There’s such a future-push within the web-world, it’s a tad crazy (and ppl know it). The fact that I’m writing this short note is more about my interest in preservation that publication. I’d like to look back one day and say “Ha! I was sooo wrong (or right!).”

So without further ado, here’s my “Future of the Web Browser.” (as opposed to his)

First off, the internet has always been more than just the web. There’s numerous protocols which fly all around, and tonnes of data packets which go to places other than IE, Firefox or Safari (or Opera, Epiphany, Midori, etc). This is the first point: the web has turned into a 3-level programming arena with AJAX: JSON packets of data + css stlying onto an html frame. It’s the browser which glues ‘em together. But do we need a browser for all of this? Greasemonkey scripts have successfully edited the CSS & the browser needs some kind of framework, but the data.. ah yes, it’s always about the information. Ever put an RSS feed into your browser’s bookmarks-bar? Yup. That’s the future. But why should this be JUST in the browser? Why not in other applications? Or in the deskbar-panel (if such exists in the future).Excel has (for years) supported database connections, and in recent years, supported them over the web with XML. Keep. going.  Is a spreadsheet the only thing that needs raw data?

So that’s my first-round of ideas: open data, in open formats, available in dispersed locations. Such diffusion of the web into separate, external apps or media seems inevitble. People can handle 5 different things on one table-top, but why cook where you disassemble your car? Maybe I’m human, and LIKE to get up from my seat, see the world from other angles.

Second, now that we have removed raw-data from the web browser (and put it to good use), we are left with the other genre’s of content: news & blogs (oh wait, that’s already in your RSS reader). Friends status updates (again, why can’t this be integrated somewhere useful, where you want it to be). Youtube and videos can also be extracted (if corporations get their grubby hands out & let content be open & free!) What’s left for a browser? Not much. Nothing unless there is novel content which is not diffused into another application.

Notice all my ideas still revolve around a “desktop” (which I actually despise). All I’m saying is that so long as the native-coded, windowed environment exists, the web can be in places other than the browser. iTunes is the prime example. Second, yes, there is a chance of “every window is a browser.” That’s already the case/possible with Prism, making your OS useful only as a windowing manager (which I debate doesn’t manage very well).

Anyways that’s the future as I can see.. Either content is re-embedded into HTML, for proprietary purposes, or it is free. Either way, it is in native code somehow.

The only other point is caching, which HTML5 brings. I love caching, I’m writing an “everything-checker-viewer” right now which survives on caching alone. Live-data is nice, but until connections across the world become faster, consistent and needless-to-mention available/accessible, caching will inevitably rule.

Scrollbars are kinda, sorta, useful. I remember teaching my mom how to use a computer & she would consistently use the scrollbar backwards. And it makes sense: moving it down is moving the page up.. wha?? What the rest of us take for standard (scrollbar-as-document-representational) is backwards to how we read any other book or piece of paper.

While writing a paper recently on a portrait-view screen, I found myself able to see the full page I was writing, AND bits of the prior & next page. This gave such a HUGE advantage for context-analysis and paper flow. I bet most prof’s suffer through poor student papers, simply because the student doesn’t realize what is missing writing in landscape (and even worse, widescreen). In contrast, this is the joy of minority report: a HUGE surface area to hold windows-of-sufficient-size that can be placed with sufficient margins instead of being in an overlapping pile, thanks to the portability requirement. I’m still waiting for short-throw, cheap, hi-def projectors that I can have overhead, to project onto my white, tilted, large physical desktop.

Our eyes are tuned (in the real world) to spatiality. There is a layout, flow, and gradual perceptive fall-off from our field of vision. Computer UI designers are kind enough to give us EVERYTHING. This is required for use, and for the most part, our eyes ignore this extra.  Yet, we never have any ‘real’ awareness of what is off-screen. Some window-managers use ‘pagers’, but that isn’t for window content, as much as window layout. I’m a content-layout fan– keeps me focused on my task of creating, instead of doing the window manager’s window-layout job.

So I propose 2 options (especially for pen-computing), both using the mouse-pointer as a grab-and-drag hand, a la Adobe Acrobat, or FF’s plugin.

Gradients

Remaining content 'edged' by gradient

Gradients are naturally-existing creatures around any real-world document. Shadows are handy, and they can tell us (representationally) how much of the document is remaining to view. Sure it’s not a mathematically precise as a scrollbar, but ‘scrolling’ through a newspaper is hardly mathematical.

Magnifing Glass

Similar (actually inverse) to the gradient, I’ve also considered having the margins of a window, where there is still more content, grow in proportion to the content remaining. This would give a ‘squished’ version of the remaining pages.

Likely scrollbars will exist forever, and likely I’ll be happy with grab-and-drag for all apps. Scrollbars do have the goodness of being able to scroll faster than anything else- instantaneous access to the top or bottom of the page. But isn’t that what the ‘home’ and ‘end’ keys do? Oh, wait, most keyboards don’t have those..

The bits missing from the Courier Video-demos:
First off, it is very, very limited:

• Left:
  • Calendar
  • Contacts
  • Web
  • Photos
• Right:
  • Maps (associated with contacts)
  • Journal (dual-pane)
  • web also (wait, which side do I view webpages on.. both?)
• Middle:
  • Clipboard-Pocket

This is it? Doesn’t sound like much. So what do I use my laptop for these days?

• Libraries of media, for which iTunes & Amazon are trying to be the end-all. For this, I likely will run a streaming media source, or keep my library on a home server, so the laptop neWed not have a multi-terabyte drive.
• Basic file editing/viewing, for which Google Documents is trying to be the end-all. And so long as your online (or whenever local caching of GDocs hits) this might work. But Google doesn’t support various Journal formats at present.

Web viewing

There doesn’t seem to be any web-organization, just a list of tabs/Safari-like window-images. My use of the web falls into the 2 categories previously mentioned: the daily views of interest & responsibility (like reading the news & checking the bank) and researching the latest idea & interest I had, or for classes.

File-Viewing

I wanna make sure this has full office & PDF file editing & annotation, otherwise, it’s not a laptop replacement. Clearly this thing won’t be creating the office-heavy/design-heavy files.. but it should have full-view of them.

Filetypes supported are likely OneNote->PDF instead of the millions of incongruent filetypes presently on any one system. Others’ ability to edit & comment on PDFs at present is limited. This is an early-collaboration tool, not a late-collaboration, final-product kind of tool.

Pocket

There is a heavy amount of visual drag-and-dropping, specifically of web/PNG images. But I deal with excel and photoshop files too. Will previews be auto-generated for these as well? How will I know which spreadsheet is which when I place it in the pocket? Sometimes filenames *are* handy.

Minor

We didn’t see any online chat-abilities with the contacts, something my nokia N800 did well.

The 2 biggest hurdles for development:

1. Metadata heavy
2. Handwriting heavy

Most of our desks are filled with frequently used items. Some bits might be more long-term, be it clutter, research or sentimental items.

However, “Computer as a Desktop” or “Desktop on the Computer” is a failed ideal of the early, by-gone days of computing. Simply: everything is not and cannot be on your personal computer *screen* let alone actually on the computer.

I’m a huge computing fan, I even PDF my readings for class (thanks to an old printer/scanner my wife’s new fancy-pants mac can’t use!). I annotate the PDFs & highlight them accordingly. I’m very document heavy, being a BachArts instead of a BachScience guy.

Compared to the image above, I’ve got only 3 piles of junk on my upper right & left corners: 2 book piles on the right & 1 ‘organizer’ on the left which holds everything from my keys & cellphone to pens, coupons & my external harddrive. And sticky-notes. Stickies can be moved, computer notes cannot so easily (yet). Cell phones & keys are much more ‘real’ objects, requiring separate spatial existence — like the kleenex box nearby. No computer will replace kleenex (I hope!).

So with all this, I’m not convinced the “desktop” metaphor is worthwhile, or even warranted. Sure there’s lots of files in folders on my computer, but (a) search has replaced some of that and (b) files and folders are in cabinets, not desks..

So, iTunes and other newer applications take on the library metaphor. You search a library; it is referenced, accessed and used. But that is just for an app, not a unifying theme between apps (though a good argument could and should be made why this shouldn’t be so!). User folders on Mac & Linux are now folder-lists of Documents, Media-types (pictures, music, videos).. Various netbook UI’s have ran with this as well. But none hold together any more than the files-in-folders theme. iTunes has it’s own folder inside your music folder. I know why this is, but shouldn’t there be a platform-independent format for these datasources?

And the desktop folder itself? Is there ever any real purpose to it? Most mobile platforms don’t have one and most linux geeks don’t place their files on the ‘desktop’. Microsoft Courier lacks a desktop as well, instead merging the desktop and the clipboard into one middle-bar for all things temporary (which most ppl use their desktop for).

Really looking at a normal, original desktop, it’s more a matter of objects which are acted upon with tools. Applications are the tool-set these days, but they fail to interact and avail themselves *on the desktop*. Desktops are loaded up with file *icons* not the file itself. Mac tries to maintain this file-window independence, but it’s just a jumble of windows overtop background pictures and icons. I cannot go from an open spreadsheet window and start making it pretty with a document design editor. These apps are mutually exclusive, and ruining any desktop metaphor.

All this critique is pointing towards a unified file format which all applications agree upon, and all have access to modify at anytime.

Clearly this is an idealism which won’t occur without limits placed. And such limits break the very nature of ‘general computing’ over which the Mac/Windows debate rages.

I’m a huge fan of limited and thereby differentiated computing: the workstation should not be running the same UI as a eJournal (Courier!). And neither of them are sufficient for the task of media center.

We once had a division of these 3 tiers with servers, desktops & mobile devices. Then WindowsNT was thrown on servers & XP(and NT variant!) was thrown on faux-tablets UMPC’s. This UI/codebase permeation is inverse of the ideal, where the low-level is the universal, and the high-level is specialized to the task.

Moblin 2.0 has a chance with this for at least one front.

UI usability is consistently my no. 1 time-suck. I’ve ‘wasted’ countless hours config’ing and reconfig’ing GNOME panels into assumed-need sorts. I’ve given up and returned to Mac & heeded again the siren call to Linux’ potential simplicity.

My latest spin *was* playing with screen-maximizing UI’s: task-bar-less environments, panel-over-top window-titlebars, then I turned wanna-be tablet with the iPen.

But I’ll not deny, the best ‘inspiration’ has been the non-existent-yet(?) Microsoft Courier. Here’s a list of the bits, related & not, which I’ve come up with:

Dual-pane

My 13″ macbook is nearly-exactly 2×9″ screens. Yes, this is very Courier-esque. Especially if I use Linux’ dragbox in between panes! But the benefit here is window management. In Xwindows-land, there’s lots of these things called ‘window managers’. They’re supposed to, y’know.. manage windows. What do they ACTUALLY do? Place windows in wierd locations & sizes which require you to use a taskbar, alt-tab or expose’ your day away. I’m not down with that. Time for a ’tiling” window manager. Yet I’m just looking for 2 panels, and on a 13″ screen, that’s plenty. If I config my window-placer-thing to throw some of my apps on the left, I (craziest idea ever) can expect them to be there. I know where they are! Which leads me to my next point..

Priority

Since I’m not fussin’ with my 9 windows that are open, wondering where to look for ‘em (something a taskbar is supposed to do, but doesn’t supply the requisite window-parallel usage scenario).. anyways, since I’m not wasting time placing windows, I can focus on what I’m supposed to be doing: being a human with responsibilities over resources and being creative and learning. Those are the categories my applications have taken: email, calendar & files on the left, OpenOffice, Journal & Web on the right.

Lists, lists, lists!

Perhaps this is more iphone-y than anything, but there’s some goodness to be had with the removal of clutter (and there’s plenty on the web!) Yet, I use google calendar & email all day long. I don’t need another cal or mail app, I just need a browser open with these bits in it. But even these apps aren’t clean. Facebook, Yahoo Mail, Google Mail, Google Reader AND Google Calendar ALL have sidebars. Why, oh why do we need sidebars? They take up sooo much screen real-estate, especially after you scroll.

Bad:

What you get when viewing web-apps half-screen'd or in portrait

Good:

Ahh, mobile: a clutter-free web-experience!

You just have to load the mobile versions of the webapps you use. To find them, I viewed them on my phone & checked the url. I’ve also found the ‘print’ versions of yahoo news to be similarly readable.

Conditions: sure forcing all my windows into 2 locations (left/right) & 2 states (half-screen or maximized) is ‘limiting’, but I’m a limited human! I need some parallels here to stay sane. I can’t be moving & resizing windows all day long.

Perhaps I’m just getting old, and will eventually regress into the old lady who only has one window open (maximized) at a time. Until then, this is a good compromise.

Everyone’s all abuzz with Microsoft’s latest leaked prototype: Courier. On an off-note, it’s hilarious to contrast the handwritten everything with the king of monospace fonts! (and previously, my latest interest has been the long-awaited crunchpad.)

But there’s 2 things I’m here to mention: my dream & the obvious rising behind Courier.

Why Courier is Obvious:

First off, how many people do you see carrying around a notebook/folio of some variety? Everyone. The business guys do it for their contacts, dates & files. College kids do it for their class notes. Artsy-kids do it for their scribble-pics. Christians have a habit of doing it for their Bible study/sermon notes. Everyone.

Second: netbook+eReader. Limited computing strikes again, and awaiting for a convergence.

Third: Intel’s atom platform (and I would argue ARM even more!) is ready for this kind of thin-and-goodness. Especially with ssd’s (heck, I’d be happy with an SDHC!)

My Little Dream:

I’m a fan of the UI-customizability of Linux. Always have been: it’s what keeps me away from MacOS. Right now, I’ve got a toolbar that has everything I need in it: time, calendar-on-click, applets & a task-switcher. All this overlays the wasted-space of window-titlebar. Most of the time I maximize my apps, so I can focus. But there are somethings that should be a sidebar: notably a tabbed filemanager (since I already have a tabbed term thanks to tilda).

And there’s no reason why a quick photo-viewer, calculator, contacts & datebook cannot also be in this sidebar (especially if cache’d & synced from Google!) All this sidebar stuff is too perfect. Why not make a Window Manager that runs specific apps in specific ‘frames’ (yes, like HTML old-skool style). The frames are resizable & collapsable. Ths is so (similar but) much more useful than a tiling WM.

Next up, I’ve never, ever understood file dialog boxes. I do however understand Delicious’ tagging. It auto-generates recommendations, and why not do this for files?  Linux is built for this: symlinks.

And while we’re at it, why not kill off scrollbars & make everything grab-and-draggable. Just use a modifier key (or both-click). Then we can have the app be like a magnifying glass, with the edges smushed to show you how much more you have down there (in preview-style).

As for all this journally stuff, Xournal is the single-best program I have ever, ever, ever encountered. Multiple-layers? yup. Print to PDF? Yup. Wanna add a new page? Click the ‘next page’ button. It could remove all it’s menus if it just had a ‘preferences’ dialog box. It is the model for any journaling program.

Lastly, mouse-gestures are very hot lately. I don’t use them because I’d like ‘em to be like Courier: context-specific AND list suggested actions, instead of always acting on its own.

Dreams, dreams, dreams.

All I’m sayin’ is if ASUS puts out an Intel Atom dual-screen netbook/eReader next year, or even this year, I’m putting my dreams to work. Just need to solve 2 problems:  the sidebar thinger (update: “devilspie” might be halfway there for me) & handwriting recognition (and I’ve got a prototype system coming this december when classes are out!).

The iPen, by Finger System, Inc.

First off, sometimes you can judge a book by it’s cover– especially when a marketing department has ahold of the cover. By this I mean the “Ages 5 & up” label. Yes, this IS a kids play toy, and I’m going to try and take it seriously. I hope that doesn’t make me a kid. Likely does.

First Impressions

first 15mins:  holy crap this is fussy! How does this work right at all?

second 15mins:  let’s try to figure some workarounds..

third 15mins:  ok, now with a basic system in place, perhaps there’s a chance.. a very small chance..

last 15mins:  It brings back the good ol’ days of owning a pocket pc.

1 month in: with the right linux-config’d system in place, the iPen has it’s place, and can even replace the need to own a full $1700 tabletPC. For $15, it’s a steal. But the whole repetitive stress thing isn’t any better or worse, but that’s my fault for being tied to this thing for 12hrs /day!

“The System” (My Linux config)

0) You really should use the old-school optical 1mm grid they provide.. graph paper doesn’t cut it. I ended up cutting a 2″x3″ square out & taping it on the corner of my laptop (nearly over the trackpad!) when I’m not at my desk. However, when running in ‘super slow mode’, i need the whole thing. Choose cutting wisely.
1) hold the mouse a bit more upright than a regular pen, especially when not using the grid-pad
2) Slow the mouse settings all the way down: xset m 0 0, (or see my ultra-slowed-down xorg.conf)
3) Use xournal to scribble notes & PDF annotate, zoomed in to full-width (150%), unless running ultra-slow, then tis fine zoomed out where the notebook lines are like college-ruled. (yes, i measured). Xournal also supports graph paper, colored paper & custom-paper sizes. Handy for when I rotate the screen. Also, I found having the default right-click to be the ‘hand’ is very handy (like acrobat’s hand/grab-drag-scrolling)
4) Use cellwriter for single character input (aka PocketPC-land!). Takes some config, but it gets pretty good.
5) Firefox plugin: “Grab and drag”!  This is when it hit me: “Don’t try and use a pen like a mouse. Use a pen like a pen!”

The Speed Test

With cell writer config’d & trained (not a hard process at all),
and using the sentence: “The quick brown fox jumps over the lazy dog”:
Typing: “THe quick brown fox jumped over hte lazy dog” is 8-9 seconds, with only 2 errors.
Cellwriter: “the quick brown fox jumped overthe lazy dog.” in 58 seconds, first try, then 44s!, with 2 errors as well.
Handwriting on paper: 13seconds
Xournal: 24seconds.

A 5-6 times increase in time is wretched. Especially if I in-line corrected about 4-6 characters thanks to a shaky nervous hand! No doubt I’ve cellwritten faster since this early test, but I seriously doubt it will ever come close to touch typing. The other variable is that when typing, I type faster than I think, so I’m always stopping to think about what I’m saying. But with cellwriter, I’m writing so slow, I can think about what I’m saying. Trouble is, if it’s too slow, I’m thinking about writing the characters & forgetting about what I’m saying.

Improvements

It’s amazing just how sensitive the nature of handwriting is. There’s 2 very strong components to pencil writing that is lacking with this digital pen.
1) Pencils are pressure sensitive across a gradient, not a binary on/off. You can swipe a pencil and get a line. When you lift a pencil while handwriting, it’s more a light drag across the paper than it is a full 1mm lift. Hopefully I’m explaining this right: the ipen requires a 1mm lifting or dropping of the pen to register a swipe. This is not natural, however oddly enough, when the mouse is super-slow, this problem dissipates, as well as with getting used to it.

2) Tracking. When we lift a pen and move it, it now resides to write in the new location. On the slowest mouse motion, the ‘mousing area’ is still 2″ wide x 1.5″ high. This is hardly a one-to-one (like true tabletPC’s have) with a 13.3″ diagonal screen! Second, this pen is very sensitive and you have to lift the pen a good centimeter off the surface to stop it from moving the mouse cursor (to get the mouse cursor where you want it, relative to the hand-on-paper position). This is also shown when setting down & picking up the pen.. a mouse or trackpad will leave the cursor & mouse, the pen must be set & picked up again.
One last comment, I’d prefer to have a scrollwheel where the right click button is. But some of that is solved by FF’s grab-and-drag plugin. If only that plugin was across all X/GNOME/KDE applications!

Since I’ve bought an ipen, I need different settings for the pen and trackpad. The trackpad should have all the fun little bits that it came with, 2-finger scrolling & a decent speed, while the ipen, to be useful for writing notes in PDFs needs to be soo much slower (a 3/4 ratio of pen-to-screen).  This setup will of course apply to any multiple input device setup..

1. backup xorg.conf (in terminal: sudo cp /etc/X11/xorg.conf /etc/X11/xorg.conf-original)
2. with the mouse plugged in, lets find which device is which:
   1. sudo cat /dev/input/pxaux   ..now touch the trackpad: see funny letters for each motion? Good. (control-c to exit)
   2. sudo gedit /etc/X11/xorg.conf, add these lines:

      Section "InputDevice"
      Identifier "Synaptics Touchpad"
      Driver "synaptics"
      Option "SendCoreEvents" "true"
      Option "Device" "/dev/psaux"
      Option "Protocol" "auto"
      Option "HorizScrollDelta" "0"
      Option "SHMConfig" "on"
      EndSection

3. Now, config the mouse by the same method, but trying /dev/input/mouse0 .. through mouse4.
4. Add these lines accordingly:

   Section "InputDevice"
          Identifier "Configured Mouse"
           Driver      "mouse"
           Option      "Protocol" "auto"
           Option      "Device" "/dev/input/mouse4"
           Option      "Emulate3Buttons" "no"
          #Option      "Resolution" "50"
          #Option      "Sensitivity"       "0.24"
          #Option      "ConstantDeceleration" "4"
   EndSection

5. You’ll notice the last 3 lines are commented out.. I used the ‘constant deceleration’ to slow down my 800dpi mouse-pen. I enver played with the other 2, though they might work as well.
6. Further, I had to add these lines as well for everything to work:

   Section "ServerFlags"
   Option "AutoAddDevices" "0"
   Option "DontZap" "off" #ctl-alt-back restart-X
   EndSection
   
   Section "ServerLayout"
           Identifier "Default Layout"
           Screen "Default Screen"
           InputDevice "Synaptics Touchpad"
   EndSection

When all is said & done, I also found gsynaptics (sudo apt-get install gsynaptics) which allowed fun things like ‘ipod-scrolling’ and corner/edge events.

Final note: I needed my pen to be slower than even xset m 0 0 could give me, though that was close.

2009-10-22 UPDATE: Instead of EVERYTHING above, I finally got this to work in HAL (which means keep the original /etc/X11/xorg.conf file clean), thanks to removing mouseemu which was always making the usb pen-mouse be reconfig’d to normal acceleration! In ubuntu 9.04, I simply created this file: /etc/hal/fdi/policy/10-x11-input.fdi:

<?xml version=”1.0″ encoding=”ISO-8859-1″?>
<deviceinfo version=”0.2″>
<device>
<match key=”info.capabilities” contains=”input.mouse”>
<match key=”info.product” string=”Finger System Inc. i-Pen Mouse FM-100BN”>
<merge key=”input.x11_options.ConstantDeceleration” type=”string”>3</merge>
</match>
</match>
</device>
</deviceinfo>
While debugging, I confirmed the settings were in lshal & /var/log/Xorg.0.log.

Holy crap! My fav toolbar calendar is more than just a calendar! It does alarms & full ICS input & output! With a little scripting & cron, I get the day’s events with one click!

#!/bin/sh
cd ~/tmp
wget http://www.google.com/calendar/ical/--your-private-gcal--url/basic.ics
mv basic.ics calname.ics

Just repeat the wget & mv for all calendars, add an hourly cron job & add the files to orage’s preferences/foreign calendars tab!

Love is simplicity!

I’m lately in love with openbox for it’s simplicity and lack of screen-coverage. Sure I could have setup my desktop within xubuntu, but I wanted to expand my horizons.

My original theory was for netbook application, maximizing screen real estate & reducing the hard drive & processor use. But lately, I’ve been toying with xcompmgr, hardly a ‘wise’ use of resources.

While Urukrama has a wonderfully extensive Openbox config page, I didn’t find the “make all inactive windows 60% transparent” like I was looking for.

A little googling gave me the solution I was after, but it was insanly processor/battery intensive! (That’s what ya get for writing an infinite loop! ) At first I tried putting a “sleep” command in the loop, and that did tremendous wonders for retaining processor use, and slowing down my frantic computer-work pace. Still, I wasn’t pleased.

So here, without further ado, is my xcompmgr+transset-df+perl script for setting all non-active windows:

#!/usr/bin/perl -w

# This script is a modification my M.Wallace of the
# original written by Andrei Perhinschi
# and is licensed under the GNU GPL license
# http://www.gnu.org/licenses/gpl.html

# Much thanks goes to Daniel Forchheimer (http://www.forchheimer.se/)
# for creating transset-df and the eutotrans
# script from which this script gets its inspiration

if ( !defined $ARGV[0] || !defined $ARGV[1] || !defined $ARGV[2] ) {
die "Usage: focustran-once <unfocused value> <focused value> <refresh value (secs)>\n";
}

# default values
$trans_val = $ARGV[0];
$opaque_val = $ARGV[1];
$sleep_val = $ARGV[2];

# grab all window IDs
@win_ids = `xwininfo -root -all`;
foreach my $win_id ( @win_ids ) {
unless ( $win_id =~ /has\ no\ name/ || $win_id !~ /0x/ || $win_id =~ /Desktop/ ) {
 $win_id =~ /\ \"/;
 $win_id = "$`";
 $win_id =~ s/\s//g;
 push @id_lines, $win_id;
 }
}

#print "ID_LINES:@id_lines\n";

# find my active window
my $active_id = `xprop -root  | grep "_NET_ACTIVE_WINDOW(WINDOW): window id # "`;
$active_id =~ /\#\s/;
$active_id = "$'";
chomp $active_id;

# set the active window to non-transparent value
system ( "transset-df --id $active_id $opaque_val" );

# make all other windows transparent
foreach my $win_id ( @id_lines ) {
# set active window to opaque_val and old window to trans_val
if ( $win_id ne $active_id ) {
system ( "transset-df --id $win_id $trans_val" );
 }
}

Then just add these lines to ./config/openbox/rc.xml (around line 302 for me):

    <context name="Client">
      <mousebind button="Left" action="Press">
        <action name="execute">
        <execute>perl ~/bin/focustrans-once.pl .6 1 0</execute>
        </action>
      </mousebind>
    </context>

Trouble is, I use mouse-raises windows, not click-to-focus! Anyone got some help for me, be it X11 active window or openbox mouse-motion-detect?