September 2004 Archives

The link dump of this site is kaputt. So no new links until I have repaired it. I do not know yet what went wrong, but after removing a test blog I am also missing part of the Links blog.

[2004.10.04]: We're back to normal again.

I have put some ska in the Jukebox. Smooth swinging horns, velvet vocals, rhythms...

Each morning before I leave home to go to the office, I take some time to see what is happening in the world. A few years ago, I used newspapers, radio or television for that, but nowadays we have internet. While eating a sandwich and drinking my coffee, I check my e-mail and scan the headlines of about 200 web sites. If I see something interesting, I read it or bookmark it to be read later.

This whole process takes me 20 minutes, thanks to my newsreader which fetches all information from those 200 sites and presents it to me in an easy digestible way. However, this list of sites I want to keep up with is growing steadily; each week I discover new interesting sites, and add them to my list. When I finished browsing to my list today, I had to run to the office. I guess I'll have to adjust my alarm with 10 minutes to give me the time to read my news.

Just before starting to write all this down, I read that Robert Scoble has exactly the same problem, but worse. He currently checks 915 feeds, and in his entry Dealing with the information flow he wonders how scalable feed are, as more and more resources use them. What happens if the number of feeds grows up to 10.000?

I guess what is going to happen is the same as happened to the web. Ten years ago I scanned the NCSA What is new? list. First I looked at this list every week, but soon new sites appeared by thousands, and I left this method to keep up with new sites. Nowadays I use search engines. They are not the same, and don't give met the same feeling, but they scale better. With feeds the same thing is happening: at this moment I follow everything published on some sites, but soon I will have to fall back on PubSub to find what I want. I propose they will build some randomness in their queries, so I can still discover some unknown sites, and don't have the feeling that I am reading the same as everybody else.

When I read about the Jukebox that Lijn created, I got this feeling inside...
I copied the two files to my server, made some changes to the configuration, and there it is: the Brain Tags Jukebox!!!
It needs a little bit of decoration, but the functionality is there. All songs will stay available for 96 hours, after which they will automagically disappear.

At this moment you can find You Learn About It from The Gathering, and of course the wildly popular new song from the Pixies Ain't That Pretty At All.

Happy downloading!

[Update 2004.09.24] I added some echo statements to the PHP code, and the results looks a lot better. It is also valid XHTML now.

[Update 2006.11.9] More than two years later, the jukebox became obsolete and has been removed from this site.

Besides collecting spam, I can also use my GMail account to invite others to use this service. If anybody still doesn't have a GMail account and wished to have one, leave your full name and current e-mail address --which won't be shown-- in a comment.

Many people starting to implement the Getting Things Done methodology start process their inboxes full of energy. The first result is usually obtained within a few days: an empty inbox! Everything has been evaluated, using the well know flow chart, and stored in reference files, someday/maybe actions, next actions... Small tasks (2--minute rule!) have been finished immediately, and a lot of junk has been thrown away. Great!

Once the big mess is out of the way, you only need to maintain it. Process arriving items as soon as possible, and start working your way through the Next Action list. You even do your first weekly review, to see whether there are any open loops left. You feel that you are getting grip on your workflow.

But usually after two or three weeks, the doubts come up. The processing goes all right, but you have the feeling that actions come in faster than you can finish them. Your N/A list is growing and growing, and looking at your N/A list gives you the same feeling you had before when you opened your inbox. You don't know where to start or how you can set your priorities. You've lost control.

At this point many people throw GTD out of the window, reasoning that if they spend less time organizing their time, they will have more time for actually doing things. This might be true for some people, but in many cases you simply need to adjust your system to fix the problem. If I have a question about GTD, I always first go to the source. The problem I described here appears on David Allen's site in the FAQ section. David's answer is: If it's 'too long' either you need to get used to a big list of still-undone things [...] or you need to make fewer commitments.

Of course this is very true if you implemented the system correctly. But in many cases, the problem is not the amount of commitments, but rather the way these commitments have been entered into the system. Before starting to use GTD, most people work with some kind of to-do list. New GTD users treat their N/A list in exactly the same way as their old list, only now they specify the context of the action. This is wrong, since a task is not the same as a Next Action. Anne Gennett explained it in an excellent way in The power of the real next action. In her example, the client needs new glasses. So he creates a N/A 'Get new glasses'. But getting new glasses is not an action! Getting glasses is something you can't do. The real N/A is to email someone for the name of an optometrist, so this is what he should have put in his list.

If you look at a long list that contains items like 'Get new glasses', you get the feeling that you will never be able to finish the list. But if you put real next actions, your list contains easy to perform steps, and you will see that you can process your next actions more efficiently. The next action is the very first thing you need to do to move on that item.

Even the best drivers sometime get into situations in which they cannot prevent damage. I am not one of the best drivers, so I might get into this kind of situations a little bit more.

This morning when I was driving to the office, I noticed a dog crossing the road. The car coming from the opposite direction stopped, and the dog reached the other side of the road without problems. I was still far away from that point, but I noticed another dog standing next to my side of the road. I imagined that this dog would want to follow his friend, so I slowed down. About ten seconds after spotting the dog, I passed the dog, driving carefully. And exactly at that moment the dog decided to cross the road. I hit the brakes, but could not prevent hitting the poor fellow. Of course I immediately pulled down the car next to the road, and went to look for the dog. However, he was gone --- I assume he ran away into the orchards next to the road. When I inspected the car, I noticed one bump and a broken spotlight. Tomorrow I will have it repaired.

I would like to use this opportunity to tell the owner of the dog that I am very sorry, but that I could not have prevented it. Sorry...

On my individual entry pages, I show two dates: the date on which I created the entry, and the date on which I modified the entry. Today, after making some changes to an entry, I noticed that the modification date was not updated.

I found some information about this problem in earlier versions of Movable Type on the Tweezer's Edge, but they state that this error has been fixed in the 3.x versions. I guessed that the same solution should work on my version (MT3.11 at the moment), I only needed to find out the right line in the scripts to fix. I tried some options, and found that the following fixes the problem:

Edit /lib/MT.pm, and add the text shown on line 598:

596 $ctx->stash('entry', $entry);
597 $ctx->{current_timestamp} = $entry->created_on;
598 $ctx->{modification_timestamp} = $entry->modified_on;
599 $cond{EntryIfAllowComments} = $entry->allow_comments;

Here is Johnny's guitar solo on the Ramones' "I Wanna Be Sedated":

E-0-0-0-0-0-0-0-0-|-0-0-0-0-0-0-0-0-|-0-0-0-0-0-0-0-0-|-0-0-0-0-0-0-0-0 B—————————————————|—————————————————|—————————————————|———————————————— G—————————————————|—————————————————|—————————————————|———————————————— D—————————————————|—————————————————|—————————————————|————————————————
A—————————————————|—————————————————|—————————————————|————————————————
E—————————————————|—————————————————|—————————————————|————————————————

(Via BoingBoing)

If you want to leave a comment on this site, you have the option to identify yourself using your TypeKey account. TypeKey is a centralized authentication service from Six Apart, the makers of Movable Type. It was developed as a measure against comment spam, and made available in version 3.0 of MT. After TypeKey was released I changed my templates to include this system. I do not force commenters to use TypeKey, since my spam problem is not that big.

This was in May. Now after some months more, I have some observations:

• Besides me, only onetwo commenters used TypeKey to comment on this site;
• No other MT--powered site I comment on has TypeKey activated;
• A Google search on the default TypeKey comment link --- TypeKey Profile Page --- returns only 21000 hits. So only 21000 pages on the whole web have TypeKey authenticated comments --- remember that some sites have many pages, with many posts.

To me it looks like TypeKey is not used a lot.

Why can that be?

I think the main reason is that most of the sites using Movable Type never update their templates. They might update to a newer version of MT, but leave their templates alone. And they are right in doing so; I remember that it was quite a puzzle to add the TypeKey code to my template.

So TypeKey is a failure?

After only four months, it is very early to make this kind of conclusions. Every new site that starts using MT will probably have TypeKey activated, so I expect more and more sites to use is. And when the base grows, other sites might benefit more from this service as well.

I have plenty of bandwidth, and have been thinking about ways to use it. It appears to me that the best way to burn up my bandwidth is to serve MP3 files.

So as an experiment I offer you Ain't That Pretty At All from the Pixies, since this is becoming a Pixies site anyway.

Keep in mind that it is an experiment; I might withdraw the file at any moment.

[Update 2004.09.17] I guess a week is more than enough for this experiment, so I removed the file. I have obtained the desired overview on how MP3 downloads affect bandwidth useage. I also found out that Pixies fans are highly interconnected: until 2 days ago, the file was hardly downloaded, but after somebody published a link on a buletin board, the link was copied to many other buletin boards within a few hours, and downloads skyrocked.

[Update 2004.09.22] With the addition of the Brain Tags Jukebox to this site, I also returned the Pixies file. It will be available for another 96 hours.

Within one week I encountered two computers showing the same problem, which I haven't been able to solve. This concerns me, since one of these machines is mine. What happened is the following:

Both machines run Windows XP, and use a modem --- in one case an ADSL modem --- to connect to internet. Everything works just fine, until I added a network card to these machines. After that, the modem seems to connect just fine, but DNS is not working. The computer has received the IP addresses of the DNS servers, and I can ping to these addresses. Furthermore, when I try a tracert, I receive the error No hay recursos --- no resources. I have checked all possible corners of Windows XP to find a place where I can disable or criple DNS, but haven't found a solution yet. Does anybody know how to solve this?

Recently my server logs show a big increase in requests for a single image --- the one shown above. Somebody must be hotlinking this image! Hotlinking is when somebody uses an image on their page, but does not put a copy of the image on their server. Each time somebody views that page, the image is requested from my server. In general, hotlinking images is not polite, since you use bandwidth of somebody else. This specific image is not mine, I found it on somebody else's page and copied it to my server to use it.

There are several methods to stop hotlinking. You can block all requests that do not come from your own site, so all visitors of that site will see is a red x. Or you can server another image instead, with a text explaining why hotlinking is bad, or with the Goatse man... I decided not to do so. First of all, since I have enough bandwidth to serve these sites, and secondly because I thought it would be better to use the image to make some advertising for my site. So I added this site's address to the image, so it will be shown on all sites using this image from my site.

I finally also have a GMail account! Even though I already have many @jeroensangers.com and @braintags.com mail addresses, I opted for beltza@gmail.com, just to see what it is and what it does. You don't have to update your address books, since I will continue to use the old addresses, and will check this one rarely. I only posted the address here, since I need some mail to see what I can do with GMail.

That's it! The upgrade went flawless. Everything seems to work fine until now. Of course I haven't tried out any of the new features yet. There are two features I would like to have a look at.

The ability to create subcategories look really sexy, however I have doubts whether my writings need a more complex category structure. I will have to review all items to decide on that.

The feature everybody waited a long time for is dynamic publishing. It simply means that the pages are created on the fly, which makes posting entries and comments more faster since you don't have to wait for all pages to generate. However, the downside is that each normal view will be slightly slower, since the pages are not cached on the server. I will have to play a little bit with this to see what works best.

Today, six apart launched version 3.1 of Movable Type, the software I use as the motor of this site. Since I am an early adaptor, I immediately downloaded the latest version, and will install it as soon as possible. I already heard about some problems in the update scripts, so we'll see how it goes...