So what's left to do? I've got time on my hands lately, and I'll be glad to help with the 2.0 stuff. I'm curious if any major changes are coming anyway, since I've been working on a 3D drop in replacement for the MapCon (This will clearly be more processor intensive then the current one).
I have GroundItems and a few other little stuff to compelete. I've been working on the packet handleing system to try and weed out bad packets. I've been running the new changes for about 6hrs now and it has only filtered out 6 bad packets so it looks like it's working now.
I might setup up Server side processing for the SpawnList and it's job would be scan the spawns and only send the spawn data that has changed. For example if your in the Bazaar there might be 600 Spawns, under this thought process it would send all 600 spawns the very first time then it would only send the spawn data for those spawns that the X,Y locs have changed. This would reduce the spawn data down to say 20 spawns or so that are running around the Bazaar. This would speed up the client a great deal while slowing the Server down a little. I don't know what the trade off on Client - Server CPU utilization would be but currently the Server is at 0% to 2% utilized. So we could increase that to maybe 5% to 10% max and might shave off 10% to 20% off the client (which is currently running about 35% to 45%).
That task may come a little later after 2.x has been released. The CPU increase with the 2.x is very minimal compared to the large amount of data being handled.
Just my 2 cents....so tell me to shut up when you would like....
I personally want the server as lean as possible...(client can be a PIG for all I care)...as I am running EQ on a box that does NOT have as much juice as I would like, seems like any addition load would not be a gerat thing...
(Course the brain-trust here might just tell me that this is a REALLY miniscule thing..and I should SHUT MY DAMNED MOUTH)...in which case I will just say...
IF ya don't hurry....the prods are gonna get BIGGER, SHARPER, LONGER...AND...ELECTRIFIED.....LMAO!!!!!
hmmmmmmmmm ouch
or should that be....
BZZZZZZZZZZT......ouch....DAY-UMN...lol
I like the abuse (wipe me, beat me call me Edna)
I personally like the balance the way it is. Most of the time I'll have the MySEQ server open, a window or two of everquest, and sometimes a window of subspace open all at the same time. If the server takes up more processor time than it does right now, my computer would be a hurting unit.
Just speaking for myself, I like that the client does more work since I run it off a seperate machine. I doubt it could handle something like a 3d map since its about 4 years old, but it takes on the current client just fine.
If I do it, it would be configureable on turning it off/on at client/server.
I agree about the light processing of the client, I personally prefer as little as possible there as I have other tasks that need to co-exist wth my EQ habit. As myseq requires a network connection (most liekely at least 10Mbit) anyway it shouldn't be a big deal in most cases.
3D is a good idea :P
Seriously, though, one thing that always miffed me about MySEQ was the Z-filter. The map doesn't Z-filter at all, and if you turn on depth-filtering, your map just shows the bottom-most areas of the map 3d-wise.
Another nifty lil' idea for the client (I know this is the wrong thread n all, but hey) would be the possibility for 2 child maps in the parents.
So you could zoom in heavily on your surroundings in one map, and on the other, either a zoomed out zone-wide map, or the immediate vicinity of your selected mob, or whatever you like.
I'd find it handy, personally, as a pulling monk and quadding wizard. But if it'd be a headache to program or a huge system drain, nevermind :P
Edit: Keep up the good work, bro. Wish I could lend a hand, but it'd take me ages to catch up to where you're at. I'm a competent C and VB programmer, but my knowledge of network programming and memory handling isn't so hot
A 3D showeq view is something I've always wanted, but it introdues other problems (which is largely why it's not done...). The EQ S3D is utterly insane and I have no clue why it was designed the way it was besides chasing off decompilers, and lets not even get into the WLD format... Wall transparency is the major problem I have at the moment, if someone knows how to pull raw texture data out of a managed D3D texture I would love to know.
Why do you need the textures ? Wouldnt it be enough for a map tool to present only a wireframe model ? That way you could orientate and lokate mobs in 3D and wouldn have to hassle with textures especially as loading up a whole zone with all textures would be a killer with quite a lot of CPU usage.
Oh and you can convert the s3d files or rather the wld files into .x files which managed DX can read in easily. I use a mixture of ZoneConverter and 3DS Max though i can by no means get that stupid .x Converter from the SDK get compiled. Thats whats hindering me atm to experiment some more with a 3D Map. Oh and the fact the DirectX 9b doesnt know at all what Managed Code is.
The Z Filter works but the issue is the map directory needs to be deleted and all the maps need to be redone from scratch hehe.
Most maps was done in a XY format vs. XYZ format so there is no depth to the maps.
You should be able to filter the spawns via Depth Filter.
There are a few ideas I have had with mapping. The biggest thing is to redo it so the zone is drawn on a offscreen bitmap once then stretchBlt it to the working copy off screen for only the area needing to be shown. This would then be copied to the onscreen area (that's called double buffering to prevent flickers). We currently redraw the map during every poll cycle with the double buffering.
I have written a Map viewer in VB 6.0 that shows the map in one pane and then allows you to Zoom In/Out in another pane and as you move the mouse around in the normal pane it copies the area to the Zoom pane. This is very quick since it's using the StretchBlt techniques. This was the beging of my personal SEQ type program. But I decided to help this project out and learn C# even tho I know VB like the back of my hand.
LONG LIVE VB!
I've always like I wanted to contribute to this project in some way (apart from money, since I'm a poor bastard.) If you can think of anything a learned VB programmer and partially trained C/Java programmer can do, lemme know. I also run an EQEmu server at times if somehow that could be any use :P
I just wished one day I could learn C# so I could program in it.
I have everyone fooled here in thinking that I know what I'm doing hehe. Reality is I probably should be put in a white jacket with the buckles in back with nice white walls with pads.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
Reply With Quote