Notpron Forums

2COOL4-U wrote: I calculated the needed bandwidth for such a thing in the past with another coder and it's quite much.

Besides there's another point that's perhaps more important: Security. It will be quite easy to fake these POST messages as all the UnrealScript code of Deathball is available to anyone. So even if we would implement some sort of encryption or other security measurement, it still would be way too easy for script kiddies to crack it.

db_10-38_4-1-2004.txt is 1KB. 1 server uploading 1KB (or even 10KB) every 20 minutes really isn't much.

Good point about people POSTing fake entries. I guess the best (which doesn't stop it) method against this is to have basic authentication/ip access lists and the website admins issue username/passwords to the deathball server admin. Any abuse from that server you delete their entries from the database and close their username/password account.

Can't think of anyway to stop it since they know the protocol.

Maybe you could do it the other way around, the website owner could query the server with a script for statistics.. makes it a bit hard to get the end of game results though. Hmm, actually, maybe the deathball server could listen for a 'send me the end of game results (also auth details)' request from a remote URL and then at the end of the game the DB server updates that URL. Sounds a bit over complex though.

db_10-38_4-1-2004.txt is 1KB. 1 server uploading 1KB (or even 10KB) every 20 minutes really isn't much.

1 server every 20 minutes, with 100 servers that'd mean 300 kb/hour, 300 * 24 * 365 = 2628000 KB = 2.6 GB worth of data traffic. This excludes the POST headers and servers which have shorter matches etc. I'd say 26 - 40 GB a year is more realistic. I didn't know the DB webpage used over 2 TB of data traffic at first. That's the reason I found 2.6 GB much(It still IS much)..

And DavidM disagrees as well. I am sorry no online statistics.

Anything on the bug were the keeper does not have keeper powers on restarts. (another player who spawned elsewere has it.)

Twigstir wrote: Anything on the bug were the keeper does not have keeper powers on restarts. (another player who spawned elsewere has it.)

This only happens in maps in which the penaltyzone contains DB_NormalPlayerStarts

todo list updated on http://2cool.free.fr, ow please note that is are only my changes

A couple of things:

- You can score a goal by walking past the outside on one of the goals in IceDemon (red goal I think)

- Stats don't always cache (Messy scored a few goals. reconnected and it said he hadn't scored any...This was in db-slamball...Haven't noticed this in other maps, but it may apply)

2COOL4-U wrote: This only happens in maps in which the penaltyzone contains DB_NormalPlayerStarts

I already know that but, it's still a bug.

2COOL4-U wrote: UScript does not have support for real file I/O. We can only write things to logs and nothing more. There's no way we could do that from UScript. If Server admins want to do this, they will have to write their own script in another language and use cron(don't know what Windows uses for this ) to schedule it.

yep... cron scheduled ssh transfers.. it's fairly easy to do for a serveradmin.

Twigstir wrote: I already know that but, it's still a bug.

It's not a bug, more a design decision we took. I'll see what I can do though

for the emitter bug, its deffinately somthing to dow ith replay. if possible i'd make it so emitters don't get activated in the replay sequence to see if that'll rid the bug. if possible anyway

2COOL4-U wrote: 1 server every 20 minutes, with 100 servers that'd mean 300 kb/hour, 300 * 24 * 365 = 2628000 KB = 2.6 GB worth of data traffic. This excludes the POST headers and servers which have shorter matches etc. I'd say 26 - 40 GB a year is more realistic. I didn't know the DB webpage used over 2 TB of data traffic at first. That's the reason I found 2.6 GB much(It still IS much)..

And DavidM disagrees as well. I am sorry no online statistics.

What about only supporting gs.net servers?

2COOL4-U wrote: It's not a bug, more a design decision we took. I'll see what I can do though

This was on the NormalPlayer_Starts --> spawn with keeper physics etc. bug (right? ;p)
I suggest resetting the person who gets highest priority on keeper physics everytime a round is scored, or maybe just link it with the quasi-random keeper spawning so that @ start of round the keeper physics are always assigned to the one that spawns in the goal.

It's not a bad thing for mappers to put spawnpoints in pzones methinx, specially on smaller maps. Hope you fix this

Onge wrote: A couple of things:

- You can score a goal by walking past the outside on one of the goals in IceDemon (red goal I think)

- Stats don't always cache (Messy scored a few goals. reconnected and it said he hadn't scored any...This was in db-slamball...Haven't noticed this in other maps, but it may apply)

- Do you actually have to have the ball?

- I scored a goal?!

edit: Several more minor things (I didn't feel the need for a new post)

- Fixed several accessed nones.

What are accessed nones? ^^;

- Fixed 3rd person bug(if you manually went into 3rd person view and the code
reset it to 1st person, you wouldn't be able to aim up)

Does this mean 3rd person aiming is always fixed/possible now? Or is only the 'respawn and not being able to aim up'-bug fixed?

- New pickup routine

Now that David is here, can he explain it? (I seem to recall jero not being able to explain since he didn't get David's explanation ) Either david or cata, or jero if he now understands, could you plz explain? ^^;

What about only supporting gs.net servers?

We'll see, since there is such a big demand for it, I'll discuss it. But don't count on it.

This was on the NormalPlayer_Starts --> spawn with keeper physics etc. bug (right? ;p)
I suggest resetting the person who gets highest priority on keeper physics everytime a round is scored, or maybe just link it with the quasi-random keeper spawning so that @ start of round the keeper physics are always assigned to the one that spawns in the goal.

It doesn't work like that

It's not a bad thing for mappers to put spawnpoints in pzones methinx, specially on smaller maps. Hope you fix this

I will

What are accessed nones? ^^;

It are errors that can make your log files very big, I won't discuss what they mean to say.

Does this mean 3rd person aiming is always fixed/possible now? Or is only the 'respawn and not being able to aim up'-bug fixed?

I only fixed the bug. Deathball is a first person game

It is me, or is crouch still bugged in 3rd person?

Originally posted by 2COOL4-U
[I only fixed the bug. Deathball is a first person game

well then or remove the 3rd person option or get it to work properly