In search of simple software
All these years, I've been saying that one of the huge differences between software architecture and structural architecture is that nobody tries to change the requirements for the building while it's being constructed.
Turns out I was wrong:
OSU Medical Center stretches hospital tower to fit $100M cancer center.
I found this sort of amusing, though I'll be the first to admit that the humor will be lost on some people. This is a clip from a web site that's promoting an infrastructure component that's supposed to be enterprise-quality and highly available. While I'm sure this is true, it's also true that having a web page that doesn't load won't inspire tons of confidence. Oops.
(yes, I blurred the text a bit -- no need to pile on for what's certainly an isolated problem)
Last week, I did a little work to track down unused stored procedures in a large-ish SQL Server database. Searching for references in source code was interesting all by itself, but I also wanted to see which procs were referenced by other procedures. This is what I came up with:
select o1.name 'referencing proc', o2.name 'referenced proc' from sysdepends d, sys.objects o1, sys.objects o2 where d.id = o1.object_id and d.depid = o2.object_id and o2.type = 'P'
According to an article on ReadWriteWeb, Q&A site Quora is blocking search engines from indexing its site -- all, that is, except for Google, Bing, and a couple other big players. If this is true, this is a really chilling development for the competitive landscape of the Internet.
Google is plenty big enough all by itself -- just look at how difficult it's been for Bing to claw out any sort of market share at all. I'd really like to think that the market is still open to someone who wants to build a better mousetrap, but this sort of behavior is going to make it just about impossible for a new player to break into search.
So take a look around -- this may be as good as search gets.
I got an email this afternoon from "VMWare Communities." You've seen emails like this, I'm sure -- here's your status, and some stuff you might be interested in, and so on. Emails like this are generally intended to strengthen a connect with customers, which is great -- normally.
In this case, though, there was a section of the email entitled "Your Content":
Activity around content you've created or contributed to
Windows Server 2008 R2 64-Bit freezes. was viewed 4,821 times and replied to by 17 people
Strange Problem with the VMware Remote Console Plug-in was viewed 511 times and replied to by 4 people
When I saw this, I was pretty excited, because I'd forgotten (for the time being) about both of these issues. I'd posted questions in VMWare's forums for a couple of issues, and I hadn't recalled seeing solutions for either of them. According to the newsletter, though, there were updates available on both questions. Happy day!
Until I went and read the forum topics.
It turns out that VMWare's little newsletter-writer counting tool was counting *all* views and *all* replies, so in both cases, it turns out that my post was the last one in the topic. Worse, yet, my posts were six and nine months old, respectively (no wonder I'd forgotten about them). So instead of reaching out and building ties with me, in this case, VMWare merely succeeded in reminding me that their open issues aren't being fixed, and that their support forums are effectively dead.
"You break up...call back in six months."
I recently helped a friend build a new PC, and I threw a few notes together along the way. If you've never built a PC from scratch (or haven't done so recently), you may find some of these things helpful. Note that the focus here is on building a high-end business machine or a developer workstation. I'll try to point out some places where you'd bump up specs for a gaming machine, but there are really plenty of other guides out there for "ultimate gaming machines".
Since we're building a business machine, we're looking for good performance without breaking the bank. The notes below are organized by component; if you work your way through the list, you should end up with a pretty solid build.
Case: A mid-tower gives you plenty of room for all but the most extreme expansion. Some crazy-big video cards might be tough to fit, but there's no reason to buy those if you're not building a gaming PC. Compact cases are nice & tidy, but might have limited drive bays, and probably won't be as easy to cool.
Styles range from conservative to gaming styles with all sorts of windows & lights. Plain black should be fine. Some will come with power supplies, and may be a bit cheaper as a result. Unless you know what kind of power supply is included, though, you might be better off getting that separate. Power supplies, believe it or not, are actually pretty important to having a reliable, quiet system.
Be sure to check out how many internal 3.5" bays the case has. A case with four or more internal bays gives you plenty of room for expansion, and usually has good options for cooling the drives. I've found that it's helpful to put a fan on hard drives, so look for a case that will facilitate this.
Good brands: Antec, Cooler Master, Thermaltake.
Target price point: $50-80.
Power supply. For most new system builds these days, 400W is just about entry-level. If you start adding extra drives and gaming video cards, you can get up in the 650-750W range. Typically, you won't need anything bigger than that unless you're going to do a gaming build with run dual video cards, overclocking & water cooling.
Look for an "80 Plus" or "Bronze" energy certification (or better). "Modular" supplies let you plug in only the leads you need, so you have fewer wires in your case. This is a nice-to-have, but not critical unless you plan to gaze lovingly at the guts of your PC. Extra wires can be tucked out of the way so they don't interfere with cooling.
Good brands: OCZ, Thermaltake, Corsair, Antec (except for low-end), Raidmax.
Target price point: $60-90.
Processor. I'd stick with Intel here. I think AMD still gives you a slightly better value, but not by much, and I think Intel's performance is a little more consistent. This choice is pretty tightly connected with your choice of motherboard and RAM, as they'll all need to match. Skip Celerons (Intel's entry-level series) and Core 2 (the last-generation mainstream processors). A Core i3 or i5 processor will be perfect for this sort of build, though. If you're really concerned about processor speed, you could go for an i7, but an i5 should be fine for most development needs.
You'll see references like "Manufacturing tech: 32nm". Newer tech is smaller, so 32nm is newer than 45nm. This isn't a big deal, but all things being equal, the smaller components will be more energy (and heat) efficient. If you see power consumption listed (ex: 73W), a lower number will run cooler and cheaper. For a development machine, get a quad-core processor.
Target price point: $150-250.
CPU cooler. When you buy a retail ("boxed") Intel processor, it'll come with an Intel heatsink and fan. In most cases, this will be fine. An aftermarket CPU cooler can cool better and run quieter. I'd skip this unless you really want to shoot for a super-quiet PC or you want to overclock the CPU.
Motherboard. The CPU socket on the motherboard must match the socket of your processor (ex: Socket LGA 1156). Not all Intel CPUs have the same socket, so be sure to check this. Other features:
Good brands: Asus, Gigabyte.
Target price: $80-150.
RAM. You'll want at least 4GB or more -- probably at least 8GB for a development machine. Dual-channel is fine -- that's what you'll get on mainstream motherboards. You'll probably want to look for DDR3.
Good brands: I like GSkill, but Kingston, Crucial, and Patriot are also good.
Target price: $75-90 for 4GB, or $185-ish for 8GB.
Video cards have chipsets by either ATI or NVidia, and the chipsets are so much a part of the card that a card made by Evga is pretty much the same as a card by Gigabyte or another manufacturer as long as they use the same NVidia chip, for instance. This is true to the extent that you'll actually get drivers from ATI or NVidia rather than Evga, Gigabyte, etc. It's tough to call whether NVidia or ATI are better. I think ATI is pretty hot right now, especially in the mid-market, but NVidia was the best for a number of years. You can toss a coin on this one.
Pay attention to the connector(s) on the back. Many now give you an HDMI connector, but you'll want to make sure your monitor can support this. If the card also has a DVI, you're pretty safe. Dual-screen support is a near must-have for a development box. Look for at least 512MB of video RAM. GDDR5 is faster than GDDR3, which is faster than DDR2. You should be able to find 512MB of GDDR5 at a pretty reasonable price point. Look for DirectX support for at least 10, but 11 would be better.
Good brands: Pretty much not applicable, per discussion above. Focus on getting the chipset you want.
Target price: $80-150.
Storage. The best setup for a new build today is a Solid State Drive (SSD) plus two or more data drives set up as mirrored pairs (RAID 1). The SSD gives you a fast boot time and fast loading for your applications, and the hard drives give you redundant storage for your data. You'll want to use Serial ATA (SATA) drives for all storage on a workstation -- avoid IDE (old and slow) and SCSI (server use only).
SSD drive. This tech is actually still changing pretty rapidly. As these drives have matured, their performance is characterized by the drive controller chipset they use (much like video cards). Right now, the best overall controller chipset is the Sandforce chipset, so any Sandforce-powered drive will tend to perform similarly.
Good brands: Intel, Kingston, OCZ.
Target price: $150-250.
When buying hard drives, it's worth looking at the warranty. Most will be three years, but if you can, find one with a 5yr warranty.
Good brands: Samsung has actually been getting really good reviews for the last couple years. Hitachi, Seagate, and Western Digital are also about the same. I know I swore off WD a few years ago, but I've seen as many failures in Seagate as I have in WD. All hard drives will fail -- that's why you're going to mirror yours and get one that's got a good warranty.
Target price: $60 for 1TB, $90 for 2TB.
NAS / Drive Arrays. If you've got big storage needs, you're going to need to build something more capacious than a single mirrored drive can provide, and at some point, you're going to end up looking at Network Attached Storage (NAS) and/or drive arrays. You can purchase something like this (ex: Drobo) or build your own. Up until a couple months ago, I would have put a Windows Home Server box on the short list for your NAS needs, but given the recent hubub over their forthcoming VAIL release, I'd be reluctant to jump on that bandwagon right now. Although focused mainly on WHS builds, there's a great forum on mediasmartserver.net that's got information on all sorts of DIY storage server builds if you're looking for ideas and/or inspiration.
Optical drive. You'll want some sort of DVD reader / writer, and for a few bucks more, you can get Blu-ray, too. This is a good place to save a couple of bucks if you don't need Blu-ray.
Good brands: Lite-On, Asus, Samsung, LG.
Target price: $20 for a DVD writer up to $100 or more for a Blu-ray burner.
Floppy drive. Skip it.
Extra fan(s). I'd recommend installing an extra fan to cool your hard drives. I've had better drive reliability on my system since I put a fan on them.
Final note: Since the shelf life on this sort of guide is pretty limited, you'll probably find that specific recommendations for brands, chipsets, prices, and so on will go out of date pretty quickly, but I hope that the general concepts here will apply for a while yet. Good luck!
As an architect, I'm sometimes asked to evaluate and/or choose platforms for software, web sites, and so on. As you might imagine, choosing a platform can be an enterprise-critical decision, and one of the big forks in this road is whether to consider open-source software.
I'm not going to touch that debate here - this tends to be a pretty religious point with a lot of people, and I'm not sure I can add too much value to a religious debate. Instead, I'm going to assume that you're at least open to open-source software and you'd like to evaluate some alternative open-source alternatives.
The factors you're evaluating in an open source platform really aren't too different from the factors for a closed-source platform:
Without a doubt, the most important factor of all for open-source platforms is the size and vitality of the community, which ends up being pretty vital for the long-term viability of the project. While it's certainly true that you've always got access to source code if you need it, you really don't want to count on that if you can avoid it. A large, active community implies a bunch of good things for the platform:
One last thought: consider where the platform gets its funding. In some rare cases, open source projects are funded entirely by donation, but when organizations reach a certain size, there really has to be a plan to pay the bills. Watch for "freemium" models, where certain functionality might be available in a "community edition" of the platform, but the latest and greatest stuff is available only to paying customers. This certainly isn't a deal-breaker, but understand that you're going to have to wait in line for new stuff, or deal with an upgrade to a paid version at some later point.
Matthew Groves has been working on a series of articles comparing some OR/M frameworks. So far, he's done some posts on Entity Framework and Fluent NHibernate, but I believe he's planning on hitting a couple more, too.
If you're a .Net developer, you know what a "lively" subject this always is. Data access consumes so much time and effort in most applications that it's worth educating yourself on the options so you can make informed decisions. I'd also recommend checking out a couple of these frameworks yourself, as merely reading about these frameworks isn't quite the same as using them on a project.
Here's a list of Matthew's posts so far:
This one is definitely worth a read:
Be sure to check out the Windows 1.0 press kit that he scanned in. I can't believe it's been 25 years, and Microsoft has managed to add -- what -- Direct-X and muti-monitor support, I think. I love how "Paint" is a big selling point, and CD technology is broken down like it's a flux capacitor or something.
Ahh, simpler times.
You've probably heard of Blogger. It's a (duh) blogging tool, and it's one of the most popular platforms available for casual bloggers. It's simple, easy to get started with, and it's hosted (and owned) by Google.
You know what else Google does? They build a browser called Chrome. Perhaps you've heard of it. I've heard of it. I've been using it for the last couple of years, and it's generally really nice. 'Cept here's the good part -- it doesn't like to work with Blogger.
I happened to see a new blog post from Roger Sessions today -- it popped up in Google Reader (yeah, I drink the Kool-Aide), and I clicked over to his site to leave a comment. So far, so good. When I clicked the link to comment, however, it popped up a comment form in a popup window (minor annoyance, by the way), but then, I noticed that there was a Captcha-style image verification panel that wasn't actually showing an image.
Hmmm....
This has got to really cut down on spam comments, right? Intent to get my message to Roger, though, I opened his blog in Firefox and clicked the comment link -- it turns out Google likes Firefox better than Chrome:
Looks fine in Firefox, doesn't it? Incidentally, I also tried in Internet Explorer and saw the same problem I saw in Chrome, but then, I hit the link a second time, and it was fine. Having hit upon a possible solution, I tried again in Chrome, and sure enough -- it worked fine in Chrome, too, the second time around.
I'm not sure what the problem is, here, but it's pretty telling that Google hasn't seen fit to update the commenting system on Blogger. After all, the native commenting system in WordPress has undergone continuous improvement since it was launched, and it's also stupid-easy to integrate commenting systems from folks like Disqus (which I use) and Intense Debate.
If you're a blogger, you may want to look at giving Blogger's commenting system the boot in favor of one of these other systems until Google gets its act together.