Devin on Earth

Dang. I just pulled off something tonight that makes me wonder if Apple wasn’t in fact smoking crack when they decided to drop the PowerPC processor in favor of moving to Pentiums. My little Mac mini — 512MB of RAM, 1.42 GHz PowerPC G4 CPU, was able to effortlessly (and I mean without a single slip, skip, hesitation, or bobble) do all of the following simultaneously:

• Burn a DVD from an ISO image using an external DVD burner
• Rip music into iTunes from a CD in the internal drive
• Play music from a separate playlist in iTunes
• Synchronize my iPod (iTunes again)
• Browse the Web with Firefox
• Read email with Microsoft Entourage 2004

Now, I’ve used a variety of systems in my day, and I’ve never seen the machine that could handle that set of tasks without at least the occasional glitch or shimmy in the works. At the very least, all of the I/O would cause the occasional pause as the machine optimized for the DVD burning operation to avoid making a toaster (if I was lucky). And my experience isn’t just in Windows — I’ve tried this on Linux and FreeBSD systems too. The unifying factor? They were all based on the Intel architecture, with its half-assed memory management and I/O capabilities.

I really want to know what kind of chipset Apple is using to do memory management and I/O, because this mini just smoked my brain. I’ve known for a while that raw clock cycles can’t be used to compare RISC processors like the PowerPC (or SPARC or Alpha) with the CISC CPUs such as, oh, pretty much anything out of Intel for modern PC use — but having a low-end system like this deal with all these high-I/O tasks without breaking a sweat is pretty damn awesome. I really hope they’re not using what is essentially typical PC chipsets, because if so, I doubt their new offerings will be able to do this, no matter what benchmarks they can cook up.

Mixing custom playlists is fun. I just created a “Sweat” playlist, intended for use when exercising. Maybe having a decent soundtrack will give me some motivation to hit the physical fitness regime with something approaching dedication again. As such, the songs don’t have to be good music, they just have to have a good beat or rhythm:

• Deep Dish, No Stopping for Nicotine
• Queen, One Vision
• Astaire, L-L-Love
• Billy Idol, Wasteland
• Many Moore, Candy
• Oingo Boingo, When the Lights Go Out
• Britney Spears, (You Drive Me) Crazy
• Falco, Vienna Calling
• Erasure, Oh l’Amour
• Bonnie McKee, Trouble
• Kenny Loggins, I’m Free (Heaven Helps the Man)
• Roxette, Dangerous
• Kelly Clarkson, Since U Been Gone
• Phil Collins, Sussudio
• Lindsay Lohan, Rumors
• ABBA, Does Your Mother Know

Yeah, I know. I should be ashamed of myself. But y’know, if I can actually stick with my exercises and drop some lard off this tubby frame of mine, my taste in music might be slightly vindicated.

Chatting with my sister just now and I realized that a lot of my problem with this current chapter is that it deals with the people side of the process. And as you all know, I’m such a dynamite people person, so clearly this is my strength.

Not.

I also realized that somewhere along the way, I’ve fallen back into the IMPS (I Make Perfect Shit) mindframe. When I’m in IMPS, it isn’t enough to write a good draft, or even a great draft. No, it has to be perfect. It isn’t enough for me to neatly and concisely distill all the information and opinon on the topic I’m working on, leaving a clear and accessible summary — no, I have to impart some unique and brilliant insight that will leave the reader gaping in awe.

Now, you may think this comes from ego, and I’m sure there’s a bit of that lurking around. However, most of it comes from fear. Fear of being seen as a young snot who has no practical experience with this topic and has no business writing about what is one of the most complex topics in today’s IT world. Fear of being seen as a phony. Fear of finding out I’m a hack. I’m afraid of my editor pushing the chapter back to me and telling me, “You know, Devin, normally your work is really good — an easy edit — but now when the rubber meets the road, this is just crap.”

[Message from Devin's brain to Devin here] Hey, idiot. It’s called a “draft” for a reason! [Message ends.]

So, now that I know what the problem is, I think I know how to solve it. It’s okay to produce material that isn’t polished and sparkling like a diamond. It’s okay to write material that lets the editors earn their pay and add value to the product. It’s okay to not be perfect.

Update: Not only is it okay to not be perfect, most of the time (if not all of the time) perfection is what really kills the effort. More on that another time, though — I have a chapter to finish.

As you may remember, the local bookstore closed down last year. Today, I found out that our charming little hole-in-the-wall CD shop will be closing down in 6 weeks or so. This bites, because Steph and I try to support small local shops even if it means paying more, as long as we can get what we want there.

On the good side, they’re holding a 15% off everything sale, and I was finally able to pick up a copy of Billy Idol’s infamous (and very hard to find) Cyberpunk album, as well as a boxed 3-CD set of Queen’s greatest hits. Still, I’d rather that they were able to stay in business.

We finally diagnosed whatever is going on with Steph’s computer — motherboard or CPU. Well, shite. I wasn’t really looking to be rebuilding machines at this point, nor to be acquiring new hardware. I think we have a gameplan, but Steph will be computerless for another couple of days. Good thing she can read her email from OWA.

Oh, yeah — chapter 4 of the DCAR ebook is kicking my ass. I just cannot find the words to put on paper. This is doubly infuriating, because there are a lot of other projects that are clamoring inside my skull for airtime. I’ve got a wonderful essay rattling around in there, tentatively titled “The Relationship Lens: Re-imaging the role of faith and the church.” This sounds like a lot more fun to write, but it’s not what has the deadline.

Got a concerned email from one of the ladies at church. Rumors are now going around that we’re leaving. This wouldn’t be so bad of itself, because we are going to be actively looking for a new church home. What pisses me off is that it sounds like there’s already a healthy load of bullshit going around about my reasons — lots of gossip and speculation. Mind you, nobody’s bothered to ask me what the truth is.

If it weren’t for the fact that we still have a great school for the kids, I’d almost be thinking it was time to seriously consider moving from Monroe.

Update: the kids are really grooving on the Queen albums. They recognize a lot of the music from the Highlander movies and TV series. Yep, they are geeks.

We’ve been getting over the flu here at Casa del Ganger. For the past several weeks, I’ve been feeling under the weather to a greater or lesser extent, so I’d slow down and try to get extra rest, etc. After a few days I’d feel better, so I’d go back to normal life, and whammo! a couple of days later, down I’d go again.

Last week and weekend, however, was the absolute bottom of this cycle. I struggled along the week slowly feeling more and more like crud scraped off the bottom of the Coney Island boardwalk, and when Friday finally arrived I almost decided to postpone the Auspex move. In the end, I didn’t postpone it because I’d already done so once. So I got up on Saturday, went to breakfast with a friend, then came back home and hauled the Auspex out and got it ready for transport. The new owner showed up right on time and right around noon, we had it loaded into his rental truck and were pulling away from the house for the trip down to Factoria. I had my new Qtek 9100 Windows Mobile PDA/cell phone with me so I could tell Steph when we were done unloading it and getting it set up. A couple more hours later and I was sitting at McDonald’s waiting for my family, entering my first blog entry on my geektoy.

When we got home (minus a detour through Costco Home), I intended to lay down for two hours or so and take a refreshing nap, since I felt like hammered crap. Instead, I woke up at midnight when Steph crawled into bed. I went back to bed and ended up sleeping most of Sunday.

Of course, Murphy was having none of this — so Steph started coming down with it too.

This week, I’ve been on the mend, although I’m still sadly lacking in energy. I’ve been going to bed by 8:30 most nights. On the upside, I’ve been getting up nice and early — anywhere from 4:30 to 7:00 — so I’ve had plenty of time to get my work done. I’m just running out of steam awfully quickly though, and if I’m tempted to think I’m completely healthy, all I have to do is plot when my energy falls off. Nobody who can tell exactly when their last dose of acetaminophen wore off by noting that the sudden loss of energy is exactly eight hours past the last pill-ingestion has any business deluding themselves about being healthy.

As an aside, reading Stephen King’s uncut version of The Stand probably isn’t a great idea when you’re sick. I felt like I was drowning in phlegm a little too often last weekend to be comfortable reading about the superflu that killed people by drowning them in phlegm. I finally realized that the reason I kept reading it was that even as sick as I felt, I took a certain perverse pleasure in not being as sick as those poor bastards.

Speaking of perverse pleasures, Steph and I scared a truck driver this morning on the way back from taking the kids to school. On our way up to school Steph noticed a Guinness truck parked at the local Canyons and idly commented that it was too bad we didn’t have our camera with us, since it would be funny to get a picture of me in front of the truck.

Enter the geektoy — the Qtek 9100 Windows Mobile 5.0 PDA and cell phone, with integrated 1.3 megapixel camera. You, being an astute reader, can probably see where this is going…

So that was a good start to the day. It was book-ended by what may be a quite awesome ending to the day — we’ll see if things pan out the way I hope they do.

[Ed: "What's an Auspex?" I hear many of you say. "Why the heck are you wasting two posts about this stupid thing?" Well, Auspex was a company in the 1990s that made very high-end file servers. The model I had, an Auspex NS5500, cost around three-quarters of a million dollars when it was manufactured and sold in 1990 and represented the pinnacle of non-mainframe computer engineering. It was a 7-foot tall, 700lbs. black metal cabinet chock full of computery file server goodness. At one point I contacted Auspex to find out what it would take to refurbish it and the tech I was talking to exclaimed, "Oh! That's where that one went!" You see, these were the kind of computers you didn't just buy, use for a couple of years, then throw away; they usually got traded back to Auspex as part of an upgrade cycle. The story about how I came to own it will have to wait for another day.]

I still owe y’all a review of my new Qtek 9100 PDA/cellphone running Windows Mobile 5.0, but in the meantime, as I’m working on some other WM5.0 projects, I wanted to share word of what is turning out to be a potentially huge problem with the new WM 5.0 devices, depending on which devices you buy and which carriers you get them from. This post from the Windows Mobile Team Blog, “Adding Root Certificates for Exchange Activesync”, doesn’t provide a lot of detail and background, but the comments give you a bigger picture of how much trouble this is causing folks.

The base problem is that some WM5.0 devices, depending on how they’ve been configured by the OEM (in many cases the carrier that sells the device), do not allow end-users to install additional root certificates. This is great if you’re only using SSL certs from a major vendor, but if you’re using self-signed certs, this becomes a problem. The Windows Mobile 2003 workaround of disabling SSL for EAS isn’t an option in WM5.0. Normally, I’d be happy about how the use of SSL is enforced — I’ve never advised using OWA/EAS/OMA over unencrypted connections — but in this case, a lot of people are frustrated because they can’t use their new devices to synchronize with Exchange. Since that was the only real functionality I was interested in for my Qtek, I have to say I’d have been extremely unhappy to find out I wasn’t able to do it.

The moral of the story is simple: be very, very choosy about which vendors and carriers you buy your WM5.0 devices from. Insist that they either provide the tool you need to install your own root certificates (if you don’t get management access to the device) or insist on having devices configured so that you have management access.

ISA Server 2004 Service Pack 2 is now available for download from Microsoft and has just gone into deployment on my home network (which I often use as a test network). Aside from the expected hotfix rollup, there’s a couple of nice new features. The SP2 biggies for me:

• Support for proxying HTTP compression. Not only does ISA support serving compressed content from its cache, it also requests compressed content from upstream sites. This is very good news, as I’ve been eyeing the necessity of enabling HTTP compression to help deal with the traffic for the Community Server and DotNetNuke websites I host.
• BITS caching allows ISA server’s cache to download updates from Windows/Microsoft Update once, then serve additional requests out of cache. This is great news for small office/branch offices that aren’t using WSUS. Updates only travel the WAN link once.

If someone at Microsoft wanted to make my life really easy, they’d include an updated auto-discovery wizard for building site-to-site IPsec/L2TP links. So far, every time I’ve gone to create one, I end up spending a couple of hours troubleshooting, then rolling back the configuration changes in disgust. ISA 2004 is an amazing firewall, but there’s still much goodness to be committed.

Of course, maybe it’s coming in theISA Server 2006 Beta. The ability to publish multiple Exchange and Sharepoint servers looks very interesting. I’m definitely going to have to give that a download one of these years.

So far, this has been an eventful, confusing, and difficult year. It hasn’t been all bad, of course, and when it has been good it has usually been pretty good — but there’s been a lot of crap to cope with.

On top of the big issues, today I just dealt with one of the smaller, more bittersweet issues. During my time at Premier1, I became the proud owner of an Auspex NS5500 server. This was seriously high-end hardware back in 1990 and it was an awesome piece to have in my collection. Unfortunately, I’ve never been able to fire it up and use it. I’ve been looking for a good home for it for the past couple years, after finally admitting I wasn’t going to have the time or energy to get it running again.

Just a few minutes ago, I helped its new owner unload it into his garage. Unlike me, he will have the time to get it restored and running. I am glad that I found a fellow enthusiast….but I miss my Auspex. I joked about getting visitation rights to cover up the fact that with all of the other goodbyes I am forced to say, this feels like one too many.

Microsoft’s Intelligent Message Filter (IMF) add-on for Exchange Server 2003 is one of the niftiest free downloads you can put in your Exchange organization that will have direct effects your users can see. However, I’ve seen a few questions and issues popping up regarding IMF that tell me there’s a mild bit of confusion out there. Let’s straighten it out.

There are two versions of IMF. IMF version 1 is the version you can download separately. It’s got the basic message filtering goodness built in, but lacks a lot of other nice features. IMF version 2, on the other hand, comes with Exchange Server 2003 SP2; you can’t download it separately and you can’t choose it as an optional component. It will automatically get installed when you apply SP2. Please note that you cannot downgrade to IMFv1 after installing SP2 and have things keep working. If you do install IMFv1 on a server that already has SP2 applied, you’ll need to reinstall SP2.

IMF doesn’t automatically upgrade from v1 to v2. It would have been nice if the SP2 installer did this for you, but you must manually uninstall IMFv1 from your Exchange server before you upgrade it to SP2.

IMFv2 is not automatically enabled. No matter your config before SP2, IMF will not be enabled by default — which makes sense, since you don’t need it running on every server in your organization. Happily, it’s easy to enable it on a per-virtual server basis. You do it from the same property page that you enable connection filters, recipient filters, and sender filters from on the SMTP virtual server. Heck, you also enable Sender ID here.

IMFv2 is worth upgrading to. If you haven’t upgraded to SP2 yet, why not? Beside all the other great features, IMFv2 includes built-in Sender ID integration, the ability to modify the word list so you can customize the filter, and auto-update capabilities.

Update: I talked a bit about some other IMFv1 to IMFv2 upgrade issues in my earlier post about SP2, so check it out too.

As part of my ramp-up for Exchange Connections in April, I’ve been working on my slide deck for Sender ID. I’ve got the slide deck written; now I need to run through it and polish the presentation.

One of the interesting pieces of information I came across concerns a bug in the Sender ID implementation in Exchange Server 2003 SP2. (As an aside, I love the Exchange community. A very kind gentleman by the name of Mike Ruman saw that I will be doing a session on Sender ID and sent email to Paul to let us know about this issue.) At the time he sent me the information, a hotfix was available but the KB article wasn’t yet available. As of February 1st, Microsoft published KB 910272, “The ‘Sender ID Filtering’ feature does not work correctly in an Exchange Server 2003 SP2 server.” If you’re using Sender ID on Exchange Server 2003, you should install this hotfix.

From the details I was given, this bug manifests when your policy uses the – (Fail) mechanism prefix combined with a reference to another domain. According to the KB, it can result both in false positives and false negatives, so this should probably be considered a mandatory hotfix for Sender ID.