Older blog entries for mobius (starting at number 9)

Well, it's been a few days. Almost enough to justify a diary entry.

Let's see... It was my bday yesterday. I stayed home, went to the doctor, found out I have bronchitis. Whee.Why is it that these things only happen at the worst times?

On the bright hand, had a hell of a good party on saturday, with quite possibly the best beer I've ever tasted. (Harp, imported from Ireland.). Much more to my liking than Guinness. (not that I don't like Guinness ;)

hacking: Sadly, didn't do any this weekend. This is a bad trend. Maybe I can motivate myself to do some today.


Spent an hour configuring/setting up Phorum on my PHP site. Quite a nice package, and fairly easy to configure. It takes a bit too much playing inside the source code, but the results are well worth it.

You may be wondering why I put closing HTML tags on the first line. This was to see if I could prevent someone else's entry from reformatting my own. Sadly, they seem to be escaped. I don't know what would happen if some bad HTML were <u>actually</u> placed in front. Oh well.

Fixed the build of Pine yesterday, about 20 minutes after posting my diary entry. I somehow hadn't installed the ncurses devel libs yet. I followed this success with a long nap.

I also hooked my new VCR into my stereo yesterday. Sweet sound. :) It didn't come with a manual though, so I'm still mystified as far as using its 'universal' remote for the tv. Perhaps I'll have a brainstorm later.

And finally, I'm at work today. 'Nuff said.

Bah, just tried building Pine on my home box(sick day, woohoo!). Either it didn't build properly or I'm blind, since I can seem to find the executables anywhere. Sigh.

I've also decided to code up a quick'n'dirty program to add a custom sig depending on who is getting each email. Should be interesting. First attempt will be in C, then maybe I'll move it to something else(Perl?)

Well, I finished reading Hyperion last night. Now I need to go get the sequel(s).

The hunt for a scanner begins soon. I'm leaning towards Umax, since they seem to have the best non-Doze support. I've also used one with no complaints before.

Made the nodeos code way less flaky. Improperly written graph files or config files shouldn't break it anymore. Had to write my own function to explode the strings, since strtok(), strsep(), and sscanf() all have their shortcomings. For once, I made the flowchart before finishing the code. Whee.

I realized yesterday what I really dislike about most MS products, specifically Office.

It's the help files. I detest them.

The reason for this: They assume that you will use the 'Wizards'. That is, go through the step by baby-step guide to making a chart, form, table, document, whatever. I admit the things are useful. I just don't like them.
This causes problems, because, in their infinite wisdom, MS decided only to put the phrase "run the Wizard, and follow the instructions on screen". Absolutetly nothing about what is actually happening behind the scenes.

I never want to do what the wizard wants me to do. So I have to change things by hand. Which is impossible if I don't know what bits do what. Sigh.

Have you hugged a SysAdmin today?

Thoughts on the trust web:

  1. Make the self-certification mean something. If I don't think I deserve to be more than an Apprentice, I set myself as such, and no amount of further certification can bump me up.
  2. Make users semi-accountable for their certifications. Perhaps add a contingency that an attack by someone you certified would affect your own level(slightly, or maybe temporarily?). Not sure.
  3. Everyone should be the root node of their own trust web. Sadly, this is _far_ beyond the scope of Advogato. For future designs it might be worth considering.
  4. make new users more accesible to those who can certify them. perhaps a "still uncertified after xx days" listing?
Just some rambling.

I really don't understand redrobin's most recent diary entry.

I certainly don't have a "sufficiently massive public rep.", nor do I know any other member, personally or impersonally.

In fact, I walked in off the street. Right now I'm certified, by people who(I hope) take the process seriously. That may change. The only reason I think I am certified is that I wrote something useful in my notes, and put entries in my diary. Samurai certainly hasn't showed up in the 'recent diary entries' list while I've been here. I think that a lot of the way to be accepted into this system is to participate in it. You can't just lurk and expect to gain Master status.

Just how do accounts get deleted on advogato? It there a limit to how long your account is active without you logging in?

It's always fun causing controversy...

Anyway, fixed the last known segfault in nodeos. One uninitialized pointer out of hundreds isn't too bad. Tried figuring out what @delete was doing. No luck yet.

Random thoughts:

  1. There seem to be a lot of Canadians on this list. I like it.
  2. It seems to be a common occurrence for people to lose their certification, when it doesn't look like much has changed. Is this because the people who certified them are no longer as trusted? Or is there a more complicated explanation?

It just struck me how much of an experiment this is. We're all volunteers in Advogato's(dammit, I keep typing Avogadro) petri dish. How much of what we say and do will contribute to future theories of group interaction, or network theory? How typical are we? Will this work on a larger scale?

I shouldn't try thinking early in the morning.


Well, I just uploaded the new version of nodeos. Yes, I'm trying to get someone to go visit, and maybe even give me feedback on it ;P

I implemented @delete late last night, it seems to create new edges at the same time as it is deleting a node. Wierd. I think it's finally time to stop adding things and just fix the bugs.

On a different note, I've started reading Hyperion by Dan Simmons. I'm quite impressed, and can't wait to get further into it.

Woohoo! I'm certified! Thanks Sej!

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!