Update and show of interest for mouse filter games??

[skarlot]

So, my KM switch that I have been working on is close to release. I expected to get working relatively easily, since I have worked on 2 others already. But it turned out hellish, since my goal this time (or rather requirement since it was always a goal) was to make it 100% reliable, not 99.5%, (with regards to it actually positioning/transitioning correctly every time). This might sound easy, but even if you fully understand multi-threading, critical sections, networking, etc, I guarantee it's a lot more messing around than you'd imagine. I am not going into it any more at this point.

Basically I have ditched Screen & Link's, and created a transition system. This eliminates the annoying issues arising from when the mouse is not returned to its original position after mouse looking in WoW (windows error) and being trapped within a screen, thus triggering links, etc. I have been using keyclone all this time too for a change and it's nice not to 'have' to worry about every single aspect of boxing software. Also It is part of my intention for it to work with keyclone.

Which leaves me in a position where Octopus has a KM, but I am not maintaining 2 KM's nor do I feel like dealing with that KM any more since my new one really is that much better for me and my complex arrangement of monitors. So that KM will be removed even if Octopus is released with a $ tag. I would be able to support a mouse filter driver, potentially, from 2 different programs, but I don't think there's enough interest in those games that need it.... to warrant the effort so I'm working with the idea of touching it up a tad without a filter driver.

On the other hand I could ditch Octopus totally, and focus with my newer broadcasting app, which sounds like more fun. I have done a fair bit of code on that but the KM had to take priority since it was annoying me to death. And, the funny thing is, no one has ever complained at all about it not being 100% accurate...

If there's enough interest I'll continue with the filter driver (which was implimented for EQ and seemingly necessary for CoH).

[Tehtsuo]

I know it's probably not any useful input for you, but I wanted to mention that I recently switched from Synergy to Octopus just for KM. I know you mentioned that the Octopus KM isn't that good (in your opinion) but it's leagues better put together than Synergy in my opinion. Being able to move my mouse to the other screen and not have to hit scroll lock seems like a luxury to me, but I like it.

I won't turn down top-quality free software, but if your new KM is as good as it sounds (I can't imagine how it could end up better than Octopus' KM feature) and you decide to put a $ tag on it, just give me a target to throw my money at.

/target skarlot
/cast "Bag O Money"

[skarlot]

:)

My setup is like, a 24" in the middle, and 4 17" around it, 2 beneath, 1 L 1 R. The 2 17" to the left are XP running a vertical span which is split in the middle. So I have links all over the place.. if something goes wrong the cursor can end up anywhere, thus I added a green border feature in Octopus to counter that :). Because these KM's I've been working on have been hiding the mouse in the lower right corner (a tradition since I MultiBox I think !) any links that are around this area just go beserk if transition logic happens out of order, or doesn't happen, etc. It wouldn't matter where the mouse is hidden really (well dropping the cursor in the middle of the desktop wouldn't look so flash even if it helped tame the issue). That's just part of the story... Windows hooks/cursor positioning is just screwy to deal with when pushed like this.

[dRiN]

I have had my share of annoyances with the KM in Octopus. But since it delivers so much more I sticked with it. It's a "you learn how to use it kind of thing".

Not really sure what your question is. And not sure if I could answer it.

But I would like better KM functionality all in one with a keybroadcasting application. No matter what it would be, a touched up Octopus or a complete new application. And preferably something that does different keymapping in an easy to implement way for the focusless, leaderless setup.

[Tehtsuo]

I have had my share of annoyances with the KM in Octopus. But since it delivers so much more I sticked with it. It's a "you learn how to use it kind of thing".

Not really sure what your question is. And not sure if I could answer it.

But I would like better KM functionality all in one with a keybroadcasting application. No matter what it would be, a touched up Octopus or a complete new application. And preferably something that does different keymapping in an easy to implement way for the focusless, leaderless setup.

I think his strategy does make sense though. All-in-one solutions are nice, but it's difficult to make sure everything works flawlessly when you do that. A 100% reliable KM that works in every situation is a very strong step in the right direction, and I think it'd be a very good thing for this community especially. Right now we have a number of solutions that work well, but each has its strengths and weaknesses. Synergy, Multibox, Octopus, etc etc. With something flawless, when someone comes on and says "I need something so I can move my mouse/keyboard across to another computer's screen, what should I use?" we can say "Use this. Don't use anything else, don't waste your time, just use this."

Note: Not trying to say anything bad about any of our current KM solutions, they're all pretty nifty in their own ways.

[skarlot]

There is no real question just my thoughts and trying to see if the filter driver is being used much... but I doubt it.

There are many broadcasting apps around, and I suspect more will surface but the likelihoods of someone else making a good KM are much slimmer because it's a lot more fiddly. Even if someone had a shot purely for profit (at something truly multi-monitor for WoW), they'd be pressed unless they were 5 boxing... consistently testing it. This is why I'm trying to make one to support all the apps, eventually. It's much better for me to code it in a seperate app, and also to have a KM running while I develop a broadcasting app too. The all-in-one factor is better for say.. mouse broadcasting, because a true p2p broadcasting app would be drawing co-ordinates from the client but thats a pretty minor factor in the end. I measured wired network latency at .7 MS two way and the speed of a mouse hook at > 280 cycles per ms. That was when I had some code in my mouse hook too, now I have virtually none. A slow hook is countered by notching up windows mouse speed. That's my logic. Time will tell how it all works out. Also all in one is a RAM chewer, at least with .NET, just for a KM.