Something a little better

Just for convenience's sake, I'm putting a link up here for my extension that now actually installs properly.
(Yeah I know, I should've fixed that a while ago. But at least it's better late than never.)

Custicon@andrew.raynier.xpi

The code's improved too, although I can't seem to find a workable solution to get images to load in both Firefox 2 and 3. The new Favicon Service for FF3 is still be worked on (The API isn't finalized yet, and there's a lot more of it undocumented).

Contributions

So earlier this week, I contributed to Simon Jung's project by sending him a list of extensions and their Id's for him to use in testing out his Metrics Server Collection. I think Simon's done quite well with such a large component to dig through and test.
You can check his project out over here if you're interested.

Simon was also very helpful to me in contributing to my project by giving me a lovely package of Favicon images (*.ico files) to use in testing out and as an addition to my extension file bundle. Although what I really appreciated from him, was a link to Cathy's extension files to use in contributing to her project, from his blog. Mainly because I had been looking to contribute to it, but her blog, wiki page, and even the project description don't have any links or real information on the canvas3d project.

Well anyways, I managed to download the canvas3d.xpi for minefield, but unfortunately it only works with specific version of Minefield(3.0b1), and mine wasn't one of them (3.0a9pre?). So I went and downloaded and built myself a 3.0b1 version of Minefield (took about 2 hours on the open lab PC), and got the xpi to install. Lucky for me, Simon also noted in his blog the simple images I needed to use to test out this extension.

....after struggling with several failed attempts to get the images to show, only know as I type this, do I realize (after reading through extension errors) that I needed to compile the build as a Gecko project, and not a Browser one.... alright then... back to the salt mines... (ugh, I wonder how many more hours that'll take to compile).

However, assuming I do get all this actually working... there aren't any instructions on what to do with the extension once I do get it working, or a project page I can list my contribution on.
Oh well, I guess as long as I post about it here for Chris to read, that will hopefully be enough.

OSD600 Course Classroom discussion

During our class discussion about this course and the CPA program, I was really intrigued to learn how others felt about the way they were taught, what they had been taught, and the things they still wanted to learn about.

After reflecting a little on what was discussed, I grew curious about what every body's high school (or pre-Seneca) Computer Programming lessons/experience was like. Namely because I think my own previous classes seemed to give me some additional insight and view of the outlook on how most of the classes at Seneca fit in to what we would need to work as a programmer.

I was lucky enough to have a few real computer programming classes when I was in high school. I got to learn how to program in Turing, VB, and Java. Although that's not all I was taught. One thing i wish we did at Seneca, is have little interactive class programming sessions in a lab, where the teacher gives coding challenge to the students (ex. do operation in this many lines; or how do you do operation with only this/these functions?), and lets them hack at it for a part of the class. Then after the allotted amount of time, the teacher writes his answer/version up for everyone to see, and they compare and explain what the students did and how it works.

That was one of the interesting and apparently unique things that I got to do in my High School (Thornlea S.S.), but I also liked learning how to write pseudo code to plan out programs and algorithms on paper, so I could code programs an the like.

I wish Seneca would've had lessons like that in CPA, cause I've met so many students who always seem so lost or frustrated when trying to program. Learning how to program with or without pre-made code (although with specifications for the structure should be used with at least 1st years, if not 2nd) and knowing the difference between good and bad code (as well as how to make it) should be essential foundations for Student's programming knowledge at Seneca.

...and that's my little addition to CPA program discussion.

Favicon Extension Release v0.2 (skeleton)

It seems that the farther I get with this extension, the harder it gets to stop it from crashing and run the way I want it to.

Unfortunately, the current stage of progress I'm at, is not one that leaves me with a working extension.

I'm currently working on an overlay to create some anonymous content within the tabContextMenu (The menu that pops up when you right-click on a tab), but the code for that stops everything else from working.

So I've commented that part out for now (in the firefoxoverlay.xul file), and I've managed to just get the command in the Tools menu and the contentContextArea menu (web page right-click menu) to go back to showing an alert when clicked.

Getting XBL bindings to work with prompt and Favicon services has got me up against a very hard and big wall, at the moment.

Hopefully a new tutorial site I've found (xulplanet.com) will help me to incorporate these two parts of FireFox properly.

For now, my files are uploaded onto the wiki at : http://zenit.senecac.on.ca/wiki/index.php/Image:Custicon.xpi

P.S.: I also wasn't able to properly build the xpi, since my current work environment doesn't want to run bash scripts, so it won't install automatically. However, all my extension files are there to look at.

Hackergotchi - test

Well after getting some advice and tips from a graphics design student, I've been able to create my own Hackergotchi. I inserted the link into the blog feed back on zenit, so now I just have to see if it will show up when I post a blog. (it hasn't shown up on my old posts on the blog planet.)

Maybe this will be one of the those super-rare instances when something will run just right after the first time you "build" it. :)

...or not. Maybe Chris has to update the feed himself?

Lab Crunch time (makefile lab)

During the past few weeks, I had been too engrossed in getting my Mozilla build working (and all the SDK software) working on my Windows 2003 Server system. Sine I've finally gotten myself through all that now, I've gone into Crunch mode to get all the subsequent labs (which I postponed for my build) done.

Well, technically I was supposed to do the makefile lab before the build, but it didn't seem as essential to the course at the time. But now that I'm doing the lab, I've noticed that I couldn't find any blogs on this one at all (I'm pretty sure I went through 98% of the individual blogs listed on the planet). The resources listed in the lab were useful in getting a makefile created though, but I appear to be having a very unexpected problem doing this on a lab PC.

The MSVC8.bat build environment doesn't recognize the 'cc' command!

Not knowing any other c compilers to use in that environ, I switched to running make using just the command prompt. Unfortunately, it didn't recognize 'cc' either. However, using $(CC) in my makefile instead of 'cc', it brings up the default compiler for the environment, which is Borland C++ for command prompt. Frustratingly, Borland went looking for .cpp source files to compile instead of nled's .c source files. Another brick to the head.

Switching back to the labs suggested build environ, now using the $(CC) command in my file instead, somehow still calls 'cc' for a compiler, even though I still get the error: "make: cc: Command not found"

Looking through the makefile examples I've found online, I'm fairly convinced that my makefile would work if not for this "command not found" error.

So for now I'm off to fill my depraved stomach, and then try out the other labs, until I can find an answer to this problem.

- Andrew R. (JM)

P.S.: Does anybody know how (or someone who does) to create a Hackergotchi in Photoshop (or some other lab default image editor program) instead of the strange one that is used in the link which Chris's "Hackergotchi how to" blog points us to?

Build Errors

After over a week of just trying to get Visual Studio 2005 installed (due to a faulty CD), I've finally been able to get down to one of the last steps in making my build. This time I've been stopped by the build error "C compiler cannot create executables.", which hits me a few seconds after running the command "make -f client.mk build MOZ_CO_PROJECT=browser".

I found the error listed in the Mozilla Build FAQ, and it says to run vcvars32.bat to get the PATH variable set up, because I need to have all the necessary directories listed in the PATH var for this build to work. Unfortunately it doesn't say what all the necessary directories are. :(
Or why my build still doesn't work after running the vcvars32.bat file.

Does anybody happen to know a solution for my problem, or maybe someone who might?

update: Thanks to some classmates, I found out Visual Studio 2008 doesn't have the kind of C compiler that mozilla looks for. But quick download of Textpad (my old favourite text editor), fixed that and statisfied mozilla to successfully complete the build.

Now all I have to do is all those labs. >.<

--(running out of time) Andrew R.

Revolution OS

Watching this documentary helped explain a lot of things to me about Linux and Gnu since I've never really touched anything but Windows. I finally learned what the relation between Gnu and Linux is, and even heard the proper pronunciation of "Linux" from the mouth of the creator himself.

One thing I don't think I'll be likely to forget anytime soon, is the unique ending to this documentary..... A group of hackers singer and playing a song about Open Source.
I can't say they were actually bad, but you sure won't see that song on my play list...hopefully ever.

- The Cathedral and The Bazaar

Although reading this article took a while longer than watching the movie, I found myself constantly agreeing, disagreeing, and reanalyzing Eric's points over several times. For the most part, I agree with the "Bazaar"'s approach and method of constantly sending out code and getting feedback on it, to be an effective way to get things to work the way people want them too.

On the other hand, from a scientific viewpoint I find it hard to believe that Eric Raymond's "Fetchmail Project" qualifies as a proper and thorough testing of Linus Torvald's approach for programming Linux with the hacker community. Granted that it would be nearly impossible to duplicate the creation of an Open source OS, since the need for one and the community has already been mostly taken up by Linux, I don't think that just working on one mail forwarding program with an open source community (even if it did become a niche filler) is what one can call "thorough testing and experimentation" for proving the effectiveness of programming with an Open Source community.