A quick Update

[Silirrion]

I just wanted to give a quick update on things going into the weekend and explain a little on the status on test-live.

One of the things we haven't mentioned yet with this update is a key part of the process and something we had wanted to keep 'under wraps' as it were until we thought it might work as it most definitely falls under the 'man, will they ever fix...' category.

Basically a little while ago one of the coders came up with a new way of the network handling 'action' requests and how they were queued in the system. The result of this should be that we finally see the back of 'already executing' messages when an action is blocked by a passive refresh, tic or forced cast that isn't under your direct control, as in auras and several other actions. This is obviously something that has been brought up many times over the years.

The fix though is risky and involved an extensive re-write of the related sections of the code. It also may not remove the problem entirely, although we are hopeful that it will.

Currently the fix is on test-live but is causing some crashes both on client side and server side and will require additional testing and work.

Given the fact it is one of the things people have most wanted fixed down the years we feel it is worth taking the time to try and include with this update. We will be continuing work on it over the next week and see what the status of the related issues is on test-live as we do so. (and on a side note obviously any additional feedback from test-live will help narrow down the issues, so there has never been a more helpful time to spend some time on the test dimension)

If we don't make too much headway on resolving the related crashes we may consider pushing it out while we work on it some more, but we will take that as it comes.

In terms of the raid interface itself, the work is progressing nicely. We are still looking at the loot distribution and exp distribution implications of raids and that's the current focus.

So there you have it, that's where we are at with the update. We will of course keep you updated as we move along.

[Contra]

w00t nice job!

[Xaun]

Awesome work Sill n co. This is the kinda stuff that gives us old fogies hope for tomorrow !

[Lukew]

The NTs will definitely make love to you if pulled off successfully.

Nice work

[Ober]

I R Luvs Yu

[Klodovik]

Great stuff.

P.S.
As long as it doesn't fux up pet pathing even more, it's fine.

[egadsrk2]

Exciting news, this is one of the big ones, that affects everyone to a degree. My first thought was 'Damn, when the name change system comes in I'm so playing my NT again!'. Was always the plan, but not one I was ever excited about.

Doctors would probably be extremely pleased as well. And anyone effected by this during PVP, where it's painfully more apparent, for casters.

[bungerman]

yes nice job and I can see you are slowly but surely working your way towards pet pathing rewrite...right??

[Darkbane]

Probably worth also mentioning that zoning is rather snappier on test these days too, another little tweak from Macro's almost infinite band of coding monkeys

[Lupusceleri]

Yarr, nice work with the Already executing bug!

Also zoning activates MUCH faster now than on live, great stuff

[Chrys]

wtf nrf aNTeeess!!!1!

[Xaun]

yes nice job and I can see you are slowly but surely working your way towards pet pathing rewrite...right??

I am pretty sure that till the new engine is released you won't see much progress on that front because of how the old engine is handled.. I could look up the various DEV posts on the topic from before but safe to say they've explained it is as optimized as it can get using this existing code base.. unless something has magically changed recently

[Disodoctor]

Doctors would probably be extremely pleased as well. And anyone effected by this during PVP, where it's painfully more apparent, for casters.

Not really, now I need to find another excuse for why I didn't heal someone

[egadsrk2]

Okay, all the doctors except Disodoctor will be pleased. It's one of their most popular complaints.

Heh, I was standing all alone when Comp Mochies got their time debuff and everyone did the ticketape parade except me who needed a new excuse not to buff. :P

Thankfully, Fixichong provided it not a day later. *Grin*

[whitey]

NT/doc lub

<3

Probably worth also mentioning that zoning is rather snappier on test these days too, another little tweak from Macro's almost infinite band of coding monkeys

i thought this was just due to lower server strain on test, due to the lower population?

[Macite]

FINALLY!!! *faints* If this get fixed, I'll be more than a crazy NT, I will be a crazy, happy NT.

[egadsrk2]

i thought this was just due to lower server strain on test, due to the lower population?

Everything about testlive zips about on hyper speed, with a grand total of 30 people on at like... peak load on testlive day when theres something really cool to play with.

But if you really want to test it, fly out to SAV and bench mark zoning to CAV. Then repeat it on Testlive. If testlive is faster on the same machine with the same settings, well then, its logical to assume something other than your computer has changed.

Myself, I always thought zoning came down to brute force; ram and hard drive speed; players in the vicinity of where you actually doing the zoning (on the other side) being a significant factor. Ie, my SAV/CAV example on both servers should be considerably faster than NL/BOR on live. NL/BOR on live should be drastically slower than NL/BOR on test. CPU usage drops sharply while actually doing the zoning, it's probably not a major factor.

[Mastablasta]

The way the game loads data can also be handled differently to make it faster. We saw many decreases in loadtimes over the years.

Macrosun for president btw <3

[trousers]

The fix though is risky and involved an extensive re-write of the related sections of the code. It also may not remove the problem entirely, although we are hopeful that it will.

You sound like me when I'm coding in my RP Bots, lol

[candle-645]

Bump for FC coder aware about "Design Patterns" and hope he isn't someone who read related book 1 week ago.