millennium | ten

I was watching some Linux desktop videos, and a number of things occurred to me...

...it seems like most folk think that the Start Menu is the greatest invention in modern application launching UI. I've blogged about this before. Microsoft has spent years following the introduction of the Start Menu in Windows 95 mitigating for the astounding crappiness of it as a convenient way to get at your programs (especially when so many applications flood it with useless submenus consisting of program shortcut + crappy release notes documentation that you never read). I guess the Linux distros have improved that aspect, but virtually none of the improvements Microsoft has made seem to have been carried over (like MRU lists, etc). I'm sure I'm overlooking MyWickedCoolStartMenu 0.1.9.7, but it's not what most people making these vids have installed.

No, my guess is that folk don't care about the usability of these app launching menus on Linux because they don't have to use it. They launch everything from the command line, so this key aspect of computer use falls to "it has a UI therefore it must be usable". (To be fair, I use a command line to launch apps too - albeit one more usable than the terminal... it's called QuickSilver - and no one ever seems to demonstrate an app like that in any of these videos so I can only assume that if one exists it's unpopular).

And the videos are most often of things like repartitioning the hard disk (which I can assure you I do three times a week), poking around in the terminal, and chatting on IRC. And XGL was cool to watch once or twice, but after that, the effect is sort of lost unless you're the one driving. What I really would love to see are people making videos of accomplishing the sorts of tasks that I think regular folk do - like attaching a digital camera and having the pictures come up in some simple photo organizer. I know this is possible on Linux, a Novell guy showed me! This is orders of magnitude more impressive than repartitioning your hard drive because Windows has not been able to recognize my new camera without complex installations and driver discs.

I know I know, I'm being the grinch, people can have fun and create content that's interesting to them - that's what Web 2.0 is about, right? The comments I'll get for this will say that Linux is about the freedom to express your identity through your partition table. Fine. I just wish folk would see that there's another freedom at stake for a wider audience - the freedom to choose something other than what gets shoved down your throat, and right now in a lot of ways what's getting shoved is still more palatable (and Windows Vista is in my opinion incredibly pretty). I'm just a little sad because I watch some of these videos and I see the fine work of projects like Tango shoehorned in amongst a mess of other ghastly looking crap, being put to use in what Apple commercials would call "dull little tasks" and I have to wonder if an opportunity is being missed? Just because some large organizations take years to produce bloated, derivative UIs (I have participated in at least one such initiative) doesn't mean that that needs to be the standard.

I harp on this because I believe that having open systems for the sorts of things we use every day - our desktop OS etc is the right way to go. But I'm a realist - people won't make the switch until they see something that's an order of magnitude (or several) better. Judging from the promotional material being turned out by users, we ain't there yet.

I've been introspective lately, and the topic I was thinking about today was Open Source, more specifically why I felt so attracted to it when others turn up their noses. I thought of a few reasons...

• that if you write something useful enough people are grateful to have it for free (as the distributions usually are) but also feel empowered by the freedoms associated with the license, even if they never use them, and that makes for a generally happier person (goodness knows you're already told enough in day to day life what you can't do!)
• that the source code lives on, to be used (either literally or as a reference) by other folk in other projects. Firefox is a great example of this - the reuse of a huge amount of code written for the Netscape 6/7 series browsers that was repurposed and modified to produce a great modern browser.
• that it's a way to give something back for all the benefit that I get from the internet today, the websites, the software, the infrastructure.
• that you can sometimes meet interesting people (I have!) and visited far flung places (yep!)
• that the people that you work with, no matter what their persuasion, are usually united in the care for the software in question, and that on average this results in higher quality work compared to places where people are paid to do work they don't enjoy.
• it's a great way to get feedback and testing from more people (and system configurations) than you could test yourself.

A lot of people have a negative opinion of open source for one reason or another, but I like to think that in general this is due to a bad experience that doesn't need to be the norm. I've had the ability to speak to a lot of people contributing to a variety of open source projects in the past couple of years; and in the truly healthy ones, the expected disagreements and flamefests aside, the contributors believe this is the best and most natural way to develop software.

What a wonderful world we live in today, with MacOS X. Lots of decent software, and a slick, simple operating system. A pity though, since the system is not free software, and as developers we are restrained by Apple's shipping schedules to find out more about interesting new developments in API land.

Over the past few years, so many pieces have come together in Linux software for building a competitive desktop operating system, but so far none of really "clicked". Ubuntu is the latest and greatest, but still has some distance to go before it reaches the elegance of the Apple system. Elegance is more than just skin deep - here are the key things I think should be done in a Linux Desktop operating system:

Standardized packaging

There seem to be a variety of different package formats, and people talk of using distribution-specific tools (command line, graphical, web site) to find software for their distribution and install it. This seems to me like a huge waste of time, compared to Mac and Windows where there's more or less one way of doing things and the convention is you go to the software publisher's site to get the latest and greatest, rather than having to wait for it to come through your distribution's package tool.

Saner application and library installation and directory layout

A Unix system typically has a lot of different places for programs to be installed. Files and folders end up all over the place. It's about as bad as Program Files is on Windows. Why not mask off the core Unix-style command line utilities behind a curtain somewhere like OS X does (still available through the Terminal app) and create a new bundle structure for applications such that their install location is obvious and installation and removal is as simple as adding or removing from the Applications folder?

(I know, this system has flaws for apps that leave behind shared frameworks, but those are usually the exception more than the rule)

HIG applied more or less consistently across all "core" applications in the system

From the web browser to the network status tool, all should feel like they were painted by the same brush.

Smaller install, with fewer packages focusing on just what people need.

The ideal distribution would ship with only one of a specific class of app. Not even two with one default and another available but not default. That's what the internet and downloading and the wonderful ease of installation afforded by simpler standardized packaging allows for.

An easy to use but powerful application development toolkit

The system software should be built with the same toolset offered to applications. It should be powerful but easy to use. It should reinforce the HIG with its very use. Simple but useful patterns repeated to give developers and users a sense of understanding and mastery over their systems. Use of standard technologies for data storage, UI etc should be the path of least resistance.

MacOS X possesses many of these characteristics, and I think it's part of the reason it's such a joy to use. The hacker in me wants more freedom though, and I look forward to the day when the open source alternative reaches the same level.

"Blu-ray disks can store more data, which will allow studios to add more behind-the-scenes information, say backers."

Yeah, I call BS on that one. Today, you can buy TV shows on DVD where the production company fits three or four whole 22 minute episodes on one DVD. DVDs can hold much more data than that right now and it's wasted so that people can use more plastic and create annoying multi-disc sets where you're always having to get up and change over. Maybe some subtle campaign to get us to watch the anti-piracy screen more often?

There are entire industries that don't seem to have much of an online presence, I guess that's because their clientelle don't shop online. Makes sense.

I have strange fetishes for certain products like switchgear. Fits in with my ranch-lust. Occasionally I get the urge to find info about old mechanical toggles online, and I am usually thwarted. No one buys consumer switchgear like this online it seems, or cares enough about noisy switches instead of today's silent models. Once we get a house to tinker with, I will have to take the obvious next step of going into a store.

I started working on websites ten years ago, this April Fool's day. Not at this URL, or even with the same site name. But since then I have used this site as a means to expand my knowledge of technology, to grow my abilities as a programmer, as a designer, as a writer. Working on this site has been one of those great mysterious magical things that inadvertently changed my life. I was able to get involved with the Mozilla project back in early 1999 because of the skills I had developed writing JavaScript for this site. Everything else has followed from there.

I want to say thanks to everyone who's helped me on my journey so far, and to everyone, happy new year.