Brandon's Blog

7/12/2005 #

IMAP

IMAP rocks.  It makes POP3 look like a kiddie toy.  I’m waiting for gmail to support IMAP.  Plus, Thunderbird let’s you connect to an IMAP server through an SSL tunnel (actually and interestingly, gmail requires SSL to connect to their POP3 server), so my password is never transmitted in the open.

It’s just weird to have a persistent connection to a mail server.  Mail is so…instant.

Subversion access to sigma is still in the works.  WebSVN is running, but I haven’t gotten the code uploaded.  I might have time tomorrow.

7/12/2005 #

A Disclosure

Now that The Cure is effectively out of the music picture, I can say this:

eye makeup on male musician == bad thing

7/10/2005 #

Linux Update

So, I finally made the big switch to Gentoo last weekend, and I’ve been incredibly pleased.  I didn’t really expect to be, as I have installed it before with less impressive results, but it has performed very nicely.

A whole bunch of real losers use Gentoo because they think they can tweak a few extra bits of performance out of their systems by compiling everything custom with optimizations.  The way I see it, you should get to know Linux on SuSE or Fedora, then move to Slackware to learn Linux.  Once you have farted around enough with compiling by hand to get dependencies, Gentoo is a logical step.  You still have the system control of Slackware, but portage handles your dependencies and configuration aspects.

Naturally, Gentoo is not time-efficient in terms of getting your packages installed instantly, but if you’re essentially dealing with a hobby computer it is a very appealing system to use.

Anyway… to sleep.

7/9/2005 #

RSS Syndication

RSS Syndication is up, which is great.  Webmail is up, too.  But I doubt you care, because I have the only account.

Boo yah!

7/8/2005 #

Farewell, Trusty Blogger

If you haven’t seen Textile, check it out.  It’s the core technology benefit behind moving over to Textpattern as a blogging utility.  It let’s you do intuitive syntax like surrounding asterisks for bolded text, then it converts that syntax to valid XHTML.  Of course, it also looks really nice and is very configurable.

Textpattern is the first addition to my brand-spanking-new paid hosting account from TextDrive.com.  It’s a full-fledged PHP and Ruby on Rails host, with MySQL databases and all that other fun. open-source server side scripting goodies.  It also has IMAP e-mail, which I’m very excited about.

It’s about time.

7/6/2005 #

(No Title)

From the Inq:

In the course of writing the book, [William Shatner] had talked to Ray Kurzweil, the father of voice recognition technology. While he was dictating a sentence something like “I have just talked to Ray Kurzweil,” the speech recognition engine translated Ray Kurzweil to Pakapakapaka.

He said he asked Ray what he should do about that and Kurzweil said, “Just say delete”. But when he said delete, it didn’t delete Pakapakapaka but just pasted delete after it. So he said delete again and now there were two deletes.

He said: “It [voice recognition software] can’t even spell the name of the person who invented it.”

“None of this shit works,” he said.

6/30/2005 #

(No Title)

I bow to the supreme 1337-ness of KDevelop.  I imported Sigma into it and it compiled cleanly, even managing to INTERPRET my handmade makefile into English summaries in the status pane.

[Pansy alert]
I am now an official Fedora Core 4 user, with KDE.  I enjoy using yum as a package manager, and I think RPM (or an equivalent like .deb) is the optimal way to manage a desktop/productivity system.  Dependency tracking rocks my face off, and I am waiting to install the Qt 4 development packages until an RPM is released.
[End pansy alert]

Oh yeah, and Oceania has always been at war with Eastasia.  Sorry, Pat.  I learned all I know from Slack, and it’s time to put the training wheels on and enjoy running Linux from day to day.

No development on Sigma yet, but lots of thinking.  How will I handle containers and inventories?  Too big of a question to run into without fully planning.  Now that I have KDevelop, I don’t have to shuffle between three rxvt terminals and a vim window.  Oh, and my Valgrind bug finder is integrated into the IDE.

When KDE moves up to version 4 (meaning Qt 4 is used as well), we should start seeing KDE applications like KDevelop being ported to Windows, since Qt 4 is finally liberated from the commercial-only license encumbrance that plagued its earlier Windows versions.  Look out GTK…

6/15/2005 #

(No Title)

Sigma Punch List

• Implement Container class (inherits from Item) with List of Item pointers for container contents
• Item weights in XML and data structures
• Weapons and Wearables (probably implemented at the Item level to avoid inheritance bloat)
  
  The Big Picture: These modifications (and statistics) lead to a combat system.  Following combat is the Lua API.  Then, it’s extension, testing, and perfecting into a 0.1 or 1.0 type release.

6/12/2005 #

(No Title)

Dig this.

This is a screenshot of me watching a Kenshin .ogm movie on my laptop over an SSH/X tunnel with ASCII art library.  The audio is patched into my television here in the apartment (the audio does not forward over the tunnel and hence must be dealt with at the Linux machine).  The source of the video is a DVD I just burned with my new Sony burner.

This is seriously dorky.

6/7/2005 #

(No Title)

CNET: “What about for business buyers? What’s your elevator pitch to those customers? Why should they buy Longhorn?”
Ballmer: “The dynamic is that the end user gets excited about it because they use it at home. And all business decision makers and IT people are end users. The end users are excited by the new shell… The new user interface-kind of sexy, kind of cool.”


So, Apple’s on x86 and Microsoft is touting eye candy as the primary motivating factor for businesses purchasing the next-generation Windows OS.  Somebody please right the world again for me.  I long for the simplistic mid to late-90’s, where Mac’s were for incompetent artsy-fartsy users, Linux was for hackers of the higher order, and MS was the only choice… and not a bad one at that.

I feel like Windows ME was underrated post facto.  Yes, it became unstable after a year or so even with de-frag/scandisk/etc, but it wasn’t as bad as people made it out to be.  It’s my belief that all computers should be refreshed at least once a year.  Windows 2000 was not as good as everybody makes it out to be (lacked nice included features of XP like drag-drop CD burning), but since businesses bought in (and smartly so), it has a good reputation.  XP Home is a pretty unfortunate system that does its job well.  XP Pro is especially beautiful, and even more so considering that once you tire of the blue taskbar you can switch back to Windows 2000 look and feel.

So why the emphasis on eye candy?  I theorize it is due to what I might term “feature saturation.”  We have DirectX, COM, DCOM, XML Web Services/SOAP, ActiveX, .Net, etc., etc., etc., ad nauseum.  But the “nauseum” is literal.  Our “Wintel” (ha ha) machines are vomiting acronym-based technologies that we have no cause to use and no balls to discard.  Because who knows if some random company is using some custom “solution” based on these technologies.

.Net was the platform to end all platforms.  Too bad MS doesn’t have the cajones to build more of Longhorn on top of it.  .Net should equate to the Java VM, but it has a core flaw (in my humble and only limitedly-experienced opinion): .Net is not friendly enough to cross-platform portability.  I find it offensive that the “common language runtime” or CLR has a Windows.Forms library when everything else is abstracted above the system configuration layer.  Why not a Qt/GTK/Swing -ish cross-platform graphics abstraction layer that could be implemented atop any graphics system?

The answer is arrogance.  And Mono hate/FUD.  But I digress.

My point is that Windows technology in itself is saturated.  Whatever .Net cannot do should be added to .Net.  The Windows kernel should not depend upon the .Net framework, but almost everything else should.  If .Net is too slow for this (it’s not, in truth), then why in the heck are they selling it?  The Mono project should provide a compatibility layer between .Net Windows apps and the Linux/BSD OS.  This allows MS to sell legal Office licenses (for some future, .Net-based version) to Linux users, which is probably more lucrative than Windows itself.  It also finally kills off the Wine project, which was noble but a bit silly.

And the end user is not learning jackedy squat any more.  What’s .Net?  What’s Avalon?  What’s WinFS?  Why do I want it?  Will it make my Internet faster?  Will it make those “security center” things stop showing up that have been there since my neighbor upgraded my CPU to SP2?  The reason they’re not learning is that it is not necessary.  The least common denominator prevails.  Even our friends the script-kiddie 1337 h4×0rz don’t really care about technologies.  It’s pointless, just a way to sell MSCE certifications and 1000 page Que books.

What I see right now is a boatload of irrelevant MIS/BCIS/whatever majors who know nothing about the work itself, care nothing about computers, and only know to call the guys who know and accept the lowest bid.

Above these morons is a technocracy of consultants and developers, which is increasingly alienated by “MS Bad Thing (fill in the blank).”  The ones who are worth their salt are already running Linux, BSD, or OS X at home and only hack ASP or C# at work because it pays the tuition and buys the Volvo.  .Net is good, but it’s not typically a whole lot of fun.  In fact, it’s easy enough that it takes the heavy challenges out of coding.  There is some great stuff out there still, but the majority of business coding is database interfaces and simple, if-then logic.  Lots of grunt work.

.Net thrives on this deterministic-type object oriented programming.  I know.  I wrote one.

Anyway, pointless rant.  Just amazed that UI is the critical factor right now, with Moore’s law chugging away and all this untapped possibility out there.