Anyone using 'extra' mouse buttons?
in [Solaris]
|
From: Richard L. Hamilton on 2 Feb 2010 02:04 In article <hf14an$ot$1(a)aioe.org>, "Colin B." <cbigam(a)somewhereelse.shaw.ca> writes: > Richard L. Hamilton <rlhamil(a)smart.net> wrote: >> In article <he468g$iiu$1(a)aioe.org>, >> "Colin B." <cbigam(a)somewhereelse.shaw.ca> writes: >>> Bit of a shot in the dark here, but is anyone using a fancy new mouse >>> (in my case, the Logitech MX620) with Solaris 10 (Sparc), and getting >>> the various accessory buttons to work? I'd love to get the side buttons >>> (forward and backward) functioning, if possible. >> >> AFAIK, Solaris 10 can usually handle 3 buttons plus (in many cases) >> vertical scroll, where up and down steps map to clicks of >> fictitious buttons 4 and 5 (which is how other implementations handle >> this also). I think that horizontal scroll functionality first appeared >> in a patch to either Solaris 8 or Solaris 9; my Sun Blade 100 still has >> Solaris 9, and the scroll wheel on the Logitech TrackMan attached to it >> works just fine. New Sun mice have a vertical scroll wheel that doubles >> as the middle button, like most mice do nowadays. (OTOH, I've seen systems >> running Solaris 8 where a scroll wheel didn't work.) >> >> Core X pointer devices can only _fully_ support 5 buttons. However, they >> can support _some_ actions on additional buttons, enough that I think >> it's common enough to map horizontal scroll steps to clicks of buttons >> 6 and 7. >> >> Still, because of the core X pointer device limitations (changing them >> would be a no-no because it would break either binary compatibility or >> over-the-wire compatibility for remote display, or both), most programs >> won't be able to do anything with more than five buttons (or given that the >> UI toolkit in use supports it, 3 buttons, vertical, and even horizontal >> scroll, as previously described). >> >> I'd have to dig a bit to get it right as to which way around up/down vs >> 4/5 (or left/right vs 6/7) go, or _exactly_ why core X pointer devices >> couldn't be redefined to handle something with a lot more functionality. >> But I'm sure I'm close enough that you get the general idea. >> >> There have been drivers for fancier devices, like the buttonbox/dialbox >> combination (8 dials on the dialbox, and I think 16 buttons on the >> buttonbox, with individual lights on each button yet). Or 3-D space mice. >> But those are treated as X extension devices, which very few apps, let >> alone UI toolkits, know anything at all about. Normally, a single device >> cannot be both a core X device and an extension device at the same time, >> but some get around that by pretending to be two devices, so that all >> programs can use the basic functionality, and those that understand about >> extension devices can look for and use them instead. So there's just enough >> infrastructure in X11 that greater functionality is not totally impossible. >> But to have apps and desktops that actually used more than core functionality >> would mean getting those smarts into widely used toolkits (like GNOME's >> GTK or KDE's Qt). I have no idea at all what if anything is happening there, >> but I think that's where one would have to look for widespread support of >> extended input device functionality under X11. >> >> I think the situation on Solaris is about on par with where it is on Linux >> regarding common pointer devices (give or take that I think the Synaptics >> touchpad driver just got added to OpenSolaris, which presumably means it's >> _not_ in Solaris 10 AFAIK). Add-ons like tablets are of course going to >> be harder to find drivers for (although I think I saw one once). >> >> But I think it's safe to say that if you can't get all your mouse's buttons >> and features to work under X11 on Linux, you probably won't ever see them >> working under X11 on Solaris, either, because the limitation is ultimately >> not the OS, but X11. > > Wow, that was...comprehensive. Thank you Richard! > > The mouse in question is a new Logitech with, as near as I can count > (because Logitech's documentation is unexpectedly crappy) eight buttons > plus the scroll wheel. The two main buttons, scroll wheel, and scroll wheel > button all work, so that would be the five you mention. That leaves five > buttons that aren't working. Pity. > > Colin BTW, I think there's a potential for things getting better, _eventually_. I saw something about (I think)I a revision to X11 input extensions. Until that, a device could at any given time be a core X device (with the limitations that had), or an extension device (which could handle something like 32 buttons and 8 axes of motion, if memory serves). But it couldn't be both at once. I think the new revision will allow a device to be both at once, so that older programs that didn't know about extension devices could use the more limited capabilities, and newer ones could use a greater range of capabilities. I wouldn't expect to see that for awhile though; even once that revision is incorporated, it wouldn't be useful to most people until toolkits and programs were revised to take advantage of it. But as long as the present situation holds (who knows what Oracle may do in terms of Solaris on the desktop?), I wouldn't expect Solaris to fall too far behind Linux (or rather, X11 on Linux or other platforms) anyway, whatever comfort that might be. My personal suspicion is that it might not be ten-button mice that move things along, but more interesting devices like affordable 3D mice, datagloves, full body motion trackers, etc.
From: Johann 'Myrkraverk' Oskarsson on 2 Feb 2010 22:52 rlhamil(a)smart.net (Richard L. Hamilton) writes: > In article <hf14an$ot$1(a)aioe.org>, > "Colin B." <cbigam(a)somewhereelse.shaw.ca> writes: >> Richard L. Hamilton <rlhamil(a)smart.net> wrote: >>> In article <he468g$iiu$1(a)aioe.org>, >>> "Colin B." <cbigam(a)somewhereelse.shaw.ca> writes: >>>> Bit of a shot in the dark here, but is anyone using a fancy new >>>> mouse (in my case, the Logitech MX620) with Solaris 10 (Sparc), >>>> and getting the various accessory buttons to work? I'd love to >>>> get the side buttons (forward and backward) functioning, if >>>> possible. I'll add that I have a logitech mouse with a single side button which works. I use it in my emacs to page down, and it wors "out of the box" as autorun in WoW (through Wine). I'm using build 115 of OpenSolaris. I have, as of yet, not found other uses for it. Johann -- who deleted all the interesting text below.
From: David Combs on 10 Feb 2010 19:50 In article <88Q9n.21848$p66.16994(a)newsfe09.iad>, Richard L. Hamilton <rlhamil(a)smart.net> wrote: >In article <hf14an$ot$1(a)aioe.org>, > "Colin B." <cbigam(a)somewhereelse.shaw.ca> writes: >> Richard L. Hamilton <rlhamil(a)smart.net> wrote: >>> In article <he468g$iiu$1(a)aioe.org>, >>> "Colin B." <cbigam(a)somewhereelse.shaw.ca> writes: >>>> Bit of a shot in the dark here, but is anyone using a fancy new mouse >>>> (in my case, the Logitech MX620) with Solaris 10 (Sparc), and getting >>>> the various accessory buttons to work? I'd love to get the side buttons >>>> (forward and backward) functioning, if possible. >>> >>> AFAIK, Solaris 10 can usually handle 3 buttons plus (in many cases) >>> vertical scroll, where up and down steps map to clicks of >>> fictitious buttons 4 and 5 (which is how other implementations handle >>> this also). I think that horizontal scroll functionality first appeared >>> in a patch to either Solaris 8 or Solaris 9; my Sun Blade 100 still has >>> Solaris 9, and the scroll wheel on the Logitech TrackMan attached to it >>> works just fine. New Sun mice have a vertical scroll wheel that doubles >>> as the middle button, like most mice do nowadays. (OTOH, I've seen systems >>> running Solaris 8 where a scroll wheel didn't work.) >>> >>> Core X pointer devices can only _fully_ support 5 buttons. However, they >>> can support _some_ actions on additional buttons, enough that I think >>> it's common enough to map horizontal scroll steps to clicks of buttons >>> 6 and 7. >>> >>> Still, because of the core X pointer device limitations (changing them >>> would be a no-no because it would break either binary compatibility or >>> over-the-wire compatibility for remote display, or both), most programs >>> won't be able to do anything with more than five buttons (or given that the >>> UI toolkit in use supports it, 3 buttons, vertical, and even horizontal >>> scroll, as previously described). >>> >>> I'd have to dig a bit to get it right as to which way around up/down vs >>> 4/5 (or left/right vs 6/7) go, or _exactly_ why core X pointer devices >>> couldn't be redefined to handle something with a lot more functionality. >>> But I'm sure I'm close enough that you get the general idea. >>> >>> There have been drivers for fancier devices, like the buttonbox/dialbox >>> combination (8 dials on the dialbox, and I think 16 buttons on the >>> buttonbox, with individual lights on each button yet). Or 3-D space mice. >>> But those are treated as X extension devices, which very few apps, let >>> alone UI toolkits, know anything at all about. Normally, a single device >>> cannot be both a core X device and an extension device at the same time, >>> but some get around that by pretending to be two devices, so that all >>> programs can use the basic functionality, and those that understand about >>> extension devices can look for and use them instead. So there's just enough >>> infrastructure in X11 that greater functionality is not totally impossible. >>> But to have apps and desktops that actually used more than core functionality >>> would mean getting those smarts into widely used toolkits (like GNOME's >>> GTK or KDE's Qt). I have no idea at all what if anything is happening there, >>> but I think that's where one would have to look for widespread support of >>> extended input device functionality under X11. >>> >>> I think the situation on Solaris is about on par with where it is on Linux >>> regarding common pointer devices (give or take that I think the Synaptics >>> touchpad driver just got added to OpenSolaris, which presumably means it's >>> _not_ in Solaris 10 AFAIK). Add-ons like tablets are of course going to >>> be harder to find drivers for (although I think I saw one once). >>> >>> But I think it's safe to say that if you can't get all your mouse's buttons >>> and features to work under X11 on Linux, you probably won't ever see them >>> working under X11 on Solaris, either, because the limitation is ultimately >>> not the OS, but X11. >> >> Wow, that was...comprehensive. Thank you Richard! >> >> The mouse in question is a new Logitech with, as near as I can count >> (because Logitech's documentation is unexpectedly crappy) eight buttons >> plus the scroll wheel. The two main buttons, scroll wheel, and scroll wheel >> button all work, so that would be the five you mention. That leaves five >> buttons that aren't working. Pity. >> >> Colin > >BTW, I think there's a potential for things getting better, _eventually_. >I saw something about (I think)I a revision to X11 input extensions. >Until that, a device could at any given time be a core X device (with the >limitations that had), or an extension device (which could handle something >like 32 buttons and 8 axes of motion, if memory serves). But it couldn't be >both at once. I think the new revision will allow a device to be both at >once, so that older programs that didn't know about extension devices could >use the more limited capabilities, and newer ones could use a greater range >of capabilities. > >I wouldn't expect to see that for awhile though; even once that revision is >incorporated, it wouldn't be useful to most people until toolkits and programs >were revised to take advantage of it. > >But as long as the present situation holds (who knows what Oracle may do in >terms of Solaris on the desktop?), I wouldn't expect Solaris to fall too >far behind Linux (or rather, X11 on Linux or other platforms) anyway, >whatever comfort that might be. > >My personal suspicion is that it might not be ten-button mice that move things >along, but more interesting devices like affordable 3D mice, datagloves, full >body motion trackers, etc. > (Not knowing what of the above to delete, I'm just leaving it all there.) Right now I'm using Windows XP with a Microsoft mouse that, on top, has a LEFT, a wheel, and a RIGHT, and one more on each side (on each vertical wall of the mouse). Those final two work for going fwd and backward in eg firefox. I stupid MS can have these things working ok, why the H. can't sun! Maybe we should complain to that guy who runs Oracle? (Oh, was a bit surprised when earlier today I typed in www.sun.com and up came Oracle!) David
From: David Combs on 18 Feb 2010 19:31 In article <N0Pdn.8391$YR1.1268(a)newsfe17.iad>, Richard L. Hamilton <rlhamil(a)smart.net> wrote: >In article <hkvk80$qd3$2(a)reader2.panix.com>, > dkcombs(a)panix.com (David Combs) writes: >[...] >> Right now I'm using Windows XP with a Microsoft mouse that, >> on top, has a LEFT, a wheel, and a RIGHT, and one more on >> each side (on each vertical wall of the mouse). >> >> Those final two work for going fwd and backward in eg firefox. >> >> >> >> I[f] stupid MS can have these things working ok, why the H. can't >> sun! Maybe we should complain to that guy who runs Oracle? >[...] > .... .... > >I would not however expect Sun (Oracle), or even RedHat to pay >for most of that though; and for sure, the makers of the fancy >input device wouldn't bother, since they make 99% of their money >from Windows users. I'd expect that most of the initial work >would be done by individuals, just because they want it to happen >and think they know enough to get it done sooner themselves. > >You could be one of those individuals. Or you could offer a >bounty to someone to do the work. Maybe you're hoping that if >you're annoying enough, someone will do it just to get you to >_shut_up_. Tempting, but I'm sorry to say it takes less time >(and is about as much fun) to tell you where to get off than it >would take to implement a solution to your problem, which is so >not _my_ problem because I don't work for Sun/Oracle, and I don't >own a five button mouse... Thanks for the clear explanation! You might not have a 5-button mouse, but once you tried one, you'd sure like it. Makes browsing so much easier -- go back just by clicking the "button" (long) along the left wall, and back forward again by the one on the right wall. Sure beats having to move the mouse up to those arrows at the upper left corner of the screen. And the wheel is *really* nice. Now, I bet (well, hope) sun DOES support that, the wheel? Thanks, David
From: Richard L. Hamilton on 16 Mar 2010 00:28
In article <hlkm5j$n1b$1(a)reader2.panix.com>, dkcombs(a)panix.com (David Combs) writes: > In article <N0Pdn.8391$YR1.1268(a)newsfe17.iad>, > Richard L. Hamilton <rlhamil(a)smart.net> wrote: >>In article <hkvk80$qd3$2(a)reader2.panix.com>, >> dkcombs(a)panix.com (David Combs) writes: >>[...] >>> Right now I'm using Windows XP with a Microsoft mouse that, >>> on top, has a LEFT, a wheel, and a RIGHT, and one more on >>> each side (on each vertical wall of the mouse). >>> >>> Those final two work for going fwd and backward in eg firefox. >>> >>> >>> >>> I[f] stupid MS can have these things working ok, why the H. can't >>> sun! Maybe we should complain to that guy who runs Oracle? >>[...] >> > ... > ... > >> >>I would not however expect Sun (Oracle), or even RedHat to pay >>for most of that though; and for sure, the makers of the fancy >>input device wouldn't bother, since they make 99% of their money >>from Windows users. I'd expect that most of the initial work >>would be done by individuals, just because they want it to happen >>and think they know enough to get it done sooner themselves. >> >>You could be one of those individuals. Or you could offer a >>bounty to someone to do the work. Maybe you're hoping that if >>you're annoying enough, someone will do it just to get you to >>_shut_up_. Tempting, but I'm sorry to say it takes less time >>(and is about as much fun) to tell you where to get off than it >>would take to implement a solution to your problem, which is so >>not _my_ problem because I don't work for Sun/Oracle, and I don't >>own a five button mouse... > > Thanks for the clear explanation! > > You might not have a 5-button mouse, but once you tried one, you'd > sure like it. Makes browsing so much easier -- go back just by > clicking the "button" (long) along the left wall, and back > forward again by the one on the right wall. > > Sure beats having to move the mouse up to those arrows at the > upper left corner of the screen. > > And the wheel is *really* nice. > > Now, I bet (well, hope) sun DOES support that, the wheel? Solaris has supported vertical scroll for a long time, the same way Linux does (steps of the wheel up/down correspond to clicks of a fictional 4th/5th button, although I may have that backwards). Not sure whether or not they implemented support for _horizontal_ scrolling, which pretends to use buttons 6 and 7 similarly. |