Weapon delay myths.

[ItazuraNhomango]

Re: Weapon delay myths.

My experiences with disconnects during battle has always been:
1. Able to move around.
2. Everyone else frozen--not moving or attacking.
3. I'm not attacking, even if I was on auto-attack.
4. If I was casting spell, the special effects loops on for a while, but no completion animation (e.g. Not lifting arms for white magic.).

Why you see something opposite? I've no idea.... ._.

You also can't say that the client side doesn't effect anything when clearly the issues of overclocked systems and other anomalies happening client side have given players an almost speed hack like behavior. Makes sense that this would be happening given the method of how the server and client communicate while in-game.

You're confusing auto-attack with other aspects of the game; my comment was limited to only auto-attack.

That speed/pos hacks are possible indicates clients sent position instead of direction/velocity vectors--at least the server doesn't check the vectors very closely if system is using vectors. Other things apparently under client control are the recast timers for abilities (and possibly spells), which would be why S-E had to ban a whole bunch of those cheaters in the last round of sanitizing campaign. But, as I indicated earlier, neither one of those things is auto-attack.

Why I actually say auto-attack has nothing to do with the clients? Because, I see people in party with the dreaded red-dot icon swinging away until fully d/c'ed. Well, if their client isn't communicating with the server, what's making them swing? Can't be the d/c'ing people's clients (since it's disconnected), and can't be my client (which has no business ordering other characters to do anything). That leaves only the server software as the logical candidate.

[Armando]

Re: Weapon delay myths.

My experiences when disconnecting are identical to Itazura's. Also, when I lag, my character stops attacking (like in Itazura's point No.3 at the top) until I stop lagging, but then my client animates all the missing swings in a very short burst. That's indicative that the client was waiting on the server's instructions and got them all at once after the lag ended if you ask me.

[Feba]

Re: Weapon delay myths.

Saying your frame rate isn't tied to your attack delay isn't completely true. Since so much of FFXI is client based, and because the way FFXI was coded in 30/60FPS (apparently normal for Japanese PS2 games from what i've read), it IS true that your attack speed will not slow down when your client does, like in besieged or lag, it IS also true that if you raise your FPS beyond the 30FPS cap, say by overclocking, your game will actually run fast, in the same way that early games would run faster when you pressed the turbo button on your CPU :p

FFXI relies on your hardware's clock, when you increase that, it throws FFXI off. This is why people who overclock their GPU/CPUs have been banned for Speedhacking on more than one occasion. This was also known to have increased delay speed.

What you are talking about, with the client continuing to attack, could be a result of the client getting R0, but still being able to send data, or SE could've changed the code since what I wrote above was true. It is possible SE changed it from a client side function to a server side function to prevent abuse.

[Lunaryn]

Re: Weapon delay myths.

Back onto the original topic, I seem to recall the Tarutaru DRK guide on gamefaqs had some useful info on delay. I've seen conflicting claims on how this works, but the general consensus holds that short spells cast well in-between swings do not adversely affect attack rate that much. Depending on your race and whether or not you have a source of refresh, you may want to stick to just the vital spells, or you may want to cast a spell after every swing. Also, the changes to signet are very kind to DRKs who actually use their MP.

As for keeping weapon skills capped, I haven't experimented extensively, but as a tarutaru I've had the best results with +DEX equipment and not too much +Acc. At one point I was getting no skillups at all for most of a level, despite being well under cap, while I had a fair bit of +Acc. After swapping Venerer Ring with Deft Ring and I think Battle Gloves with Wonder Mitts, I started getting skillups at a decent rate again.

And I have to agree, Marksmanship isn't worth the bother. You don't need to be able to hit a mob to pull it, and you'll never be particularly effective with ranged. I don't even bother bringing ranged weapons to my static trio (our BLU pulls). If you want Marksmanship skill, consider leveling RNG.

[ItazuraNhomango]

Re: Weapon delay myths.

Saying your frame rate isn't tied to your attack delay isn't completely true.

/sigh [Deleted most of the the long explanation on how "above cap" frame rate affects auto-attack delay while low frame rate does not could be possible.]

What you are talking about, with the client continuing to attack, could be a result of the client getting R0, but still being able to send data,

Here are two ways to look at the possibility of client controlled attack speed in this scenario:

1. When the network connection is degrading, and if the client is really directing every swing, when watching other people d/c in the middle of auto-attack, we (those with good connection) should notice much longer than normal period of not swinging (packet lost), and much shorter than normal swings delays (packet resent and timely packet for next swing) from time to time. This, I've not seen; Auto-attack tends to be fairly regular for everyone in party, as long as I maintain connection myself.

2. If client controls the attack speed, like the movement speed, why hasn't S-E banned anyone hacking that, to our knowledge? The apparent lack of such hacks is evidence (though not proof) pointing toward server side control of attack speed.

or SE could've changed the code since what I wrote above was true. It is possible SE changed it from a client side function to a server side function to prevent abuse.

I don't know what it was like once upon a time, but for my entirety as a FFXI player, what I saw indicates the server side controls auto-attack speed. YMMV.

[Feba]

Re: Weapon delay myths.

People have been banned for speedhacking, this falls under that. Generally, actual hackers don't bitch much when they get banned.

And I do distinctly remember a time when I was around lv30, I went to take some food to an LS group partying in Boyahda tree, one of the characters (A RNG or THF or NIN, some mithra) was happily swinging their daggers about, and stopped. DCd 30 seconds afterwards.

I'm not saying it's always been like that, and i'm not saying that couldn't have been something else (crab just a few feet in the wrong direction, I think we've all been annoyed by that at some point or another), but I have heard people who have overclocked reporting their characters attacking faster, and given how FFXI is frame and client based, it is possible. It would be a huge oversight on SE's part, and they would fix it, but that doesn't mean it didn't happen.

[ItazuraNhomango]

Re: Weapon delay myths.

People have been banned for speedhacking, this falls under that.

Don't know what you mean...

Generally, actual hackers don't bitch much when they get banned.

I suppose those in prison claiming innocence were falsely accused, to a man.

but I have heard people who have overclocked reporting their characters attacking faster, and given how FFXI is frame and client based, it is possible.

Placebo effect, maybe? If you expect the game to "run faster" (or other nebulous concept like "perform better") by this "overclocking", and you do see some other changes, it becomes easy to fool one's own mind to see more than there is.

What do you mean by "FFXI is frame and client based"? To me, that's like saying web sites are web browser and HTML rendering engine based.

[Feba]

Re: Weapon delay myths.

FFXI movement is based on frame rate. FFXI in general is based on the client, the server doesn't have control over a lot of things it should, in a lot of ways. This is why so many hacks that shouldn't be possible are, such as teleporting. If the server controlled movement, it could easily slow down speedhackers, or move teleporters back where they should be.

[ItazuraNhomango]

Re: Weapon delay myths.

I give up. Feba wins. EVERYTHING is frame rate based.

After all, when programmers write routines, they wouldn't do a system call to calculate elapses time. Heck, maybe no such system calls exist. Instead, the code monkies sync everything to frames. Opps. Excuse me; I mean "frame rate", not "frames." EVERYTHING--spell, movement, ability, weapon delay--yes WEAPON DELAYS--despite observational evidence to the contrary!--are based on how fast things are displayed on the client.

Hopefully, Feba knows what a "frame" is when he make these kinds of statements about "rates".

Done with this (off) topic now; sorry about the large amount screen space I took.

[Feba]

Re: Weapon delay myths.

Raising your max frame rate causes your character to move quicker, and it has been reported by various people that other actions are sped up. SE might've changed this.

There's no point being pissy, it is a fact that FFXI is based on a 30FPS cap in at least some ways, and it used to be that attacks were one of them.

You might have observational evidence to the contrary, however, so do I. It's more a matter of updates and SE's coding than what either of us see.

[Macht]

Re: Weapon delay myths.

My experiences with disconnects during battle has always been:
1. Able to move around.
2. Everyone else frozen--not moving or attacking.
3. I'm not attacking, even if I was on auto-attack.
4. If I was casting spell, the special effects loops on for a while, but no completion animation (e.g. Not lifting arms for white magic.).

Actually the stuff you stated here I've seen happen as well. But these I've only been able to relate to when my end lags. It's totally different when it's a hangup on my end vs. and hangup on the server. These do not happen when the server side has it's "hiccups".

When the server hiccups the game still appears to run like normal. Yes there are certain things that do not happen if the client doesn't get the UDP packet from the server. If you initialized the WS then you will see it on your client it is not guarenteed the other clients will get the UDP and see it, this is the same for ending sequences on spells and ranged attacks (except these are global, not even your client will see the ending sequence if it doesn't get it from the server).

The client however doesn't wait for the server to ok an action back to it that is obvious. Obvious being if a player is engaged with a mob in a melee fight that player will be stricking the mob at set intervals. You can see that the client does share some of the control with the server by just watching a running player hitting a wall. On the player's client he's still running and see's his model bouncing around, to everyone else it appears he's walking. Two totally different views of the same action only because it's client info vs. server info.

You also can see that the client runs that attack part by watching things like Range attacks being engaged. On multiple occasions I've activated my range attack exactly at the same time that I auto-attack. It results in a graphical messup on my end were it'll look like I'm about to fire my dagger I had in hand, but to everyone else they see the attack happen cleanly and right after the bow is drawn.

In this case I believe SE intentionally made it were server had to respond back so that the bow shooting would end the same on everyone properly setting my auto-attacks on their client to continue engaging like normal without an offset. You can actually create an offset effect intentionally with the worms, by stepping just far enough away when they cast a spell. You'll see the server responds that the spell was cast, but the client didn't end the worms spell casting. You can repeat this action with such control that there is no way you missed the UDP packet that many times, it just makes it painfully obvious that the server isn't sending the client that info. That info is actually controlled by the client and when you step far enough away the client doesn't get it right.

[Gilge]

Re: Weapon delay myths.

I thought the game was capped at 30fps?

lol, could you imagine a 16 second delay on a 480 wep?

and you thought guns had a long delay as is...

[VZX]

Re: Weapon delay myths.

FFXI movement is based on frame rate. FFXI in general is based on the client, the server doesn't have control over a lot of things it should, in a lot of ways. This is why so many hacks that shouldn't be possible are, such as teleporting. If the server controlled movement, it could easily slow down speedhackers, or move teleporters back where they should be.

I think that's caused by "Server always trust whatever clients say". When it's possible to fake the message being sent by client, time-cheating is possible.
I remember one of my friend attempted and succeeded to remove the delay between synthesis.

[Feba]

Re: Weapon delay myths.

lol, could you imagine a 16 second delay on a 480 wep?

He is correct. While Delay rates are based off 60FPS, the actual graphics (output) will never go over 30

[Feba]

Re: Weapon delay myths.

I don't think <pos> hack can be one example/prove for the statement of "ffxi movement is based on frame rate"

If you read my post, you'd know that isn't what i'm saying.

They are two very seperate points.

FFXI movement is based on frame rate. FFXI in general is based on the client, the server doesn't have control over a lot of things it should, in a lot of ways. This is why so many hacks that shouldn't be possible are, such as teleporting.