Spring in a browser window

[Beherith]

Is this even possible? Because I would like to try to implement minimods running like this, with no install needed. Since the dl can be trimmed to about 30 megs, which is totally acceptable by todays dsl standards.

[imbaczek]

quake did it, i don't see why spring couldn't do it.

of course, then there's a question of actually _doing_ it

[CarRepairer]

Start with a flash-based replay viewer.

[imbaczek]

unpossible

[aegis]

not quite... flash has a 3d engine... given the time and motivation, I could make a replay viewer.

[imbaczek]

remember a replay is only network traffic saved to disk, you'd have to implement all of spring's sim code down to the fpu instructions. it's possible but extremely unfeasible.

[aegis]

yep

[Tobi]

Wasn't there a C++ compiler outputting flash bytecode?

[aegis]

As an example of the complexity of this project, the engineers ported the C++ version of Quake 1 engine into Flash Player and it worked perfectly and performed well. It is an extreme case but showcased the potential of this project.

http://brajeshwar.com/2008/cc-to-actionscript-compiler/

[BrainDamage]

As an example of the complexity of this project, the engineers ported the C++ version of Quake 1 engine into Flash Player and it worked perfectly and performed well. It is an extreme case but showcased the potential of this project.

http://brajeshwar.com/2008/cc-to-actionscript-compiler/

good luck in replicating all the haxes that boost relies on

[CarRepairer]

Assuming the spring sim was was implemented in flash (I can understand the point that this is a ridiculous task), I was imagining just a 2D replay viewer, no need for 3D. So you could view replays as dot wars, isometric, or some other 2D way. It would be cool.

[aegis]

you'd still need to implement all of the physics

[zwzsg]

You're still missing the point that replays only stores players input.

To know where the points are, you have to resimulate the whole game. Completly, in 3D, down to every cob instruction idiosyncrasy, down to zero error on 3D models vertice positions, because if that sumo was 1/30 smaller, then that plasma shot would have missed it, crashed into the morty behind, which would have died sooner, so wouldn't have been able to finish that fusion 3 mins later, which in turn ....

[aegis]

not missing the point at all. it would basically need to be a reimplementation of all of spring's simulation code.

[zwzsg]

I, of course, wasn't replying to you, but to Otherside.

[CarRepairer]

I, of course, wasn't replying to you, but to Otherside.

A) Why are you calling me otherside? The spelling of my screen name isn't remotely close to it.

B) Do you often skip the first two lines of posts when you read them or only this time?

Assuming the spring sim was was implemented in flash (I can understand the point that this is a ridiculous task),

I feel doubly insulted.

[zwzsg]

A) Why are you calling me otherside?

I don't know. Sorry.

Maybe because both of you have a round avatar, expect yours is square?

Assuming the spring sim was was implemented in flash (I can understand the point that this is a ridiculous task),

Looks like I missed the word "sim" too, sorry.

I was imagining just a 2D replay viewer, no need for 3D. So you could view replays as dot wars, isometric, or some other 2D way. It would be cool.

That point about going 2D made me feel that you thought there was a way to simplify the engine, so I went on explaining why the sim has to be in 3D. Maybe I should have instead talked about how unlike most commercial games, Spring is more CPU intensive than GPU intensive.

I'll go sleep before I type any more mistake.

[CarRepairer]

It wouldn't have been such a big deal if it were any other name but that one!

[el_matarife]

Use Google's native code sandbox plugin: http://arstechnica.com/open-source/news ... plugin.ars I'm pretty sure it can make OpenGL calls judging by the Quake port.

[MidKnight]

It wouldn't have been such a big deal if it were any other name but that one!