November 27, 2004

Linux Crawls Towards Ease-of-Use

From the Blogging of the President:

The Blogging of the President: 2004: In the past week, I've downloaded and run two Linux distributions: Knoppix and Ubuntu. Both are "live CD" distros, meaning that you can boot off of the CD, without installing anything on your hard drive or screwing up your Windows-based PC.... [W]hat's important is that I was able to get, for the cost of a blank CD-ROM, an OS that installed almost flawlessly on a computer. It comes standard with an office suite that includes a word processor, spreadsheet, presentation program, games, instant messengers, email (Evolution, which replicates most of Outlook's functionality and adds more)...and unlike the Linux of a few years ago, is as user-friendly as Windows and as easy or easier to set up. It's also better protected against viruses, and did I mention that it was all free?

Now, you're on the Internet, so that means you've probably been exposed to Linux evangelization before if you aren't a Linux user. I have been too. The important thing about Ubuntu is that it just works. It was easier to get running than any version of Windows has ever been. It was easier to get running than the Windows that came pre-installed on my laptop. And yet, everyone uses Windows--and either pays through the nose for it or steals it.

I have good reasons to have Windows, Office, and a bevy of other Microsoft products sitting on my laptop--I use them because everyone else does. In my case it's not that I'm a sheep, but because it facilitates the transfer of data between me, my employer, and my customers, and because I use some specialist tools that don't have great Linux equivalents.... [But] look at the pricing trend of Office over the years. The trend is not upwards. Nor is it upwards for Windows. Standard economic theory would suggest that Microsoft should be in a position to jack software prices through the roof, but in fact the cost of their applications is falling in real terms. Meanwhile, innovation is largely stagnant. One of the reasons that Open Source software is becoming viable, and even competitive with commercial applications, is that those applications are no longer moving targets.

Microsoft has not introduced any significant innovations in any of their products, from a consumer perspective, since about 2000 or so. Windows and Office 2000 were the last upgrades that introduced real value in their products. Since then, Windows has been a relatively stable operating system, Office has done everything people need, and many, many companies have stopped bothering to upgrade. Microsoft has had to move to a policy of forcing users to upgrade to newer versions of its applications to maintain profitability. And that's a lose-lose proposition for them; sacrificing backwards compatibility is likely to hurt them in the long run, because it makes it just as easy to convert away from Windows....

There are two theories about open-source. The first is that while it has always been possible for a charismatic leader to call forth immense team effort and accomplish great things by force of inspiration and example alone, such enterprises are never stable. In the long run you need either the stick of potential punishment--call it the authority of the state--or the carrot of material reward--the market--in order to maintain a large social division of labor and preserve a project across any substantial length of time. Call this process the routinization of charisma.

On this theory, those parts of open-source that will survive are those that get successfully routinized: substantial companies with big revenue flows have to say to people, "We've hired you to work on open-source, specifically Linux." Why would a company do this? Well, why was Microsoft so eager to make sure that the hardware specification for the IMB PC remained open? Anyone with a substantial market position in a neighboring segment can profit enormously from the expansion of the market that free-as-in-beer product provides. This is what IBM is doing now by supporting Linux: trying to remove or at least reduce the Windows tax on computers, and so grow the market in the business-services segment that IBM is close to dominating.

Think of it this way: Microsoft is like the late Roman Empire, IBM is like the Huns, and the Linux programmers are like the Goths. IBM's support of Linux is the analogue of the Huns driving the Goths before them to soften up the Roman Empire at the end of the fourth century.

The second theory is that it has been difficulties in communication and organization that have prevented the running of persistent, large-scale divisions of labor off of other human motives than fear and greed, but that now--thanks to the IT revolution--costs of communication and organization have dropped far enough that simple enthusiasm, curiosity, or the desire to demonstrate technical skill can attract enough part-time workers who can be coordinated enough to make meaningful contributions to an ongoing project.

It will be interesting to see what the answer is.

Posted by DeLong at November 27, 2004 02:16 PM | TrackBack
Comments

Like so many articles of its kind, there was no mention of internationalization issues. Support for double-byte characters (Japanese, Chinese, etc.) is still vastly better in Windows and Mac OS than with Linux distributions.

Also not mentioned is that Unix is in many ways a step backwards for computers. It was designed in an age when only computer pros used computers. Now that everyone and their grandmother is using computers, there have been many attempts to put a friendly face on Linux and other Unix derivatives; but other than Mac OS X, they are only partially successful. The Unixness still pokes through in too many areas. Those of us who run our own businesses and must be our own computer administrators, even though our expertise may be in other areas, tend to be overwhelmed by the arcane administrative requirements in setting up Linux.

A third issue is compatibility with others. As a translator from Japanese to English, I must deal every day with MS Office files in Japanese, including complex PowerPoint and Visio files, and must translate directly on those files before returning them to the customer in a usable form. Sending these files through some alternative office program even on the same platform, let alone across platforms, has never, ever produced satisfactory results.

Someday, maybe. But we're not yet there.

Posted by: John de Hoog at November 27, 2004 03:13 PM

There seems to be a huge - but so far missing - wildcard in the open source arena: The needs of third world government bureaucracies coupled with the increasing availability of very good local programming prowess.

Governments of countries like India, China, Brazil, Russia, etc have to purchase hundreds of thousands of software licenses. They could instead set up teams of a few dozen software engineers to provide open source substitutes for expensive Microsoft (and other proprietary) software.

By keeping the software open source, they get to leverage the efforts of a worldwide community, who in turn profit from the systematic resources these country teams would provide. Quite a strong synergy, one would think. Could easily spell the death of much of proprietary software. Definitely of software like wordprocessors, spreadsheets, os, databases, etc - that are used by the majority of users.

Posted by: jimvj at November 27, 2004 03:29 PM

Well, some of the late Roman Emperors were Goths. I think that we should work Emperor Gustavus I into the allegory somehow.

Theodoric's Gothic church still stands in Ravenna. It's Romanesque. All the other Gothic cathdrals are fakes.

Posted by: John Emerson at November 27, 2004 04:02 PM

Well, some of the late Roman Emperors were Goths. I think that we should work Emperor Gustavus I into the allegory somehow.

Theodoric's Gothic church still stands in Ravenna. It's Romanesque. All the other Gothic cathdrals are fakes.

Posted by: John Emerson at November 27, 2004 04:04 PM

> This is what IBM is doing now by supporting
> Linux: trying to remove or at least reduce the
> Windows tax on computers, and so grow the market
> in the business-services segment that IBM is
> close to dominating.

Ah, but you have to factor the GPL in as well. If 10 or 20 large entities start making public contributions to Linux it could become self-sustaining, as: (1) there is no way for those contributions, once made public, to be retricted (2) but the total cost for the entities (and I am assuming here they are consumers, not IBMs) remains smaller than the cost of purchasing propriatary software.

So there would be a positive feedback loop encouraging those entities to continue contributing and for others to join.

One also has to consider the long-term effects of Digital Restrictions Management (DRM), which will be built-in to the next generation of Windows. It is very hard to see that as anything other than a long-term strategy by Microsoft to seize and tax corporate data.

Cranky

Posted by: Cranky Observer at November 27, 2004 06:09 PM

> Also not mentioned is that Unix is in many
> ways a step backwards for computers.

Wipes tears from eyes. I have used many different computer operating systems over the last 30 years, and have spent the last 10 helping clients convert to Microsoft-based systems at their insistance.

With that perspective, I can say without much fear of being contradicted that the ascendency of Windows was not only the largest step backwards ever taking in computing, but possibly also the largest step backwards ever taken in the history of human technology. And I would nominate it for largest step backwards in human history, although admittedly it has some competition in the Black Death.

Oh well. Off on Monday for a 3-day trip to a client site to excise spyware from their Microsoft Windows desktops. All the safeguards were in place so it must have come in through those unpatched Internet Explorer bugs that Microsoft finally admitted on Wednesday do exist. Maybe after this $3500 bill arrives the client will agree to convert to Mozilla ;-)

Cranky

Posted by: Cranky Observer at November 27, 2004 06:18 PM

Brad, why do you think the answer is either one or the other?

Posted by: Chris Hanson at November 27, 2004 07:23 PM

Cranky Observer wrote:

"With that perspective, I can say without much fear of being contradicted that the ascendency of Windows was not only the largest step backwards ever taking in computing, but possibly also the largest step backwards ever taken in the history of human technology. And I would nominate it for largest step backwards in human history, although admittedly it has some competition in the Black Death."

I'm not saying what kind of step Windows was, but rather that looking to the ancient Unix base for the answer is not a forward-looking move. The BTRON project in Japan is more the kind of thing I have in mind. A computer architecture needs to be premised on the reality of who uses it. Unix was premised on professional use; the Mac was premised on use by people with specialties in non-computer areas. BTRON is premised on universal use. (If you have been using computers for 30 years and helping people for ten, you are obviously a professional and biased toward Unix.)

Posted by: John de Hoog at November 27, 2004 08:00 PM

"Those who don't understand UNIX are condemned to reinvent it, poorly." Henry Spencer.

John de Hoog seems to be conflating computers (operating systems) with applications: the applications are what we interact with for most things -- browsers, word processors, spreadsheet apps -- while the operating system, be it Windows, Mac OS, or some variant of UNIX gets much less of our attention -- unless it misbehaves.

I submit that no one gives a rat's tail about the architecture of their operating system or their hardware as long as they can write, calculate, look stuff up, read and print reliably with a minimum of hassle.

UNIX's strength is that it offers a variety of small discrete tools instead of a monolithic all-inclusive approach. Many of the UNIX (and now OS X) file utilities and tools do one or at most a small number of things well and can be made to work together.

Word or any other MSFT application offers to do everything but darn your socks, whether or not you want it to. The notion that a system should be designed around its users describes UNIX-like systems pretty accurately: I would be hard pressed to find the user who finds the latest versions of Office to be just exactly what they need.

Separate the OS from the applications and re-think your argument: consider how you interact with the systems you use and how much you enjoy the experience (or do you just endure it?).

Posted by: paul at November 27, 2004 08:53 PM

"Those who don't understand UNIX are condemned to reinvent it, poorly." Henry Spencer.

John de Hoog seems to be conflating computers (operating systems) with applications: the applications are what we interact with for most things -- browsers, word processors, spreadsheet apps -- while the operating system, be it Windows, Mac OS, or some variant of UNIX gets much less of our attention -- unless it misbehaves.

I submit that no one gives a rat's tail about the architecture of their operating system or their hardware as long as they can write, calculate, look stuff up, read and print reliably with a minimum of hassle.

UNIX's strength is that it offers a variety of small discrete tools instead of a monolithic all-inclusive approach. Many of the UNIX (and now OS X) file utilities and tools do one or at most a small number of things well and can be made to work together.

Word or any other MSFT application offers to do everything but darn your socks, whether or not you want it to. The notion that a system should be designed around its users describes UNIX-like systems pretty accurately: I would be hard pressed to find the user who finds the latest versions of Office to be just exactly what they need.

Separate the OS from the applications and re-think your argument: consider how you interact with the systems you use and how much you enjoy the experience (or do you just endure it?).

Posted by: paul at November 27, 2004 09:03 PM

<quote>If either is to survive and prosper, it will be the one, which is willing to cut off its own "air supply." In this regard, I think Microsoft's adoption of XML in the format of its Office documents is most interesting. The XML format is as a "open" as the HTML of a webpage; Microsoft's previously proprietary file formats have been partly displaced by a file format anyone could produce with little effort. The advantage to Microsoft is in the server arena -- IBM's world.</quote>

It remains to be seen how "open" MSFT's XML file formats are: lock-in is as key advantage for them and they're not giving it up anytime soon. Doo a Google search for "office XML file formats lockin" (I'd make a link but HTML doesn't work in comments).

And now to press Post just once and take it on faith that everything works . . . . sorry for the earlier duplicate.

Posted by: paul at November 27, 2004 09:34 PM

I don't claim that the small simple tools approach beats the "Swiss army knife" approach for everyone. It depends on the task, obviously. I would never claim that UNIX is user-friendly anymore than I make that same claim for electricity.

What I find interesting is the use of terms like "computer pros" as if most people use computers vs applications. Most of the work of the past 20 years has been directed at levelling the playing field: it sounds like it hasn't worked all that well. As for people going "outside the box" and messing things up, that's a lot harder to do in UNIX-based systems than in Windows or old school Mac OS.

I don't know problems people who take that position think are worth solving: internationalization is underway, as noted above. Standards compliant data (XML) is available with OpenOffice. Standards compliant browsers are available for all platforms. sure there are bloated and flaky applications and operating systems but I think there are a lot of choices, more every day, and the tools to make your own solutions are getting easier to use as well.

Maybe this isn't the right venue for this discussion: Professor DeLong can weigh in on that. But I would like to know where the shortcomings are in desktop open source tools for a given task. What does OpenOffice (for example) not do that the Leading Brand does?

Posted by: paul at November 27, 2004 10:35 PM

> But it's hard to think of a software category
> where open source wouldn't benefit any number of
> large companies sufficiently to entice them into
> backing the effort. Consider the price of 10,000
> seats of OpenOffice...

I think an Open Source ERP project would really put the cat among the pigeons...

Cranky

Posted by: Cranky Observer at November 28, 2004 05:59 AM

Linux has been natively utf-8 since glibc-2.1 (it was in 2.0, but not compulsory.) That was RH8.0, I think. The glyph sets are not complete, but nobody's are.

Windows uses utf-16, except where it's ucs2, or a codepage, and it isn't quite utf-16, and many older apps do strange things, including writing xml in utf-16 without adding an 'encoding' attribute to the xml header line, which is a wonderfully reliable way to make a standards-compliant XML parser emit gibberish.

So I'm not sure where the 'lack of localization support' is coming from; I have to deal with lots of localization issues at work, and Windows has lots of quirks in that respect.

(Tangentially, here is no way to make something that does complex things less complex than the complexity of the things themselves. You can expose only a simple subset of things as being adjustable, but that's not quite the same thing, and it's a limitation on utility in any case.)

Posted by: Graydon at November 28, 2004 07:22 PM

Graydon writes:
"Linux has been natively utf-8 since glibc-2.1 (it was in 2.0, but not compulsory.) That was RH8.0, I think. The glyph sets are not complete, but nobody's are.

[snip]

"So I'm not sure where the 'lack of localization support' is coming from; I have to deal with lots of localization issues at work, and Windows has lots of quirks in that respect."

First, "natively utf-8" does not make Linux magically internationalized. Most of the Japanese and Korean text out there, for example, is still iso-2022-xx encoded.

But another problem is that there is no universal method in Linux for the conversion needed to enter CJK text, or for switching to CJK input. It also seems that individual applications need to be called using obscure command lines just to get them to use, say, Japanese. And then there are the awful font problems. As bad as things may be in Windows, they are far worse in Linux in this respect.

Posted by: John de Hoog at November 28, 2004 08:06 PM

Brad: Unless it's an act of censorship, I would like to point out that a bunch of comments seem to have disappeared from this thread.

Posted by: cm at November 29, 2004 01:45 AM

Amiga. Remember?
If the morons who followed their IBM corporate blue 'compatibility' line so long ago (has it BEEN 20 years now?) would have known a damn thing about computers, we wouldn't be in the mess we are now. The most advanced personal computer was Unix-based. Why? Because the logical arrangement of commands and directory structures had been researched by decades of university professionals, instead of the one night stand known as MS-DOS.
Translation, voice recognition, graphics, user interfaces would all be better if they didn't have to contend with Microsoft's lame-ass ideas of data handling and Intel's manure spreader interrupt architecture.
Not to mention the secret hidden directories capturing all of your searches and email addresses.
Linux may not be perfect, but you can make it so if you think you are so damn smart. Just shut up and code.

Posted by: auntiegrav at November 29, 2004 09:00 AM

Gradon wrote:
"Tangentially, here is no way to make something that does complex things less complex than the complexity of the things themselves."

Drag a window with a pointing device. You have accomplished something complex with a less-complex action. A good user interface is always designed to do just what you say above cannot be done.

Or, run a Google search. It's extremely simple for the user, but an extremely complex operation by the service provider.

"You can expose only a simple subset of things as being adjustable, but that's not quite the same thing, and it's a limitation on utility in any case."

There are complex and less complex ways to adjust things. A good user interface provides a control model that is both efficient and easy to grasp without special knowledge. Most of the things that are still difficult to do in Linux are that way because no one has yet bothered to make them easier. Then we get people saying things like "Just shut up and code." That's exactly the reason Linux is used by only a tiny segment of computer users. That's not necessarily bad, just the reality that the original post at the top of this page glosses over.

Posted by: John de Hoog at November 29, 2004 03:39 PM

> Or, run a Google search. It's extremely
> simple for the user, but an extremely complex
> operation by the service provider.

Except that when you _really_ need to find something, you still have to go to Lexis or a similar fee-based database and have someone who understands searching (if you don't) find it for you. Because searching is in fact a very difficult problem.

Google has put a friendly face on the 80-20 rule. But even Google is struggling with making its results relevant as too many searchees have gamed its results.

Cranky

Posted by: Cranky Observer at November 29, 2004 07:41 PM

Sounds like a tired discussion about yesterday's news.

Kodachrome gives you the nice, bright colors.

Posted by: BangAGong at November 29, 2004 08:25 PM

> There are two theories about open-source.

Well, allow me to advance a couple of my own then... :)

1. Open Source is a labor market signalling facility.

Back in 1973, Michael Spence showed that under certain conditions, well-informed agents can improve their market outcome by signaling their private information to poorly informed agents. Applied to the open-source software development, this can mean that participatoin in an open-source project can be a singal to a potential employer showing that an applicant without an employment record may nevertheless possess valuable development skills, which may include version control, managing a geographically dispersed development team, or whatever else a particular employer is looking for.

The value of such a signal should be the greatest in a tight job market, when employers are cost-conscious and applicants are numerous. Acquiring the experience, on the other hand, is the easiest while in school (quite a few open-source pursuits grew out of term projects and dissertations). Hence, a testable prediction: open-source development should be concentrated in places where tight job market (and especially high youth unemployment) coexists with affordable technical education/training (some of which, in turn, may be youth unemployment in disgiuse). Anecdotally, this sounds plausible. Finland, for example, is home to both Linus Torvalds and Monty Widenius; neither lives there anymore...

2. Open Source development is an option on related business

If you believe that giving away a piece of software can increase your chances of being paid for related products or services, the cost of developing the software can be thought of as buying an option on that related business. Given a high degree of uncertainty surrounding those would-be revenues, the option can be quite valuable (meaning, a lot of time and/or money will go into a piece of software that may or may not bring in any related revenue)...

Posted by: Nikolai Chuvakhin at December 1, 2004 09:13 AM
Post a comment









Remember personal info?