Jeffrey A. Miller: Open Source

Mon, 30 Jun 2003 22:43:43 GMT

The Register: Open Source Prepares to Kiss EU Patent Ass Goodbye [Linux Today]

Mon, 30 Jun 2003 22:41:55 GMT

CNET News: Group Claims Linux Advance on Xbox [Linux Today]

Mon, 30 Jun 2003 22:40:31 GMT

TechWeb: Will Linux Revive Novell? [Linux Today]

Javascript Obfuscator

Wed, 14 May 2003 03:05:46 GMT

Stunnix JS-Obfus 1.2 [freshmeat.net]

A Project Communication Tool

Wed, 14 May 2003 03:00:15 GMT

Outreach Project Tool 1.0.0 (Max) [freshmeat.net]

My company has been yearning for a project "dashboard" application that helps customers and developers, testers, project managers stay in good communication with each other. I'm adding this to the list to come back and review.

MySQL security patch

Tue, 29 Apr 2003 23:12:52 GMT

Red Hat Linux Advisory: MySQL [Linux Today]

Tue, 22 Apr 2003 19:43:19 GMT

FoxPro On Linux, Drama Ensues. bltfast32 writes "I don't know how many people have been following this, but this is definitely worth keeping an eye on. Whil Hentzen, prominent FoxPro and ... [Slashdot]

Mon, 21 Apr 2003 19:51:47 GMT

Video Disk Recorder 1.1.28 (Development) [freshmeat.net]

Mon, 21 Apr 2003 19:45:29 GMT

dirdiff 1.6 [freshmeat.net]

Mon, 21 Apr 2003 19:43:44 GMT

Atomic Photo Album 1.1.0pre1 (Development) [freshmeat.net]

PHPWebSite

Tue, 15 Apr 2003 19:59:13 GMT

TeamSite 0.1.0 [freshmeat.net]

A sports roster plug-in for the PHPWebSite content management system. At first, I thought this was going to be a business-oriented plug-in. But it's not. Oh, well, at least that reminded me about the PHPWebSite CMS.

Tue, 15 Apr 2003 13:27:24 GMT

Debian GNU/Linux Advisory: kdegraphics [Linux Today]

Mainframes Never Die?

Wed, 09 Apr 2003 14:18:24 GMT

TechRepublic: CIO Update: Future of the IBM Mainframe Looks Surprisingly Good [Linux Today]

Hmmm...

What I hated about the IBM mainframes was that a lot of the organizations that relied on them were staffed by stuffy, old-school programmers unable or unwilling to see the value in doing anything new with technology. Many times it was obvious that they were hoping to coast on their current skillset in their current job until retirement.

Forget trying to teach an old COBOL jockey about object-oriented techniques. Heck, even event-driven programming was a stretch after eating-drinking-sleeping top-down procedural programming all your life. I know COBOL programmers who insist that too many separate paragraphs leads to excessive performance overhead. You're kidding, right? What about code maintainability? Oh, right, I forgot. You're only interested in job security.

I once made the mistake of complaining about a COBOL program that I had to maintain and modify for Y2K. I said to my manager, "Man, this code ugly, and there's no documentation or comments to help." It turns out that he was the primary author of the code. He said, "Jeff, if I'd have documented all of this stuff, I'd be making $20,000 less than I am."

Don't get me wrong. I understand COBOL. I used it for close to six years between college and three years of dreadful jobs that included it. Now, I don't even include it in my advertised skillset. Yeah, I could still go back and write a PICTURE clause if I had to. Ironically, I wrote a utility in Turbo Pascal and then Quick BASIC 4.5 Professional to generate PICTURE clauses for me.

What I don't like about COBOL is its monolithic style. All variables are global. Encapsulation and modularity are hard to come by. There is no support for parameter passing to in-program functions. Dynamic arrays are a pain. And, if you're data is not fixed-length in every instance, you may as well tear your hair out manipulating variable-length strings. Don't believe me? Just try presenting a "Last Name, First Name" on a report. You'll have to munge it up with the STRING statement DELIMITED BY SPACE (or some similar junk that I've thankfully forgotten). Sure, you can call other programs by way of a LINKAGE SECTION, but please! That's a lot of crap to go through just to introduce some modularity and reusability.

Now, we could argue that some "innovative" (not) companies have revamped COBOL to put a new face on an old language. But I say that it's not really COBOL any more. Micro Focus tried to introduce Object COBOL in the mid 1990s. Now Fujitsu has introduced a .NET compatible compiler. No thanks. I don't feel like arguing about the merits of COBOL. I hope Fujitsu sells a ton of the product. More power to them.

Anyway, as far as the platform itself goes, the IBM mainframe is very robust and well-engineered. A system administrator friend of mine told me that they took an IBM mainframe and carved it up into about 1,000 virtual machines and ran virtual Linux servers on a couple of the VMs. Now that's cool.

Wed, 09 Apr 2003 13:59:33 GMT

2003 Webby Awards. We're pleased to announce that Movable Type has been nominated under the Best Practices category in the 2003 Webby Awards.... [Movable Type News]

Samba Security Hole

Tue, 08 Apr 2003 14:17:14 GMT

Samba Exploit Discovered, Fixed. An anonymous reader submits: "Digital Defense reported a remote root vulnerability in Samba that has existed in Samba source code for over 8 years. ..." [Slashdot]

Though found and fixed quickly, this vulnerability is present on countless existing systems until patched. Download and update soon!

PHP-Nuke 6.5 is released

Wed, 02 Apr 2003 16:22:25 GMT

PHP-Nuke 6.5 is out with phpBB stuff included.

Fri, 07 Mar 2003 11:43:19 GMT

PHP *is* a toy.

PHP Everywhere: Is PHP a toy?:

Yes

Maybe when PHP 5 comes out it'll be better. PHP's neat because it has all these modules available for everything, but right now, the language is a toy language.

[Keith's Weblog]

Fri, 07 Mar 2003 11:41:55 GMT

Java sucks.

Via Adam, Advogato: Why Sun is right that Java sucks

Lots of neat stuff in the article and ensuing comments. There was some red herring discussion of garbage collection, but a link to a very good presentation by Hans Boehm (PS) came out of it.

[Keith's Weblog]

Drupal

Fri, 21 Feb 2003 23:45:49 GMT

URLs are what matter.

You know, I've realized that the only thing that's really important to me is meticulous control over my url-space. That's the only thing that really matters. As long as your URLs stay constant, you can rip out your backend as many times as you want and it doesn't matter. So, I might just say screw it all, move over to Drupal, and set up a simple gateway that'll translate named URLs into Drupal nodes. Hmmmm...

[Keith's Weblog]

Drupal seems to be built PHP, XML, and perhaps other technologies. Need to read more here.

A Bold Declaration

Wed, 12 Feb 2003 03:39:55 GMT

GvR: Strong vs. Weak Typing.

Guido van Rossum: Strong versus Weak Typing:

Strong typing is one reason that languages like C++ and Java require more finger typing. You have to declare all your variables and you have to do a lot of work just to make the compiler happy.

[Keith's Weblog]

I'm sorry, but I can't sympathise with someone who is too lazy to declare variables. Period.

The disgust with excess typing I can understand. I feel your pain. I don't want to worry about the type of a variable on every statement but omitting variable declarations is without excuse.

Serendipity at Work?

Sun, 02 Feb 2003 21:31:13 GMT

<RANT>

OK, I've written several times about Sam Ruby's description of Manufactured Serendipity. The depressing part about all of this is that most of my traffic on this site is truly serendipity (the unmanufactured kind). It comes from people using Google to look for a "PlanPlus crack." The only reason this happened is because I arbitrarily chose to split the word "crackpot" into two words: "crack" and "pot." Gee, maybe this post will attract people looking for illegal drugs! ;-)

I'm sure this entry will draw even more attention from search engines and those folks out there who insist on not paying for software. Well, excuse me, but go jump in a lake! Pay for your own stinking software or start using open source software for its most touted benefit -- it's free.

Hey, I shell out a lot of money trying to stay largely legit, even personally. I think it's part of working in an occupation that is based on intellectual property rights. I hope to someday right a somewhat successful shareware application that is registered by a throng of willing folks and thereby supplement both my lifestyle and my retirement. But until that time, I can only do the ethical thing and behave in a reciprocal manner by PAYING FOR LICENSES!

</RANT>

Meta Keyword Tags are Dead and Syndirella is Alive and Well

Sat, 18 Jan 2003 23:41:17 GMT

Thanks to Dean for including this tidbit from one of my favorite sources: Search Engine Watch.

Apparently, nearly all significant search engines ignore the <META NAME="keywords"> tag. It's a "spam magnet" and has been largely dismissed since the 1997-1998 time frame with regard to its importance in search engine relevance. I was once badgered by so-called expert when building a site (in mid 2001) because I didn't include the META keywords tag. The site won't be promoted properly, blah, blah, blah. Well, I think the most important factors are keyword usage in page titles and content, search engine submission, and external links leading to your site. Except for Inktomi, the major search engines don't use the META keywords tag. So it's virtually dead, unless you have some other purpose for your own Million Dollar Markup.

If you're interested, here's a few other things you can do with a META tag. The article taught me something new today. Some META tags (the ones that use the HTTP-EQUIV attribute) correspond to headers found in HTTP messages (see IETF RFC 1945, section 4.2). Some web servers translate the values stored in the META tags (the HTTP-EQUIV kind) into the HTTP headers when they deliver content to the browser.

Dean also noted the arrival of an open source, Windows-based desktop news aggregator called Syndirella by Dmitry Jemerov. Excellent! I like the idea of the ability (more research required here) to scrape regular web pages for news. This means that I might be able to digest the Search Day newsletter from Search Engine Watch without visiting the page directly. Maybe.

I was using email subscriptions to get most of my news in the past, and I possess many dedicated mail folders for particular sources that fill up with unread messages. Pretty useless.

So, as I move along in my blogging adventure, RSS collected by a news aggregator seems to be a much be better solution. If I miss six months worth of news, I don't have six months worth of news staring me in the face during my next session--only recent, relevant stuff.

Now, Syndirella goes one step better. Instead of having an extremely long web page to read, Syndirella carves up the news into each feed and allows in-client reading without bouncing all over to other windows. It still feels a little clumsy and basic, but my hat is off to Dmitry for creating such a useful tool and releasing it as a free, open source product.

Not only that, but it's written in C# running on the .NET framework. I was wondering what the most effective way to distribute a .NET desktop application over the web would be. The Framework is a 20MB+ download for those that don't already have it. That's a long wait for a lot of users. But Dmitry is doing it. He warns the user about the size of the download and gives instructions about where to acquire it. This seems to be the best way I've seen so far, short of mailing out a CD. No, I don't want to do that.

Is Movable Type Free?

Tue, 14 Jan 2003 14:24:04 GMT

Yes. Mostly. Here's the scoop from the Movable Type FAQ.

MT version 2.0 and higher is free for personal or non-profit use. We're following a donate-ware model in which we ask that you send us a paypal payment for the amount which you feel the product is worth to you. In return, for every $20 that you donate, you will receive a Recently Updated Key. For donations of $45 or more, you are entitled to additional support. See here for more details.

Here are two good finds about Movable Type:

Radio vs. Movable Type
http://blogs.salon.com/0001111/2002/08/14.html#a194

Making the Move to Movable Type
http://www.meryl.net/articles/archives/000947.php

Tue, 14 Jan 2003 05:28:17 GMT

"Linux Business Week - SCO Threatens to Press IP Claims on Linux" [Daypop Top 40]

PyPerl - Python and Perl

Wed, 08 Jan 2003 17:18:24 GMT

I sometimes wonder what the ultimate culmination of all of these seemingly disparate open source technologies. While that remains to be seen, here is yet another instance of two (actually three) popular open source technologies coming together.

'pyperl' is an extension module that bridges the gap between Perl and Python. It allows Python code to invoke Perl code and operate directly on Perl data and permits Perl code to do the same to Python code and data. This provides for almost seamless integration between the languages and thereby greatly expands the library of modules available to each of the languages.

ActiveState and Digital initially developed the pyperl extension Creations as the foundation for supporting Perl in the Zope application server.

[O'Reilly]

I certainly haven't researched these (PyPerl, Zoperl) very much, but they have at least made my open source mental watchlist.