Hide full article
The first issue of TidBITS is dated 16-Apr-90. I sit here, four years later, working on our 222nd issue, and think about all that has changed and all that has remained the same. Permit me a rambling and decidedly non-sequential recollection.
Tonya deserves credit for the concept for TidBITS. We were living in Ithaca, NY, after having graduated from Cornell University the year before. I was working as an independent consultant, and Tonya had the impressive title of "New Technologies Consultant" in the part of Cornell that sold computers. Unfortunately, Tonya's title translated to "Seller of Macs and DeskWriters," a task which she did along with several others. She thought a weekly newsletter of sorts might help her co-workers keep up on the industry. We figured that we could easily create such a summary, given that we read MacWEEK, PC WEEK, and InfoWorld weekly, and I regularly scanned the nets.
Tonya had an ulterior motive. Her degree from Cornell is in Communications (more appropriate than my double major in Hypertextual Fiction and Classics), and while we were students, she edited the newsletter for the local users group, MUGWUMP. But after passing that on, Tonya felt her skills in PageMaker were rusting away, and thought this newsletter might provide some lubrication, though she was concerned with potential waste of paper.
At the time, I was heavily involved with HyperCard, so my immediate reaction was that we should use the same text to create an electronic HyperCard version to distribute freely on the Internet. And so it was decided, although after only two weeks it became clear that electronic distribution was the way to go; trees would be safe from TidBITS.
You can go back and look at the early issues - I cringe every time I do. We started out summarizing the top stories in the trade rags, but quickly became uncomfortable with the legalities involved. We weren't concerned about copyright, since everything we wrote was in our own words, but there's another legal concept called misappropriation that might or might not have applied. That concern pushed us in the direction of writing our own articles, using the magazines only as sources (which we cited carefully, being good little academics).
The first few weeks of distribution were... interesting. I posted a note on the nets announcing TidBITS and a mailing list for it. I knew how to set up a mailing list on one of Cornell's IBM mainframes, so we stuck with that list for three weeks until it hit about 300 people. It was after the third issue that Tonya got the phone call saying that something with her name on it was crashing Navy computers in California. The Robert Morris Internet worm was still recent history, but a few panicked calls and email messages later, the truth came out. Certain old versions of the BSD mailer used by Unix boxes had a bug that prevented them from dealing properly with headers containing more than several hundred recipients, so when one of these machines received the issue of TidBITS (remember, it was a HyperCard stack, stuffed and BinHexed), that machine bounced the issue back to Cornell's main email computer, which looked at it, saw that there was nothing wrong, and sent it back again, repeating the entire cycle. You can imagine what this did to the Internet, but it all ended well.
We then went looking for alternate methods of distribution. The first people to come through were the net heavies who run major Internet sites to this day. After talking over the issues with them, they allowed us to post TidBITS to the moderated Usenet group comp.sys.mac.digest, which had around 30,000 readers or so back then. Not bad, from 300 to 30,000 in a week, although not all of those people downloaded and defunked each issue to read it in HyperCard.
Around this time we started uploading to the commercial services as well, although I only had an account on America Online, so other people handled redistribution for us. I no longer remember the chronology of when certain people came on, but Dennis Cohen (then, and perhaps still, of Claris) uploaded to CompuServe, Masato Ogawa moved issues to NIFTY-Serve in Japan, Jean-Philippe Nicaise redistributed issues to Calvacom in France, Riza Nur Pacalioglu (who lives in Turkey, making for a roundabout path) and later Eric Apgar of Apple uploaded to AppleLink, and Jay Vollmer and later Randy Simon took care of GEnie. Denise Petersen puts issues on Delphi, and before I was given an account, Larry Loeb and then Paul Raulerson uploaded to BIX for us. These are merely the people whose names rise to the surface of my memory - I cannot count all the folks who have helped spread TidBITS around the networked world of the Internet, commercial services, and BBSes. We owe every one of them a massive debt. Since those early days we've traded accounts for uploading TidBITS each week, and working with Chris Ferino on America Online, Ben Templin and Ric Ford on ZiffNet/Mac, Charlie McCabe and Arwyn Bryant on AppleLink, and Paul Raulerson on BIX, has enabled us to spread TidBITS far and wide.
Right after we started (he's first mentioned in TidBITS #06 and wrote about Macworld Expo in TidBITS #36), Mark Anbinder began writing articles for TidBITS. Mark graduated from Cornell with a degree in Linguistics the same year Tonya and I did, 1989, but went to work for Baka Industries, the main Macintosh dealer in Ithaca. Mark later became the president of MUGWUMP, the local users group, a post he holds to this day. He continues to write for TidBITS frequently and is the only person to whom we've ever given an editorial title. Other regular, though less frequent, contributors have chipped in as well. Matt Neuburg has written extensively on various programs, including a massive 90K review of Nisus, a special issue on the hypertext editor Storyspace, and reviews of several outliners. To close the loop, Matt was my Classics professor at Cornell before ending up in New Zealand, and his Greek Composition class taught me more than any other class in those four years. We've also published a number of articles from Ian Feldman, who created the setext format. The first year of TidBITS I wrote 90 percent of the articles, but that percentage has thankfully been declining, because I never pretend to be an expert on everything, and would far prefer to have someone who is an expert write about what she knows. Oh, if you're wondering, all those who are Pythaeus prefer to remain unnamed - "Pythaeus" is one of the names of Apollo at his oracle at Delphi.
In May of 1991, the world changed. Tonya accepted a job with Microsoft supporting Macintosh Word. We married in June and moved to the Seattle area in July. I had no consulting contacts in Seattle, so I devoted my time to TidBITS and frankly, my Internet contacts kept me sane during those first few difficult months. We realized that although we could live on Tonya's salary, just barely, it would help if TidBITS could bring in some money as well. That's when we came up with the corporate sponsorship program that has resulted in various select companies such as Nisus, Dantz, and APS providing information to interested readers.
Back to the HyperCard stack. One reason I originally used HyperCard was that my stack could merge its contents into another copy of itself, creating a single archive. Information is useless if you cannot find it, and the single stack archive helped solve this problem. Unfortunately, my stack proved equally problematic. Programming quirks caused the archive size to grow too rapidly, but I fixed that after 25 issues. The stack also devoted too much room to background decoration and navigation controls, reducing the text space. After the first few issues, Ian took me to task for the stack, and we started discussing issues surrounding the dissemination of electronic periodicals, and those discussions resulted in Ian creating setext, or structure-enhanced text.
This all took time, and in fact, we published the first 99 issues of TidBITS in HyperCard format. My master archive of all the issues had increased to well over 10 MB, and merging an issue took a long time. It didn't look as though we would have a HyperCard browser for our setext files any time soon as 1991 drew to a close, but I couldn't live with HyperCard any more. TidBITS #100 was our first issue in setext format, and in one of the chronological conjunctions we like so much, it was also the first issue of 1992.
Switching to setext format was terribly important. Every previous issue had to be stuffed and BinHexed before being sent out, forcing everyone to jump through hoops to read it. This limited readership to those who could download to a Mac. Once the issues were in setext format, everyone who subscribed to comp.sys.mac.digest could easily read the issues without additional processing. Currently estimates place comp.sys.mac.digest's readership at about 75,000.
Setext format opened distribution doors in other ways. Alvin Khoo of Simon Fraser University set up a mailing list that garnered over 1,000 subscribers before his home-brewed mailing list software and the SFU machine had trouble with the volume. Luckily, Mark Williamson of Rice University saved the day with the Rice LISTSERV, so we transferred everyone over. The LISTSERV list has grown steadily since the spring of 1992 and now serves about 8,000 people. Also because of the setext format (which looks like plain text but is implicitly structured for decoding by special programs), TidBITS appeared on some Gopher servers and Ephraim Vishniac of Thinking Machines created a WAIS source for it, enabling anyone on the Internet to search the complete text of all issues. A World-Wide Web server is up as well, and I'll announce that officially soon, probably next week.
We still had no browser for setext, though, and no way of creating an archive of all the issues, which was one of my original design goals. I've used a Nisus macro to encode issues since TidBITS #100, but my Nisus macros for decoding setext never worked right. In August of 1992, Akif Eyler released Easy View 2.1 with the capability to browse setext files. Easy View not only had all the features of my simple HyperCard stack, but it could do things like extract all articles that contained a search match. Since Easy View worked on the original setext files, we didn't have to modify our distribution at all, although over the years we've tweaked the format of the issues to make them more attractive for reading in Easy View.
Software reviews were a major step for us. I remember the first time I was sent a commercial program to review, Now Utilities 2.0. Being in poor college-student mode still, I couldn't believe my good fortune and wrote an in-depth review for TidBITS #45. Other products slowly followed suit, including the long-gone Kennect Drive 2.4, and MacInTax back when it still came from SoftView (then purchased by ChipSoft, which recently merged with Intuit). Needless to say, we've looked at many other programs over the years, but I think I'll always have a soft spot for Now Software for that day when Now Utilities 2.0 arrived on my doorstep. It's easiest to talk about products we use regularly, and of all the programs we've used over the years, the constants have been Nisus, uAccess (now UUCP/Connect from InterCon, a full-featured UUCP-based email program), and QuicKeys.
Hardware-wise, we've evolved slowly. The first TidBITS issues were produced on a 4 MB double-floppy SE with a 30 MB home-built hard drive. It eventually transmogrified into an SE/30 with an APS external 105 MB drive and 5 MB, jumping to 8 MB relatively quickly. My strategy was to keep that SE/30 viable, so I added an APS SyQuest drive for backup, a Micron Xceed video card and an Apple 13" color monitor (and since then have refused to use any single-monitor Mac other than a PowerBook). Our second Mac was a Classic with a 40 MB drive that we actually bought for the floppy drive - our SE/30 only had a single 800K internal drive and it was dying. A new SuperDrive was only slightly less than a Classic without a hard drive, although we weren't able to resist the hard drive model. On the whole, the Classic was a mistake - we seldom use it and it's painfully slow. In August of 1992 I jumped the SE/30's memory to 20 MB and adopted my seemingly unusual technique of launching all my standard applications at startup, which makes scheduling easier and simplifies single-key program switching with QuicKeys.
Our third Mac was an extremely cute 8 MB PowerBook 100 with a 20 MB drive that we got during the PowerBook 100 fire sale. I use the PowerBook for most of my serious writing - I wrote some of Internet Starter Kit and all of Internet Explorer Kit on the PowerBook. The hardware purchases that made writing the first book possible were an APS 1.2 GB hard drive and an APS DAT drive for backup - books suck hard disk space and nightly scheduled backups with Retrospect have eased my backup paranoia significantly.
When she started writing her book on Microsoft Word, Tonya bought a Duo 230, which I'm basically forbidden to touch. Eventually, in November of 1993, I broke down and replaced the SE/30 with a Centris 660AV. Tonya snagged my Apple 13" color monitor for double-monitor use on her Duo with a MiniDock, and I switched to the combination of an NEC 3FGx 15" color monitor and an Apple 12" monochrome monitor that I bought used. The only major thing I regret about the 660AV (other than the fact that the speech recognition doesn't really work) is that the Curtis MVP Mouse trackball and footswitch that I used stopped working. I've switched to a Kensington TurboMouse trackball but still miss the footswitch.
The pointing devices remind me of perhaps the worst problem we've faced and continue deal with daily. In early 1992, Tonya injured herself and ended up with tendinitis in her hands and arms. Shortly thereafter, I was diagnosed with carpal tunnel syndrome, which has similar symptoms. We were pretty pitiful for a few months, wearing wrist braces in bed and to go grocery shopping, but we've gradually recovered. The most important factor in our recovery was the realization that a repetitive stress injury is related to extreme amounts of stress that must be reduced. In addition, these silly lycra gloves called Handeze Gloves have made an incredible difference for both of us.
I could ramble for a lot longer, and if I went back and read through my outgoing mail I might remember even more of the stories that make up our history. But what's important about TidBITS is people - the people who have redistributed issues, the people who have contributed articles, the people who have read the issues, and the people who believed in us for years before electronic publishing was conceivable to most publishers. I've always said that I write to the person behind the personal computer, and as my interests lean more and more toward the Internet, I believe all the more in the importance of the individual. This is why we avoid corporate-speak and distribute TidBITS for free. It's worked for four years and 222 issues and although I never predict anything more than a year or so in the future, another four years seems no more inconceivable than the first four were.
