Dear linux.

Mar. 6th, 2008 03:26 am
ibneko: (Default)
I hate you.

Namely hacking you. Stupid structs.

If I add one line:

unsigned long window_size;

to the struct net_device_stats in includes/linux/netdevices.h, the kernel fails to boot completely. I get memory errors and shit.

But noooo, looking at how someone else does it, that's EXACTLY how we're supposed to do it. My code works perfectly in UML too. :P WHAT THE FLYING FUCK?
ibneko: (Default)
I got a compliment today that consisted of this link. :) I feel all warm and fuzzy inside now~ bwee.
ibneko: (Default)

Useful, I think. Posted here so I won't lose the thing.


Mar. 8th, 2007 05:37 pm
ibneko: (Default)
I've been looking around for various web "front ends" for svn - essentially allowing comfortable browsing of an svn repository. The catch, however, is it must support remote repositories. As in, accessed through http:// or https://. Most, like Trac and ViewVC (was ViewCVS) either are working on handling it, or don't handle it at all.

So WebSVN [] handles it. And runs in PHP. That is good. (Trac and ViewVC are python based, a bit more of a hassle to install, at least for me.)

"XML error: no element found (3) at line 3 column 0 byte 45"

This error is due to a custom certificate. If you hit this, you'll need to find some way to tell svn to accept that certificate. Since my certificate was signed by CACert, all I had to do with add their root file to the svn config:
(in ~/.subversion/servers, on the bottom, after the [global])
ssl-authority-files = /etc/ssl/certs/cacert-root-ca.crt

And then I hacked (and modified
Diff as follows:
Diff, generated by svn )

Note: the configuration location set function must be called BEFORE setSVNCommandPath. ie;
$config->setSVNConfigPath("/Path/To/Subversion/config/.subversion ie. /Users/benjamin/.subversion");

Also, yes, it does use your subversion settings. I see no problems with this, but I may be sadly mistaken. If you want to play it safe, recreate the folder somewhere and set the "servers" file there.

Side note: accessing an SVN repository remotely is quite slow. If you have alternatives, use them.
ibneko: (Default)
....I think they're the dumbest thing ever. Along with the lack of initializing variables.

I just spent over an hour trying to figure out why:
$userid = -1;
was overwriting the value in $_SESSION['userid'].

::really disliking PHP now::
ibneko: (Default)

Microsoft's IUI guidelines. IUI is "Inductive User Interface". Essentially, it covers how normal users, those that don't code, and don't understand the programmer's "mental model of the product".

My favorite is here:
Users don't seem to construct an adequate mental model of the product. The interface design for most current software products assumes that users will understand a conceptual model that the designers carefully crafted. Unfortunately, most users don't seem to ever acquire a mental model that is thorough and accurate enough to guide their navigation. These users aren't dumb — they are just very busy and overloaded with information. They do not have the time, energy, or desire to wonder about a conceptual model for their software.

And it's true. Most of what they've said in the first... maybe, page or so, that I've read. In CITES OnSites, I watch the older generation stumble over the Mac GUI as well as the Windows GUI. They just don't have the time/energy/whatnot to absorb the information and quickly filter out what they want.

It doesn't only affect the older generation. Today, (or rather, yesterday, since it's saturday now) during physics lab, I watched one of my group mates stumble over instructions for operating the lab software. Granted, the software was rather confusing. But the instructions were clear enough, at least to me. He, on the other hand, often spent at least a minute looking for whatever button the instructions called for.

I don't know, I'm not done reading the entire IUI guidelines, but I find them very interesting.


I learned quickly that business executives didn't care about usability testing or information design. Explaining the importance of these areas didn't get us any more work. Instead, when we're in front of executives, we quickly learned to talk about only five things:

How do we increase revenue?
How do we reduce expenses?
How do we bring in more customers?
How do we get more business out of each existing customer?
How do we increase shareholder value?
Notice that the words 'design', 'usability', or 'navigation' never appear in these questions. We found, early on, that the less we talked about usability or design, the bigger our projects got. Today, I'm writing a proposal for a $470,000 project where the word 'usability' isn't mentioned once in the proposal.'s sad. Again, I guess, partially, it's connected to the whole.. stupid-but-not-really-stupid users. That the management and execs. fall under. 'cause they don't use, or realize the need for usage... I dunno. Those 5 are important, but really, wouldn't it be easier to get more customers by making things more, well, usable?

I think that's what Apple has accomplished, in a way. Things are.. much more userfriendly... all of their i* apps (iTunes, iPhoto, iDVD... omfg, have you seen iDVD? So... userfriendly. "Drag stuff here." "Click here to burn.")

That would explain why I dislike the iLife and iWork applications. It doesn't follow the programmer's model, and most "pro" options that one usually finds has been stripped out or left out. Hrm...

Finally, unrelated to IUI, but still to Apple (as well as Sony)

French Bill Threatens To Slice Into Apple's Pie
By Laurence Frost
Associated Press
Saturday, March 18, 2006; Page D01

PARIS -- Apple Computer Inc. faces a serious challenge in France, where lawmakers have moved to sever the umbilical cord between its iPod player and iTunes online music store -- threatening its lucrative hold on both markets.

Amendments to an online copyright bill, adopted early Friday, would give rivals access to the hitherto-exclusive file formats at the heart of Apple's music business model as well as Sony Corp.'s Walkman players and Connect store.

Thanks to the success of the iPod models -- in the United States, the players accounted for 72 percent of the portable media player market in 2005, according to NPD Group -- iTunes has become the global leader in online music sales. The iPod is currently designed not to play music from rival services.

According to the latest amendments, however, copy-protection technologies like Apple's exclusive FairPlay format and Sony's ATRAC3 "must not result in the prevention of the effective application of interoperability."

Companies would have to share all "information essential to the interoperability" of their copy-protection formats with any rival that requests it. If they refuse, a judge can order its delivery, on pain of fines.... )

I don't know whether to cheer or worry over this. In a way, it'd be nice.. very nice, actually. But on the other hand... I dunno, something's bugging me, and I don't think it's because I think Apple's stocks will go down as a result...
ibneko: (Default)
In my attempts to convert the Engineering Council website ( from coldfusion to PHP, I've also taken it upon myself to not use tables, and convert it over to pure CSS.

This, I find, is somewhat of a pain in the ass. While tables are clunky, and tend to uglify code, CSS isn't implemented in the same way for all browsers.

And that annoys me. Only ONE browser, out of the five I keep on my laptop, handles the div the way I want it to.

Here's what I'm trying to do:
See the rather ugly peach-colored navigation bar at the left? And how there's that extending empty chunk at the bottom (it extends, trust me; just try the program/events link.), 'cause it's based on a table, and that's the result of the table row-height extending?

I can't replicate that with CSS. I'm pretty sure I'm doing it right, but it only displays perfectly in OmniWeb, of ALL THINGS. Opera probably handles it correctly, but I have a bug elsewhere that's causing it to offshift a bit too far upwards.

Here's how I'm doing it:
<div class="main-area-wrapper">
<div class="left-filler">
<div class="main">
<?php include "main.php"; ?>

CSS code:

.main-area-wrapper {
/* Dummy wrap */
margin: 1px;
width: 160px;
height: 100%;
background-color: #ffcc99;
.main {
position: absolute;
left: 0px;
top: 71px;
width: 575px;
margin-left: 170px;

Essentially, the idea here is that, due to the fact that stuff is wrapped in the main-area-wrapper, this will expand as far as there is text, thereby stretching the left-filler so that it matches the height of the main content area.

However, most browsers read the height:100% in the left-filler incorrectly, and stretch that to 100% of the browser window, which I DON'T want. XP

So I have two options. Keep poking around to see if there's some sort of CSS hack to work this out, code up a javascript hack to calculate the size of the text, and insert that in, or use the faux columns method (where you have a vertically repeating background).

None of those methods are particularly appealing... Why the fuck can't things be implemented correctly?!?
ibneko: (Default)
It took way too long to figure this out. A whole... 2+ hours.

Here's the problem:
In XCode, due to their use of header maps (.hmap files), they confused system files
ibneko: (Default)
You know, at this rate (0 wpm, 0 lpm, 0 thoughts-per-min), the base layer for my layout will never get completely written.

The job's too big. I'm too disorganized.

I have no interest left in it.


I ought to reapply the old one, mess with the code just enough so my journal looks presentable. After all, I look at it every day. Making it look good would probably make it easier to read and generate feelings of happiness. Hah. Maybe. Not.

Yeah, whee, sarcasm. Sleep.override($body); . . . .
ibneko: (Default)
Starting my work on modifying my layout to something more original. Currently aiming for a floating sidebar, like previously done. Although my current focus is on the calendar on the sidebar - I want to be able to move back and forward without going to the "month" page (ie. )

Mmm, and some of the code's broken or icky. Bleh. The code? Yes, I'm using the layer I wrote for xella's layout - not the user layer, but the one underneath.

[edit] "move" rather, not "more"... >.> and it doesn't look like it'll work - there's no way, as far as I know, to access the entry count for a certain day from within the layer.

Expand Cut Tags

No cut tags


ibneko: (Default)


RSS Atom

Most Popular Tags

Style Credit

Page generated Apr. 28th, 2017 10:15 am
Powered by Dreamwidth Studios
November 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 2016