The Hell?

[NeuroToxin]

Okay, so, Just an update, Hiveworkshop is no longer accepting KT2 Spells.
I'd love to protest this if anyone is with me.

 

[tooltiperror]

KT2 is little more than deprecated at this point, really.

Just use TimerUtils (Or T32 at low frequencies).

Like, what are you doing anyways? Using a worse timer system hoping that it will for some reason be better?

 

[Sevion]

I believe the poll is misleading. I don't hate KT2, but I don't use it nor do I see a reason to use it.

For most purposes T32 is fine. For everything else a simple timer or wait is fine.

Because let's face it, with short periods (0.0325) T32 is the best, IMHO. Accuracy with T32 is great. You should really only need/care about accuracy when you have short periods.

With longer periods (periods 1 second and longer) that have very lax needs on accuracy can use waits for convenience.

With longer periods that have a slight necessity for accuracy can use a simple timer and if need be, a simple attachment.

In short, I think KT2 is pretty much deprecated.

I feel Hive's rejection of KT2 is deemed right.

As I said earlier as well, the poll is very skewed towards your (the original poster's) opinion.

"Yes I want KT" versus "No I hate it".

Just because I don't use/condone the use of KT2 doesn't mean I hate it.

And the poll question is also misleading, "Do you approve?". Approve of what? The use of KT2 or the rejection of the use of KT2.

The thread title is also very crappy.

Do read: http://www.thehelper.net/forums/showthread.php/101956-How-not-to-write-a-thread-title

Not exactly the same, but the principles apply.

 

[NeuroToxin]

I use Key Timers all the time. I like functions more than methods and such. I like using two periods, one for a regular move, and the other for a knockback. My point is, They shouldn't reject a resource because its move period is 0.000000003 off.

 

[tooltiperror]

>I use Key Timers all the time. I like functions more than methods and such.
You can use any timer system with methods and functions other than T32? Someone could write a T32 function wrapper anyways?

>I like using two periods, one for a regular move, and the other for a knockback
Can't you do this with TimerUtils also?

>My point is, They shouldn't reject a resource because its move period is 0.000000003 off.
Cite your source

 

[NeuroToxin]

Also, Read here: http://www.hiveworkshop.com/forums/faq/resourcerules/#faq_resourceunique
It simply says that the resource must be unique and functioning. It functions. So why do they have to bi**h.

 

[Sevion]

Using something deprecated can be considered an "error" of sorts.

 

[Uberplayer]

I think hiveworkshop is slowly getting corrupted by Nestharus and Bribe (Nestharus bribed Bribe and got him over his right palm, I see no other reason for this madness).

 

[NeuroToxin]

I think hiveworkshop is slowly getting corrupted by Nestharus and Bribe (Nestharus bribed Bribe and got him over his right palm, I see no other reason for this madness).

Seriously? Cause I feel like they're taking over also. It used to be an awesome site, now it's like getting more messed up.

 

[Laiev]

Seriously, you do not need others to approve what you do, if you did, like and use, that's enough. Ignore the comments of others.
There is no reason for you to think that just because someone does not like or does not approve, it is bad. The question is, is bad for you?

 

[tooltiperror]

Seriously? Cause I feel like they're taking over also. It used to be an awesome site, now it's like getting more messed up.

rofl THW is a joke.

When TriggerHappy became a mod most of the modding world stopped taking them seriously.

All WC3 modding sites are pretty bad, we lack coordination with other sites, WC3C is way too inactive, and THW is a joke. But WC3 modding itself okay.

 

[NeuroToxin]

rofl THW is a joke.

When TriggerHappy became a mod most of the modding world stopped taking them seriously.

All WC3 modding sites are pretty bad, we lack coordination with other sites, WC3C is way too inactive, and THW is a joke. But WC3 modding itself okay.

True. True. Although I feel like we are the best of them all.

 

[Laiev]

True. True. Although I feel like we are the best of them all.

and we are, IMO...

 

[kingkingyyk3]

I think hiveworkshop is slowly getting corrupted by Nestharus and Bribe (Nestharus bribed Bribe and got him over his right palm, I see no other reason for this madness).

Agreed.
Anyway, I think we need some powerful coders here.

IMO, the approaches are different for each sites.
THW - Speed freak, no functionality.
TH - Efficiency and functionality are balanced.
WC3C - Functionality, less care about speed.

For me, I prefer functionality over speed anytime.
KT2 is fine, I don't see any problem with it.

 

[Azlier]

Now, now.

Bribe (what an unfortunate name, considering the accusations here) may agree with some of Nestharus' views, and I do not discount the possibility of Nestharus being a madman, but I doubt he is a mere pawn. No need to shout about corruption and bribes while foaming at the mouth; THW is fine, albeit messy.

Jesus4Lyf himself has said that KT2 is essentially useless, with T32 serving for high-frequency events and TimerUtils for low-frequency. While Bribe's judgement in this case is harsh, I do think it is a step in the right direction. Might as well try to have as many standards as we can. It's for the good of all of us.

 

[BlackRose]

I think hiveworkshop is slowly getting corrupted by Nestharus and Bribe (Nestharus bribed Bribe and got him over his right palm, I see no other reason for this madness).

Bwahah. I don't know if corrupted is the word, but the JASS section is definitely being overrun by those two.

 

[emjlr3]

For me, I prefer functionality over speed anytime.

wow some one who actually gets it! especially considering whats being argued over is the least of your worries when coding periodic intervals. (and you can quote me on that)

to quote Bribe

emjlr3, KeyTimers2 will lag with just a few hundred instances. 500 instances of KeyTimers2 is more than 500 times slower than 500 instances on Timer32.

not sure if the more instances you have the greater this speed decrease becomes, but again, I ask

so keytimers 2 is 500 times slower?

i'd like to see a realistic test case scenario demonstrating real life observed slow down. aka. 500 times more fps reduction, or 500 times more lag, or the equivalent

i don't know how either work, really, and if I tried to figure it out I may never get it - but I do know that once you get a few spells going, T32 is probably faster then a struct loop/spell (uncited source), which means its damn fast (but at that point who cares anyway)

could someone explain how each of these systems work? that I might understand where the sudden need for rejection of the system exists? if it works, and it doesn't terribly suck,what's the problem?

I can see where you are coming from, one system to rule them all and such. Makes sense, and it would work if not for the fact that there are a dozen systems out there, each with their own pros and cons, each fast enough, and each appealing to various sects of scripters.

I say, forget about the needless banter, and accept good spells for what they are.

@Neuro, you could always use a struct stack, as its the universally accepted standard. or, if THW decides this a permanent "fix", just forgo their website in favor of greener pastures

off note - no one here probably remembers massive maps running on handle vars, which was as slow as it gets, without a complaint in the world - and we are concerned with the weeds in a jungle.....

 

[Sevion]

I don't see why some people never understand... It's not the fact that X amount of Y will lag, but the fact that having Z amount of systems like that will compound the "added" amount of lag and create a lot of lag.

No one complained because before that, no one knew better (And actually, lots of people complain about lag).

It's not weeds in a jungle, it's weeds in a neighborhood. If each house has weeds, the total weed count in the neighborhood increases.

Pruning deprecated systems is like each household goes out to pick weeds. The entire neighborhood is now better because there are less weeds.

Don't think of it as "This system doesn't lag for the first 500 instances and no one will ever really use 500 instances" but rather "If this system contributes this much lag, how much total lag will there be in a final map?"

It's not the fact that this SPELL doesn't lag a ton, but rather how much will it contribute to ALL of the spells and ALL of the systems in a FULL map.

 

[emjlr3]

When the pruning in question contributes to the net effect of <.001% of a neighborhoods landscaping, it’s still in the weeds - no one seems to understand THIS.

I repeat, struct retrieval is the least of your worries when executing script within periodic intervals.

when benchmarks are run, people say - well let’s see how much faster this data retrieval is then this one - but no one says, let’s say this complicated spell is running 25 times (realistic representation of code execution in a map), what is the net effect of changing data storage systems, I guarantee you will stop seeing quotes of 75% speed improvement!!! I'd be surprised if you saw anything above 1/100th of a percentage point.

My point is this

Why use something that isn't the fastest? Well Neuros reasoning for one - he needs multiple timer intervals outside of what is provided through T32. If KT2 is ideal in this scenario, regardless of whether it is slower than other alternatives, namely TimerUtils, which it shouldn't be given realistic testing conditions, it has broad spectrum application for Neuros needs, and he has chosen practicality as his #1 priority, which in the end it was really matters.

Is KT2 a legit system? Sure... Is it the best for what Neuro needs? Maybe not, but it gets the job done nonetheless. Should every system other then what you consider the best be barred? Well I guess that is your prerogative.

In a perfect world, we have one universal standard. If that is your goal, start with KT2, and work your way around, because you've got a lot of work to do.

 

[tooltiperror]

I think Neuro would be better equipped with a timer queue, but let's not open THAT can of worms.