Archive of UserLand's first discussion group, started October 5, 1998.

Re: Some open questions for Doc Searls

Author:Eric Kidd
Posted:5/1/1999; 5:46:23 PM
Topic:Linux' wide open spaces
Msg #:5514 (In response to 5512)
Prev/Next:5513 / 5515

Our plan is not to "go non-technical." It's to add non-technical content, in any one or more of several ways: 1) within LJ, resulting in something like Byte at its best; 2) in a separate publication that would perhaps originate as a section of LJ; 3) on the Web; 4) some combination of the above.

I'm glad to hear that one of the best geek magazines in print will keep its technical content! I'll have to go renew my subscription; the LJ has always been a good buy.

Don't worry about my either/or logic. History suggests that you can't have a half geek, half non-geek magazine for ever, but you can keep it up for quite a long time.

Let me put it this way: I have been told that I had best deliver my editorial matter written in VI. And learning VI for a client-with-a-mouse guy like me is a major bitch; although I'm glad to do it.

Oh, just smile and tell the vi guys to shove it. I never touch vi and I do just fine. My normal editor is a heavily customized Emacs, and my "emergency rescue editor" is ed.

There's a lot of good text editors out there, and you should feel free to choose any one that suits your style. As long as it can save a plain text file, you'll be fine.

As with many corporate creatures in the Linux world, there is an extreme alergy within LJ and its parent, SSC, to anything from the non-Unix world, especially from Windows and Mac (unless it's really cool, like the BeOS).

Some of this is simple bigotry, granted. But have you heard the popular Microsoft/Netscape phrase "eating your own dogfood"? The developers of Windows 2000 and Mozilla use the pre-release products for day-to-day tasks so that they have a real incentive to make things work.

Many Linux companies will try to use Linux for the most unlikely of tasks. Quite frankly, the Linux desktop is still "dogfood"--we know this. But by trying to use it anyway, we lay pathways for less adventurous users.

Case in point: if you want to do serious publishing, you have to use Photoshop. While the open source GIMP is a marvelous tool, and very much a replicant of Photoshop, it won't create the CMYK files that are necessary to do color separations and publish magazines.

I used to do a lot of prepress work on the Macintosh, and I'll agree that Linux is a lousy choice for most of these tasks. The GIMP is a huge step in the right direction, but we still need replacements for QuarkXPress, FreeHand and a dozen other programs.

LJ's graphics people have pleaded with the GIMP development community to give the world CMYK compatibility, and nobody is interested.

I'm interested. My friend Jason is interested. I could find you another half-dozen people who are interested. But some of us are short on time and most of us are already working on other projects.

But first, could I ask a question? Is CMYK useful to you without color correction?

If not, there's no point in implementing CMYK. As I discovered at a previous job, there's a firm which holds patents on color-correction. They don't actually write software; they just sue people. Most recently, they attacked Apple over ColorSync.

Several members of the GIMP community want to add prepress features to the GIMP. But they assume, rightly or wrongly, that prepress features are useless without color-correction. Since they're legally prohbited from implementing color-correction, they haven't done work on any prepress features.

Quite frankly, the color-correction patent in question is bogus. The USPTO is massively incompetent, and software patent law is broken beyond belief. But none of the GIMP volunteers want to get sued, so you'll have to live without color-correction for a long time to come. :-(

But hey, it's early. And I'm excited, no matter where this thing goes.

Yup. Linux isn't just a great OS, it's a fascinating political and economic phenomenon.

I sat down the other day with a bunch of source code and tried to estimate the number of volunteer hours invested in Linux and GNU. My best guess: they represent over 5 billion dollars worth of donated time, counting only the major projects. That's "billion" (1,000,000,000 for you Brits out there).[1]

This means that annual productivity of the Linux community exceeds the GNP of many small countries. :-) For comparison purposes, Microsoft spends a couple of billion per quarter on R&D. The active Linux community is huge.

Cheers, Eric

[1] I calculated this number by counting lines of source code and looking in my software metrics books. Then I multiplied the estimated numbers of man-years by the annual salary of a good Unix programmer (as reported by ZD's Smart Reseller) and a standard overhead rate. At each stage, I chose the most conservative values available. So US$5 billion is a pretty reasonable, order-of-magnitude guess.

There are responses to this message:

This page was archived on 6/13/2001; 4:49:41 PM.

© Copyright 1998-2001 UserLand Software, Inc.