Bring back clientside bombs...
 

I want clientside bombs back.... instead of having serverside bombs bring back the old clientside bombs just add more protection.

They made the bombs serverside so they would be harder to hack but they just opened up new flaws...

Yeah I hate the new serverside bombs, they have a huge delay making pking with them near impossible, even with low latencies like 40 MS.

The people who made the iPhone servers are some of the worst scripters out there. There are zero lag compensation techniques on any of the servers and it's way important to have those on a mobile platform. You can make serverside bombs that feel and behave identically to clientside ones without being easily hackable. This takes a little effort and knowledge of networking protocols and lag compensation/prediction algorithms.

That being said, the serverside bombs are actually LESS secure, use more networking overhead, and are very awkward to use in normal play. Whomever made them is half ******ed.

I completely agree.

Have no idea what any of the mumbo jumbo meant but I think I agree.

then... make better ones?

I'm sure Downsider has better things to do than help a game in which he doesn't play.

Hi mom, I'm on TV!

Wow I wasn't expecting so much support! Thanks guys!

I don't understand why they would change it either way there will always be hackers and glitches

serverside bombs is the most ******ed idea graal online has accepted after p2p restrictions.

The problem is default bombs are insanely limited in what they offer and what can be done with them. Serverside bombs are the simplest way of achieving alternatives, and the only person who has ever made the claims that Downsider has made were Downsider. I doubt you'd find many other scripters other than him that when presented with the problem: "Hey, I need to make it so players can place bombs that explode and sync with other players" wouldn't say "putnpc2 and join to a bomb class(serverside)."

I'm not saying Downsider is wrong, but he's been the only person I've ever seen make such claims to such techniques and even the best scripters around tend to not come to the same conclusions... so I don't think it's right to claim that it's bad scripting to use the method used.

Simplest way in game networking is nearly always worst way. The original Quake used absolutely no prediction in multiplayer, but it was an elegant and simple solution. The client acted as a termnial, sending only input to the server and receiving entity positions, then drawing the world and the state of the game. All of the game logic was performed on the server. However, because of that design choice, it was only playable over local networks because of dropped packets and latency over the internet. Imagine if every action you did in a game was delayed by your latency. Now mix that in with Graal's bombs, over the iPhone's wifi or 3G network, on Graal's single threaded, crappy ass servers, and it's fucking ******ed.

When NetQuake came out and added some prediction algorithms, Quake's multiplayer scene soared in popularity. Prediction, masking lag = better user experience. How can you even argue that?

putnpc2, by design, is totally counter-productive to anything prediction related. Something as performance critical as PK'ing should have its own entity system.

It's bad scripting because it doesn't work well in practice. What more is there to it other than the user experience?

Sure, writing your own clientside-serverside predicted and sync'd entity system is probably outside the grasp of your average Graal scripter. However, from the perspective of someone who has worked on multiple game engines from scratch, with a ton of experience in making multiplayer games over networks that are highly volatile, lots of packet loss and latency, the proper way to do it is not iClassic's approach.

That's why I think differently. I'm always concerned about the end user and presentation. Something as performance critical as PK'ing shouldn't be done like this. If you're going to make something serverside, do it fucking right and write a prediction system. Otherwise, if you're a terrible programmer, then fuck it and do it clientside with events that are sent out to other players like the original, builtin bomb system.

yea I think I'm going with downsider...