The Shareware Revolution

In 1982, software came in boxes. Shrink-wrapped, retail-priced, sold at computer stores next to the hardware. A word processor cost $300. A database cost $400. A spreadsheet cost $500. And before you bought it, you had no way to try it. Two programmers, working independently, thought this was insane. Andrew Fluegelman wrote PC-Talk, a communications program, and gave it away. He called the model "Freeware" (he trademarked the term). Try it. If you like it, send money. If you don't, keep it anyway and give copies to your friends. Jim Knopf (writing as Jim Button) did the same thing with PC-File, a database. He called it "user-supported software." A year later, Bob Wallace, a former Microsoft employee, released PC-Write and coined the term that stuck: shareware. The model was simple. The problem was distribution. ─────────────────────────────────────────────────────── The distribution problem ─────────────────────────────────────────────────────── If your software is free to try, people need a way to get it. In 1983, the options were: ▓▓ User group meetings. Local PC clubs passed floppies around. Reach: dozens of people per meeting. ▓▓ Mail order disk vendors. Companies that copied shareware onto floppies and sold them for $1 to $5 per disk, covering media and postage. You'd find their ads in the back of computer magazines. Reach: thousands per month. ▓▓ Computer shows. Swap meets and trade shows where people traded disks. Reach: a few hundred per event. None of these scaled. A shareware author with a great program could reach maybe ten thousand people in a year through these channels. A retail publisher could put a box in every CompUSA in the country. Shareware needed an infrastructure that was free, fast, and nationwide. ─────────────────────────────────────────────────────── BBSes were the answer ─────────────────────────────────────────────────────── By the mid-1980s, thousands of BBSes across North America had file areas. Organized directories where callers could upload and download software. A SysOp maintained the collection, wrote file descriptions, deleted junk, organized categories. When a shareware author released a new program, the distribution path looked like this: 1. Author uploads the ZIP to a few well-known BBSes. 2. Callers on those boards download it, try it, and upload it to their other favorite boards. 3. SysOps see a popular new file and pull it onto their boards via FidoNet file requests or by calling other boards directly. 4. Within days the file is on hundreds of BBSes. Within weeks, thousands. The author paid nothing. The SysOps distributed it because good files attracted callers. Callers spread it because sharing was the culture. The entire distribution network ran on voluntary effort. ─────────────────────────────────────────────────────── Upload/download ratios ─────────────────────────────────────────────────────── Many BBSes enforced upload-to-download ratios. You couldn't just leech. If you wanted to download ten files, you had to upload one. This created a constant pressure to find and share new software. It turned every caller into a distributor. The ratio system was brilliant. It meant the file areas were self-curating. Bad software didn't spread because nobody wasted their upload credits on junk. Good software spread fast because people wanted to share it and earn download credit at the same time. ─────────────────────────────────────────────────────── The economics ─────────────────────────────────────────────────────── Compare the two models in 1992: ▓▓ Retail box: Publisher pays for packaging, manuals, disks, shelf space, distributor margin, retail margin. The store keeps 30-40%. The distributor takes another 15%. The author sees maybe half the retail price. ▓▓ Shareware via BBS: Author uploads a ZIP. Distribution cost: zero. No packaging. No middlemen. A registration form inside the program says "send $25 to this address." The author keeps everything. The margins were incomparable. A retail game at $40 might net the developer $15 after channel costs. A shareware registration at $25 was $25 in the developer's pocket. ─────────────────────────────────────────────────────── The legitimacy problem ─────────────────────────────────────────────────────── Early shareware had a reputation problem. People assumed free software was bad software. In 1987, Jim Knopf and other authors founded the Association of Shareware Professionals (ASP) to change that. ASP members agreed to standards: respond to registrations promptly, provide support, don't use crippled trial versions. The ASP gave shareware credibility. An "ASP Member" logo in a program's about screen meant it was real software by a real developer who would answer your mail. ─────────────────────────────────────────────────────── CD-ROMs and the golden age ─────────────────────────────────────────────────────── By the early 1990s, shareware had outgrown floppies. Walnut Creek CD-ROM and other publishers started pressing entire BBS file archives onto CD-ROMs. Hundreds of programs on a single disc, sold for $20 at computer stores. This was the peak. Shareware was everywhere. The distribution network that BBSes had built was so effective that it had created its own retail channel, but one where the software was still free to try. ─────────────────────────────────────────────────────── What BBSes actually built ─────────────────────────────────────────────────────── BBSes did not invent shareware. Fluegelman and Knopf invented the model. What BBSes provided was the distribution infrastructure that made the model work at scale. Without BBSes, shareware stays a niche thing. Disk vendors and user groups could not have grown it past a few thousand copies per title. BBSes gave shareware authors access to millions of potential users for zero distribution cost. Every freemium app, every free trial, every open-source project on GitHub sits on a foundation that was poured in the 1980s by SysOps who ran file areas and callers who uploaded software to earn download credits. The first app store was not iTunes. It was a BBS file area with a ratio requirement and a SysOp who cared about curation. --- END OF MESSAGE ---