The Mandelbear's Musings

Recently I started reading this Ruby on Rails Tutorial by Michael Hartl. It's pretty good; very hands-on, and doesn't assume that you know Ruby (that's a programming language; Rails is a web development framework). It does assume that you know enough about software development and web technology to be dangerous. And if you're not dangerous yet,...

It points you at a web site where you can learn enough to be dangerous. Starting from knowing nothing at all.

It's the author's contention that Tech is the new literacy [and] [l]earning the basics of programming is only one piece of the puzzle. LearnEnough to Be Dangerous teaches [you] to code as well as a much more powerful skill: technical sophistication. Part of that technical sophistication is knowing how to look things up or figure things out when you don't know them.

There are seven volumes in the series leading up to the Rails tutorial, giving you an introductory course in software development. I haven't gone to a bootcamp, but I'd guess that this is roughly the equivalent. More importantly, by the end of this series you'll be able to work through and understand just about any of the thousands of free tutorials on the web, and more importantly you'll have learned how to think and work like a software developer.

The first three tutorials lay the groundwork: Learn Enough Command Line..., Learn Enough Text Editor..., and Learn Enough Git to Be Dangerous. With just those, you'll know enough to set up a simple website -- and you do, on GitHub Pages. You'll also end up with a pretty good Linux or MacOS development environment (even if you're using Windows).

I have a few quibbles -- the text editor book doesn't mention Emacs, and the author is clearly a Mac user. (You don't need a tutorial on Emacs, because it has one built in -- along with a complete set of manuals. So you'll be able to try it on your own.)

The next three books are Learn Enough HTML to Be Dangerous, Learn Enough CSS & Layout, and Learn Enough JavaScript. The JavaScript is a real introduction to programming -- you'll also learn how to write tests, and of course you'll also know how to use version control, from the git tutorial.

At this point I have to admit that after starting the Ruby tutorial I went back and skimmed through the others; I'll probably want to take a closer look at the JavaScript tutorial to see if I've missed anything in my somewhat haphazard journey toward front-end web development.

The next book in the series is Learn Enough Ruby to Be Dangerouse. (If you skip it on your way to the Rails tutorial, there's a quick introduction there as well.) Ruby seems like a good choice for a second language, and learning a second programming language is important because it lets you see which ideas and structures are fundamental, and which aren't. (There's quite a lot of that about JavaScript -- it's poorly-designed in many ways, and some things about it are quite peculiar.)

Another good second or third programming language would be Python. If you'd like to go there next, or start from the beginning with Python, I can recommend Django Girls and their Tutorial. This is another from-the-ground-up introduction to web development, so of course there's a lot of overlap in the beginning.

Another fine post from The Computer Curmudgeon (also at computer-curmudgeon.com)

NaBloPoMo stats: 593 words in this post, 1172 words in 3 posts this month.

Yesterday was the last show of the season for the Lamplighters, the mostly-Gilbert-and-Sullivan company that we've had season tickets to for over three decades. So the flower_cat, chaoswolf, B. (WINOLJ) and I headed up to San Francisco a little before noon. (The Y.D. doesn't usually like to go, for some bizarre reason, so we take a friend.)

The show was Gilbert & Sullivan Straight Up, With a Twist, written and directed by the redoubtable Barbara Heroux, the Lamplighters' artistic director. It's basically a series of selections from all the operettas, in chronological order, staged between two actors playing the roles of G and S. Their dialog is taken from diaries and letters, and provided a fascinating look into their occasionally-troubled collaboration.

There were nine singers: five men and four women. I really didn't know what to expect going in, but found it delightful. The singing was first-rate, of course, and the whole thing hung together surprisingly well as an artistic biography. The ending, after their death announcements, was especially effective: "The world is but a broken toy" from Princess Ida, followed by an ensemble rendition of "Once more gondolieri" from The Gondoliers.

I'd say, "Go see it!" but the season's over, so you lose. The program notes point to the Gilbert and Sullivan Archive if you want more information, and of course the complete libretti. Next season is The Mikado and Iolanthe.

We came home via I280 and stopped at Buck's for dinner; a delicious end to a pleasant afternoon out.

I finished off the day practicing some songs that aren't on the Baycon setlist but will probably come up in circle -- I'm sadly out of practice. And, a little surprisingly, the start of a setting for "The Collar-bone of a Hare", which started tickling my mind a couple of days ago while I was working on a (not totally unrelated) new song of my own.

Conclusions and developer recommendations

I'm through hoping that the next version of Fedora Core will fix all of the problems with the previous release. Fedora's identity has gradually eroded over six releases, finally ending up as a second class clone of Ubuntu. On the other hand, Red Hat Linux was never really all that easy to install, configure, and use, so I guess this is just the natural evolution of a product that was destined to be eclipsed by more complete distributions like Mandriva and more easily configured distributions like SUSE.

I appreciate the fact that distributions like Fedora Core are still focused on free-as-in-rights software, but today's Web content requires more proprietary browser plugins than yesterday's did, and today's hardware is increasingly designed to be dependent on proprietary binary blobs in the form of firmware and driver packages. Programmers are not falling over themselves to write free replacements for these things (or they are unable to because of a lack of documentation from hardware manufacturers), and the projects that do exist are non-operational and/or several generations behind current technology. Users do not want to hear reasons and excuses for why the operating environment doesn't work with their favorite Web sites or computer hardware -- all they know is that it doesn't work, and making it work is not a simple or obvious process. It is possible to keep the distribution free-as-in-rights while making it easy to add proprietary extras, but the Fedora Project is not willing or able to do it after six releases.

The Fedora Project has failed six consecutive times to produce a viable desktop operating system. I say pack up, move on, and let Fedora Core die, but remember it fondly as the last of the holdouts from an era when desktop GNU/Linux meant missing out on most Web media while struggling to get network drivers installed and configured. It's nice that my video cards worked with the 3D desktop effects with little effort, but wobbly windows and the cube desktop switcher don't make up for a lack of basic network functionality and ease of configuration.