In developing my Twitter application, I decided that I needed to know how many users the average Twitter user follows. To accomplish this, I started a poll which was a pretty sad failure. Only 5 people responded and two of them were myself and my wife. However, all was not lost. As part of my Twitter application, I cache lists of who a user follows. (This helps reduce my impact on Twitter’s servers.) I checked and I had over 10,000 users’ information cached. So I wrote a quick script to query that information. The results were quite enlightening.
- 90.33% of users follow 2,600 people or less.
- 75.67% of users follow 1,400 people or less.
- 51.14% of users follow 500 people or less.
Using OpenOffice.org, I graphed the results and came up with this interesting chart:
As you can see, the percentage of users climbs sharply until around 2,500 followers. Then it levels off for a slow ascent to 100%. This really helps me. If I make my application cut off checking the Friends list at 2,000 people, it will handle the complete lists for 85% of users. I’m willing to have partial functionality for the top 15% of users (who are likely to not need my application as much) to preserve functionality for the bottom 85%.
If you’re interested in more data from this, let me know in the comments below and I’ll see what I can do.
I haven’t talked about it much (or at all) on my blog, but I’ve been developing a Twitter application for the past few months. Every time, I think I’m done, my intrepid beta tester, HighTechDad, does what any good beta tester should do. He finds where my application is severely lacking, sending me back for some furious code-rewriting sessions. My latest versions seemed to work well enough for my Twitter username, but I ran into a problem when I attempted to "stress test" the application by running it through guykawasaki‘s name. The web server runs out of memory trying to process the 100,000+ people he follows.
I was trying to think of a way around this, but then I thought: Perhaps it is best not to. Perhaps Guy is such an edge scenario that I would be better off pulling the first 5,000 he follows and leaving it at that. Not to give too much away, but my application is intended to analyze your following list in the hopes of finding more people for you to follow. So the audience is most likely not someone already following over 100,000 people, but people following 1,000 or less people.
This led me to question what is the average number of people that Twitter users follow. Is it 500? 1,000? 5,000? More? Less? To help me answer this, please answer the following poll. It is easy to determine how many people you follow. Just stop by your Twitter page and note the number above "Following." Then choose the answer in the poll that best matches how many people you follow. NOTE: Do not use the number above "Followers." I’m not looking for the number of people that follow you. I’m looking for the number of people that *YOU* follow.
Security is important in today’s online world. A big part of that security is what password you use for the various websites you frequent. It is not uncommon for a person to have dozens of different sites, each with their own username and password. But how to you manage them all? Setting all of your passwords to be the same (e.g. "12345") isn’t safe. (I have the same combination on my luggage!) After all, if someone guesses *one* of your passwords, *all* of your accounts will be compromised. Writing them on paper and sticking it to your monitor isn’t very safe either. Anyone passing by your office/cubicle could see the list and gain access to your accounts. So what should the security conscious web surfer do? Why, put them in a safe of course! No, not a fire-proof safe (though that might be a good hard-copy backup method), but inside an application called (appropriately enough) Password Safe.
Using Password Safe is quite easy. First, you set up a master password. Make sure you remember this. It’s very important. Next, you add passwords to your list in Password Safe. You can even group them together to make them easy to find. When you’re done and you save the list, Password Safe encrypts it. The only way to decrypt it and to get at the passwords again is to enter in your master password. It is much easier to remember *one* password than it is to remember *one hundred* passwords.
In addition, Password Safe can help you come up with secure passwords. In the Add Entry screen is a button labeled "Generate." Click that button and you get a password like "3mGrsgfXEN" (an actual generated password). This is much more secure than "fluffy7". You can even store notes about the website in the (where else) Notes section.
There’s even a Portable version of Password Safe. Stick this on a USB key along with your encrypted password list and you can view your passwords on any computer, no further software required. I recently used this on a trip to Long Island to see my parents so I wouldn’t need to remember the passwords to my various accounts. It worked beautifully and now I carry the USB key along with me every day. You never know when you’ll need to check on someone online and will need to retrieve that password that you just can’t remember at the moment.
Password Safe has become a must-have application on any PC that I use. I can’t recall how I kept track of all of my passwords before I discovered it.
I’m going to start a new feature here on Techydad: Freeware Friday. Every Friday, I’ll post a short review of a piece of freeware I’ve found that I think is useful to have. I’ll try to avoid the obvious ones that nearly everyone knows about like OpenOffice.org and FireFox (at least for now) in favor of applications that might be lesser known.
This week, I’d like to highlight KIDO’Z. KIDO’Z is a web browser aimed towards kids. Why do kids need a special web browser, you might ask? Well, young children, just being introduced to computers, might not know that they can go to the address bar and type in (for example) Noggin.com to get to games and activities on Noggin’s website. Even if they do, they might misspell it as Noginn.com and wind up at a completely different site. KIDO’Z, in contrast, is extremely easy. Children merely select Games, Web Sites, or Videos and then select what they want to see. Except for any Games/Websites that require keyboard input, all controls are handled by the mouse. Nice, large, easy-to-understand icons are used for controls (with text for kids who know how to read).
The KIDO’Z team is constantly striving to improve the program. They’ve recently added Parental Controls so parents can better control what their child sees. You can block sites that KIDO’Z would normally show to the children (useful if, like me, you have a philisophical opposition to all things Barney) and can add your own sites (like a webpage of family photos).
It’s not perfect, mind you. Due to the limitations of the platform it was designed on (Adobe AIR), it can’t block kids from CONTROL+TABing to another program. It can’t keep kids from closing KIDO’Z, firing up Internet Explorer, and browsing to a site you’d rather they didn’t go to. However, it is very good at giving children an easy to use first browser that puts fun and informative content a few clicks away. I would definitely recommend this for any computer that a child would use.
I didn’t intend to do another Facebook/Breastfeeding post so soon, but a new wrinkle in the whole controversy just occurred. Facebook recently changed their Terms of Service. Previously, they claimed rights to use your content as they see fit, but you could terminate those rights by having your account deleted. Now, not only can they use your content if your account is closed or deleted, but they can sublicense your content as well.
How does this impact the Facebook/Breastfeeding controversy? Well, suppose a mother uploads a photo of herself breastfeeding. Facebook, regarding this content as sexually explicit, deletes the account. However, Facebook retains rights to all content uploaded, including the breastfeeding photo. Facebook can now sell that photo to a stock photo agency who can sell it to someone else to be used in an ad campaign. This mother might find her nursing photo being used to advertise baby formula. Imagine the outrage the mother would feel to see herself in the ad with the tag line "It’s just as good as mother’s milk."
What would the mother’s options be? Pretty limited. She could try to sue Facebook, but she likely wouldn’t get very far. You see, that same Terms of Service also states that you’re agreeing to Mandatory Arbitration. In short, she will have to travel to a state that the company chooses, face an arbitrator chosen by the company, who likely rules in favor of the company over 90% of the time, won’t be able to subpeona any documents from the company, won’t be able to appeal any decisions to a real court, and – even if by some quirk she wins – won’t have any legal teeth to get any money from the company. In other words, Facebook has all but declared themselves the winner if you decide you want to challenge them legally.
This might solve the "Breastfeeding Photos on Facebook" controversy, but not in the expected way. I don’t know why anyone (especially breastfeeding mothers) would want to upload photos to Facebook knowing that those photos could be sold by Facebook to some other company for any possible use without any compensation or recourse given to the user. Perhaps it is time for another social network, one with better Terms of Service and better operating practices, to rise up and displace Facebook.