In previous articles, I've described some of the ways that the BSD systems differ from mainstream computer software such as Microsoft. Many people see these differences as the main obstacle to BSD becoming used by a wider audience, in particular on the desktop. I disagree: I believe that as computer users become more sophisticated, they will appreciate the different approach. There are other things that need to be done before any of the BSDs can truly compete with, say, Microsoft, but they're more in the area of fine tuning than in concepts.
This month I'd like to discuss a different problem, the one that I personally think is the greatest obstacle to the spread of BSD. So I should be recommending FreeBSD to as many people as possible. Do I do it? Well, maybe. The question is, what is possible? Let's look at a few examples.
Over the last few years, Pam has investigated every known Paso horse in Australia, about 60 of them. She's a fountain of knowledge about the breed and thus the obvious choice as a registrar. Now maintaining a horse registry is not quite a typical office job: the main office activity revolves around pedigrees and transfers of ownership. Clearly an application for a database. It would also be nice to be able to print the certificates completely on a computer, thus saving on pre-printed forms. Nothing that we couldn't do with BSD.
In addition, there's the question of cost: we need to get a computer for Pam. Where do we get one? If all Paso owners in Australia became members, we would get an annual revenue of about $AUS 1500, about $US 900, out of which we need to print a newsletter and pay registration fees. What kind of Microsoft system could we get for $1500, anyway? Microsoft Office alone would cost about half that sum, and it needs a pretty hefty machine.
On the other hand, I have an old 486 lying around that I could contribute to the Registry, along with FreeBSD and enough ported software to do the job. It makes sense, doesn't it? Will we do it? No.
The problem is that Pam isn't a computer freak. She needs the computer to do a job, and she knows how to do it with Microsoft. We'd have to teach her all over again with FreeBSD, and she doesn't want to do it, and I don't have the time to do it.
The problem is, Casio has decided that they don't want anybody to know the protocol they use to transfer the data. Instead, they supply a comprehensive software package which runs under–you guessed it–Microsoft. Just for this camera, I need to maintain a copy of Windows 95, I need to reboot a machine twice, and then mess around with a complicated program where I have to set lots of flags before I can start. Finally I have to ftp the images to the machine where I want them.
To the best of my knowledge, no printer understands the JPEG or GIF formats that are commonly used for storing digital images: you need to convert them. I chose PostScript as the best intermediate format in view of the fact that I would often need to mix text and images, and that Netscape uses it for print output. Ghostscript supplies a driver for some Epson Stylus models, but not for the 740.
Never mind, Epson supplies two CD-ROMs with the printer, and they're both full. There's bound to be some technical information on one of them–I thought. Unfortunately, the most technical information on the CDs is a data sheet. I don't know what a lot of it is, because it's designed to be executed from a Microsoft environment. A lot consists of flashy images with not too much content, a sad indication of where the industry is going.
Epson isn't alone in this. In 1997 I decided that my old Panasonic KX-P 4450 laser printer, which I had had since 1990, was a little long in the tooth, and it was time for something newer. I bought an HP LaserJet 6 MP, with PostScript, which was a great improvement in speed and resolution over the aging Panasonic. In one area, though, the Panasonic beats both of the newer printers hands down: it comes with documentation, over 100 pages with exact descriptions of the control sequences that the printer understands. Another significant difference is that the control panels of the newer printers are very simple compared to that of the Panasonic: they perform most of the functions of the control panel in software. Of course, the software only runs under Microsoft.
Obviously, I'm not happy with the situation. What can we do about it?
Unfortunately, there's no single answer. Each problem has its own potential answer:
Of course, the Paso Registry won't be an all-Microsoft effort. The web site is and will remain a BSD machine, as is the system which handles Email. We're making very sure that the database will be in a format which will allow easy export of data to database systems which run on BSD, and it's likely that all added value software will also be UNIX-based.
The HP printer isn't really a problem at all. It does everything I want it to do, and in fact a manual is available–for a price. Considering the drop in laser printer prices over the last 10 years, and the fact that the number of purchasers who want technical information is almost certainly less than 1%, this is not unreasonable. There's a subtle difference from Epson here: Epson could have put the text of the manual on the CD-ROM for free.
The real problem is that there's no single solution: each little problem has its own character, and you need to solve them individually. In some cases, for example a large proportion of the software which runs on Microsoft platforms, the only thing to do is to give up and not use the software. But writing to the vendors can help as well. I suspect there's hardly a Microsoft-based software vendor out there who isn't currently evaluating Open Source. If you find a package you want, but which doesn't run BSD or Linux, write to the vendor and ask them if it's available for Linux.
Huh? Linux? Have I turned traitor? Isn't Linux the enemy?
No, I'm primarily a BSD user, not a Linux user. But all versions of BSD can emulate Linux, and we're more likely to get a Linux version of any specific package than a BSD version.