notice how he didnt say the game which has already pushed the boundary and makes his statement false when it comes to open worlds. EVE Online has one sandbox. a space boundry thats huge and just crazy in imagination. polys are higher then sto and video cards are still stream lined. Eve online has been doing this for hmmm over 7 years with a huge sandbox. comaring wow to sto is not true facts as wow is on land and sto is in space or the portion he was talking about is space wise.
also in the borg simulation ed really does not know the game that well as when you die you just respawn. there is not much penalty to this attriubute.
voice chat was not needed or rather that ball was missed and should have been put on the back burner as to content and better features.
Eve is not an open world. It's instances with cleverly camouflaged loading.
It's an interesting interview. Dstahl is focusing a lot on shooter mode at the start, which I guess is cool (albeit not all that helpful to anyone who actually plays the game since the mode speaks for itself well enough). There's a good discussion here on design where he says (wrongly) that the old style is still there. It isn't. Pacing effects the style of play - a lot. Where before you could play the game with 1 -1.5 second reaction time you simply can't do that and survive anymore. That may or may not be a good thing, but both parties seem to skim over an interesting problem. Too tolerant of lag is boring, especially for power gamers or younger twitch type players, but changing from a game where you could spin your chair 360 degrees between abilities to one where you can't wait more than about 300 ms between abilities is a major change. Not to draw too deep a comparison but that was part of the NGE experience in SWG, that suddenly people who could play the game before couldn't. I'm not sure if that's the case with STO yet, but certainly you've locked people out of content with it, it just might be worth the trade off to have a much more engaging experience.
There's a great question about do you still get 'one shotted' (or as he puts it 'insta party death'). The answer to this question is "yes" and it's worse than before, yet Dstahl works hard to dodge the question and talk about how bosses are in general harder, and there's a more diverse range of challenges. Again there's interesting design here. WoW cataclysm intended to be an interesting counter point. You can have faster paced, more intense, but less one shotting - more health and so on. It tends to make combat more 'averaged', you push more buttons, and they're all important, but it's less bursty and more tolerant of one wrong button press. STO went from averaging over spikes to just spikes, bosses are decent for player dps to average, but they can burst very hard back on players - which requires a lot of pro-active setup. A very different style of play than it was before.
He makes an interesting (and as far as I can tell factually untrue) statement about WoW deciding to use less polygon resources to allow for seamless worlds. That isn't how it works, at all. Rather than loading an entire level at once (as one tends to do), you use your define a volume around a player, and stream level data into and out of that volume. From a content creation perspective distinct levels are easier, you don't have to stitch them together after, everyone gets there own little area. But it's not that it uses less polygon data. It streams polygon data, you can have higher detail in your volume because you're necessarily only storing the visible volume + a bit (you can calculate the + bit part based on how fast the player is able to move and how quickly you're moving data), or you can use less detail if you want. That has very different performance properties than defined levels though. Dedicated GPU ram with high bandwidth works well, but it can be an out of order read if you're not careful which causes some grief. There are lots of good algorithms here to solve this problem. I'm not sure STO would get a lot out of them with a whole lot of radical changes though. The space overview is where it belongs, and that would be by far the easiest to implement because it's no longer a volume but an area, and you can define small chunks and stream small chunks (heck, some people could probably fit nearly every area into memory and not need to stream much). Spatial subdivision algorithms are good. WoW uses less polygons because they wanted to hit a wide system requirements base (compare to say EQ 2). Getting rid of local and global memory lets you put a more diverse range of stuff in memory (think the 80 or something hunter pets, 200 armour 'tier' sets a few thousand more general armour pieces etc.) all at once, but you have less there for level geometry in hub places and so on. It requires a lot more sophistication to do well, but if you want a character focused experience it's pretty cool. STO of course can use a lot more memory than WoW which gives you some other options too.
The borg stuff is good. We really need a sticky post somewhere with a translation of "Remodulate" in all the various client languages since we can't explain to everyone how to do it. But he gives a good overview of how it works.
The voice chat question was funny, because it was a giant softball and Dtahl hit it out of the park. Well done. There's a bit of work to be done with it (and maybe some default settings to be tweaked) but definitely something that was needed.
The queue question was interesting, because well, no offence, it doesn't work right. It's good to hear the thought process, but there's also nothing than hasn't be hashed to death on the boards. It's a great concept, assuming it can be designer and developer tweaked, but it's locked a lot of people out of content they'd been doing, and enjoying for ages, which is bad. I think the follow up question belongs as "I spent 40 minutes in a queue with no idea how many other people were waiting, never getting to do content I could solo or do with 3 people last week, what are doing to get that functionality back quickly?". Triage in game development, sort of a depressing topic though.
Edit: changed a "no" to a "yes" (the meaning should be implied from the rest of the sentence but I apparently screwed it up. the actual question in the interview is a "is this problem fixed" whereas I am referring to as "is this problem still there").