.comment: Ain't Anti-Aliasing Amazing? - page 4
The Kindness of Strangers
First, there's scarcely enough praise available for Keith and Waldo, who one imagines had better things to do this weekend than answer user questions about why something wasn't working. In my experience in Linux this is not especially rare--which tells us that the developers of open source software that they give away take their work even more seriously than do the wide range of commercial developers. I've noticed this especially among the KDE folk, though I've spent a lot of time in KDE so they're the ones with whom I've most dealt.
Too, there have been a lot of "mere" users who contributed suggestions, even on Christmas. I suppose that this has something to do with the community and the fact that the people with whom we associate online are just as big a part of our lives as those we see face to face. The news reports are full of stories of online "friends" who meet little girls (or boys) for terrible purposes. There's never a mention of the tremendous good that our community does.
So any discussion of the fact that it just didn't work here needs to be illuminated by the light of a lot of people I never met offering suggestions and advice, all motivated by a genuine desire that somebody they never met succeed.
Beyond that, though, it's time--no, past time--for a comprehensive convention for the handling of typefaces in Linux. When I wrote about the problems of XFree86 a couple of weeks ago, one of the notes I received in response was that IBM should be encouraged to port it's excellent Work Place Shell to Linux, in large measure because it contained a great printing engine. True. It also contained a magnificent typeface installer. There was a single typeface repository whence applications got their screen and printer typefaces and fonts. Adding new typefaces happened there, easily. Likewise, the printer queue was both understandable and functional. (There are advantages of ease of use beyond beloved relatives having the opportunity to grow tired of one's eccentricity during a holiday weekend.)
The WPS would certainly be an answer. There's probably room in the open source world for such an approach to be improved upon. But here again we're up against the big open source test: what are we trying to do? The estimable Eric Raymond has written that to code is to scratch an itch. True enough, but that only goes so far. Programming to satisfy oneself is fine, but Linux seems to have graduated to the point where the programming is now largely designed for others--the great unwashed masses, in many cases, those in the thrall of Microsoft who seek an alternative. I know serious Linux people who have never hooked a printer to their machines, and kernel hackers who wouldn't dream of putting XFree86 on their machines at all.
This is where the distributors, at least those who suppose that Linux ought to be a desktop operating system, need to dump some of that IPO money that was left after the party. It isn't that tough. It just requires some work on the unglamorous task of establishing a good typeface system that serves both screen and printer, with an API that's available to applications one and all.
That having been said, Happy New Millennium! Sources tell me that Linux-2.4 will be released before the following millennium arrives.