Last week's poll asking about your oldest regularly used program proved fascinating in a number of ways, not the least of which was in the enthusiasm it generated on TidBITS Talk, where we heard about the many old programs still in regular use throughout the Macintosh world. I was surprised that messages to TidBITS Talk had little duplication - most people seemed to have different favorites.
<http://db.tidbits.com/getbits.acgi?tbpoll=32>
<http://db.tidbits.com/getbits.acgi?tlkthrd=969>
The Numbers -- We weren't surprised that few TidBITS readers use only recent programs - only 9 percent of respondents said their oldest program was less than 5 years old. We also weren't surprised by the number of people who use programs that were 5 to 10 years old - a total of 68 percent of respondents. For instance, the oldest program I regularly use turns out to be Peter Lewis's Finger client, which remains my preferred method of doing whois lookups on domain names despite its last revision in 1994. However, a whopping 23 percent of respondents still used programs that were 11 or more years old. Considering that the Macintosh has only existed since 1984, we were stunned that such elderly programs were still in regular use.
To be fair, had we broken out the years past 11, the answers probably would have fit a standard bell curve with the top of the curve at about 7 years, so perhaps it's not so surprising. Still, the mere fact that programs from 1984 run in Mac OS 9 on today's G4-based Power Macs is testament to how well Apple has handled backward compatibility over the years and to the code created by those early Macintosh developers.
The Rationale -- The reasons why people continue to use these programs vary widely.
Many programs, particularly small, single-purpose utilities such as the Prairie Group's DiskTop that Matt Neuburg writes about in this issue, have no need for major updates past the occasional bug fix, because they do everything the programmer intended from the beginning. Such programs tend to be quite simple, of course, but there's no shame in writing simple and effective software. I'd say this category made up the majority of the notes in TidBITS Talk.
In cases where updates had been made available, the old programs offered the precise feature set that the user wanted, making updates unnecessary or at least not worth the expense and trouble of upgrading. Quite a few people have stuck with Word 5.1 or Canvas 3.5 for this reason.
The software business isn't an easy one, and many programs have been orphaned over the years for reasons unrelated to the quality or utility of the program. Many older programs are still in use purely because updates were never forthcoming, such as the ever-popular outliner MORE. We've bemoaned this fact in TidBITS in the past in Matt Neuburg's "Long Day's Journey into Night of the Living Dead Software" in TidBITS-494, and the seemingly unanswerable topic of where you can still legally get old software is going on in TidBITS Talk.
<http://db.tidbits.com/article/05519>
<http://db.tidbits.com/getbits.acgi?tlkthrd=971>
Some users stick with older hardware, at which point older software is more likely to match the processing power and available resources on the Macs in use. A IIci paired with WriteNow 3.0 probably still provides all the basic word processing power most people ever need. Though the fact that this is possible is high praise for Apple and software developers, it's also the bane of their continued financial existences, which relies on people buying new Macs and software updates or new packages.
Games tend not to receive significant updates, and when they do, those updates often change the play enough that many users stick with the original version. Quite a few people seem to play the 1984 version of Missile Command still, and Tetris from 1988 also garnered a few mentions.
Lessons -- It strikes me that there are some important lessons for software developers and for the Macintosh community at large, but I continue to have trouble winnowing them out. Should developers carefully avoid including certain features in programs to make upgrades compelling? Probably, and there are also trade-offs in development time versus feature completeness. Are upgrades often over-priced? No question. Do software companies rely on upgrade revenue to survive? Absolutely. Do software companies also rely on the splash of an upgrade to stay in the eye of the market? Indeed. Should users stick with versions of programs that meet their needs? Yes. Does the Macintosh community have some sort of an obligation to support software companies, or should survival go only to those companies that can appeal to a sufficiently large number of users? Good question.
These issues have been hashed over in the past in TidBITS Talk, but in the end, I think we end up with an uneasy synergy, where we in the Macintosh community rely on software companies to provide the quality programs we need to make the Macintosh experience compelling, but those companies in turn rely on us for financial support. Everyone's happy when the synergy works, but if either side fails to live up to its side of the bargain, the relationship fall apart. That's what happens when software companies release buggy software, fail to provide necessary tech support, or release more for-pay upgrades than seem warranted. And the Macintosh community is quick to complain when a favorite program is orphaned, be it MORE or Emailer; in the end, however, those orphaned programs must not have sold sufficiently well to overcome other obstacles to continued development.
<http://db.tidbits.com/getbits.acgi?tlkthrd=122>
<http://db.tidbits.com/getbits.acgi?tlkthrd=816>
<http://db.tidbits.com/getbits.acgi?tlkthrd=857>
Alternate business models have been tried, but none have proved sufficiently successful to convert existing companies to a new way of thinking. Perhaps there's room for improvement in that area, but until a new approach can be shown to work, the Macintosh community and software developers will just have to agree to abide by the tenuous contract that ensures our mutual future.