July 2008 - Posts
Quite the debate going on from today's editorial: Stored Procedures Reconsidered. I chimed in late, and going on the "pro-sprocs" group. I think stored procedures are helpful, and a better idea, and most importantly:
Don't take that much more effort
You should be able to auto-gen trivial procs, and you should be able to auto-gen the stored procedure calls from .NET. Actually that's a little utility that we've considered writing before. Something you can point at a server, give it a proc name, have it sniff out the procedure parameters, types, etc. and write the code. How hard can that be?
I don't think that stored procedures are required everywhere, but I think they do provide benefits and ignoring those benefits because you don't want to write them (laziness) just doesn't make sense.
At least this project is moving forward. We deployed a new design to the site this week, and turned it live this morning. Instead of altering the Dot Net Nuke framework that it was on, we built a new site, mostly because it's a fairly simple type of site and really is based on the newsletter only, not people visiting the site.
There are a few minor issues, but they're mostly on the administrative side, not the public side. The front page shows the news from the previous week and you can click links to follow them. On the admin side...a few issues, none really major, although without developers this weekend I suspect that the newsletter will be a day late.
I have to say that Database Weekly, evolved from Database Daily, has changed somewhat in it's mission. I was against the idea of the site 4 years ago, but Andy convinced me to give it a try. He said that there was value in having an editor pick through news and put the newsletter together and he was right, I was wrong, and it's a valuable web property that added value to the bottom line. Our plan was to eventually try and go daily ot twice a week with the newsletter, but we never go the process down to that level.
When Red Gate purchased the site, they decide that Database Daily was too confusing (I agree), bought Database Weekly, and now we have a new site. Eventually I'm hoping we find a way to automate most of the functions of DBW, or I get to hand it off to someone else. It helps for editorial ideas, but it's also a PIA each week and now that I'm stuck with it for the foreseeable future, it's a chore every week to walk through hundreds of links.
We've had a project underway to upgrade these blogs, moving to a newer version of Community Server, better skinned, etc. The project is scheduled to deploy in a couple weeks.
A couple weeks ago, I heard the project would be ready in a couple weeks. I've heard the same thing for months, possibly going back to the beginning of this year. At this point it's gone on so long that my memory is failing a little on when we even started it. It wasn't my project, an outside consultant was used because all of us here in the publishing group are fairly busy, and our consultant appears not to have put this as a high priority. To be honest, I'm not sure how high a priority I put on it as well.
I've seen a few posts (part 1 and part 2) on blogging from Andy Warren lately. I like his assessment of blogs, and we've debated offline quite a bit on the value of blogging. Early on Andy wasn't sure there was value, but he decided to try it and commit to a year to see how it went. I think he sees some value in it, but it's a nebulous value. Most blogs are fairly low exposure unless you can write something sensational, or it becomes your publishing arm. But if you have another publishing arm (like SSC), then why blog?
My view is that you should have a professional blog that you use to blog about work. What you've done, what you've read, what you think of things. This is in addition to any publishing someone pays you for. You want to blog regularly, at least a couple times a month, and show what you are doing. It becomes part of your resume and something you can use in interviews. If you want to blog about your life, band, cat, whatever, do that elsewhere.
So do we need blogs at SSC? It's a nice add-on, and I think there is some value for people, but I'm not sure we get a lot of value out of it. It's almost a good karma thing for us to offer this as a free service to people, let them blog. Of course it has to work, including offline posting using the various blog APIs from LiveWriter, Word, etc.
However with the time, $$, and resources this has consumed and continues to consume, I'm not sure if we're not better off getting some type of "group blog" at Blogger or something. Hey, maybe I need to check on that :)
The other day I fired up my old desktop, intending to use it as background in the podcasts. It gave me a "Activate Windows" error for the W2K3 installation on there. I know I was thinking that I would use it as a file server, but downstairs it has no Internet access, and I'm not in the mood, so that will go back to XP.
What does that have to do with VMs? Nothing really, but it was foreshadowing for me. Today I went to build a new VM to hold RC0 for SQL Server 2008, which will let me do some tech reviewing of a new book. Since I hate installing the OS, it's simple, it's the same every time, and it tends to waste my time, what I did last year was install a copy of W2K3 as a VM, shut it down, and copy it to a file called "BlankW2K3". Now I have a blank VM and building a new one is as simple as copying the files over, renaming them, and then starting them up. I did that today and ...
"This copy of Windows needs to be activated"
Grrr, not what I wanted. Especially since I didn't get a W2K8 VHD downloaded last week because bandwidth was moving slow around here. That's something for today.
I typically install TechNet or MSDN versions of the OS's for testing. I have licenses for them, so that's not a big deal. My VM has internet access, so I activated and then moved on to installing SQL Server 2008. Unfortunately it didn't go smoothly.
The first thing needed is the .NET Framework 3.5. And that needs SP2 on W2K3. Enough acronyms for you?
So setup shut down and I set about installing SP2. In this case, since we're far enough along in the process, as soon as SP2 downloads and installs, I'll shut down the VMs, copy the files over and rename them as W2K3SP2, and then bring my new SS2K8 VM up and put RC0 on it.
It's going in the background now, aren't VM's great? If this was my desktop I'd be done for now.
I signed up for the SS2K8 beta exam, the maintenance and implementation one, and had it scheduled for this morning. I didn't really prepare or study at all, counting on the limited work with CTPs and Betas, as well as reading articles to help me out.
The beta format is 70 questions, 3 hours, and then a 30 minute comment period. As with most other exams, it took me about an hour, with about 25 minutes for comments, since I like to give feedback on the exam.
The exam wasn't too bad, definitely a few things I wasn't aware of with spatial indexes and monitoring performance. I took issue with a few questions that I thought were poorly worded, not giving enought detail, and a few others that were looking for paths to files. Why ask about paths? Those are silly memorization questions that don't require skills.
There were some good questions as well, some that made you think, especially with backup and recovery scenarios and I gave positive comments on those.
I was surprised by a few things, like the Extended Checks for DBCC, and the Resource Governer questions, and came home to look things up.
However I'm a little concerned about the exam process. This is the first time I've taken a beta exam before the product was released. I think that's a mistake as there are bound to be problems and issues identified in the first few months that will be fixed in SP1 (at RTM + 6 months) and should be incorporated into the exam. Also, the exam writers are people like me and I'm not sure I could devise a good exam without spending more time with the product.
I missed it last year, but I just booked my flight for the Business of Software 2008. It's being held in Boston on September 3rd and 4th. I'm flying in on Tuesday, arriving at 4pm, dropping my bag off at the Hotel (I'm going to be at the Seaport) and then grabbing a cab over to Fenway Park for the Baltimore/Boston game.
I've been to Boston quite a few times, even rowed in the Head of the Charles, but I never saw a baseball game there. I've walked around Fenway, so this will be a treat, possibly even more than the conference itself.
Neil Davidson, co-CEO of Red Gate, my employer, started this conference last year with Joel Spolsky and I thought it was a fantastic idea. I had planned to go last year, but then had some conflicts with my wife's schedule and was just dreading travel after flying to the UK over the summer. When I saw he was putting it on again this year, I was begging quite early to come. I offered to blog during the event, and must have done a good job since I talked my way in to this year's event. I'm still not thrilled with the traveling, but I am excited to to go Boston and I'm looking forward to the event.
I've worked for 2, technically 3, software companies over the years. First was JD Edwards, then they were purchased by Peoplesoft, and now Red Gate. However in each case the business of software wasn't my business. For the first two companies I worked in the internal IT group, managing production servers and people as I did for most of my career. Now I talk to the software people at times and we try to work with them on software releases, but we don't do a lot in terms of what the software business means.
All my other jobs were producing software for the internal company to use or for our clients to use in a controlled environment. So much of the technical issues with widespread clients, as well as the sales and marketing efforts of shrinkwrap companies is lost on me. Or it wasn't necessarily. I've learned a few things from working with people over the years at SSC, but for the most part I'm a novice.
When I was younger I wanted to write software. I've written various utilities and pieces of software for myself and others, but wasn't great at it. Over the years I've learned that you don't have to be great, and that you need to be effective and market well. I've also learned that support calls are a pain and I haven't wanted to get into that business. Over the years we've debated about various pieces of software that might make sense for SSC to sell, but we never got anywhere on things since our primary business was building a community.
I think software is a hard business, and there can be a lot of hassles, but as I watch one of my business partners try to sell his software company, I rethink that there might be something to building software. Especially as we have a programmer and other resources and could potentially build something profitable.
So I'm interested to see almost everyone at the conference. From experienced software developers like Joel Spolsky and Erin Sink to Richard Stallman's unique view to the marketing and money people that will be there. My goal is to grab lots of notes and then think a lot about what the business is like.
If you want to come, ping me and we'll definitely meet up for a beverage of choice.
It's been a long time since I saw a vulnerability in SQL Server released. And to date, there haven't been any for SQL Server 2005, nearly 3 years after it's release in Nov 2005.
That changed today I saw a note from Secunia just some in: Microsoft SQL Server and MSDE Multiple
Vulnerabilities, which lists a few issues with SQL Server 7, 2000, and 2005. That's a major release, and I'm surprised.
I shouldn't be, after all there are security issues with all software and I knew that one would come in SQL Server 2005 a tsome point. I'm just hoping that this is fixed in 2008 and that we don't start to see more and more.
I went away in the mountains for July 4th to a camping facility with the family. I was an unwired weekend, no computer, limited media, and a few books I'd packed along. And it was the first time that I wished I'd had a Kindle with me and it wasn't convenient to go get another book.
Here at home I can always drive to the bookstore or library, going down there quite a few times a week, plus I have a large collection of paperbacks and books that I can always re-read or dig into. However at the campsite, no books for sale, none even close and the truck was connected to the trailer quite a bit. Plus packing up to go somewhere wasn't necessarily the best social thing to do with kids around.
I'd taken 3 fiction novels, all of which I'd already started, a technical book (Grant's Execution Plan book) and a landlording book. I finished off the first three on Friday, which left me Saturday with no books. Not exactly the best situation for me to be in. I started Grant's book, getting about 80 pages in and while it's well written and I learned a few things, it's not a relaxing, enjoyable book. It felt like I was working and I kept thinking of Question of the Days based on his book. Same for the landlord book.
Now it was a good and bad thing. I was forced to read some stuff I might put off for awhile, but I also didn't enjoy the vacation as much as I would have liked. If I'd had a Kindle, I would have been able to purchase a book or two to read when I ran out. Not to mention the lack of space for packing up books.
However I've been emailing with Ted Pin, and he bought one for his wife. They were ready to return it, not because they didn't like it, but because she was buying multiple books a week. Eventually Ted decided with his wife to set a budget for books and keep the Kindle. I think I might need that as well.
Tia and I talked about more vacations like that and about being on the road once the kids grow up. I think that a full-time, or often on the road life would definitely be great for the Kindle. I could get books as needed and not have a need to store them. Of course, I'm sure I'd need a budget and I'd probably need to find a way to get older books on there, maybe even moving to try out more free or "borrowed" content.
I saw an interesting note about how Intel is warning developers that they need to learn to program with multiple cores. They're saying 100 cores on a chip isn't far away. I'm sure there's a lot of truth here and it worries me. I think we'll see some really poorly build applications as most programmers struggle here. It's probably similar to what we saw when people tried to move from a command line / DOS environment to an event driven model like Windows. I know I had issues learning some concepts there.
I'm wondering how much this would apply to DBAs, however. We have an OS (SQLOS) on top of another OS, and together, these two systems need to manage the parallelism and threads we use. In most cases, we don't care about threads, though in SQL we sometimes want to limit things.
Moving to a parallel, multi-threaded world will be tough. I struggled in college and barely understood some of the concepts, though to be fair, there weren't a lot of places to practice and it seemed to be a very specialized skill I didn't want. Now I wish I'd done a little more work, but perhaps it doesn't matter. I can go on writing my SQL queries and let the SQL Server platform manage the cores.
As expected, I forgot someone again in my MVP announcement. Last time I announced Kathi Kellenberger and Andy Leonard, forgetting Jacob Sebastian. This time I announced Michael Coles and Jeff Moden, but forgot Gail Shaw! MS doesn't make it easy to figure out who's renewed, so I kind of depended on somone to ping me, which Michael and Jeff did.
I've corresponded with Gail for a few years, and finally met her at the PASS conference in Denver last year. She's a great SQL wizard and has really helped site grow over the years.
Congrats! and apologies.
Today's the day that new MVPs are notified for Q3 and two longtime community members, Jeff Moden and Michael Coles, were awarded. I didn't see announcements on who dropped off, but I know of three people I see in the community often that are new.
Jeff has answered so many questions, and provided some great advice, in addition to quite a few articles, so I'm not surprised that he got it. I nominated him and definitely think he's an MVP in the SQL community.
Michael has written quite a few articles for me and worked on 2 or 3 books in the last year, so I'm not surprised at his award either. He's a smart guy in NYC that's taught me a few things along the way.
I'm up for renewal in Jan and I have no idea if I'd get it again. I have to just do my thing and see what happens.