<fills count="2">
  <fill>
    <patternFill patternType="none"/>
  </fill>
  <fill>
    <patternFill patternType="gray125"/>
  </fill>
</fills>

If you create three custom fills, and do not include the “none” and “gray125″ as the first two fills, Excel 2007 will ignore the first two fills and treat them as “none” and “gray125″.

It only took me about two and a half hours and playing with why my two custom styles were either coming out with a no background or a gray125 background before I added a third style and that one worked. I should have been clued in when documents saved by Excel 2007 always inserted the “none” and “gray125″ fills.

I don’t know if it’s bad reporting, or miscommunication from those involved, or weird thinking by those involved, but this story leaves be a bit confused.

The stated problems are 1) third graders are not at the proper reading level, and 2) students are not scoring well on tests (with the implication being that kindergarten students are the ones scoring poorly).

The solution: an iPad for each of the school’s 250 kindergarten students. Clearly this will not help those third graders, but maybe those third graders have been deemed a lost cause? I’m not suggesting the kindgarten students shouldn’t received potentially better schooling methods. Whether the iPad will improve things, however, the school won’t know until they get test results.

One teacher says the iPads “give her 19 students more immediate feedback and individual attention than she ever could.” I’m guessing this is because if they do assignments on the iPad, the iPad can let them know what answers they got correct and incorrect. I do believe there is a place for software in the classroom for this reason, if the software is done right. I’d also like to see free, open source software, but I’m expecting they’ll be using expensive learning software.

I wonder, do the iPads follow the kindergarteners to first grade?

The problem with the latter is that it renders files such as robots.txt as a wiki page at http://example.com/Robots.txt. This can be avoided by using MediaWiki’s extensions capability.

To get around this issue, I wrote an extension.

Note: This license is released under the Creative Commons Attribution License.

<?php

if( !defined( 'MEDIAWIKI' ) ){
    die( "This is not a valid entry point.\n" );
}

#################################
## ADD FILES TO PASS THRU HERE ##
#################################

// A list of files MediaWiki should load from file.  These must be in the same
// folder as MediaWiki executes from.
$passThroughFiles = array(
    'robots.txt',
    'favicon.ico'
);

#################################
## DO NOT EDIT ANYTHING BELOW  ##
#################################

$wgExtensionCredits['parserhook'][] = array(
    'path' => __FILE__,
    'name' => "Pass File Through",
    'description' => "Passes specified files through, rather than trying to render an article.",
    'version' => 1.1,
    'author' => "Christopher Fritz",
    'url' => "http://kurifuri.com/mediawiki-passthroughfile-extention",
);

$wgHooks['BeforeInitialize'][] = 'passFileThrough';

function passFileThrough( &$title, &$article, &$output, &$user, $request, $mediaWiki ) {

    global $passThroughFiles;
    define( 'NonPassReturnValue', false ); // should this be true or false?

    $file = $title->mUserCaseDBKey;

    // Skip this page if we didn't specify to load a file for it.
    if ( !in_array($file, $passThroughFiles) ) {
        return NonPassReturnValue;
    }

    // Don't try loading this file if it doesn't exist.
    if ( !file_exists($file) ) {
        return NonPassReturnValue;
    }

    include($file);
    exit;
}

?>

Copy this text and save it in the extensions folder as pass_file_through.php.

Add the following to the bottom of LocalSettings.php:

require_once( “$IP/extensions/pass_file_through.php” );

By default, the extension only supports robots.txt and favicon.ico, but others can be added.

My PHP is rusty, and I’m not read up on MediaWiki’s programming guidelines for whitespace and layout. If you see anywhere that this extension can improve, feel free to leave a comment and let me know!

The e-mail was still on in my ~/Mail folder, but not in KMail. I was able to find it again by going into “Settings” > “Configure KMail”, and to the “Accounts” section.

Here, under “Receiving”, I had “Gmail” with all three “Retrieval Options” checked, but “Local Folders” had only “Include in Manual Mail Check” and “KMail Folders” had nothing checked. I checked “Include in Manual Mail Check” on the latter, and “Check Mail on Startup” on both, then closed and re-opened KMail. All my missing e-mails are back!

I’m sure I could have just clicked on “Check Mail” rather than restarting KMail, and I might want to uncheck some of those options I checked (for better performance), but for now I’m glad to just have my missing e-mails accessible again.

I decided today to try the whole tagging thing again. I added “Strawberry Marshmallow” as a tag to some images. After closing the tag window, the files were not tagged. Opening the tag window again shows the following:

Closing Dolphin and reopening it removes this “resource” tag.

I wish I could go in and investigate the problem and fix it. I’ve tried at least five times in the past six years to get a KDE build environment together, and I’ve failed every time. Came close once on my PC while trying it simultaneously on my PC and laptop (doing the same steps on both, and failing miserably on the laptop).

As the situation is right now, I can’t even get enough information on the problem to write up a usable bug report… I can’t just leave a bug report that says, “When I tag a file in KDE SC 4.4.5., it doesn’t work.”

The information I have gathered is:

• KDE SC 4.4.5 on Kubunutu
• Tags added in Dolphin become “resource” tags as seen above.
• Tags added with Dolphin do not remain associate with image.
• Closing and opening Dolphin removes the tags from the list of known tags.
• Tags added in Gwenview become “blank” tags.
• Tags added with Gwenview do remain associate with image.
• Closing and opening Gwenview removes the tag from the image and from the list of known tags.

Once I had things up and running, I found I can only transfer data wirelesses so long before it loses its wireless connection, and seems to have some data transfer errors. If I’m not transferring a lot of data, the wireless connection seems to be stable.

And it gets hot.

“GuruPlug Server Plus : don’t waste your money on it!” is a good overview of why to avoid this product.

“GuruPlug Server Plus Heat Problems” includes photos of various “solutions” of taking the device apart and bulking it up with a fan or heat sinks.

I’ve also read about problems with the SheevaPlug (GuruPlug’s predecessor) and its power supply.

Plug computers sound like a great idea, but the GuruPlug (and its 30 day warranty) are one to avoid.

It’s just as well. Some other images I hadn’t tagged yet, I mass-renamed them from the command line, and that would have wiped out the tags if they had any.

The reason for going the electronic book reader route is to decrease the amount of Stuff I’m accumulating. I’ve even started looking into getting rid of part of my anime DVD collection (listed on my for sale page, and available to anyone in my area who can buy in person).

Going digital is something I’ve only started recently with music. I’ve bought a handful of music online, but until recently had been buying CDs. It started with some Japanese music, where a one dollar song was a lot cheaper than importing a $25 CD containing two songs and their two karaoke versions. With CDs cheaply available in the US, it’s harder to justify the cost for digital download versus what the CD offers. Digital means instant MP3s. CD means a few days way, but the ability to copy music as FLAC (my preferred format), and a contribution to the Stuff filling my bedroom (which I’m trying to cut back on). If the CD is cheaper, then buying digital means I’m essentially paying for getting music without the Stuff.

Back to books, I’ve wanted an e-reader since around 2003 or 2004 for this same reason. There are many older books (out of copyright) I’d like to read, but reading them on the computer simply isn’t comfortable for me. Computers are fine for short reading sessions, or skimming, but to really get into a story (or non-fiction reading material), the computer screen just doesn’t work out for me. The “e-ink” technology solves this problem, and as a bonus I can store a library of books in only the space taken by the e-reader.

Not all books I would like to read are out of copyright. Unfortunately, most of those that are copyrighted are not available electronically. I’m hoping that this turns around, with more new publications being available electronically. And like with music in its digital infancy, those books that are being published electronically happen to be DRM-encumbered (which I do not expect to ever go away, so here’s hoping I’m wrong). Before making a purchase, I had to make certain (like with the songs I bought from iTunes some years back) that I would be able to strip off that DRM, and read it on my reader (which does not support DRM).

Today I decided to buy my first e-book. I went with Barnes & Noble because I wanted to go with the Epub format, and I found tools for circumventing it. Signing up for an account requires creating a password requiring one to “Use 6-15 letters, numbers, or numeric symbols. CaSe SeNsItIvE. No spaces.” which means I can’t use my normal type of password (often over 30 characters long), which isn’t any fun, especially when my designed password could only be compressed from 34 characters to 16. Luckily it turned out that a period isn’t allowed, so that took it down to 15.

This particular book I planned on buying is listed at about $25.00 (or $20 bought online) for hardcover, $15.00 (or $10.00 bought online) for paperback, and $10.00 for e-book. I am one of those people who feels the e-book version should be cheaper than paperback, and the online price of the paperback makes it feel as if the e-book and paperback are the same price. I am not one of those people who feels every creation that can be copied at no cost should be free. I also feel that, based on reviews of this book on Amazon.com, that $10 is a justifiable price. I also expect that I will finish this book if I own an electronic copy, whereas paperback books I buy I sometimes finish and sometimes do not finish (due to various inconveniences of paperback books for me).

Other than the password issues, creating an account and buying the book was easily enough. The “thank you for purchasing” page has link, “Go to Your eBooks Library”, at the bottom of the page, which would take me to a place to download my purchase. This area included some free books (Dracula; Pride and Prejudice; Merriam-Webster’s Dictionary; Little Women), with my purchase down at the bottom of the list. At first I thought I was looking at a listing of books available to buy, wondering where my book was, but I figured it out.

Left-clicking on a “Download” link left Opera (my web browser of choice) thinking (with the mouse pointer showing the “waiting” icon), and after a minute of this, I resorted to middle-clicking the link instead, to open it in a new tab. The new tab prompted me with a save/open dialog. I don’t know why simply left-clicking wasn’t working out, but I was able to download the book (and the four freebies, just in case I need something to read and don’t have anything else new with me).

Ten dollars down, book downloaded, DRM removed, and onto the reader it goes, and the book is accessible. (The free books were not DRMed, which makes sense as they’re likely all out of copyright.) It feels weird not owning a physical copy of the book, but the digital copy isn’t any worse (I wouldn’t be lending or selling it), so I’m certain I’ll adjust easily.

Amarok 2 seems to have been a more dividing release than KDE 4 was after the stability and expanse of KDE 3. While the KDE Software Compilation 4 has grown into a fairly stable system (with Plasma a strong part of the workspace, and tagging/search improving quickly), Amarok has staggered around a bit.

With a recent release of Amarok, perhaps Amarok 2.2, I noticed two things missing. First was that I could no longer modify the design of the playlist. The icon for it was no longer below the play list. Second, I could no longer set whether to play tracks randomly, or otherwise.

After playing around for a long time, I finally found that a weird icon at the bottom of the playlist, with the tooltip “Track Progression”, had nothing to do with the progression of a single track, but actually was the missing method of setting the playlist to play random tracks or albums, and to repeat a track, album, or playlist. Apparently a good amount of discussion and consulting with usability experts determined that modifying how the playlist progresses from one song to the next should be in the play list area. This makes sense. But why remove it from the menu? I haven’t yet figured out how to access this icon using they keyboard, nor do I see anything for it in the shortcuts setting.

Looking over the Amarok Userbase article (which is overly fanboyish at the time of this writing), I found where the icon for editing the playlist’s appearance went. It’s a menu now. I can only determine that a usability expert pointed out that because this relates to the playlist, it should be removed from the playlist, and placed under a menu item. I’m not complaining, though. I have easy keyboard access to it now.

I’m still looking for options such as “Overwrite a saved playlist file with the current playlist,” and perhaps, “Remember that I loaded this set of tracks from a saved playlist, so when I add and remove tracks, and save the playlist, you can either overwrite the one I loaded it from, or ask me whether I want to save or save as.” Might just require a feature request if they’re not there already.

I’m going to keep an eye on Clementine, and hopefully Clementine and Amarok can even share some common libraries, but maybe Amarok 2 isn’t so bad anymore. Almost. Now if only I knew why setting it to “random tracks” stops playing after every track on the playlist has played once…