Ian Lewis' Blog

I recently saw an interesting article on the "Loudness War". Basically it's a sort of loudness arms race between competing music labels and radio stations to increase the loudness of recordings so they stand out. Wikipedia has an extensive article on the subject.

The loudness of the entire recording is increased, but the maximum volume of a particular part of a CD/Vinyl recording is fixed so what results is that any sound dynamics are lost. The louds are loud, and the soft parts are loud.

When I saw this video, I thought this is exactly what I felt was missing from the more recent Rush recordings that the previous ones had, sound wise. It was just a feeling as I was listening to the music, so I assumed it was basically overprocessing with expensive equipment. It's nice to see it codified in the nice, simple to understand example above.

I just read this essay by Richard Stallman. While I think it's kind of old and while Richard sometimes seems like a conspiracy theorist his argument is compelling. Especially when reading documents and plans put forth by the Trusted Computing Group (the term trusted is a misnomer. The FSF uses the term treacherous computing but the phrase "Trusted Computing" itself is ambiguous because the subject is left out. The person doing the trusting is not you, the user, it's the computer manufacturer).

In any case, I would hate to live it that sort of world. If that sort of thing, god forbid, becomes pervasive and programs like Linux, and free software will not run legally (circumventing a Trusted Computing chip would be illegal in the U.S. under the Digital Millennium Copyright Act) I might have a change in career. Maybe environmentalism?

Recently I added a small applet which displays news stories that I am sharing from Google Reader. With Google Reader you can share news stories and blog entries that you think are important publicly. I thought it was pretty cool so I added it to the page here.

The news stories that I tend to share will probably concern three topics, Environmentalism, Technology, and Japan related news. If you have a chance, check it out.

Somewhat recently, Google released their online version of Microsoft Word and Excel. Essentially, it is a document editor that can import and export various formats and allow editing of documents on the web. I originally thought that the AJAX approach of Google Docs could never touch the feature set, speed, and usability of a native application. After all, documents can be big, and native applications like Word, and Open Office have huge feature sets that couldn't possibly be replicated easily and efficiently on the web using AJAX.

So I thought. But in reality Google has replicated many of the important features in Word and Excel already with surprisingly good results. But even still that's missing the point. The point is being able to manage all of the documents for you and/or your organization collaboratively online. And then once finished being able to publish the results to the web. The power of such an idea only hit me fully today. Once you publish it on the web, your users and customers can view it. If you were able to integrate this functionality into your organization, namely writing documents online collaboratively and publishing the results to the web, think of the things you could do with such functionality. You could release financial data and reports this way. You could release and manage revisions of software documentation this way. Many things are possible with this approach.

Managing revisions of anything is a hard task. Maintaining what versions of the software are installed where and what versions are published is an annoying and hard task to manage. This sort of approach would make it easy. Publishing documents on the internet this way could allow customers and/or users to view new documentation or new versions of documentation from your organization's website instantly. I feel that, more and more, integrating standards based applications to create working systems that help people manage their businesses (including publishing) and access data from anywhere via the internet is where the future will be heading.

I'm not sure how much of all the great new things Google is putting out were the result of Google's own inventiveness or whether Google simply read the writing on the wall (which is not such an easy task itself) but it's exciting to see the new companies direction first hand as many of the applications are available for free for essentially non-commercial use.

Google of course wishes to remain an innovator but history has shown that companies, just like people grow up and gain responsibilities (to employees and customers) like people gain responsibilities to their families and that the innovation really rests in the hands of those who are young and free to explore their whims. I just hope that Google can maintain it's innovation well into the future. It's quite amazing to watch.

I was surfing the web earlier and found a link to elephantsdream.org. It's a project to create a movie short created with only open software like Blender. The production files are also available under the Creative Commons Attribution License. The content itself was created by the Orange Open Movie Project studio in Amsterdam and financed equally by the Blender Foundation and Montevideo in the Netherlands.

I checked out the movie and it's actually really good. Though it's only a short about 10 minutes long or so, it has a pretty interesting premise and obviously had high quality production, art, writing and direction. I would urge anyone who hasn't seen it to check it out sometime. They also have a documentary about the making of the short which I haven't watched yet but no doubt that it would be interesting. They also have a DVD which contains everything, the movie, documentary, subtitles in 34 languages, commentary, production files, and original script and storyboards.

It's a good example of the quality of content that can be created with open tools and techniques. I'm awaiting the time when a full length film could be developed this way.

Just recently Google Calendar has allowed calendar users to display their calendars publicly. This has been something I've been looking forward to since now people I know can see when I'm busy. You could do this previously with Yahoo Calendar and was lacking in Google's version.

Though Yahoo allowed users to publicly publish calendars, Google went a step farther and allowed users to imbed calendars in webpages, so I went ahead and imbeded it here. It uses a >iframe< tag to imbed the calendar in the webpage. Which I'm having trouble getting to render well in Internet Explorer. In fact I'm having trouble trying to get Internet explorer to not lock up when browsing the imbeded calendar. Strangely enough, if you press the forward arrow on the calendar to move forward a month, it seems to do ok. But, if you hit the back arrow, it locks up Internet Explorer. Internet Explorer is also not rendering the width of the calendar correctly. It seems that it's making it too wide within the box that it's contained in. Firefox of course renders it perfectly.

Anyway, the imbeded calendar is pretty cool. If you are logged into google calendar it even shows it with the correct permissions.

I use b2evolution for my blog on my home page. Recently someone from work tried to leave me a comment here and got the message

MySQL error!

Unknown column 'comment_author_ID' in 'field list'(Errno=1054)

Your query:
INSERT INTO evo_comments( comment_post_ID, comment_type, comment_author_ID, comment_author, comment_author_email, comment_author_url, comment_author_IP, comment_date, comment_content) VALUES( 156, 'comment', 1, NULL, NULL, NULL,'68.100.33.135','2006-05-22 20:36:39', 'test comment' )

While this seems simple enough to fix, I wonder why b2evolution wouldn't recognize that this field is missing and create it when upgrading. Perhaps I was upgrading from too old a version and it didn't think it should have to update this field. I dunno. I suppose this has been a problem since I upgraded b2evolution sometimes late last year. In any case, all I had to do was use PHPMyAdmin and do the query

ALTER TABLE evo_comments add comment_author_ID int(11)

and now it works. So visitors should be able to add comments again.

Today I'm messing with Google Calendar. I've been using Yahoo's Calendar which is nice and has features for everything I use and ties nicely into Yahoo Groups. It has the capability of providing 2 reminders, which I like. These reminders can come to you as e-mail or go to your cell phone. You can provide a location, along with other data such as a phone number and location name. You can schedule repeating events using odd schedules like "The second Friday of the Month" (which works really well for our Japanese Meetup Group because it ensures that the repeating event happens on the same day of the week every month), or 'Every other day' or 'every third day'. You can enter multi-day events, and it handles them ok, creating a new box for each day. So a three day event looks like 3 identical one day events.

Google Calendar doesn't allow you to put in as much data as Yahoo groups. You can put things like directions, information, or a phone number in a description but it doesn't show unless you look at the event details. It allows you to put in the location of the event which is normal. But I've found that if I put in the location name in with the address of the location, sometimes it doesn't open up the location correctly in google maps. It also doesn't support 2 reminders, and the reminders seem to come up as a popup by default when I have the calendar open rather than as an e-mail. It's also not integrated with Google groups, which is in-and-of-it-self designed as just a mailing list, and doesn't hold a candle to Yahoo groups which integrates a calendar and photo album along with the mailing list for your group. Sometimes it's also a bit unresponsive and gives you no indication that it's doing anything because it's all AJAX.

That said, when comparing the two Google Calendar looks much better. There are no ads. It handles multi-day events well, displaying them as only one box, showing the title only once. It also allows you to repeat a meeting on Mon/Wed/Fri, or Tues/Thurs which Yahoo doesn't do. This is good for me as I have Tues/Thurs events. It also would be easy for students I imagine. Though it does have a weekly option with checkboxes to select the days of the week to repeat on. It also allows you to change the time and date of the event easily using drag-and-drop and normal gui (however, I'd add that events for me don't change that much and it's no problem to go to the event to change it. Drag-and-drop is more likely to be done accidentally than on purpose for me). Another thing that it does that Yahoo doesn't do is display Japanese/UTF-8. That's a big plus for me.

Overall, there are things I like about both but Google wins for it's good looks and ability to do everything I want. But even though I like Google Calendar, it seems like a work in progress and I'm looking forward to new versions of the calendar.

Since gtk 2.0 has been out for like 3 years I figured it was about time for me to learn some of it. So I've been playing around with it, creating some demo programs. Now I'm looking for a project to work on and thought that I might try to make a gtk implementation of the keepass password safe. I wanted to do a strait gtk-2.0 implementation of this in C but it looks like some of the code that is used for the cryptography is done in C++ which makes it a bit difficult to do something like that. So I thought about using gtkmm which is a C++ wrapper for gtk-2.0. I figured that would suit my needs so I think I might try that. Though, I still want to brush up on my C skills. Maybe I'll port gorbital into C and gtk-2.0 as an excercise.


Of course Reiko is coming next week and I still have some stuff to take care of so I don't think I'll be able to do very much with this until after March.

Yesterday I released jsXe 0.4pre3

Hopefully I'll be releasing new versions more regularly in the future. It's hard to maintain focus on Japanese study and jsXe and everything else all at once.

I really need to redo the website jsXe website soon. It's pretty crappy. The guy that designed it did an Ok job but he didn't work on it for very long. Maybe I can come up with some design requirements and have someone work on it. I'm could do it but I'm not a web guru and the website could use a bit of php scripting probably.

Anyway, I have study group with friends tonight so I gotta study a bit.

Been busy this week but I managed to add cut/copy/paste support to the tree view in jsXe cvs. Though I pulled my hair out a bit and there are some limitations, it works pretty well.

I just wish there was more interest in the project. It would help me justify the time that I put into it. I think jsXe could be a cool project and I have some good ideas for it, like a schema editing view, I feel like I don't have enough time to devote to it to make it a really cool, useful program. I guess the folks that don't need a professional XML IDE just use a text editor but jsXe can't compete with professional tools like oXygen and XMLSpy with their XSLT debuggers and professional tools. :-{

I guess I've thought about making jsXe a platform for developing plugins that allow you to do develop easily with multiple XML applications at once , and I've implemented some features in that direction already, but again that's rather ambitious and I don't think I have the time. And besides then jsXe wouldn't be the Java Simple XML Editor anymore. It would be a huge, featureful development environment.

Well, I didn't get much of what I wanted to do finished over the weekend. I was looking at what the best way to install X.org and noticed that they have two current versions. This puzzled me a bit but apparently they created 6.9 with their regular build system and 7.0 with the standard autoconf build system. So I guess that since they say that 6.9 and 7.0 are the same 6.9 would be adequate to get the GATOS features that I want (though the gatos website makes no mention of 6.9). So after reading about it for a couple hours I concluded that either I should just go ahead and go through the process of building X.org 7.0 using the normal autoconf builds or I could try to install the slackware-current package for X.org 6.9. The slackware approach would be easier and I haven't checked but I'm pretty sure it will be built with a version of glibc that exceeds mine thus making it not runnable on my system (I'm NOT planning to upgrade glibc though I'll research it). The building approach may work but will surely be more painful and require much more time to install and configure properly (proably why most people just go the route of using packaging systems like rpm).


Once I've got X.org figured out installing the new kernel should be cake by comparison.

Hey all, this weekend I'll be spending quite a bit of time on the computer. This weekend I'll be adding a few new minor features to jsXe. I haven't worked on it seriously in months. Almost as long as I've been talking to Reiko. I wonder if that had something to do with it. Anyway, I've added some DTD/Schema validation features and I'd like to expand those a bit. I'd like to add in some built in completion features for some popular XML applications too (Schema, XSLT Stylesheets, Apache Ant etc.). I also have to start seriously thinking about working on other plugins/views for jsXe. Namely the Schema View. Though sometimes I feel like it's all for naught since noone really uses jsXe but, even still, I do enjoy working on it.

I'm also trying to get back into C programming a bit. I've been mostly looking at programming with GTK+ 2.0. I've been creating some demo programs and as a learning exercise I may upgrade gorbital to GTK2. However, learning this stuff is lower on my priority list and I may not get to it this weekend.

I also would like to do a bit of upgrading on the machine as well. For a while I've been using the 2.4 series of the Linux kernel and XFree86, the second of which has been replaced by X.org in the community. Part of the reason I've been running legacy software is because I use GATOS which allows the use of some of the TV input/capture functions of my ATI All in Wonder Radeon. But I noticed recently on the GATOS site that newer versions of the Linux kernel in the 2.6 series and the new release of X.org 7.0 handle the All in Wonder Radeon's hardware acceleration and TV input correctly out of the box. I also recently had some problems with using virtual terminals on my current version of the Linux kernel and XFree86 so I think I'll be working to upgrade to both the Linux kernel 2.6 series and X.org this weekend. Needless to say (or maybe not), I do everything the hard way on my system so this may or may not be a trivial feat.

Anyway, If you want to send me some words of encouragement then I'll probably be online all weekend.

Been playing with Last.fm recently. It's a website that records your "musical profile" (mine is here) Basically you can install a plugin that will send the website information about the track you're playing. So it records how often you listen to a particular artist or song.

This in and of it self is just a cute little toy but what's pretty interesting is that they are taking that info and, Amazon.com style, they give you recommendations on what new music you could try based on what other people listen to.

Even more interesting than that is they also have streamable audio now. So basically you can either enter a number of artists, or just use the radio based on your profile, and it creates a customized "radio station" that plays music similar to the artists you like. They even stream a number of Japanese artists that I wouldn't have guessed they would be able to stream.

Anyway, I thought the website was a little cute but useless at first but they seem to actually be providing interesting stuff now.

After going back and looking at the debug log for my Windows build of gaim at work I noticed that google does send a set of mechanisms that is the same but the response that the Windows build sends is different. The response causes google to re-send a set of mechanisms which has the PLAIN mechanism added.


Sending:
<?xml version='1.0' ?>

<stream:stream to='gmail.com' xmlns='jabber:client' xmlns:stream='http://etherx.jabber.org/streams' version='1.0'>
Recv:

<?xml version="1.0" encoding="UTF-8"?><stream:stream from="gmail.com" id="EC3C8EBE" version="1.0" xmlns:stream="http://etherx.jabber.org/streams" xmlns="jabber:client">
<stream:features>

  <starttls xmlns="urn:ietf:params:xml:ns:xmpp-tls"/>
  <mechanisms xmlns="urn:ietf:params:xml:ns:xmpp-sasl">

    <mechanism>X-GOOGLE-TOKEN</mechanism>
  </mechanisms>

</stream:features>
Sending:

<starttls xmlns='urn:ietf:params:xml:ns:xmpp-tls'/>
Recv:

<proceed xmlns="urn:ietf:params:xml:ns:xmpp-tls"/>
Sending (ssl):

<stream:stream to='gmail.com' xmlns='jabber:client' xmlns:stream='http://etherx.jabber.org/streams' version='1.0'>
Recv (ssl):

<?xml version="1.0" encoding="UTF-8"?>
<stream:stream from="gmail.com" id="EC3C8EC4" version="1.0" xmlns:stream="http://etherx.jabber.org/streams" xmlns="jabber:client">

<stream:features>
  <mechanisms xmlns="urn:ietf:params:xml:ns:xmpp-sasl">

    <mechanism>PLAIN</mechanism>
    <mechanism>X-GOOGLE-TOKEN</mechanism>

  </mechanisms>
</stream:features>

Sending (ssl):
<auth xmlns='urn:ietf:params:xml:ns:xmpp-sasl' mechanism='PLAIN'>

***auth string edited out***</auth>
Recv (ssl):

<success xmlns="urn:ietf:params:xml:ns:xmpp-sasl"/>


So it appears that something is wrong with the tls and ssl in my build at home since my build at home doesn't send the starttls tag back to the server alhough I have both ssl support and mozilla nss support compiled in. In fact the about window, which shows what support was compiled in, for the windows build shows almost identical information as my build at home. So I'm still a bit clueless as to why it's being wierd.

I tried to get Google Talk to work with my build of gaim 1.5.0 yesterday but it didn't really play nice. I get a cool error message, "Server does not use any supported authentication method"

I was able to connect at work so I thought that not being able to connect from home was a bit strange. After ensureing that I compiled in support for SSL and correctly linked the mozilla NSS libraries I took at look at the debug dialog in gaim. Since Google Talk is based on Jabber I could look at the protocol and I noticed that at home Google Talk's server sends a stream tag that only contains one authentication mechanism called X-GOOGLE-TOKEN but at work it had two, X-GOOGLE-TOKEN, the method the standard Google Talk client uses, and PLAIN, which is what gaim expects.


<?xml version="1.0" encoding="UTF-8"?>
<stream:stream from="gmail.com" id="EC4FFD5B" version="1.0" xmlns:stream="http://etherx.jabber.org/streams" xmlns="jabber:client">

<stream:features>
  <starttls xmlns="urn:ietf:params:xml:ns:xmpp-tls"/>

  <mechanisms xmlns="urn:ietf:params:xml:ns:xmpp-sasl">
    <mechanism>X-GOOGLE-TOKEN</mechanism>

  </mechanisms>
</stream:features>



instead of


<?xml version="1.0" encoding="UTF-8"?>
<stream:stream from="gmail.com" id="ED36A600" version="1.0" xmlns:stream="http://etherx.jabber.org/streams" xmlns="jabber:client">

<stream:features>
  <mechanisms xmlns="urn:ietf:params:xml:ns:xmpp-sasl">

    <mechanism>PLAIN</mechanism>
    <mechanism>X-GOOGLE-TOKEN</mechanism>

  </mechanisms>
</stream:features>



I currently have no idea why it's doing that but hopefully I'll figure it out soon. It's not a high priority or anything I just have 1 friend so far that's using it. :p

I was recently thinking about open source projects to contribute to and bouncing around in my head the idea an open source integrated library application. Something like a card catalog, database, check-in/check-out system. So I started pinging sourceforge to see if anything like that existed and was surprised that I found it nearly void of projects like that.

So I went to the wider world and found the Open Source Systems for Libraries which pointed me to the Evergreen and Koha projects. They both seemed pretty interesting. The Evergreen project seems to be just getting off the ground with an alpha release and is being maintained by the Georgia Public Library Service to be used by member libraries of the Georgia Library PINES Program.

Koha seems to be a bit farther along and has a number of users. It seems like a pretty cool project and I think I'll do a bit more research.

I like the idea of these open database applications but both Evergreen and Koha seem to base their web parts on mod_perl which I probably wouldn't do. I guess I'm biased against perl but I think I'd use something easier to use like Tomcat or even mod_python. Maybe they chose mod_perl because it's faster

I feel like I have a need to work on an open source database application that actually has users and a bit of structure. I want to actually help people acomplish goals. Interacting with people locally would be even better (Koha has some users in PA). Anyway, they seem like cool projects and could be a potential project to get involoved with.

Once again Bill Thompson echoes my own feelings. Given how many folks have been talking about how even open source programs have bugs, he offers his insight into why open source bugs will always be better than closed source bugs.

Basically it's just a better system where money isn't a factor. Only quality.

I've been working a lot on jsXe recently. Mostly on the schema view, bug fixes, better namespace support etc. The schema view is going to look pretty cool when it's farther along. I've got an early screenshot of the schema view up on the jsXe website. Hopefully I'll have a releaseable version in the next few weeks.

There is literally too much work to do on jsXe in general. Just the normal tree view, source view, and schema view will keep me busy for a good time to come. And I've got ideas and more views planned. Plus, writing an XML editor in general is hard. Lots of quirky XML gotchas. I'd like at least one other decent developer to work with but no one is very interested in the project. I suppose I've got some ways to go before I'll have folks willing to use jsXe and test it out, submit bug reports, and submit code.

So the European Commission recently passed the "Computer Implemented Inventions Directive" which basically means that software patents in the EU are back on their feet. According to MPlayer's website, in order to stop this initiative " The European parliament will now be taking the last stand against software patents in a voting for which an absolute majority is needed. Such a majority is hard to come by in a parliament with a low attendance level."

As many bloggers and folks quoted in news articles have said, this looks like a breakdown of democracy. Zdnet's article states "The Foundation for a Free Information Infrastructure (FFII) accused the EU Council of ignoring the views of both the national parliaments that spoke out against the directive and the European Parliament, which demanded that the directive be restarted." So both parliments of elected people were circumvented with this directive.

I'm not going to go much into why software patents are bad but Free Software, Free Society by Richard Stallman and Free Culture by Laurence Lessig are recommended reading to get started.

Five posts! I'm on a roll.

I finally broke down and got a new monitor the other day. It's a 17 in. Samsung SyncMaster 710N. The one in the picture is grey but mine is black. It's pretty cool. I'm running it's native resolution, 1280x1024 and it looks pretty good. I did have to play with the sharpness and adjust the settings a bit to make it look better but overall I'm happy with it.

Whoa, four posts in a day. New world record!

Software patents in Europe hit a big snag. The law that would allow software to be patented in Europe has a long way to go now to become law. Thank goodness that some European lawmakers saw the harm in software patents.

BTW, I made some chicken the other night. It was good.

This last week I had been playing around with trying to get a wireless Netgear WG311 to work in Linux. The TI chip that powers the card isn't supported by TI i.e. TI won't give technical specs to the developers. This has resulted in a driver that works but is slow and tends to loose the connection until I reboot (not sure why I have to do a full reboot, normally unloading and reloading a kernel module will do it right?). Possibly caused by the router doing strict 108.11b compliance checking as hinted to in the README.

So I decided to return the card and get a wireless ethernet bridge instead. This has turned out much better (though quite a bit more expensive) as it's faster and doesn't require any software drivers on my system. Just the good ole regular ethernet driver I've been using.

Course that doesn't help when the cable goes out . Possibly due to windy conditions during the last day or so. So the cable folks are coming out on Saturday to check it out. Given that the cable has gone out and the heat where I'm staying is on the fritz (BIG PROBLEM) one might wonder whether anything in the house works. *Sigh*

I've gotten pretty annoyed with music players in linux recently. My gripe? They don't display japanese characters properly and do what I want at the same time.

XMMS is great for every one of my needs except displaying japanese file names and metadata from a music file. First it's written in GTK 1.x which doesn't support UTF-8. I could use a different encoding but I would rather use UTF-8, and that wouldn't help a whole lot anyway because XMMS is a skinned player. The skins themselves contain the characters that can be displayed on the main window. It's not a font. So the way the skins work it's impossible to display japanese characters properly. All you get is ########.

BMP, Beep media player. A GTK2 port of XMMS but it's still skinned. And it's buggy it seems. I didn't even bother to install it.

Rhythmbox. When I saw this I thought, great! a player written in GTK2. So it will support UTF-8 and Japanese characters by just being a GTK2 program. The problem here comes with how the design direction that Rhythmbox has taken. It's going for the big bloated media program that will store all the data about your media collection in a database sorta thing. This might be ok if the program wasn't so buggy, and bloated and slow.

• First it stores the music metadata in an XML file which gets pretty big. It takes a full 10 sec to bring up Rhythmbox on my 1.3Ghz Athlon with my modest 7Gb or so music collection.
• Second you can't seem to create playlists on the fly very easily. In XMMS or Winamp you open it up and use the playlist editor. Add a song and there you go. With Rhythmbox I have to go to File->Playlist->New Playlist or something and create a new playlist. Then I can modify it and whatnot while I'm listening but I have to delete it later when I don't want it anymore. It's a little cumbersome to say the least.
• Third, it's built with the GStreamer library in mind. GStreamer is a nice idea but to me it's just another cumbersome bloated library I have to install that nothing else but Rhythmbox is going to use.
• Fourth, what's up with the shuffle and repeat checkboxes?!?! They seem to flicker and do wierd stuff for me. Sometimes they make the program lag.
• Fifth, Did I mention it was slow?

So I've almost decided to write a port of XMMS using GTK2 for the interface and getting rid of the skinning. That or I'll extend the command line ogg123 player to support mp3 and playlists. I'd really like a better alternative to these. They're a lot of work. I just want a simple music player that plays mp3 and ogg vorbis and displays japanese characters. Is that so much to ask?!?!

So building the new version of mozilla turned out to be much harder than I hoped. I'm trying to compile it with gtk2 support and it comes up with compile errors. Ok, so I play with some files and include directories and such and I get it to compile. Ok cool. BUT, it then crashes when I click on a radio button on a web page. Any radio button. So it's anything but a stable build of mozilla.

After a few days of tinkering I still haven't got a stable build. Also it takes forever to build so when I'm not successful it means another build which is about 45 min or so.

I've found that when compiling new programs for use on my machine I almost have to have a test environment where I can test the program before I put it into production (start using it)! Even though I would only have to do that with big and high use programs like mozilla, it's a little frustrating.

I sometimes think it's not worth it to install everything from source but then I remember how streamlined, and fast they are when I don't compile in features I don't need or use. And how much better they look, or how much useful they are with features you don't get from pre-built installs.

Oh well, maybe I'll fix it tonight. Maybe it'll be fixed by upgrading to gtk-2.4. Who knows.

A while ago I started looking for a stand-alone calendar application to replace gnomecal. I've been using gnomecal as long as I've been using Linux, as a simple stand-alone calendar where I can enter events I want to remember and such. However, it's showing it's age and I've grown out of it. It uses old gtk 1.x, and as such doesn't have the advantages in display and input of Japanese characters. The newest version is also downright buggy, having problems rendering the year view properly and remembering events.

I really didn't want to use evolution since it's a big bloated thing that integrates mail, calendar, contacts and such. It's great for a company or organization where you would have meetings and there is a need for shared calendars and shared contacts. Like Lotus Notes/Domino, or Microsoft Exchange/Outlook.

After a good deal of seaching I was surprised to find a mozilla calendar project. A stand alone calendar based off of mozilla. I can't believe I didn't see this before. Anyway, it got me to upgrade my previous build of mozilla, which was 1.5 I'm sad to say given the security problems recently (like the image buffer overrun vulnerability).

I shouldn't have to tell you that building mozilla from source is no small task, even though it's easy. Building it takes me over half an hour (I usually don't time it but it usually takes about 40 min or so I think). By comparison I can compile my Linux kernel in about 7 min. The source directory takes up over 1 Gigabyte when it's finished compiling. I'm really excited about this calendar since it looks good. Proof is in the puddin' though so I'll have to see how much I like it after I try it out.

I finally got Avview to work. The silly audio cable connecting my sound card to my all in wonder video card wasn't connected right or something because when I reseated it Avview started magically working. Bums me since I spent a while compiling and recompiling Avview trying to get it to work.

I just got into checking out Wikipedia again. It's a free (as in freedom and as in beer) online collaborative encyclodpedia. I contributed some articles on Design Patterns a while ago and was pleasantly surprised that the articles have remained largely intact for 2 years. It's nice to know that people think your contribution has enough merit to keep it there almost untouched. Additional content has been added to the articles of course but the text that I wrote remains.

I also started checking out Wiktionary also. The dictionary counterpart to Wikipedia. It's a little newer so there's much content to be added there.

I recently read the following articles and found them pretty interesting. They're rather simple concepts but I'm not a database guru but I know enough to be dangerous. I work for a company that sells database applications so that's how I got into it.

The parts that I found particularly interesting in "The Fundamentals of Mapping Objects to Relational Databases" included the part about what the author called "shadow information" and mapping metadata to datastructures, and the part about mapping object relationships.

Database refactoring seemed like a task that you wouldn't be able to generalize very much. Since a database is a single point of contact for all of an organization's data, used my potentially many applications from reporting tools to 3rd party interfaces, it would be a daunting task to refactor data in the database. The solutions for refactoring metadata would be so varied that generalizing methods for doing it seems like a waste of time.

The Object-Relational Impedance Mismatch
The Fundamentals of Mapping Objects to Relational Databases
The Process of Database Refactoring

I have a All in Wonder Radeon and I use Gatos to get the video capture working. But the silly AVview stopped working. Well it works but no sound. No sound for just playing from video input, let alone capture. I know to do capture you need to have the km module and alsa installed but alsa is optional if you're not doing capture (right?). I never really used capture (and for whatever other reasons) so I never went ahead and installed alsa. And with all the patching (lm_sensors among other things) I've done to the kernel it's more than a trivial amount of time to upgrade to 2.6.x where it's standard. I swear it worked with my current (Linux) kernel version (2.4.26) and setup before.

I recently played around with the CVS stuff to try to get video capture to work (including installing alsa) but didn't get all the way to a fully working build. So I guess somewhere along the line I screwed up my stable working build of AVview. Grr. Now I can't play Mega Man 3 on my Dreamcast using AVview with sound.

I just finished separating the gui component for the tree view in jsXe. I wish I had done more today. There's always tonight to make up for it.

I'll probobly start watching Heibane Renmei tonight tho. I'm not sure if I'll like it but we'll see. Right now I'm going to head to the store to buy some vegetables to make food because I've waited too long and I'm starving.

Spent much of the day getting finances up to date and cleaning but I had a cruise over to Lawrence Lessig's website. I recognized the name when his new book popped up as a recommendation for me on amazon.com. He's a lawyer that is active in copyright law and computers an is very influential in the free software community... so I added his books to the long wishlist (hint, hint) I have.

Being on the software development team for reporting at work makes me somewhat interested in reporting in general. Also given my interest in GNU software I naturally looked for open-source or free software for reporting.

In my search I came across JasperReports, which is an open-source framework for creating printable reports. It uses XML as a way to define reports and has some nice viewers for the reports. It seems like an interesting project and might apply well as a future plugin for jsXe.

G'day.

I'm sitting at home wasting another perfectly good day trying to figure out how to get GTK-2.0 and pango to render the east-asian unicode fonts that I have installed. I've tried a number of things but no GTK-2.0 programs seem to want to render non-western characters. As with everything I install, I installed GTK-2.0 and pango from source. Not sure how that would affect how pango and/or Xft and/or freetype find fonts.

I'm not sure that any of the few people that read this would really be able to help me or really even know what I'm talking about but I'll pose the question anyway. Has anyone had a similar problem? What am I doing wrong?

Yes. So I got a new (to me) laptop from work. It's old but hey, it's a laptop.

Now I'm having fun installing Linux and compiling programs. The laptop is a Toshiba Tecra 8100. Linux seems to work ok on it so I'm happy. It'll be a learning experience to get the more advanced power saving features to work.

On another note I found a pretty cool GNU project the other day called JSwat. It's a graphical java debugger. I was excited to get it working because now I can use it at work instead of (ugh) JBuilder. I don't know what it is but I just have problems using a full blown IDE to do work in. I prefer to use free tools that do one thing well. JEdit works well as an editor, JSwat works well as a debugger, Ant works well as a build tool. JBuilder just takes forever to load (and my alternatives are java programs. You know it's bad), insists on compiling every file in the source tree unless I exclude it (as opposed to just compiling only the files necessary to run a particular class), and detracts from editing by having wierd indent rules and an inflexable framed window. Garbage.

And still another note, Datatel's Christmas party is this weekend. It's a pretty formal shindig at the Ritz in Tyson's corner. This is the 3rd time I've gone to this thing and the 2nd time I'm asking myself why. It's not particularly fun, I don't really like to dance, and I don't really have anyone I want to take. The only redeeming feature is that I have friends that are going but they'll probobly all have dates. So why am I going? I don't know. I think, unless someone else gives me a pretty compelling reason to go, I'm not going to go next year. What do you think? It would be nice to know someone reads this thing besides me.

Finally, Hasweb decided to give me my account info for my hosting with them. So my website is up! Today was pretty boring otherwise. I've been mostly working on getting this blog functioning which hasn't been a small task. Hopefully I'll get this thing updated and such in the next week or so. I have quite a bit of customizing to do.

Went out to lunch with Kerri today. She's an old friend I met freshman year in college. We went to the new Sakura Japanese Steakhouse in fairfax near work. That proved to be somewhat of a mistake since Japanese steakhouses are JAPANESE STEAKHOUSES and they have these huge tables that lots of people sit at and it was just the two of us so they of course sat us with other people. Anyway, I didn't think that one thru.

Also, Talked a little online to my friend Spiff (That's just a nickname I have for her). She seemed somewhat short with me today. She said she was fine but I wonder... Hope she's ok.

How's that for a first post? Later.