Older blog entries for Waldo (starting at number 153)

Girlfriend, Meet Mac
My girlfriend, who I've gradually been converting into a Mac user for the past few years, has asked to borrow one for a while. She's currently running a rather-nice Dell and dual-booting between Windows 98 and Mandrake Linux. Now that she's finished with school, she mostly needs a computer for Internet access and digital photography. (Mostly pictures of friends and family and pets and such.) Now that she's seen iPhoto on Mac OS X, the deal's as good as sealed. I'm loaning her my Rev. A iMac (which I got the moment they came out, as a gift from a good friend.) I just wiped Yellow Dog Linux from that system last week and did a clean install of Mac OS X, the first time I've had a Mac sans OS 9. I think it will be perfect for her.
31 Jul 2002 (updated 31 Jul 2002 at 04:42 UTC) »
I got my replacement Asus A7V-266e in the mail today from Asus, 28 days after I express-mailed the previous one back to them. This is the replacement for the replacement, the third thus far. Maybe this one will work?


Tomorrow (well, in 30 minutes EST) I turn 24. I'm having a hard time even caring. I don't know why. I think it's because I'm in the beginning of a long chain of uninteresting birthdays. I guess I get to rent cars at 25, but that's not particularly interesting. It's not until 30 that there's any notable digit change. I've decided to celebrate the occasion by taking a flying lesson at the local airport, and then Amber and I will go up to a Dave Matthews Band concert in upstate Virginia, for which I have appropriately-fabulous tickets.

E-Mail Filtering: In Danger of Being Commercialized

The first e-mail filter that's on the right track has been announced. The software product, made by Banter doesn't yet have a name. The idea is that it sorts e-mail in your in box based on a whole mess of criteria on a scale of 1-100. It's the criteria that's the trick, of course, but they claim to have licked that. This is obviously necessary, as the current state of e-mail is quickly rendering it useless, as I'm sure many of you can sympathize with. The problem? This company has a patent on this type of natural language technology, and they're selling this software as part of a $75,000 package. I'm sure when they break it out as its own package, it'll be an enterprise plug-in to Microsoft's mail server. (IIS, I guess it's called. Or maybe that's just HTTP. I don't know.)

I'm worried. I lack both the time and the proper skills to launch an open source counterpart, but I'm gravely concerned that nobody else will do so. It's essential that an RFC be created to describe a public-domain, standards-based approach to a similar system such that everybody can benefit from this mail system. I'd hate for this system to be the defacto standard 24 months down the line, and find OSS advocates once again playing catch-up.


I've been doing a lot of PHP recently. Not only because it's necessary, but as a warm-up for getting involved in PostNuke. I need to get back into the swing of things.
I used ICQ in 1997ish, and I didn't like it, mostly because it meant that Dave Matthews Band fans could interrupt me at any time and beg me for backstage passes, something that I'm not even capable of providing them with. However, I've read so much about Jabber in the past year or so that I just had to download it. I'm running Fire.app on my PowerBook, and I'll run some sort of a Jabber on my Linux box when Asus finally sends me my motherboard. So far I've only done the whole IM thing, but eventually I'll get into the guts of the system. From what I've seen so far, it seems quite elegant.


I've joined the PostNuke team, more or less. I've made an account on the developer site, joined the mailing list, and e-mailed them to notify them that I intend to contribute. I've been reading the mailing list (which doesn't appear to be available as a digest, but Procmail is doing the trick right now), and it'll take me a week or two to find my sea legs, I'm sure.

OSS Advocacy

Today, I visited a large insurance company that shall remain anonymous. Let's just call them Insurance Company. Insurance Company is working with Proprietary Software Developer, who makes policy issuance software. The firm that I work with, a surplus lines brokerage, has offered to beta-test Proprietary Software Developer's new Web-Enabled(tm) version of Proprietary Software. So I spent five hours in Richmond, Virginia (USA) today in a meeting with The Powers That Be at Insurance Company, going over the basic concepts of Proprietary Software to see how it can fit in at my company. There are so many flaws in this system that it borders on silly, but the biggest is that they've more or less locked themselves into this Proprietary Software.

Now, Insurance Company has been around for a long time. So this move to the Web-Enabled system is not something that they take lightly. What would lead them to transition themselves from paper to digital is fairly obvious to all of us, I imagine. However, what would make them think that it's a good idea to become entirely dependent on Proprietary Software Developer is a complete mystery to me. Proprietary Software Developer has been in business for about a decade. They've shown a strong bent towards gauging their customers, with licensing fees upwards of $40,000/year. I put this question to Insurance Company: Who is going to be in business longer, Insurance Company, or Proprietary Software Developer? They felt, of course, that they would be. Which made me wonder why they would make their company entirely dependent on Proprietary Software. They didn't have an answer to this.

After about four hours of discussion, I delivered an impassioned speech on the merits of open source, free software, attempted to convince them to release their XML standard into the public domain, set up an XML-RPC server and allow brokerages to develop the software. This was lost on them in a tremendous number of ways. Why would they give away their valuable (in their eyes) XML standard to their competition? Why should they cooperate with people that they'd like to defeat? Why would brokerages develop software?

Try as I might, I made little more than a dent in their two basic misconceptions. And these are the misconceptions that most BigCos (credit to Dave Winer) have:

  1. Their idea/technology/standard is so incredibly valuable that they can't possibly share it with anybody under any circumstances.
  2. Cooperation with competition is inherently damaging.
If they would take their largely-pathetic innovations and combine them with similar innovations of just a few other insurance companies, then they'd have something close to a standard. A standard would allow software developers of all sizes and shapes to develop software that would work for many insurance companies, brokers and agents and, if the standard continued to grow, perhaps a tremendous portion of the industry as a whole. A rising tide raises all boats, after all. But I guess Insurance Company didn't get that memo.

I'm so eager to get my Asus motherboard back -- I've just sent back the one that they sent me to replace the one that I bought that was broken. (Phew.) I'm desperate to have a system that I can use for Python and PHP development, notably to be able to start making contributions (beyond bug filings and such) to PostNuke. My motives are largely itch-scratching; the bugs on cvillenews.com are really very irritating. But rather than fix them for myself every version, I'd like to get the changes into the source.

Anyhow, just gripin'.

I upgraded one of my sites from PHP-Nuke to PostNuke this evening. Boy, do I have a lot of bug fixes to submit. All of my old URLs broke; they're now hundreds of characters long. It ignored my old design and put on this nasty beige thing. User registration sucks just as much as ever. I'm baffled as to how to set the items that appear on the main menu. There are errors ("Block Type userbox Not Found") that mystify me; why would there be errors on a stock install? Moderation reasons don't appear: it just says "Score: 3" without any "Insightful" or anything. The whole system is really very confusing.

That said, I'm very happy with it. It's a fine project that is developing nicely. I've tracked it since day 1, and I have high hopes. But it's sure got a long way to go between v 0.714 and v1.0 if it's to be any good.

Note to PostNuke team: if you want an excellent, excellent program on which to model things, check out Gallery.

Home Network
I've got a pretty crazy home network now. I recently moved my office into my apartment, and meshing the two separate networks proved challenging. Here's the setup now.

I've got a DSL with two static IPs. The DSL comes into a 10Mbit hub. The hub breaks off into two directions: one goes to my primary server with its own IP (Tux), the other to a firewall. The firewall hides the rest of the network. It routes most traffic to the main home server, Rez. Tux takes care of DNS, some mail, and some HTTP. Rez handles most mail and some HTTP. It also routes traffic from 8080 to Rusty, my mailing list server. Mail gets to Rusty via Rez, who works as an incoming mail proxy via Procmail to reduce the load on poor Rusty, who's aged, to say the least. There are a few other machines behind the firewall as well, mostly clients. Brodie, the iMac; Woot, the PowerBook, Vig, the Linux box (temporarily motherboardless), Johnny, the office fileserver / MP3 server; and a few other machines always coming and going.

I'd like to replace these hubs with switches to increase the speed. I could use a few five-port Netgear 10/100s, rather than wasting all of the ports on these ten-port ones. I'd also like to get an 802.11a hub, which I'll likely do before the year's over. I can't justify it right now, because Woot the Powerbook needs a new AC adapter, so for now it's tethered to the wall. And eventually, gigabit Ethernet would be nice. But I'll hold off on that right now. Too expensive.

Somehow, it doesn't look like the apartment is wall-to-wall with computers. It helps that the three machines that function headless -- Rusty, Johnny and Tux -- are tucked behind a mammoth desk, invisible in the living room.

I think that, for now, I may finally have enough computers.

I finally threw in the towel on this Asus A7V-266e motherboard. This is that Asus ("aye-SUSE," for the curious) motherboard that they sent me last week to replace the previous one that was broken when I bought it. It's unclear, but it appears that this motherboard is used. In any case, it doesn't work at all. The fans spin, but that's the extent of its functionality. So, once again at my expense, I'm going to mail this motherboard back and hope that the third one that they send me will work. If it doesn't, I'll be asking for my money back and getting a new motherboard. After three months, I think it'll be time for a new brand entirely.
25 Jun 2002 (updated 25 Jun 2002 at 21:03 UTC) »
I got my broken motherboard back today. Asus mailed it back. They don't appear to have done anything with it. It came with a printout attached to it, with various form fields such as "Missing Part:", "Replaced Parts:", "Error Location:", etc. All are blank.

It appears that I could have saved a lot of time by simply mailing this motherboard to myself. Thanks, Asus.

Update: I called Asus. Turns out it's not my motherboard: it's a used one. That's right: I bought a brand new motherboard that was DOA (Dying on Arrival, return a few weeks later), I sent it back to them and they replaced it with a used motherboard. The woman that I talked with had such a thick Korean accent that I was entirely unable to understand what she was saying. I'll have to call back later and hope that I can somebody else, so that I can ask them what the hell they're trying to pull.

I've been spending more and more time reviewing the code of PostNuke, and I've started to submit bug reports and such. I'm trying to slowly warm up to it rather than just start submitting patches. :) I think that it has the potential to be very, very good. At the moment it's extremely ugly, primarily in that it lacks any sort of grace.
I've switched my mail mail server from Sendmail to Postfix. I like it better. I've also started using SpamAssassin instead of Spambouncer on that server. I think I like SpamAssassin more. It worked faster, with less tweaking, and appears to be more actively updated. I wish it came with a plug-in to filter out Klez, though.

Verisign is a horrible, horrible, horrible company. I hereby resolve to move all of my domains that remain with them to DirectNIC as soon as I am able and never suffer through their "service" again.

144 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!