Team UnrealSP's mappacks: "Déjà Vu - Gryphon Revisited" & "Battle For Na Pali"

[Hellscrag]

Reading more about BFNP makes it really sounded just too massive to be done. Plot seemed quite super complicated, 30+ NEW maps, new weapons, new monsters (making one new monster is already an over the top thing). I honestly have no idea how could you have even finished this, it sounds more like something that an indie gaming dev team would do (and iirc indie gaming wasn't still a thing 12 years ago or so).

I'd like to disagree with you, but on the evidence I guess I can't.

Looking at the number of people who were on the team, I do think that, if everyone had prioritised BfNP and truly committed their time to the project then we could have finished the maps at least. I remember being a bit shocked when RD was announced just after we did a lot of the brainstorming. I had thought the community was going to pull together to build BfNP and create the sequel that Unreal deserved, and had I known about RD beforehand then maybe we would have been a bit less ambitious. This may sound like a personal jab at Prophet but it's not really, as nobody besides myself made a lot of progress with their maps, endlessly committing themselves to too many other things or simply not doing anything much.

Modelling and coding was always going to be the greatest challenge, but these are elements that could have been scaled back without scuppering the whole project. Even one or two new creature classes would have provided a feeling of freshness.

Could it have been finished? I think it could have if everyone had done their bit (and I'm not pointing the finger at any individual here). Huge amounts of textures and some great music was produced. Enigma was well on the way to totally revolutionising the 3d engine itself. Just look at what Turboman has achieved on his own with Firestorm. But, in the end, I guess it just wasn't meant to be.

[UBerserker]

Enigma was well on the way to totally revolutionising the 3d engine itself.

What was this about actually? I don't remember most of BFNP's features but was it already in effect in some of the shots you posted? (some just looked way too good)

[Hellscrag]

Enigma was well on the way to totally revolutionising the 3d engine itself.

What was this about actually? I don't remember most of BFNP's features but was it already in effect in some of the shots you posted? (some just looked way too good)

Well with the shots, that could just be because we were using more detailed textures. As a prelude to adding S3TC textures, our new textures were imported to display at half scale, so for example a 256x256 texture would actually have a detail level of 512x512. This wasn't an engine mod, just a standard feature of Unreal texture files.

(EDIT: Well, actually, no, some of the shots do also show the distance fog support.)

Enigma knew we wanted distance fog support and he also wanted to see if he could optimise the engine so that the fog could occlude distant polygons and thereby improve framerates. Well, he found of course that these optimisations were not possible with the standard engine. UE1 was originally built to work with 3DFX Voodoo accelerators using a Glide renderer, and as such a large part of the rendering work is done on the CPU rather than the GPU. Even the most enhanced OpenGL / D3D9 renderers can only accelerate the software rendering. With an engine that worked this way, the more advanced features of OpenGL that Enigma wanted to use for his optimisations weren't available.

Well, Enigma is a coding genius - it's no surprise that he now works professionally for a game design studio - and somehow (don't ask me how), he managed to actually bypass the parts of the engine that were CPU-bound and create a full OpenGL renderer. At the time work on BfNP stopped, he had overcome the engine's node limit, allowing us to build larger and more detailed maps; he had introduced occluding distance fog; and I believe he was also going to investigate adding support for greater light radii, making it easier to add realistic sunlight and shadows to large outdoor areas. However, the renderer was not finished to a releasable state at the time the project was shelved.

Enigma got involved because he was a friend of mine at University and we actually shared a house for two years. We were extraordinarily lucky to have him on the team, and it's yet another reason to be sad that BfNP was never completed. Our legacy would have been a great array of new code and tools for future mappers to draw upon.

[Sat42]

Enigma was well on the way to totally revolutionising the 3d engine itself.

What was this about actually? I don't remember most of BFNP's features but was it already in effect in some of the shots you posted? (some just looked way too good)

Well with the shots, that could just be because we were using more detailed textures. As a prelude to adding S3TC textures, our new textures were imported to display at half scale, so for example a 256x256 texture would actually have a detail level of 512x512. This wasn't an engine mod, just a standard feature of Unreal texture files.

(EDIT: Well, actually, no, some of the shots do also show the distance fog support.)

Enigma knew we wanted distance fog support and he also wanted to see if he could optimise the engine so that the fog could occlude distant polygons and thereby improve framerates. Well, he found of course that these optimisations were not possible with the standard engine. UE1 was originally built to work with 3DFX Voodoo accelerators using a Glide renderer, and as such a large part of the rendering work is done on the CPU rather than the GPU. Even the most enhanced OpenGL / D3D9 renderers can only accelerate the software rendering. With an engine that worked this way, the more advanced features of OpenGL that Enigma wanted to use for his optimisations weren't available.

Well, Enigma is a coding genius - it's no surprise that he now works professionally for a game design studio - and somehow (don't ask me how), he managed to actually bypass the parts of the engine that were CPU-bound and create a full OpenGL renderer. At the time work on BfNP stopped, he had overcome the engine's node limit, allowing us to build larger and more detailed maps; he had introduced occluding distance fog; and I believe he was also going to investigate adding support for greater light radii, making it easier to add realistic sunlight and shadows to large outdoor areas. However, the renderer was not finished to a releasable state at the time the project was shelved.

Enigma got involved because he was a friend of mine at University and we actually shared a house for two years. We were extraordinarily lucky to have him on the team, and it's yet another reason to be sad that BfNP was never completed. Our legacy would have been a great array of new code and tools for future mappers to draw upon.

That makes me shiver just thinking of what could have been - plus Enigma is just a great handle, especially given the circumstances...
...any chance the renderer could be recovered? Not finished to a releasable state - but you got to use it for the distance fog support, and you say he overcame the node limit - for UT! That would rival Unreal 227... if I remember correctly, there was also someone working (still is, I think) on something that does to UT what patch 227 does to Unreal - maybe he could make use of this unfinished renderer (or not, but others might!)

Just really curious about this now

[Semfry]

One Enigma thing that did get finished was something that increased the vertex limit, which was related to issues with having lots of movers in a level. I just checked it (since I still have the file for it) and it seems it's linked to the BFNP OpenGL renderer that was being made though, as it seems to just ask for that now (Guess I still have that around somewhere too).

There was also BFNPDepends for making sorting file dependancies simple, which I already "unofficially" released a while ago: viewtopic.php?f=3&t=3484

[editor Dave]

I found the release of BfNPs soundtrack back in 2014 just now by chance and I was surprised how much of the story you gave away there already. viewtopic.php?f=4&t=3308

So far, this thread has been an interesting read, and I will comment more in detail when I might have re-played Deja Vu later this month!

[Teridax]

Gryphon Revisited felt disjointed at times with all the different mapping styles, but it's still pretty great.

That BfNP stuff is really interesting, especially that stuff about the vertex limit and distance fog.

[Sat42]

One Enigma thing that did get finished was something that increased the vertex limit, which was related to issues with having lots of movers in a level. I just checked it (since I still have the file for it) and it seems it's linked to the BFNP OpenGL renderer that was being made though, as it seems to just ask for that now (Guess I still have that around somewhere too).

There was also BFNPDepends for making sorting file dependancies simple, which I already "unofficially" released a while ago: http://www.unrealsp.org/viewtopic.php?f=3&t=3484

Hey thanks for the reply Semfry, if you find the BFNP OpenGL renderer together with the file increasing the vertex limit, could you make it available? (I guess Hellscrag is cool with that now) I would be interested in taking a look - and I am sure there are others too. Someone might even pick up the slack for finishing up the renderer if there are still major issues with it. No idea how hard that would be, just a thought!

[Hellscrag]

One Enigma thing that did get finished was something that increased the vertex limit, which was related to issues with having lots of movers in a level. I just checked it (since I still have the file for it) and it seems it's linked to the BFNP OpenGL renderer that was being made though, as it seems to just ask for that now (Guess I still have that around somewhere too).

There was also BFNPDepends for making sorting file dependancies simple, which I already "unofficially" released a while ago: viewtopic.php?f=3&t=3484

Hey thanks for the reply Semfry, if you find the BFNP OpenGL renderer together with the file increasing the vertex limit, could you make it available? (I guess Hellscrag is cool with that now) I would be interested in taking a look - and I am sure there are others too. Someone might even pick up the slack for finishing up the renderer if there are still major issues with it. No idea how hard that would be, just a thought!

The fixes weren't finished. The vertex limit was raised, but IIRC vertices that exceeded the limit weren't rendered.

My initial reaction would be that it would be very difficult for anyone to pick up where Enigma left off, even if they did have the sources. Any coder working on something that advanced has a thought process that it would probably be difficult to replicate / break into.

[Sat42]

One Enigma thing that did get finished was something that increased the vertex limit, which was related to issues with having lots of movers in a level. I just checked it (since I still have the file for it) and it seems it's linked to the BFNP OpenGL renderer that was being made though, as it seems to just ask for that now (Guess I still have that around somewhere too).

There was also BFNPDepends for making sorting file dependancies simple, which I already "unofficially" released a while ago: http://www.unrealsp.org/viewtopic.php?f=3&t=3484

Hey thanks for the reply Semfry, if you find the BFNP OpenGL renderer together with the file increasing the vertex limit, could you make it available? (I guess Hellscrag is cool with that now) I would be interested in taking a look - and I am sure there are others too. Someone might even pick up the slack for finishing up the renderer if there are still major issues with it. No idea how hard that would be, just a thought!

The fixes weren't finished. The vertex limit was raised, but IIRC vertices that exceeded the limit weren't rendered.

My initial reaction would be that it would be very difficult for anyone to pick up where Enigma left off, even if they did have the sources. Any coder working on something that advanced has a thought process that it would probably be difficult to replicate / break into.

I see, maybe it can't be helped at this point. Distance fog might still be salvageable though!

[UBerserker]

So question

the infamous Deja Vu artifact - the fact that you never find it, was intentional or people just completely forgot about it? I'm assuming it was next to Triskaden Station but the excavation failed hard.

[Mister_Prophet]

Honestly...I don't even remember serious discussions about it. I made the last map and I never thought about artifacts when I made my level. I think it was a McGuffin and nothing more. The way the final version of Deja Vu ended has nothing to do with it and ends on a sort of cliffhanger anyway.

[Hellscrag]

I covered that on page one!

I'll throw one thing into the mix re Deja Vu - the story of the 'artifact' was meant to be a bit further developed, but the map that was going to feature it - Nuk'Ratha Village - was never built. Willis' map was meant to feature both the valley and village, but he made the decision to split it into two maps and only had time to finish the first one. Nuk'Ratha Village was going to be a battleground, with Skaarj and Mercs vying for possession of the artifact. Having said that, I don't think we ever developed a very clear idea of what the artifact was going to be. Remember - the first version of Deja Vu was speedmapped over the course of one month to celebrate the anniversary of the founding of UnrealSP.Org.

[UBerserker]

Oops I missed that I guess. Truth to say while I think Map 14 is where things go a bit downhill in DV2, Map 13 felt like something was missing - it truly had a filler aura to it. It was also extremely linear and the boss fight location didn't work at all. Wasn't Willis also working on something else at that time?

[Hellscrag]

Oops I missed that I guess. Truth to say while I think Map 14 is where things go a bit downhill in DV2, Map 13 felt like something was missing - it truly had a filler aura to it. It was also extremely linear and the boss fight location didn't work at all. Wasn't Willis also working on something else at that time?

I also struggle with the last few maps of Deja Vu, and quite a few of my play-throughs have faded away after map 12 or map 13. Willis' map is decent but is lacking in purpose without the accompanying village. Maekh's map is strong in many ways but doesn't quite capture my imagination the way a Nali fortress should (maybe the return to broad daylight doesn't help). EBM's map is just a bit of a slog to get through due to its format, and Prophet's map, while very good, feels like an Xidia map rather than a Deja Vu map. With the exception of Prophet's map, none offer much in the way of subplots or story development.

EBM and Prophet's maps diverged a lot from my (very sketchy) brief. They were meant to be part of the same location, an outdoor map with ruins leading to a more intact temple that had been converted into a Skaarj base. The small area of ruins close to the beginning of EBM's map is the only real clue to this intended continuity. And, of course, the maps were most certainly not meant to be connected by a teleporter - a mechanism I'm often not keen on and didn't like in RtNP, as it feels like it invalidates much of the player's journey (I make an exception for the teleporter that links Outpost 3J to Velora Pass, as the Sunspire is clearly visible from the landing pad so you're still in the same general area as you were before).