ucontrol-list Mailing List for uControl (Page 3)
Brought to you by:
semios
Menu
▾
▴
ucontrol-list — General ucontrol mailing list
You can subscribe to this list here.
| 2003 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(21) |
Dec
(35) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2004 |
Jan
(14) |
Feb
(13) |
Mar
(22) |
Apr
(19) |
May
(5) |
Jun
(5) |
Jul
(5) |
Aug
(21) |
Sep
(13) |
Oct
(7) |
Nov
(20) |
Dec
(10) |
| 2005 |
Jan
(10) |
Feb
(51) |
Mar
(19) |
Apr
(52) |
May
(54) |
Jun
(11) |
Jul
(9) |
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
(8) |
| 2006 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
(1) |
Jul
|
Aug
(1) |
Sep
(2) |
Oct
(4) |
Nov
(7) |
Dec
(15) |
| 2007 |
Jan
(7) |
Feb
(6) |
Mar
(22) |
Apr
(22) |
May
(8) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Matt A. <ma...@ap...> - 2005-12-12 19:23:52
|
On Mon, 12 Dec 2005, Stephan Schulz wrote: >I'm using uControl to put Ctrl where God, Unix and Emacs intend it to >be (i.e. left to "A"). I don't use any other function. I still very much miss the scroll-wheel-with-modifier. I'd pay a bit for a program that would bring back that behavior. >I'll finally order Tiger today, but I'd like my Panther installation >to be useful until it arrives. In Tiger, you'll be able to change the evil caps lock key to control through the normal UI. There may have been 'defaults' ways of doing it even pre-Tiger. But at least on my main machine, I use a better method -- I use an "Apple Keyboard" with an ADB<>USB converter. Not only does it have the control key in the right place, it has the best key feel on any keyboard I've used. -- Answer: Because it makes conversations flow in a nonsensical order. Question: Why is top-posting wrong? |
|
From: Stephan S. <sc...@in...> - 2005-12-12 09:29:33
|
Hi All,
I'm using uControl to put Ctrl where God, Unix and Emacs intend it to
be (i.e. left to "A"). I don't use any other function.
I'm running Fink 0.72 on MacOS-X 10.3.9 (Panther). I recently
installed Apples latest security patch and updated fink to the latest
stable packages. That includes a new version of X11 (unless I'm
mistaken, it is XFree 4.4.0.10 - my powerbook is at home). uControl
is "recent" (see above) ;-)
After the update, uControl still works fine in Aqua, but not in
X11. If I disable uControl, I can use xmodmap to remap Caps Lock to
Control, but I get sticky control. If I use uControl, xmodmap has
"weird and undesireable" effects (No usable Control at all). Also,
Meta (the Apple key) seems to cease working in X11.
Does anybody have useful suggestions? Please? ;-)
I'll finally order Tiger today, but I'd like my Panther installation
to be useful until it arrives.
Bye,
Stephan
--
-------------------------- It can be done! ---------------------------------
Please email me as sc...@ep... (Stephan Schulz)
----------------------------------------------------------------------------
|
|
From: Victor E. <eij...@cs...> - 2005-10-11 22:52:07
|
Ok, so fkeys doesn't handle the new powerbooks yet. Here's for the people who do everything in X: %% cat ~/usbbook.modmap clear Control clear Mod1 keycode 63 = Alt_L keycode 66 = Control_L add Mod1 = Alt_L keycode 84 = Control_L add Control = Control_L (I just got a new powerbook, and my previous xmodmap didn't work anymore.) Victor -- Victor Eijkhout -- 865 974 9308 eij...@cs... http://www.cs.utk.edu/~eijkhout/ |
|
From: Flynn <fk...@ko...> - 2005-07-29 16:01:50
|
> =C2=A0 Great ;-)=C2=A0 Haven't had time to test 0.0.1, now I get =
0.0.2.=C2=A0 Not
> that I mind.=C2=A0
<laughter> Well, I hope 0.0.2 works for you, at least. [ ;) ]
> Could the underlying keyboard mapping part of fKeys basically be plugged
> directly into the uControl source, or would that not work due to how the
> uControl prefPane and kext communicate?
It would be tricky, but only indirectly because of the preference pane.=20
The basic difficulty is that the architectures of fKeys and uControl are=20
extremely different: fKeys is an I/O Kit driver and uControl is not.=20
Therefore, the mechanisms that fKeys uses to locate and reprobe keyboards=20
are not available to uControl. So you really can't just take the keyboard=20
management from fKeys and dump it into uControl.
It would be possible to make uControl an I/O Kit driver, of course, but=20
then you run into the communications issue you asked about. uControl uses=20
the sysctl interface to communicate between the preference pane and the=20
KEXT -- and I've been warned that sysctl and I/O Kit don't play nicely=20
together. So you need to change that when you change uControl into an I/O=20
Kit driver...
...in the end, it seems like it's probably simpler to add runtime=20
configuration to fKeys, rather than how to get fKeys' keyboard management=20
into uControl. (For the curious, yes, I talked to Shane before going this=20
route. [ ;) ] uControl has been a big help for me, both in its=20
functionality on Panther and as a learning tool just now.)
> I noticed when installing 0.0.1 that because the fKeys disk image is
> read/write, it tries to prebind... on the *disk image itself*. . .
Ooops! Sorry about that. You're correct, for both 0.0.1 and 0.0.2 the=20
only application on the DMG is the installer itself, so god only knows what =
Tiger's trying to do. I'll make 0.0.3 a read-only DMG; thanks!
-- Flynn
fk...@ko...
--
God is a comic playing to an audience that's afraid to laugh.
(Voltaire(?))
|
|
From: Jim W. <js...@bl...> - 2005-07-29 08:13:05
|
> Having just announced fKeys 0.0.1 yesterday, I'm even more pleased
> to announce the availability of fKeys 0.0.2, a barebones keyboard
> remapper for Tiger.
Great ;-) Haven't had time to test 0.0.1, now I get 0.0.2. Not
that I mind. Could the underlying keyboard mapping part of fKeys
basically be plugged directly into the uControl source, or would that
not work due to how thte uControl prefPane and kext communicate?
I noticed when installing 0.0.1 that because the fKeys disk image
is read/write, it tries to prebind ("optimize system performance") on
the *disk image itself* (Can it do that - I thought it could only
prebind applications - and fKeys just has the installer package, right?
Or is it that little "program to see if it can be installed" (the
new "Do you really want to delete this file" nuisence dialog on
MacOSX [and therefore totally useless from a security POV..)
Jim
|
|
From: Flynn <fk...@ko...> - 2005-07-29 03:48:25
|
Ladies and Gentlemen and Children of All Ages:
Having just announced fKeys 0.0.1 yesterday, I'm even more pleased to
announce the availability of fKeys 0.0.2, a barebones keyboard remapper for
Tiger.
fKeys 0.0.2 is basically a complete rewrite of the keyboard management part
of fKeys: it can now successfully load on boot and handled hotplugged
keyboards. At least, uh, on my machine... [ ;) ]
As always, fKeys can be found at
http://www.kodachi.com/fKeys/
PLEASE read the web page for more information. If any of you choose to try
it out after dire warnings like that one, please let me know how it goes --
email to fk...@ko... will always work. Thanks!
-- Flynn
fk...@ko...
--
Two people / and a machine / make one animal.
(Motorcycling through the eyes of Keiko, from David Mack's _Kabuki_)
|
|
From: Flynn <fk...@ko...> - 2005-07-28 04:20:28
|
Ladies and Gentlemen and Children of All Ages:
I am pleased to announce the availability of fKeys 0.0.1, a - very
barebones! - key remapper for Tiger, at
http://www.kodachi.com/fKeys/
PLEASE read the web page for more information -- fKeys is currently
_extremely_ young and _extremely_ limited. It's best to think of it as an
early beta at this point -- don't expect too much yet, other than (with any
luck) fairly rapid improvement.
If any of you choose to try it out after dire warnings like that one,
please let me know how it goes -- email to fk...@ko... will always
work. Thanks!
And, last but most definitely not least, my sincere thanks to both Shane
Celis and Alex Harper for their help in this effort. It's been a steep
learning curve so far and they've done a nice job helping me along it. [
:) ]
-- Flynn
fk...@ko...
--
The price of freedom is eternal vigilance. (Thomas Jefferson)
|
|
From: Nathan H. <ger...@ya...> - 2005-07-05 18:25:16
|
Since Microsoft's keyboard drivers that were written for 10.2 or earlier are still working and able to do the Alt-Cmd switch it is obvious that some kind of legacy keyboard magic is still working. > This all changed in 10.3 with Panther (I'm the guy who changed it in > uControl). > > - a |
|
From: Adam M. <ad...@me...> - 2005-07-05 17:20:40
|
This all changed in 10.3 with Panther (I'm the guy who changed it in uControl). - a Heiko Hellweg <hh...@sn...> writes: > > Is there a technical manual (other than the source code) anywhere > > describing how uControl worked on Pre-Tiger systems, and which hooks > > are necessary which are gone in Tiger? > > There is one basic overview document on the mechanism iJect used on > OSX 10.1.1: > > http://www.obdev.at/developers/articles/00001.html > > iJect was - to my understanding - the GPL anchestor of all the later > OSX HID hacks like DoubleCommand, CommandAndControl, PowerDelede and > also uControl. > > I think, the technique described there was used by uControl thru OSX > 10.2 but > stopped working on Panther and uControl switched to something different. > > maybe this helps a little (at least to understand old code fragments > in the sources)... > > Heiko > > > ------------------------------------------------------- > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > from IBM. Find simple to follow Roadmaps, straightforward articles, > informative Webcasts and more! Get everything you need to get up to > speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click -- |
|
From: Brian L. M. <bl...@bl...> - 2005-07-01 16:56:56
|
At 9:53 AM +0200 7/1/05, Heiko Hellweg wrote: > I have yet to find an official Apple >Documentation on these (i'm sure, there is one, ...) I remember Inside Mac used to have these. I don't know if they kept them up-to-date or not though. Brian |
|
From: Heiko H. <hh...@sn...> - 2005-07-01 07:53:18
|
Hi Dave, At this level, you can not use Character codes like ASCII or ISO-Latin-1 (that's where backspace os 08) but rather the internal ADB Codes and I have yet to find an official Apple Documentation on these (i'm sure, there is one, ...) You can see them if you have Classic and run good old ResEdit: create a new dummy ressource file and create a new Ressource of type 'KCHR' (classic keyboard layout). The built in KCHR Editor shows you a mockup of your keyboard and if you mouse-click on one of the keys, it shows the hex ADB code in the status area. In your case: the ADB code for the upper-left key is 0x0A and Backspace is 0x33 good luck - Heiko |
|
From: Heiko H. <hh...@sn...> - 2005-07-01 07:35:11
|
> Is there a technical manual (other than the source code) anywhere
> describing how uControl worked on Pre-Tiger systems, and which hooks
> are necessary which are gone in Tiger?
There is one basic overview document on the mechanism iJect used on
OSX 10.1.1:
http://www.obdev.at/developers/articles/00001.html
iJect was - to my understanding - the GPL anchestor of all the later
OSX HID hacks like DoubleCommand, CommandAndControl, PowerDelede and
also uControl.
I think, the technique described there was used by uControl thru OSX
10.2 but
stopped working on Panther and uControl switched to something different.
maybe this helps a little (at least to understand old code fragments
in the sources)...
Heiko
|
|
From: <in...@ka...> - 2005-06-30 09:10:46
|
このメールは下記事業者・送信者が送信しております。 ------------------------------------------------------------ <事業者><送信者>(株) ＧＳＳ inc 住所：東京都世田谷区等々力5-20-12 TEL 03-6425-2181（担当：木村） MAIL gss...@ya... ------------------------------------------------------------ 特定商取引法施行規則 受け取りを希望しない場合の連絡方法 関係のない方に届きました場合お詫び申し上げます。今後一切の配信を 拒否する場合はお手数ですが、下記URLのクリックしてください。 http://www.qmsys.net/stop.php?id=2&sid=2c1f51e9189c0c5c256f8259a9779686 上記URLのクリック処理でエラーが起こった場合は、お手数ですが、 下記のアドレスに「受信拒否」の件名で送信してください。 st...@qm... ------------------------------------------------------------ ☆未経験者歓迎！完全保証のSOHOワーカー募集ニュースです☆ ☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆ 《楽マガVol.121》 速報!オススメ楽笑マガジンニュース ＜2005/6月発行＞ ☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆…☆ 【↓今月のオススメ情報/ビジネス】 1,935,654部発行ー☆ 三┃大┃ 保┃ 証┃で┃守┃ ら ┃ れ┃た┃Ｓ ┃ Ｏ┃ Ｈ ┃Ｏ ┃ワ┃ ー┃ ク ┃で┃ ━┛━┛ ━┛ ━┛━┛━┛ ━┛ ━┛━┛━┛ ━┛ ━┛━┛━┛ ━┛ ━┛━┛ 人┃生┃を ┃イ┃キ┃イ ┃キ ┃！ ┃ ┃ ━┛━┛━┛━┛━┛━┛ ━┛ ━┛ ┃ ┃ ┃ ┃ ★保証その1★業務提供保証→必ず稼げる在宅ワークの募集です。 ┃ 一日2時間で月5万〜 ┃ ┃ ★保証その2★業務サポート保証→未経験者・初心者の方でも大丈夫です。 ┃ ┃ ★保証その3★生涯業務提供保証→登録1回のみ!続けられる限り業務提供が保証されます。 ┃ ┃ ┃ ★まずは資料請求・お問い合わせを！ ┗━━━━━━━━━━━━⇒⇒⇒⇒⇒http://www.kawara.rulez.jp/ ||||||||||||||||||||||||《在宅業務募集》|||||||||||||||||||||||| ☆資料請負ホームページ☆ …………………☆ 在宅業務お探しの方に朗報!! ☆…………………… ▼空いた時間で少しでも余分に稼げれば・・・ ▼子供が小さくて家を空けられない・・・ ▼地方だとそうそう仕事がなくて・・・ ▼人間関係がうっとうしくて・・・ ▼先行き不安なので今のうちに・・・ ☆☆☆ おまかせください!全部解消されます・・・・ ☆☆☆ 〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜〜 ▼「特別な資格がない…」 ―――――→ 『SOHO』に資格は必要ありません。 ▼「育児が大変だから…」 ―――――→ １日２〜３時間程度なら十分あるはずです！ ▼「自分はもう歳だし…」 ―――――→ 在宅ワークに年齢は関係ないし定年もありません！ ▼「文字入力程度しか出来ない…」――→ 実務研修制度があるから安心です！ ▼「本当に仕事はあるの？」 ――――→ 私共は常々、登録者が100％確実に安定収入を 得られる事を最優先しております。 ☆彡------☆彡------☆彡------☆彡------☆彡------☆彡------☆彡------☆彡 これにより登録者には常に安定的な仕事を供給することが可能となっております。 収入は簡単なデータ入力（ベタ打ち）を1週間10時間程度で月収５〜６万円程度から、 中・上級レベルの会計・検品・校正業務、Webインストラクター・HPデザインなど 月収18万〜32万円迄、長く堅実に安定収入を得たいとお考えの方であれば、 私共が責任を持って万全のサポート体制でご支援致します。 現在、多くの主婦・会社員・年配者・学生・フリーター等が活躍しています。 業務はどなたでも出来るデーター入力・文書入力などまったくのパソコン未経験者でも マニュアルに沿った基本操作さえ覚えれば大丈夫です！ ********興味のある方は下記から資料請求(無料)ください****************** ⇒⇒⇒⇒⇒ http://www.kawara.rulez.jp/ 人数枠の都合がありますので、なるべくご希望に添った対応を するためにもお早めのご連絡をお願いします。 ***************************************************************** ---------------------------------------------------------------- ∞∞∞∞∞∞∞∞∞∞∞∞◇資料請求手順 ◇∞∞∞∞∞∞∞∞∞∞∞∞∞ ---------------------------------------------------------------- １． 『http://www.kawara.rulez.jp/』へアクセスする。 ２． 『トップ画面』→『資料請求』 ３． フォームにしたがって必要事項を入力する。 ４． メールアドレスはお間違えのないように入力してください。 ５． 必要事項をすべて入力し、『送信』をクリック。 ∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞∞ ※追伸 ------------------------------------------------------------------------ ●突然の広告メールで失礼をお詫び申し上げます。 迷惑メールになりましたらご容赦の程、お願い申し上げます。 ●この広告メールはウィルスチェックソフトにて万全なチェックをしております。 ●弊社はマルチ商法や悪徳業者とは一切関係ございません。 お問い合わせ頂きましたお客様のみにきめ細かな対応をしており、執拗な営業は一切致 しません。 - 未経験者歓迎！完全保証のSOHOワーカー募集ニュースです!----------------- 掲載内容についての詳細、サービス内容については、(株)GSS inc. ではお答えすることができません。メール中でご案内している各団体・ 企業等のお問い合わせ窓口をご確認ください。 |
|
From: Michael E. <mic...@gm...> - 2005-06-30 02:43:41
|
Yes, I've now filed a bug report with Apple. Thanks all, ME |
|
From: Matt A. <ma...@ap...> - 2005-06-27 21:18:33
|
On Mon, 27 Jun 2005, Michael Edwards wrote: > Now I'm in the horrible position of having to open the control panel and > remap the keys whenever I move my PowerBook from my home office to any other > location (and then one keyboard is useless). You have written up a bug detailing your problem, right? bugreport.apple.com -- Answer: Because it makes conversations flow in a nonsensical order. Question: Why is top posting wrong? |
|
From: Michael E. <mic...@gm...> - 2005-06-27 17:17:34
|
Harris, Thanks so much for the suggestions ... I will investigate. I've been grovelling around in the prefs files myself and I've found the one I need to swap. I can do what's required with a small bit of shell script or Applescript ... I already need to do manual location changes for network settings (thanks to some DHCP stupidity in the office), so I can wrap the network changes and the keyboard changes into one little script ... Peripheral Vision could be very helpful (just had a quick peek, and I may be able to use it to fix my network problems too!). Hmmm ... Thanks again, ME On 27 Jun 2005, at 23:48, Harris Upham wrote: > If you have the powerbook internal/external keyboard mapping > problem (as I do myself) I suggest that you look at Prefs Commander- > > http://www.stone.com/PreferenceCommander/PreferenceCommander.html > > I have not had time to experiment with it, but I'm of the > understanding that this tool can switch out prefs files easily... > so you could do two sets of keyboard prefs and have this thing > switch between them for you. I also thought of automating it by > triggering it with Peripheral Vision, which visually identifies > peripheral equipment onscreen as you connect it- and can optionally > open files/run scripts etc. when certain gear is recognized. > > Assuming that it works, and that you buy both licenses, you've > solved the problem for about $27. Not such a bad tax for putting a > tiger in your tank, now is it? Wait, don't answer. > > I don't use my powerbook very often anymore so I haven't taken the > time to experiment, but I thought I'd share my idea with others... > > bon chance! > > Harris > |
|
From: Harris U. <ha...@es...> - 2005-06-27 16:48:03
|
If you have the powerbook internal/external keyboard mapping problem (as I do myself) I suggest that you look at Prefs Commander- http://www.stone.com/PreferenceCommander/PreferenceCommander.html I have not had time to experiment with it, but I'm of the understanding that this tool can switch out prefs files easily... so you could do two sets of keyboard prefs and have this thing switch between them for you. I also thought of automating it by triggering it with Peripheral Vision, which visually identifies peripheral equipment onscreen as you connect it- and can optionally open files/ run scripts etc. when certain gear is recognized. Assuming that it works, and that you buy both licenses, you've solved the problem for about $27. Not such a bad tax for putting a tiger in your tank, now is it? Wait, don't answer. I don't use my powerbook very often anymore so I haven't taken the time to experiment, but I thought I'd share my idea with others... bon chance! Harris |
|
From: Michael E. <mic...@gm...> - 2005-06-27 16:05:59
|
ME TOO! I have been using the keyboard remapping with uControl in exactly the same way as Nathan H. At home I have a USB keyboard plugged into a small hub full of devices which is attached to my laptop. The Option(Alt) and Command(Windows) keys need to be swapped for the external keyboard to be any use to me. Using uControl I was able to set up a mapping which applied only to this keyboard. When I plugged in my various home harness cables (USB, power, audio, network), I automagically had a remapped external keyboard ... and my PowerBook keyboard continued to behave normally. Now I'm in the horrible position of having to open the control panel and remap the keys whenever I move my PowerBook from my home office to any other location (and then one keyboard is useless). I'd also be willing to pay something for this. Heeeelp!!!! Michael Edwards |
|
From: Nathan H. <ger...@ya...> - 2005-06-23 21:34:23
|
I use a powerbook. At home it's connected to a KVM to a microsoft natural keyboard. With uControl each keyboard could have different settings so I could have the Cmd/Alt keys switched on the external KB but not when I was using the pb's built-in keyboard. Using Tiger's Cmd/Alt switching affects the built-in keyboard as well. So without ucontrol I'm constantly switching the keyboard mapping. This sucks! I'd pay good money to have this back. |
|
From: Victor E. <eij...@cs...> - 2005-06-10 18:42:46
|
So I finally upgraded to Tiger. So I finally lost my right control key.... At the start of an X session, run xmodmap on this file: [~] %% cat tibook1g5.modmap clear Control clear Mod1 keycode 67 = Alt_L add Mod1 = Alt_L keycode 60 = Control_L add Control = Control_L The key 60 is the Enter, and now it is a control. Since I only need the right control in emacs, which I run from X11 (Xdarwin, not Apple's), this is enough for me. Victor. |
|
From: Alex H. <bas...@fo...> - 2005-06-02 22:17:03
|
> =A0 Are these errors both tied to the ObjcUnit.h file, which seems to be mi=
ssing
> in xCode 2?
OCUnit is jUnit for Objective-C, it was never bundled with Xcode. This is
just Shane being professional and trying to add unit tests.
The important thing to understand is that the unit tests do not exercise
anything that actually is broken in Tiger. The unit tests cover the
datamodel for how the configuration is stored and pushed into the kernel.
This is actually all old code that probably should have been replaced
already. Sysctl is a bit evil and better techniques exist. But hey, it
works, so no one ever bothered.
In any case, don't fight trying to get the tests to build, it won't prove
anything. The broken parts are in the driver which have no unit tests. Buil=
d
the driver target, look at the code and compare pre-10.4 HIDSystem to the
new version.
> Why doesn't xCode just have a "Build All Targets" setting?=A0
Make a target that is dependent on the other targets (an aggregate target).
The project wasn't set up with one already for no particular reason, it jus=
t
never was something that anyone bothered to add.
> If it does, I can't find it - and it's not clear at all how to set a targ=
et
> without going to the Build menu - I'd expect a "Set active Target" option=
in
> the contextual menu for items in the Targets "folder"..
You can add a target selector popup to the project window using the usual
Cocoa "Customize Toolbar..." magic.
Alex
--=20
Alex Harper bas...@fo...
"Generally speaking, things have gone about as far as they can reasonably
go when things have got about as bad as they can reasonably get."
- Tom Stoppard
|
|
From: Alex H. <bas...@fo...> - 2005-06-02 22:05:11
|
> Is there a technical manual (other than the source code) anywhere
> describing how uControl worked on Pre-Tiger systems, and which hooks
> are necessary which are gone in Tiger?
Nope, no documentation like that ever existed.
> It was my general impression (not having looked at the code), that the KEXT
> worked by intercepting the key at the level of the HID driver level, then
> changing the raw key code in the event record, or in the case of virtual
> scrolling, creating an appropriate NSEvent to send up to Foudation (don't
> know how of if it worked with Carbon/Classic).
That's not really correct. UControl patched the HIDSystem object, which is
the chokepoint for events flowing from all HID devices. Event records are
abstractions that exist outside the kernel. Inside the kernel its a much
different model. UControl never talks to Foundation or any of the UI events,
which is why its effects were universal.
> Or was it that the old Darwin architecture allowed a kext to sit
> between the HID driver (whichever one was being used, for a keybaord,
> mouse, or trackbal) and the higher levels of the stack, and modify
> what came out of the HID driver regardless of what it actually was?
"Allowed" is a strong word :)
But your basic description is close. It was possible, pre-10.4, to hack your
way into the HIDSystem that let us modify events prior to the
kernel->userspace bridge.
Alex
--
Alex Harper bas...@fo...
"Generally speaking, things have gone about as far as they can reasonably
go when things have got about as bad as they can reasonably get."
- Tom Stoppard
|
|
From: Jim W. <js...@bl...> - 2005-06-02 00:10:20
|
Well, I'm mucking around with the source, after find what (look
like) the correct header files IOHIDSystem.h and ev_private.h from
the 164 version of IOHIDFamily in the Darwin tree.
I'm building the 'All Tests' app now, and getting one error that
'ObjcUnit.h' can't be found (is this something that used to exist in
the xCode tree that they got rid of? The other error is a parse
error concerning the instance method. The original code is
#import <Foundation/Foundation.h>
// The AllTests.h file
#import <ObjcUnit/ObjcUnit.h>
@interface AllTests : NSObject {
}
+ (TestSuite *)suite;
@end
I get an error 'parse error before + (TestSuite *) ..' and a warning
'@end must appear in implementation context (how else will the
compiler know where the end of the interface is?)
Are these errors both tied to the ObjcUnit.h file, which seems to
be missing in xCode 2?
Jim Witte
<xCode rant />
Why doesn't xCode just have a "Build All Targets" setting? If it
does, I can't find it - and it's not clear at all how to set a target
without going to the Build menu - I'd expect a "Set active Target"
option in the contextual menu for items in the Targets "folder".. |
|
From: Matt A. <ma...@ap...> - 2005-05-31 22:15:21
|
On Tue, 31 May 2005, Jim Witte wrote: > Is there a technical manual (other than the source code) anywhere describing > how uControl worked on Pre-Tiger systems, and which hooks are necessary which > are gone in Tiger? It was my general impression (not having looked at the > code), that the KEXT worked by intercepting the key at the level of the HID > driver level, then changing the raw key code in the event record, or in the > case of virtual scrolling, creating an appropriate NSEvent to send up to > Foudation (don't know how of if it worked with Carbon/Classic). AFAIK, it call(ed) SPIs, and those SPIs have changed. (SPIs being private APIs.) What I think someone really needs to do is to write a bug (yeah I know I say that a lot) asking for an official way to post a scroll-wheel event. I haven't gotten around to this, but of course, external requests for this are good too. -- Answer: Because it makes conversations flow in a nonsensical order. Question: Why is top posting wrong? |
|
From: Jim W. <js...@bl...> - 2005-05-31 17:34:44
|
Is there a technical manual (other than the source code) anywhere describing how uControl worked on Pre-Tiger systems, and which hooks are necessary which are gone in Tiger? It was my general impression (not having looked at the code), that the KEXT worked by intercepting the key at the level of the HID driver level, then changing the raw key code in the event record, or in the case of virtual scrolling, creating an appropriate NSEvent to send up to Foudation (don't know how of if it worked with Carbon/Classic). Or was it that the old Darwin architecture allowed a kext to sit between the HID driver (whichever one was being used, for a keybaord, mouse, or trackbal) and the higher levels of the stack, and modify what came out of the HID driver regardless of what it actually was? Jim Witte Indiana University CS |
53 messages has been excluded from this view by a project administrator.