Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Cassandra Lag
#41
Leaves Wrote:Could anyone confirm whether this issue exists (or has existed previously) in KMS? I'm sure this isn't a system exclusive to GMS, so...

KMS has the exact same data in the files for Gaga/Cassandra. Also, I have an i5-2500k @ 4.2 and an SSD. I experience no issues, of course.
Reply
#42
Halloween Cassandra and Gaga changed at midnight too as far as I know. (Most of them changed at midnight)
Reply
#43
Fiel Wrote:I also think their implementation would be better by using a Unix Timestamp rather than what they currently use. It would just be so much better.

Wait, are they one-time events or annual things? If it's one-time, it's an even stupider setup than I'd thought. Check every second for something that happens once, ever... woo.

I mean, why not just have the server broadcast the NPC's style when you enter the map or it changes, and then have the client keep using that recorded value to know which sprites to use.

Sigh.
Reply
#44
Stereo Wrote:
I mean, why not just have the server broadcast the NPC's style when you enter the map or it changes, and then have the client keep using that recorded value to know which sprites to use.

Sigh.
Because this would require far too much competence on the part of the dev staff.
Reply
#45
They have so many useless characters in every town (Gaga, Cassandara, Ms. Approperiation, Inkwell, PVP guy, etc). I've always thought they should make a town just for them and we can use the mirror to go there whenever we feel like it. It really ruins the individual feel of a town to have it filled with NPCs that are everywhere.
Reply
#46
Wait, then why is this issue "GMS-exclusive" only?

I don't recall people reporting this issue in KMS forums, nor did I see any complaints in MSEA forum.

*GMS Cassandra causes lag on my old computer, but not MSEA and KMS.

Anyway, GMS client memory usage always make me wonder what are they doing with that much memory.

Running MSEA right now with 78MB memory (stationary), 200~300MB while training. It might shoot up to 400MB in certain bosses but that's it. My GMS client could use up to 800~900 MB.
Reply
#47
Killmeplsok Wrote:Wait, then why is this issue "GMS-exclusive" only?

Obvious answer - Do the other versions have these time-based-changes programmed in?
It's not having what you want - It's wanting what you've got.
Reply
#48
Eos Wrote:Obvious answer - Do the other versions have these time-based-changes programmed in?

Not sure but...

Fiel Wrote:KMS has the exact same data in the files for Gaga/Cassandra.
Reply
#49
Killmeplsok Wrote:Not sure but...

Then clearly it is the poorly written monkey code of the GMS localization.
Or the clients just treat it completely differently. Or B via A.
It's not having what you want - It's wanting what you've got.
Reply
#50
Killmeplsok Wrote:Wait, then why is this issue "GMS-exclusive" only?

I don't recall people reporting this issue in KMS forums, nor did I see any complaints in MSEA forum.

*GMS Cassandra causes lag on my old computer, but not MSEA and KMS.

Anyway, GMS client memory usage always make me wonder what are they doing with that much memory.

Running MSEA right now with 78MB memory (stationary), 200~300MB while training. It might shoot up to 400MB in certain bosses but that's it. My GMS client could use up to 800~900 MB.

I think you sort of answered your own question.
The other versions use much less CPU and memory (I presume because GMS is bloated with anti-hack crap).
So maybe being near Cassandra does use more CPU in the other versions, but they are not pushing the limit beforehand, so it makes no noticeable difference.
Similarly, other versions may eat up a lot more memory bossing, but it doesn't cause them problems since they are not already near the limit before they start.
On both my old comp and new, memory usage is around 900M before I even get logged in.
My old comp only has 1.5G total, so I got problems. New comp 8G, no problems.

EditEdit:
The real mystery to me is why does GMS require three times the memory. Are they making multiple copies of everything to help detect hact alterations? Seems really poor coding whatever the reason is.
Reply


Forum Jump:


Users browsing this thread: