Archive for March, 2004
I ran across a link to this flash animation. At most sites, I’m annoyed by any flash animations. This kind of animation I can deal with.
Ned Batchelder posts about punishing programmers that break the build.
In our shop, we don’t have a daily build. However, each developer frequently refreshes their code with the latest from SourceSafe. Occasionally, somebody checks in code that won’t compile. This causes unnecessary frustration. I can understand the urge to form a lynch mob when this happens.
A number of weblogs that I read have been posting maps of where they’ve traveled. I’ve not been a world traveller, but I’ve been to most of the US. Here’s my map.
You can make your own map here.
Paired programming is an aspect of extreme programming that many people find hard to swallow. I must admit that I am somewhat skeptical of the practice.
However, this article about paired programming at the Software Factory (why I am I hungry for cheesecake?) is intriguing. I haven’t read through the whole thing, but it certainly asks some interesting questions.
I do believe that pair programming produces better quality from the outset. Even without a formal process, I think every software shop has ad hoc sessions of pair programming when two developers are looking over some troublesome code. Whether I’m the one at the keyboard or the one looking over the shoulder, the discussion always involves analysis that wouldn’t have occurred otherwise.
The other idea that has me interested in this article is the question about focus drifting away from the necessary task. When I am working alone, I do tend to drift away from the task at hand. It’s hard to concentrate on difficult tasks for an extended period of time. Programming well is a difficult task. I can understand how having a partner will help keep me on task.
In school, we occasionally did group projects or pair projects. I wonder if the output of these projects was better than what would have been done if the students worked alone?
This Slashdot post points out that the REXX language has been around 25 years. I note this because it allows me to fondly reminisce about the days when I operated a Fidonet mail hub. This was in the days of Windows 3.0.
Windows was not effective on a 386 for running a BBS and writing papers for school. I used DESQView for a while, but it was prone to crash as well. A fellow SysOp pointed me to the wonders of OS/2. Specifically, OS/2 2.1 started me on a lengthy relationship with IBM that I didn’t abandon until the Internet became more prevalent and BBSs became a memory.
After finding all the OS/2 versions of the necessary software, which were readily available in those days, I discovered the joys of REXX scripting. I used REXX to automate all the management tasks of the BBS and Fidonet systems. I was using IPC and pipes to launch and communicate between tasks. It was magical stuff at the time.
My REXX skills are nonexistent today. But I can still say — Those were the goold old days.
David Berlind, a columnist for ZDNet, writes about a personal DDoS attack that took down his cable modem connection for several hours. This was an attack perpetrated by an administrator of an online gaming server. The article doesn’t say whether the server was a commercial server or run by a private individual. I sure hope a commercial gaming service wouldn’t tolerate such behavior.
On what I think is a related topic, this Slashdot posting asks the question, “Broadband Access Leading to Internet Breakdown?” The argument is made that with more internet connections using higher bandwidth solutions, the infection rate of viruses and the delivery rate of spam also increases, which possibly explains why these problems seem to be on the rise.
On yet another related note, a somewhat disputed survey claims that 3/4 of the United States population has internet access. This access isn’t limited to the home. People were counted if they said they had access at home, work, school, or a library. Some people have access at all four locations. Do they get counted 4 times?
I thought it interesting that all these reports came out within a day or two of each other. I don’t think there’s a conspiracy afoot. I just think they were timely and they all point to a growing problem. Specifically, I think that behavior on the internet is becoming like a mob mentality. When mobs form, the old saw about a few bad apples spoiling the bunch rings true. People will do things in a group environment that they would never do on their own. I think that part of this is because people lose their individuality and identity in a group setting. This loss of identity releases inhibitions because they are not acting as an individual any longer.
You can see this behavior in small behaviors. Rush hour traffic, when it’s not hampered by accidents, travels at a faster speed than other times of the day. The loss of inidividual identity ties into this. People rationalize that they probably won’t get stopped by the police since everybody else is driving so fast. If you put those same people on the same highway, but with nobody else on the road, most of them will drive more slowly.
The mob mentality on the internet does not come from people being physically gathered together. But the internet does promote a loss of individuality and identity. Most internet activity is practically anonymous. Yes, as the RIAA has shown, people can be tracked down. It is not easy to track down rogues, and even when they can be tracked, it often doesn’ matter.
David Berlind’s DDoS attack came from an individual that he could track down. Yet he has little recourse for the damage that was done. The relative impunity of inflicting an attack surely encourages the perpetrators. It’s much like the kids who TP a house or turf somebody’s yard. They know they probably won’t be caught.
The Slashdot posting doesn’t mention it, but having higher bandwidth on a lot of computers makes it easier to execute a DDoS attach. A couple compromised computers that are attached to a T1 line can easily overwhelm a cable or DSL line.
The increased availability of computers, as surveyed by Harris Interactive, means that there are more potential computers that can be compromised. We had a court case in our area a few years ago that involved a number of young men who repeatedly forced girls into the basement of a home and would rape the girls. When the boys were caught, it was discovered that none of them lived at the house. The owner of the house said he had no idea what was going on down in his basement.
The owner’s story was hard to believe, but that same scenario is playing out on computers all across the internet. Rogue software has been installed in the “basement” of numerous computers without the knowledge of the computer owner. This software is used to carry out DDoS attacks, spread viruses, and send spam.
What can be done about it? Like most societal problems, I’m not sure that we’ll ever have a thorough solution. Will it hurt the value of the internet? It’s something that I might start worrying about.
Erik Porter is looking for a convenient way to sync databases. In the comments, some people have mentioned AdeptSQL, SQLDiff, and SQLDelta. These products range in price fom $225 to $300 for a single user.
In our office, we use SQLCompare. We’re still using version 2.x. Version 3.x is supposed to be much, much faster.
We’re very pleased with the version 2.x product. I’m sure the 3.x product is even better!
Stuart Laughlin writes that he was blocked from Google by the corporate proxy server. Like many developers, I think we only need a few things to conquer the world (or at least conquer that pesky bug that’s been nagging at us for the last two hours). I can sum up the list like this:
- Internet connection
- Caffeine (by any means)
- Access to Google
Nice to have, but optional, are things like a desk, a chair, lighting, and solid food.
I list Google right up there because it is indispensible when researching computer-related issues. Chances are, if you’re having a problem, somebody else has had the same problem. Learning the art of advanced Google search techniques is invaluable to any serious programmer. I would be breathing down the admin’s neck if I couldn’t get to the Google page.
I’ve configured mod_perl on the box that hosts this blog. That took quite a while, which explains the time of this post. But, the results will be worth it, I think. Adding a new entry is so much faster than before. I hope comments will pop up more quickly, too.
One of the frustrations I had is that every Perl module I install creates directories with the permissions set to 700. This prevents any user other than the owner from being able to search the directory. I have to manually find all the directories created in the perl library and issue a ‘chmod 755′.
If I could figure out a way for the directories to default to 755, that would have saved me four hours of time. Even a script that “fixes” the directory settings within a tree would be very helpful. Are there any bash script kiddies who could help me out?
One of my favorite bloggers, Raymond Chen, posted this entry that is related to spam and viruses. Lke all of Raymond’s posts, he provides some interesting insights into the problems we face as programmers.