Usenet's own message standard set the rules for posting code

The workload

Posting to a moderated Usenet newsgroup such as comp.sources.misc meant following header and formatting rules that the network's own technical standards, not any single archive's house style, defined. RFC 1036, dated December 1987 and written by Mark Horton and Rick Adams to update RFC 850 from June 1983, states that any message posted to a moderated newsgroup required an Approved header line, normally added by the moderator, before other sites would relay it. A poster submitting source code to a moderated group therefore sent the submission to a human moderator first, rather than posting directly, and the moderator's approval line is what let the message propagate across the store-and-forward network of independently run sites.

What the documents show

Verified, from RFC 1036 and RFC 850 themselves, both standards documents rather than any single publisher's house rules: the specification defines the article header fields, including Newsgroups, Distribution and Approved, and describes the control-message syntax a site used to create a new group as moderated rather than the unmoderated default. Verified, from the Internet Archive's own catalog record for a preserved mailbox archive of comp.sources.misc: a bulk backup of that newsgroup's postings survives in the Archive's usenet collection, confirming that a record of the distribution mechanism, not just its rulebook, has been kept. Neither document names a download count for any individual program ever posted to the group; Usenet's transport model, relaying copies between independently administered sites, kept no central log that would support such a figure.

The operating cost

Neither RFC states a subscription price, because Usenet was not sold as a product; the cost these standards impose is administrative and technical, a site operator's own hardware, storage and phone or leased-line transport, plus, for a moderated group, a volunteer moderator's unpaid time reading and approving submissions. Neither document estimates that labor in hours, so a specific cost figure is not established here.

The stop condition

The RFCs describe protocol, not policy, and set no expiration for the standard or any newsgroup; RFC 1036 explicitly obsoletes RFC 850 rather than retiring the underlying idea of moderated distribution. Editorially, a given newsgroup's practical stop condition was the point where its administering site, or a critical mass of relaying sites, stopped propagating it, a threshold these two documents do not define because they describe message format rather than newsgroup lifecycle.

• Who is accountable when a distribution channel has a moderator but no company behind it?
• Does a preserved raw-message archive tell you what a poster claimed, without telling you whether it was true?
• What would a modern equivalent of unpaid newsgroup moderation actually cost a one-person software business in hours per submission?

RFC 850 and RFC 1036 are the closest thing comp.sources.misc had to terms of service: rules for headers, moderation and propagation, written for the whole network rather than one archive, with no pricing and no download counter attached.