Hooray for /dismount!

Great changes, thanks slouken! (and thanks iriel for all the effort put into these threads!)

Edit: slouken, is there any way that /castsequence could advance to the next step on a failed spellcast? Or is there possibility for exploitation? Take for example the following macro:

/castsequence Conjure Mana Ruby, Conjure Mana Jade

Each of the conjured items is unique, so, first time through the macro works beautifully, conjures both gems. But then say you consume the mana jade, and still have the ruby in your inventory, the sequence will not advance past the ruby because the spell fails.

WTB a "Fixed a bug with tooltip:SetUnitBuff()" ;D

And can we have a /castingsequence the continues even if a spell fails to cast? Or maybe an [oncooldown]/[nooncoolddown]?

Would it be possible to get the [modifier] conditional expanded to include all keys (ie z,x,c).

Right now the modifier conditional is only used for mouse clicks (ie ctrl-leftclick) and not hotkeys (ie ctrl-1, which casts pet-skill 1 by default). Even setting both 1 and Ctrl-1 to the same hotkey in keybindings will still evaluate as if the ctrl modifier was not used. There doesn't seem to be any way to accomplish this with keyboard presses right now.

By allowing any button to be a modifier the player could for example set Z as a modifier allowing Z-1 to perform a different action then 1. Some functionality for keyboard users to modify macros would be greatly appreciated.

Please note that this thread is to discuss the upcoming changes and any clarifications or features that are a direct result of those changes, or things which we've been asked to remind slouken of.

IMPORTANT: Off-topic or entirely redundant posts are liable to get deleted.

This isn't, strictly speaking, a suggestions thread. A continuing-through /castsequence and a [cooldown] check in generic forms are also not within Slouken's power to give—you'd be better off in the suggestions forum.

---

Q u o t e:

And can we have a /castingsequence the continues even if a spell fails to cast? Or maybe an [oncooldown]/[nooncoolddown]?

---

Discussed (ad nauseum) and not going to happen.

---

Q u o t e:

Would it be possible to get the [modifier] conditional expanded to include all keys (ie z,x,c).

---

Can be done with addons if I'm not mistaken.

One thing to consider with that is this - how do you seperate a modifier from a keypress?

When I press shift then ` do I want "Shift-`" or am I about to press another key?

If I press "Shift-`-1" do I want 1 modified by "Shift-`", ' modified by "Shift-1"?

Slouken had to put a precedence system in place to take care of just 3 modifiers, I think the worry here is that if you generalized it, the processing becomes much more complex and asking macro users to understand which press they are actually getting is harder.

On 2.0 live, debugprofilestop() is still returning CPU cycles instead of milliseconds. (Intel Core 2 Extreme 2.93ghz CPU on WinXP SP2 OS)

Is this intended? If so, perhaps consider listing it as a change? (and adding the corresponding debugfrequency() function so that we can convert it back to human time units easily)

I know I was quite confused when I booted into 2.0 and started getting huge performance numbers.

---

Q u o t e:

One thing to consider with that is this - how do you seperate a modifier from a keypress?

When I press shift then ` do I want "Shift-`" or am I about to press another key?

If I press "Shift-`-1" do I want 1 modified by "Shift-`", ' modified by "Shift-1"?

Slouken had to put a precedence system in place to take care of just 3 modifiers, I think the worry here is that if you generalized it, the processing becomes much more complex and asking macro users to understand which press they are actually getting is harder.

---

Since actions are activated when the key is released it should not be difficult. You should only need to do a few extra things, like testing if a key is still pressed when another is released. As for the precedence, unless you set very complex and overlapping combinations a simple key order would be sufficient. You would also need to pass keypresses that are overlapping but not used as modifiers as independent keypresses; implementing something like this could also make binding to one of the actual modifier keys (alt, control, shift) by itself possible.

The problem I see with this is different. Lots of keyboards can get phantom keypresses if you press some specific keys together, and this could create quite difficult to find bugs into player's interface configurations.

---

Q u o t e:
(Extra space to discuss subsequent bug fixes or additions)

---

if the last patch is anything to go by, you will need another 3 or 4 reserved spaces to account for this particular point...

Slouken, can you please make one more API function for cursor grabbing a macro.
HasCursorMacro()
and/or
type = GetCursorType()
where type is "spell", "item", "macro" or nil.

I have made an addon named BindPad which is GUI to set keybindings for spell/item/macro
and it need to detect which icon is grabbed by cursor.
I'm using securehook on PickupSpell/PickupMacro etc. to detect spell/item/macro on cursor,
but we can not detect when these icons removed from cursor by right button click.
Currently only way to detect cursor change is CURSOR_UPDATE event, but it triggerd in way too many situations.

To resolve this problem, new API GetCursorItem(), GetCursorSpell() are very good for spell and items, and I wish to use one more new function for macroes too.

For reference:
http://www.wowinterface.com/downloads/fileinfo.php?s=&id=6385

I don't see any thing about fixing problems with right clicking and selecting menu items with certain addons installed. Specifically getting script blocking message despite not being in combat.

For example if I have Map Notes installed (Cosmos) and right click someone's portrait and select duel (when not in combat) I get a script blocked message. If I disable Map Notes I get the same thing with !improvedErrorFrame, the Titan (which uses hooksecurefunc) and so on and so forth.

Is this not being fixed or is it not a bug?

Please god dont make me log in again & retype this from scratch.

I have 2 issues with SecureActionButton:
1) ActionButton_Update is specific to ActionBarButton, and not the more generic ActionButton which can have more than just type = "action". Any chance of making this work for other types?

2) Onclick handler needs an itemId or itemLink option:

	elseif ( type == "item" ) then

		local bag = SecureButton_GetModifiedAttribute(self, "bag", button);

		local slot = SecureButton_GetModifiedAttribute(self, "slot", button);

		if ( slot ) then

			if ( bag ) then

				UseContainerItem(bag, slot, unit);

			else

				UseInventoryItem(slot, unit);

			end

		else

			local item = SecureButton_GetModifiedAttribute(self, "item", button);

			if ( item ) then

				UseItemByName(item, unit);

			end

		end

As is this code will not let me use multi stack items unless by name, but using by name does not allow discrimination between differently enchanted / bejeweled items. You can write some frankencode to switch between the two modes as needed but then still get hosed if the user shifts inventory during combat.

So something like:
elseif ( type == "itemId" ) then
local itemId = SecureButton_GetModifiedAttribute(self, "itemId", button);
if ( itemId ) then
UseItemById(item, unit);
end

where itemId can be just the plain number, or the more complete string describing all 8? properties as well. The magical UseItemById() is probably already inside UseItemByName.

Is this the place to post suggestions/requests for inclusion in the patch?

With the new restrictions, it's no longer possible to switch pages on the 4 built-in MultiBars.

I'd like to suggest that the ChangeActionBarPage() function take an optional second parameter of the name of the multibar (such as "MultiBarBottomLeft") whose page should be changed.

Thanks!