times is a more important factor than level of details for one thing. I was willing to bet differently when I released Dijon, but wouldn't make the same mistake again.

Originally posted by ultranew_b
Where Dash?

No, I always vertex edit (wayyyyyyyy faster and also avoids texture scaling),albiet brushwork is sloppy in a couple areas in the map as Furyo pointed out. Carving is perfectly fine using rectangles.

It is a beta map, keep that in mind. I made it within 3 weeks, so it was rushed.

Ah ok, thanks for clearing it up. it just seems some of the vertex looks like its been done by carving. My bad!

People will download something larger if what they are downloading is good. If they wait 5 mins to download a empty cube, they will not be happy.

I believe a good custom map should have good custom content (textures and models) in order to stand above the crowd. You can only see the dod:s textures in so many villiage maps. The maps all end up looking the same. My new map, dod_push has a lot of new textures (and possibly models). The custom content is what makes the map unique. I will continue to make maps that utilize new content. Sora did not suffer from the size of the download it suffered from not great fps with 32 people.

However, I will caution mappers about making night maps though, they DO NOT go over well (regardless of looks and gameplay). If dod_pier were daylit it would be much more popular i'm sure.

Just an FYI, UN....... dod_push was also a custom 1.3 DoD map......



That is all.

:)

Originally posted by ultranew_b
However, I will caution mappers about making night maps though, they DO NOT go over well (regardless of looks and gameplay). If dod_pier were daylit it would be much more popular i'm sure.

i agree.
but the best DOD map of all time was a night map

dod_koln

:)

what's that Neutrino?? You're doing koln for source??



o wait til I tell my friends!!!!!!!!!!!!!!!!!!!!!!!!!


;)

im patiently waiting for Waldo to take up this project. i know hes got great ideas, just hopefully he gets time too

dod_koln would be awesome!

then i'd only love to see one more, namely dod_cherbourg.

:cool:

koln was good.

*cough* zafod FTW *cough*

We had dod_sora on our server (with fastdownload), but the people will not load it. i think, it´s a fantastic map. But when you play this map with just 5 or 6 players it gets sh....

Same problem with dod_tiger_source. Very good map, but ppl seem not to like it.

"Die Heckenpenner"

83.133.81.174:27415