Category Archives: Science

Let’s do a science!

Posted on by
Reply

I’m conducting a sociological experiment. On days when I remember, I’m going to count how many students in my class use mobile phones, portable game consoles, multimedia players, and other handheld gadgets during class. I’ll do it in only one class per day and not necessarily the same one. I’ll try as much as possible to do it in the same class each day but if I forget to do it one class I’ll do it in the next instead. My hypothesis is that, though young people today are, in general, uncouth, barely literate brutes unfit for middle school, let alone higher education, female students are more disruptive during class, at least, than male students.

The experiment will not, admittedly, be very scientific because I will, of course, be paying attention to the content of the class as well as looking out for slackoff behaviours. This means it’s almost inevitable that some students will whip out an iPod or PSP or something and it’ll slip under my radar. I expect this to happen. Whatever my results are, the reader should realise that the actual figures are probably much worse. It’s also not always possible for me to see everyone in the room from the seat I’m in. I try to sit near the back in the centre of my classes so that I can see everyone but there will surely be times when I won’t have a good view and will fail to spot offenders because of it. There’s also the fact that I don’t want to look like an insane person, so I’m not going to go nuts looking around the room like a psychoneurotic squirrel.

I speculate that if there were simply a sign at the entrance to the school that asked people to turn off cell phones and other electronic devices before entering most people would comply. It works for the most part in movie theatres.

I’m not counting kids with laptops or netbooks as offenders because it’s possible to use those things to type notes even though this is actually not allowed at my school for some backwards reason. I’m giving them more than the benefit of the doubt here since most students I see with laptops are playing computer games or otherwise goofing off.

The classes are all an equal length — 75 minutes — and are both supposed to be the same size. I’m not going to try to count how many students are actually present in the class though. The only calculation I’ll attempt to make is the percentage of students who used one of these devices at least once during class. To calculate this I’ll just assume nobody is absent and all 50 students are in the class. The fact is, however, that there are actually something like 20 students absent almost every single day. There are only about 30 chairs in the classrooms and there are usually one or two empty seats left over, so the reader should, by all means, assume that the percentages are much higher. I’ll also do an “adjusted” percentage of students present who are offenders based on the assumption that there are 30 students present every day. Obviously, this is not always going to be accurate.

I read somewhere that the students in my school are about 60% female and 40% male, which I suppose is its legacy as a former all-women normal school. This means there will be a tendency no matter what towards female students having a greater number of offences. I predict, however, that the unequal distribution of offences between the sexes will outweigh the bias introduced from the unequal distribution of sexes among the student population.

Because so many students begin using these devices as soon as class begins, I will be noting how many students who do use electronic devices begin their usage within the first five minutes of class as measured by my trusty wristwatch sidekick, Benoit Garibaldi. Since some students use these gadgets for nearly the entire duration of the class, whereas others only use them intermittently, it would be nice if I were able to measure the duration of usage for electronic devices, but I don’t think I could divert from the content of the classes the amount of attention required to record that information. Then I’d be just as guilty as the boors I’m observing. Yes, I’m not a particularly objective observer.

Days 1 and 2:

 

Day #Offences for male students (within 5 minutes of start of class) #Total offences for male students (inclusive)
1 0 0
2 2 2
Day #Offences for female students (within 5 minutes of start of class) #Total offences for female students (inclusive)
1 5 8
2 9 11
Day #Total offences(Male + Female) %Offenders [%Offenders(adjusted)]
1 8 16% [26.667%]
2 13 26% [43.333%]

My CPU usage keeps shooting up to 100%

Posted on by
Reply

This picture is obviously old, but this is more or less what the graphs looked like for days


I’m not sure what’s causing this. I assumed that I got “hacked” again and that there were some malicious files somewhere in this website that I hadn’t known about. That wouldn’t have surprised me; this site is probably already in a lot of blacklists. I have to disable the Network Shield in Avast when I want to visit this site. Interestingly, it doesn’t bother me if I visit the homepage or access the site via FTP, but if I try to visit the WordPress dashboard or access cPanel, the alarm bells go off. I know some of those web-based tools in which you can type in a URL to see if it’s in any blacklists always say this site is associated with malware. robtex used to say that, but it looks like this site has been removed from whatever list it was in because robtex now says it’s not in any of the blacklists it checks in. All of that is well and good, but it didn’t answer what was eating up those CPU cycles.

Rather than fix the source of the problem I installed W3 Total Cache. That stopped the CPU spikes. I’m still worried though because I get absolutely no visitors other than bots from Google, Baidu, and Yandex so there’s no reason I should be using much CPU at all. I installed a fresh copy of WordPress, replacing everything except the config files I’ve manually edited. That should have put my fears to rest, but it hasn’t. While I was still worrying about all of this, the site went down for two days due to DDoS. I sure hope that’s not my fault. I just want everyone to get along. Maybe I’m just paranoid. As though that wasn’t bad enough, I noticed just now that my IP address has changed. Now I have to go and change some DNS records. I really enjoyed having my IP being with a 96. Now it begins with 71.

I can’t tell if this is working

Posted on by
Reply

mplayer2 works in the version of SMplayer I’m using. This was easy in Windows because I just downloaded the precompiled build from the site linked to at the mplayer2 website.

Doing stuff in *nix is hard though, especially since, despite the fact that I’ve used *nix operating systems for many years, I’m the worst type of noob; one who’s too lazy to learn. On Debian I had some trouble but I think I eventually got it working by using the latest version from here and trying various things until it seemed to be working.

But the fact of the matter is that I know so little about these things that I can’t tell if it’s playing the Hi10P files that I’m trying to open with it properly or not. For all I know, they’re being decoded improperly and I just can’t tell. I think everything is working fine though because when I used a version of mplayer that I knew for certain didn’t support 10 bit encodes I got the following error message and the video wouldn’t load at all (unlike in Windows with ffdshow and MPC-HC which wouldn’t complain at all but would play the file with various artifacts here and there that were pretty obvious):

Unsupported PixelFormat 72

But when I used the latest mplayer2 version the file did play and I didn’t see the obvious weirdness like I did when I used a version of MPC-HC and ffdshow that I knew wouldn’t play the file properly (I had done this intentionally so that I would know what it would look like when it wasn’t working, to help me be certain that I had gotten it working). But I’m really not so sure…just because there aren’t any error messages doesn’t mean things are hunky-dory.

Is this what things are supposed to look like? I really don’t know. It looks all right to me, I suppose, but I don’t have an eye for this kind of thing. The second Yuru Yuri one has some obvious banding by the lower left star, but this was apparent in 2 other versions of this episode that I’ve looked at as well and it’s a black background so it may not necessarily mean that I”m doing something wrong.

Actually, the thing that annoys me even more than being too dense to be able to tell if these files are playing properly is that I can’t figure out how to play files from SMB shares using mplayer2. I have nearly all of my video files on hard drives in a computer running Windows XP and I access them via the network from my TV with DLNA support as well as various other computers. If I try to open a video file from an SMB share in mplayer I get an error like this when I click Options->View Logs in SMplayer:

Playing smb://server/Video/filename.mkv.
No stream found to handle url smb://server/Video/filename.mkv

Exiting... (End of file)
ID_EXIT=EOF

I tried KMplayer too and got the same results. So I got Smb4K and mounted the shares and then tried playing some files again. It still didn’t work. I compiled mplayer with the –enable-smb option too. So I’m giving up for now and just copying whatever file(s) I want to play to the local drive and playing them that way. The irritating thing is that Totem Movie Player has no problem playing files from the SMB shares, whether they’re mounted or not. I want to use mplayer2 and SMPlayer though, not Totem.

I’m conflicted about which media player to use

Posted on by
Reply

But first some background information.

I acquired an iMac G5 this week for no cost. It’s got a 1.6 GHz PPC processor and it came with 512MB of PC3200 DDR RAM. At some point a year or more ago I was successfully running OS X Leopard on my AMD-based PC with an A8N5X motherboard. The HDD in that machine which had OS X on it failed and I never bothered reinstalling it, since it was so much trouble getting it working in the first place. As a result, I had nothing running OS X at all except for my G4 iBook and G4 snowball iMac, neither of which are the most practical of devices to use. In other words, I was very happy to have a machine that could run OS X that, though undeniably dated, is still useful.

I wish I had a Mighty Mouse to go with it


I looked around on search engines and found a PDF of the instruction manual. It says that this computer takes PC3200 DDR RAM, that it has 2 memory bays and, the best news contained in the manual, that it’s expandable up to 2GB. I said to myself, “I have all these computer parts lying about. I simply must have some unused PC3200 RAM that I can put into this G5″.

Unfortunately, I looked through all the Dell Optiplexes and similar that I have in my closet, and Yune, who I also keep in there, popped out and informed me that the closest thing I have are two 512 MB PC2700 modules. “2700 and 3200 are close enough to each other, right? That’s only a difference of 500”. Such thoughts drifted through my unsophisticated mind. I looked around on some more search engines to see if I could find out if this PC2700 RAM would work in the G5. I found mixed answers online so I decided to give it a try. In so doing I found out that it’s very easy to open these iMacs and you can’t lose the screws that hold the white outer shell together since they’re not actually completely removable. The RAM works fine. I tried putting in two sticks of 512MB PC2700 and it worked. It also worked when I mixed the 512MB PC3200 module with one 512MB PC2700 module. I did go ahead and buy some PC3200 since I feel odd about using the wrong type of RAM indefinitely, but it’s been working fine like this with one module of each type for two days.

The point of all that is to emphasize that my results are not at all scientific since I neither made actual measurements nor am I even using the type of RAM meant to be installed in this computer. With two PC3200 modules I would assume things would go 500 mystery units faster than they’re going right now.

So after installing Leopard on the machine and updating to 10.5.8 I went ahead and downloaded Mplayer OS X Extended, which is the media player I always use on OS X. I used it on my A8N5X setup and watched lots of 720p videos with it.

The conventional wisdom seems to be to use ffdshow-tryouts on Windows and mplayer on *nix and OS X. That’s fine by me. I’ve used xine and gstreamer on *nix but since I’m stupid and don’t like to read instructions I failed to figure out how to configure them to do what I wanted and ended up in a pattern according to which I always use mplayer frontends on *nix. So that’s more or less why I also use Mplayer OS X Extended on the rare occasions when I have a chance to play around in Mac OS X.

But when I tried using it on this G5 I was a little bit disappointed by the results. I first tried playing SD H.264 videos and had bad results in which the audio would play just fine but the video was decoded so slowly that it’d get stuck on a single frame for over 30 seconds at a time, while the audio, meanwhile, chugged away smoothly, ignoring the dilly-dallying video.

I checked the option in the Preferences menu to drop however the hell many frames are necessary but it still was no use; the playback was so slow it was unwatchable. Not that it’s surprising, but I got the same results with 720p material as well.

So I installed Perian and tried watching some SD material in Quicktime. I had similarly bad results. It would drop frames like crazy. It dropped frames to the extent that I would sometimes stop and wonder if it was even playing at all, seeming as it did that it had frozen on one frame. But sure enough, it was playing, just ridiculously slowly.

I was going to give up and conclude that this computer would be useful for word processing, web browsing and email and not much else when I remembered that troglodytes like me who cannot be bothered to properly configure things are often recommended VLC.

Ordinarily, I would not use VLC unless I had a video encoded with an extremely arcane codec that I could not play in any other player. I also use it sometimes to make Firewire video captures from my STB, because, unlike CAPDVHS, you can set it to play the material while it records. I wouldn’t actually watch a completed recording with it though. But I tried VLC on the G5 here and not only did it play my SD H.264 files perfectly, it surprised me and also played 720p material almost perfectly even with some really fancy karaoke (at least 3 lines of stylised karaoke: Japanese, romaji, and translated English lyrics). Artifacts like in the above image only actually occurred two or three times per 24-minute episode and even so, only for a second or two each time. The audio sync is also perfect, as far as I can tell. The video doesn’t look jerky either, which would have been an indication that at least some frames were being dropped to maintain speed. 720p plays with VLC on this G5 almost as well as it plays on my Phenom X4 9950 2.6GHz system with MPC-HC. It’s more than completely watchable.

I’m conflicted though. Cool people aren’t supposed to use VLC to actually watch videos. At least, that’s the impression I get. Maybe if I were more patient and willing to actually read documentation I could get Mplayer OS X Extended to play 720p or merely 480p as well as VLC does. It’s just that VLC accomplished all of this out of the box without any configuration by me at all. Like I always say, I suppose I should just enjoy this without complaining, but as usual when something nice happens, I’ve found something to be unsettled about in this otherwise fortuitous turn of events.

Well, that was harder than I thought it’d be

Posted on by
Reply

My face after finally figuring this shit out

I just switched (tentatively) to the Twenty Eleven theme for WordPress. I obviously don’t know anything about web design or CSS, but I found it annoying how narrow the content column is and thought, naïvely, that fixing it would be easy. I assumed I could just go ahead and change one or two lines in the style.css file and be done with it, so I looked at this bit:

#content {
margin: 0 34% 0 7.6%;
width: 58.4%;

and changed the 58.4% to something higher. I refreshed the page and it didn’t look any different so I also looked at this bit and reduced the width by the same number of percentage points that I increased the content width by:

#secondary {
float: right;
margin-right: 7.6%;
width: 18.8%;

That didn’t seem to do anything either. I then thought that maybe I had to create a “child theme” of the Twenty Eleven theme for things to work, so I restored the original style.css file for the Twenty Eleven theme and proceeded to create a child theme. The style.css file for the child theme was completely empty except for modified versions of the above two lines. It looked like this:

/*
Theme Name: Twenty eleven Child
Description: Twenty eleven child theme
Author: asdfgh
Template: twentyeleven
*/

@import url("../twentyeleven/style.css");

#content {
margin: 0 26% 0 7.6%;
width: 66.4%;
}

#secondary {
float:right;
margin-right: 7.6%;
width: 10.8%;
}

This didn’t seem to work either. I gave up and decided that I should try to use the Twenty Ten theme and see if I could make it wider. I didn’t really care which theme I ended up using; I was just sick of the “Classic” theme that I’ve been using for a long time and wanted something different, so long as it wasn’t really, really narrow. So I went ahead and looked at the style.css for the Twenty Ten theme, trying to find the lines that I might need to change. I tried changing a few bits here and there that seemed, as far as I could tell, to be the lines that controlled the width of the page, content column and side bar area. Again, none of this seemed to work, regardless of whether I changed the values in the style.css for the Twenty Ten theme itself or if I made the changes via a child theme.

I then read a thread somewhere in which a user complained that he or she was trying to change something in a child theme but that no matter what he or she changed the blog looked the same. The person who responded to the post gave a very curt reply, saying, “clear your browser cache”. The OP responded to that message saying, of course, that all of his or her modifications were now showing up. Of course, after reading that I realised that this was my problem too. It really is as easy as just changing the percentages in the style.css file, but I ended up spending a few days wondering about this because I overlooked such a mundane factor as my browser cache. Dabbling in things you’ve got no knowledge or experience in sure is a good way to make yourself feel like an idiot!

Now this site looks decent without so much wasted white space when viewing it on my 1280×1024 monitor or my 1920×1080 monitor. I need to get a better header image though. This one doesn’t really fit.

Incidentally, I missed Mawaru Penguin Drum because I was so distracted by this. I ended up writing down the time incorrectly. Why the hell do Japanese TV stations have to list times in the early morning as though it’s part of the previous day? I was going to watch the show live streaming at 26:25 on July 07 (it’s weird saying that, since it’s still in the future in my time zone) but I missed it because I wasn’t paying particularly close attention and wrote down that it would be on at 26:55. I tuned in in time to see some commercials and then a rerun of Denpa Onna to Seishun Otoko.

Best mouse in the world? Or best mouse in the universe?

Posted on by
6

The Logitech MX310 is, without any semblance of doubt, the best mouse I’ve ever used.

I’ve always had a problem finding mice I like. I think I have larger than average hands or something, since my most frequent problem with mice is that they have too small a surface area. If the body of the mouse is too small, what ends up happening is that the buttons are too close to where my palm would naturally rest, so I end up having to bend my knuckles at an unnaturally sharp angle, which is uncomfortable. I use my index finger for the left button and my middle finger for the right button, so the knuckles on these two fingers should only have to bend slightly to operate the buttons. The index and middle fingers should be almost straight. Ideally, if you ask me, the buttons on a mouse should be situated so that your knuckles bend at a very gentle 150 degree angle or so.

The Logitech MX310 is long enough so that this is exactly what happens naturally to my hands. It’s also got an attractive color scheme and six configurable buttons: standard right and left, one on the right side, one on the left side, one just below the scroll wheel and finally, the scroll wheel itself, which serves as a button.

I’m fairly certain that the mouse is no longer being manufactured. When it was still in production though, I believe it was marketed as a “gaming mouse”. The type of gaming I do on the computer doesn’t require a fancy mouse, so I don’t actually care all that much about the configurable buttons or how its precision will help me to get a headshot. No, what I like most about it is its weight. There’s just the right amount of resistance when you move it around. It has a very smooth feel when used with a mousepad. I’ve never used another mouse with such a satisfying degree of inertia.

How did I discover this unsurpassed mouse? By accident, of course. Prior to my discovery of the MX310, I had been using Dell 0D1161 mice. You know, these office surplus ones:

Despite their ubiquity, these are pretty good mice. As I said, I don’t actually care about the six buttons on the MX310 — that’s not why I love it so much — so the fact that there are only three buttons on the Dell mice doesn’t bother me. These Dell mice match the MX310 in every respect except for weight. The Dell mice are not too light, but, ideally, they should be slightly heavier. The MX310 is a bit heavier than the Dell mice and, in my opinion, the perfect weight.

Those Dell mice stop working after a relatively short while. I suppose they’re not really made to last. I’ve gone through four of them in as many years. I buy them on eBay several at a time. One day, when another mouse broke and I was down to my last spare, I placed an order for three more of those Dell mice so that I would have them in stockpile, ready to be swapped into service as soon as needed. When the package arrived, instead of three 0D1161 mice (or whatever they’re actually called) I found a single Logitech MX310. I contacted the seller, who acknowledged his mistake and immediately sent out the three correct mice that I had ordered. He told me not to bother sending him back the MX310. That was a fateful bit of apathy on his part, because if it weren’t for that eBay seller, I never would have had an opportunity to use the MX310.

Even once the three 0D1161 mice arrived I didn’t put them into service; I used the MX310 until it stopped working properly. After some time it would intermittently stop responding. Eventually it got to the point where I’d have to unplug it and then reconnect it every minute or so. In other words, it became unusable, as inevitably happens with mice. I didn’t get rid of it though; I just dumped it in my huge box of broken electronics and computer components that, for various reasons, I hadn’t yet disposed of properly.

Since then, I’ve been using those three 0D1161 mice. That is, until a week or two ago when the last one of them that remained stopped working. Since I needed a replacement, I decided that I would try to replace the beloved MX310. I bought one cheaply on eBay for USD20, compared to the USD40 to 50, which was the price on Amazon and other online stores at the time.

Ironically, shortly after it arrived I found this article about how to easily replace the USB cable in an MX310. I followed it and fixed the first MX310 that had been sitting around broken for at least a few years. Now I’ve got two and I couldn’t be happier.

I’m sure I’m not the first person to think of this,

Posted on by
Reply

but I’ve found a new way to delevel characters for SP farming. I have a level 53 wizard that I’m using to farm SP for a new character I created. A level 53 wizard is not able to one-hit kill party Ongs, so I thought it would be best to do the SP farming at Chakji Workers. For this to work properly, I decided to simply leave the mastery levels for all skills on my new character at zero. Ideally, the character should be about level 8-10 to be farmed at Chakji Workers. The spawn rate there is fantastic and I can one-hit Chakji Worker party mobs. It’s also good because Chakji Workers are non-aggressive and won’t attack the character getting farmed, so you don’t even need to worry about protecting him or her.
Continue reading →