Manual for PUB (a markup language in 1971) – Larry Tesler | Hacker News Reader

Manual for PUB (a markup language in 1971) – Larry Tesler

nomodes.com

80 pointsfanf222 comments

Pub getting attention made my day. I wondered how everyone forgot about it and how no one ever mentioned that it actually gave birth to TeX and Web. Good job hacker news

troff (1972) and eqn (1974) also existed. I don't know if they were influenced by PUB. SCRIPT (1968)[1] had macros but it's unclear whether they were present at the beginning or added later. RUNOFF (ancestor of troff) doesn't appear to have had macros. I've heard that Forth documentation was produced using a formatter written in Forth, that let the user switch between documet text and Forth code, but idk the dates or specific for this.

Anyway I'm sure PUB was influential but the idea of markup languages with macros certainly wasn't unique to it.

[1] https://en.wikipedia.org/wiki/SCRIPT_(markup)

I actually came across this manual two months ago and studied it in deep. Your comment was greatly appreciated, I will research all of them. Thank you!

I understand Scribe also started as a set of PUB macros.

This article reminds me of two experimental ~20% projects we worked on at one of my companies in the mid-2000s. Some unreleased projects age like fine wine, offering a chance to inspire others years later.

The first project was a wiki with active scripting capabilities. By “active scripting,” I mean that the wiki could execute embedded code snippets and display the results directly on the page. This concept feels somewhat akin to modern tools like Project Jupyter or Wikipedia’s feature that calculates ages and other data dynamically.

The second project was more radical and linked with the first, aiming to introduce a new kind of collaborative scripting. We developed a small proof of concept (PoC) by modifying Jython [1] to enable shared namespaces among users. This allowed users to access each other’s functions, objects, streams, variables, and more—in real time—creating an experience reminiscent of the Matrix in Neuromancer [2]. All in Python.

[1] https://www.jython.org/

[2] https://en.wikipedia.org/wiki/Neuromancer

In case anyone has copy they'd like to typeset as monospace plain text, I made a tool to do this recently. If you might find it useful, feel free to play with it at https://monotext.pages.dev/

My grandad worked on a computerized press in this era. He primarily printed short-run text books for southern Ontario universities. Whatever markup they used had a syntax that allowed emboldening and italicization, but you had to specify the end of the formatting (obviously). Professors never would, and the first run of a book would always be all bold and italic. The profs would get mad, and tell him “the computer should JUST KNOW.”

The cover image is a Hieronymous Bosch woodcut of an English pub.

I’m not sure the author has seen the rest of Hieronymous Bosch’s oeuvre.

Edit: seems I phrased this badly, my point was that a Georgian woodcut circa 1810 is unlikely to be from Bosch (1450 – 1516).

Larry was a worldly guy with wide-ranging interests. Were you thinking that if he knew about The Garden of Earthly Delights he wouldn't have chosen a woodcut of a pub for the PUB manual? Bosch has been thought of as a heretic [1] and that could have been part of the appeal for Larry, who pushed against orthodoxies throughout his career.

[1] https://en.wikipedia.org/wiki/Hieronymus_Bosch#Interpretatio...

Doesn't really seem out of the ordinary; even in high school in the 1980s the Bosch paintings were well-known.

... from album covers.

This 2020 document says that it's a “cut-and-paste” from a Thomas Rowlandson (1756-1827) woodcut print. Which seems much more likely. https://www.saildart.org/simple/booklet/SAILDART_PREVIEW_202...

NSO’s ForcedEntry (https://en.wikipedia.org/wiki/FORCEDENTRY) showed how anything scriptable creates potential security vulnerabilities.

That website format is very interesting. The original text (while also hyperlinked) on one side, and notes/explanations on the other side.

The irony is that this rendered pretty poorly on my phone. Ignore me, though, as I like the idea.

"The PDP-10 operating system was TENEX, later called TOPS-10." (PUB manual, page 2).

What.

Should be “The PDP-10 operating system was TENEX, later called TOPS-20”, no?

Um no.

It was (after porting to DEC's simpler paging hardware and purchase from BBN in 1973), but if that's not what your comment referred to, then what?

There seem to be two mistakes here.

First, the SAIL PDP-10 didn't run TENEX. SAIL maintained their own operating system based on the PDP-6 Monitor, called WAITS. I guess you could say its closest relative might be TOPS-10; certainly many programs could easily be ported over back and forth.

Second, as noted here, TENEX was not "later called TOPS-10". It also wasn't called TOPS-20. DEC took TENEX, and with extensive modifications and updates produced TOPS-20. Meanwhile TENEX continued to be developed and maintained separately at sites that elected not to switch to DEC-20 machines.

Reply on news.ycombinator.com