On Wed, Sep 29, 2010 at 07:51:56PM +1300, Jason alavaliant wrote: > Some of you may vaguely recall a small python/gtk front end for configuring > wacom settings called wacom-config that was around about 4 years ago. > After a length of time that's turned the promised wacom-config v2 into a bit > of joke (at least until it got forgotten about 2 years ago) I'm finally back > and the complete rewrite of wacom-config is almost done and I'm hoping with > the much better code base and my improved coding skills since 4 years ago I > can release something of much better quality this time. However there are > some lingering points with the interaction between my code and the results I > get from xsetwacom that I'd be very appreciative if the developers of > linuxwacom could help me with so I can get things into a safe state for end > users. (hence why I'm writing to the devel list - I need official opinions > and I'm trying to avoid too many normal users getting into the app at this > point since it's not quite ready for normal use just yet). > > Right now if you checkout the svn from > https://sourceforge.net/projects/wacom-config/ and actually try using > wacom-config v2 you'll find a few rather interesting things go wrong with > your tablet settings. (and if you do run the code beware, when it saves > settings it creates login items in ~/.config/autostart/ to apply it's > settings on startup and runs a small script called wacom-hotplug-detector > you'll need to kill to completely disengage it from making changes to your > tablet). I'm not sure if it's my code using xsetwacom wrong, or if they > are bugs in xsetwacom and if so what the final behaviour will be so could > use some advise so I know if I should be altering the wacom-config code, > looking at what can be done with patching linuxwacom etc.
Haven't checked out the code because right now I'm at an airport and I'm not even sure when I can send this email. Please don't use xsetwacom. We have properties, they are the driver's official API. xsetwacom makes no guarantee for consistency or backwards compatibility on its commandline interface. It's intended for testing of settings at runtime, but not as a base layer for any other client. > 1) The biggest one that's causing problems for me currently is that in > 0.10.8/git the value for TPCButton seems to be backwards ( > https://sourceforge.net/tracker/?func=detail&aid=3075826&group_id=69596&atid=525124) > Since my code has TPCButton off by default as soon as anybody enters > wacom-config and saves their settings they end up with TPCButton acting as > if it's on due to the current off=on status. see the other email, but genererally if the setting for TPCbutton now works reverse to what it used to, it's a bug. > I can change my code to reverse the values I have assigned to TPCButton but > that will cause trouble for me if the reversed values get fixed in a later > linuxwacom release since then things will be backwards in ever-bodies > settings again (and by that point most people at the company I work at will > be using wacom-config v2 causing mass confusion). So I'd love > clarification on if the reversed on/off values for TPCButton are here to > stay or are something that will be fixed just as soon as somebody writes a > patch? > > > 2) TVResolution - This one I know is in a bit of a state of flux with > TVResolution being broken in 0.10.8 and only starting to be fixed in git > since the 0.10.8 release. I've been following the git changes and testing > and trying to get a working build it's looking much much better but there > still seems to be a few issues that are making it hard to support properly > from wacom-config ( > https://sourceforge.net/tracker/?func=detail&aid=3075831&group_id=69596&atid=525124 > limits you to limited functionality on one monitor). Then what is also > confusing me (but I've not logged a tracker item for it yet) is that if I > just set TwinView and TVResolution I can move from one twinview monitor to > the other by touching the edge of the screen as per usual. But if any > setting is used that effects tablet or screen mapping, ( say setting TopX > even if I set it to the current value it already has ) that stops the cursor > from moving between the screens, it stays locked to the left screen and > disconnecting and reconnecting the tablet seems to be the only way to get > out of that state that I've found. > > I'm unsure if I'm being premature logging bugs about the git TVResolution > support since it's still not part of on official release (and there might be > a developer who has their own big list of TwinView stuff they are still > working on?) or if this stuff is only being fixed as people point things > out/offer patches? my plan is to remove all multimonitor handling from the driver. this needs to be handled in the server, not in an input driver. > 3) the xsetwacom settings Accel and SpeedLevel no longer exist according to > xsetwacom, I'm unsure if I've missed an email on the matter when searching > the lists or what the story is. Should those settings just be ignored at > this point and removed from my code or is there something else I should be > using instead that replaces them? We use the server's standard acceleration code now. See http://xorg.freedesktop.org/wiki/Development/Documentation/PointerAcceleration, I hope it's still current. > 4) probably a dumb question but what is the difference between AbsWUp/Dn and > RelWUp/Dn? I'm trying to add support for bamboo devices to my code but I > don't actually own one so I'm having to make guesses as to how to support > them and not sure on the why and the how for Abs vs Rel for their > touchwheels. The Intuos4 has an absolute wheel. The wacom mice have relative wheels, that's I guess the difference. Cheers, Peter > Hopefully this makes a bit more sense of why I'm suddenly posting a bunch of > things to the linuxwacom bug tracker, I was mostly just checking my code > wrote what I thought was the write xsetwacom lines during the rewrite but > now I'm trying to move towards release status I'm finding all the points > where I think there are bugs in xsetwacom or other places where my knowledge > of the matter doesn't match up with the behaviour. > > > Thanks for your time. > -J > > p.s. There is also > https://sourceforge.net/tracker/?func=detail&aid=3075823&group_id=69596&atid=525124 > on my list of oddities but I've worked around it in the wacom-config code by > changing how I set buttons. ------------------------------------------------------------------------------ Beautiful is writing same markup. Internet Explorer 9 supports standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. Spend less time writing and rewriting code and more time creating great experiences on the web. Be a part of the beta today. http://p.sf.net/sfu/beautyoftheweb _______________________________________________ Linuxwacom-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/linuxwacom-devel
