Live data from Hacker News

Artillery is building a hardcore RTS with HTML5 and WebGL

blog.artillery.com

101–110 of 151 posts

Re: Artillery is building a hardcore RTS with HTML5 and WebGL

#101
post #19
post #11

This looks really great. I just think the claims of "console quality" or "AAA quality" (which I heard associated with Day[9]'s announcement) are a bit disingenuous. I'm excited that you guys are pushing browser gaming, and what you've done so far looks great. Just please don't fall into the trap of over hyping and then disappointing customers.

CEO here: Thanks for the feedback. 'Console quality' and 'AAA' are very subjective and we totally understand your points. Our job is to make Atlas and anything on our platform objectively feel amazing and please keep in touch with us.

> Everyone deserves empathy and respect.

Respect for putting that on your company page. It's good to see core values like these!

Re: Artillery is building a hardcore RTS with HTML5 and WebGL

#102
post #99

I will be exceedingly impressed if they pull this off. RTS is probably the gaming genre that offers the most "interesting" problems in programming, and thus is the one that needs the most computing power beyond simply showing stuff on the screen. Pathfinding is Hard. Maintaining sync is Hard. Unit AI is Hard. And keeping those performant is doubly so, especially using JS. So yeah, I'll be impressed if they get anythi…

I'm not sure I'd that path-finding and maintaining sync are hard problems, especially since they are problems that in an RTS aren't quite as sensitive to latency.

Re: Artillery is building a hardcore RTS with HTML5 and WebGL

#103
post #99

I will be exceedingly impressed if they pull this off. RTS is probably the gaming genre that offers the most "interesting" problems in programming, and thus is the one that needs the most computing power beyond simply showing stuff on the screen. Pathfinding is Hard. Maintaining sync is Hard. Unit AI is Hard. And keeping those performant is doubly so, especially using JS. So yeah, I'll be impressed if they get anythi…

I'm not sure I'd that path-finding and maintaining sync are hard problems, especially since they are problems that in an RTS aren't quite as sensitive to latency.

Ask anyone who's ever worked on an RTS and they'd disagree with you. It's not even difficult to find evidence to support them being tough problems by googling it:

https://www.google.com/search?q=rts+pathfinding

Lots of people asking for solutions, lots of different solutions, lots of caveats.

Re: Artillery is building a hardcore RTS with HTML5 and WebGL

#104
post #82
post #74

Earlier quoted context omitted.

This is exactly what they said about Java ten years ago (and I still hear it today): "Java is way to slow for games/anything performance-critical". The choice of Java in your argument is kind of ironic to me.

Minecraft isn't exactly a very demanding game. Also 10 years ago Java was nowhere near as fast as it is now. If you are saying that Javascript might be an option 10 years from now, I am not disagreeing with you.

Its actually very demanding, CPU and memory wise.

Re: Artillery is building a hardcore RTS with HTML5 and WebGL

#105
post #31
post #20

This is impressive and cool, but I think web-based gaming will never catch on with the 'hardcore' for a few reasons. For one, C++ and the native toolchain is still more efficient. But it mostly comes down to perception and convenience. The perception of web-based games is that they're casual and cheap. And most hardcore gamers like to have a copy of the game on their computer - they paid for it, they want a copy that…

I'm not convinced that web-based is automatically doomed among the hardcore. I think that Quake Live the most popular game in it's subgenre (Arena FPS), and it's still hanging on as an esport, with a slot in Dreamhack as well as weekly tournaments. Granted, Quake Live has the massive advantage of being a resurrected version of Quake 3 Arena--but it does make a precedent in people's minds of a hardcore web-based game.

QL uses a native binary that you install, no? Like BF3, the only components handled in the web browser are matchmaking, clan management, etc. For the actual matches/gameplay, it invokes the native client

Re: Artillery is building a hardcore RTS with HTML5 and WebGL

#106
post #99

I will be exceedingly impressed if they pull this off. RTS is probably the gaming genre that offers the most "interesting" problems in programming, and thus is the one that needs the most computing power beyond simply showing stuff on the screen. Pathfinding is Hard. Maintaining sync is Hard. Unit AI is Hard. And keeping those performant is doubly so, especially using JS. So yeah, I'll be impressed if they get anythi…

I think the reasoning behind using an RTS is likely that the company wants people to write inherently online games on their platform as part of whatever monetization strategy they have - and the other big alternatives are the MMO and MOBA genres, which would need a lot of content to be made. Having a couple of bright programmers on call - the designer is both a Starcraft celebrity and a mathematics nerd, so he likely can chip in too- would no doubt be cheaper than hiring enough artists/character designers or voice/motion capture actors to make a respectable game in some known hardcore online game genre...

Re: Artillery is building a hardcore RTS with HTML5 and WebGL

#109

Wow. I wish I had even just half the talent of HTML/CSS/Javascript to pull something off of this calibre. I am really impressed, I can't wait to get my hands on this and deconstruct how they did it the best I can and play it too.

I think you should pickup openGL/webGL instead of that other stuff, if you are serious and not trolling. Because I really can't tell at this point in the thread.

Re: Artillery is building a hardcore RTS with HTML5 and WebGL

#110
post #29

Earlier quoted context omitted.

We may be seeing the same HTML5 demos on HN, but we're living in two different universes. In my world, I always open up these demos in a browser that's incompatible. If I do manage to get them working, my reaction is always like wow, I saw better native demos 10 years ago, and those just worked. Let's be honest. JavaScript, DOM, and CSS are terrible technologies to base any gaming system off of. As a gamer, performan…

It's easy to talk trash on a technology that is in its infancy. Wait... WebGL games are still not as good as game engines that have been developed for 20 years? SHOCKER. The technology is only going to get better. The whole point of these demo's is to show what is possible even now when the specs aren't finalized and the technology is barely adopted. Give it 3-5 years and I would be willing to bet you'll need to eat…

> Wait... WebGL games are still not as good as game engines that have been developed for 20 years? SHOCKER.

That's the point, isn't it? We already have all these fast, battle-tested game engines Heck, we have 15-year-old game engines that run on the browser (or do Flash games no longer count?). Is it really worth it to toss all that out and switch to WebGL?

Post reply on HN