People love to get super creative and name their blog the most bizarre or domain specific phrase possible. That is bad.
I might not be your average reader. I read about 190 blogs every day. I found these blogs because somebody was talking about them, a friend told me about it, or they seem to have similar objectives as mine. In summary, of the 190 blogs that I read I only know personally about 20 of the authors.
Why does it matter? Because most of the blogs I don't even know the author name or his company, or what he does. I think that is bad.
People give funny names like "Greg Corners", or "Ctrl-X", or "Cat & Clam" (all fictitious names, don't know if they really exist). Why? What is your point? To prove that you are clever?
Now comes the worst part: The description. All blogs can have a short description associated with them. This appears under the blog name on most feed readers. These are some gems that you find on the web: "Rants from within", "About everything and anything", "Yada, yada, yada"... Again, why?
Finally, there are the worst offenders of them all. Blogs that have a clever meaningless title, clever meaningless description and the person writing configured their own blog and the author name doesn't show up or appears as "Administrator" or "Team". Triple "why?"
When naming your blog:
Choose your name, or your company's name, or a nickname that people will know and remember.
Write what your blog will be about (this also helps search engines).
If context is relevant, add a few words about you;
If a company blog, at least, write the company's name and website.
Good descriptions:
"The official blog of Sampa Corp."
"P-I reporter John Cook explores startups, venture capital..."
"Richard MacManus on Next Generation Web and Media"
Bad descriptions:
(no description)
"Our company, our vision, our passion."
"A running commentary of occasionally interesting things"
"Confessions of an Old Fogey"
Finally, don't forget to add at least your first name as the blog author. You are losing opportunities every time somebody thinks you wrote a great post, but they don't know who you are.
There are those that swear by Open Source and those that swear against it.
Being such a polarizing force, it is hard to get good information if Open Source is bad or good. Add to the mix that most people associate Open Source with big projects like Linux or Apache and you get a pot of confusion.
First of, most Open Source projects are tiny. Tiny as in just one or two developers (go to SourceForge and check for yourself).
Second, the good developers, I mean, the really good ones will not be spending their time in a small library or project, they will go to big high-profile projects like Linux, Firefox and Apache. Which leaves the average or below average developers doing the majority of the Open Source projects.
This week I had to deal with two Open Source issues. A former developer wrote an email server mostly based on two Open Source projects.
Last week I went to test the feature and the first (and simplest) Test case that I wrote didn't work. Ok, do the routine: debug, try, debug, try. Found the issue. The SMTP server had a small bug. Fixed it. Nope, not working still. Investigate a little more. Well, turns out that the Open Source project implemented the original version of the SMTP protocol (circa 1982). Virtually nobody uses this protocol anymore (now we use ESMTP). Then I found a good Open Source version of ESMTP (called NMail), but it was big. It was huge. Basically, I end up implementing my own version taking clues from NMail. The result code is less than 1000 lines and does what we need.
The second piece of Open Source was used to parse attachments from email messages (multipart mime stuff). This works flawlessly, if it wasn't for the stupid way that it was implemented throwing and catching exception all the time to check values syntax (hint: this can slow down performance by 100 times). Then I found a killer bug on that one, it had a humongous memory leak. In technical terms "humongous" means it happens every time the API is called. I found a workaround to fix it.
Now, don't take the wrong way, I go to SourceForge and CodeProject all the time to see how people implement things. It saves me a lot of time with the small but dangerous details. However, when you are using an entire library from Open Source on the hopes that it will do what was advertised... Be prepared, very prepared.
Subscribe to this Site
January, 2008