Comment by dredmorbius | Hacker News Reader

Comment by dredmorbius | Hacker News Reader

The key lesson here is not about a specific image format or company, but the fact that even a very light assertion or even hint of proprietary restrictions can kill or handicap a nascent technology or protocol in the bud.

The age of early hypertext was awash in such cases. The University of Minnesota only mentioned that they might introduce a charge for gopher servers, effectively killing the protocol. Various early conferencing and hypertext systems, including one from Sun Microsystems (NEWS?), and a few models from Microsoft, existed. Somewhere, Tim O'Reilly has commented about the early days of O'Reilly & Associates where he was looking at various options and made the early decision to not get locked into any one proprietary system, a choice which has paid handsomly.

The whole phenomenon of prortocol creation, promotion, evolution, and quite often, stagnation, is a fascinating and understudied one IMO. The role of non-market players, particularly government standards organisations, the military, and academic fields, is extraordinarily underappreciated.

Completely agree with and the sad thing is I don't even know where to point someone to learn about bout the roll that "government standards organisations, the military, and academic fields" play, except to talk to leaders of individual W3C groups.

A lot of discussion of early web technologies, like MOSAIC and HTML, took place in newsgroups that are still recorded and displayed online.

Not sure I've seen protocol discussions there (just never bothered to look) but it shows a whole lot of academics and government employees in the mix, and some commercial actors as well.

It's disappointing since those programs succeeded beyond the dreams of avarice but were not designed to gather publicity. The fact that so much of our economic growth is undergirded by public research funding is forgotten even by direct beneficiaries.

I was going to suggest the IETF and RFCs as a trove of standards-setting activity. There should be a few other similar ones.

There's an interesting transition as those groups moved from small (able to fit into a VW Bug, bus, or small bar) to large (auditorium, or multi-campus groups), and the relative frictions in new developments.

Past a certain scale, the easiest way to change a standard is to move forward, either abandoning or ignoring backwards compatibility.

Both software languages and various social networking protocols (there are ... many) are interesting cases in point. Mastodon is in many ways IRC v. 2.0 (and there are any number of competing contenders). There actually was (and/or is, depending on your judging criteria) a Usenet II, though it's seen limited popularity.

Another good source on much of this would be the late Pieter Hintjins. See for example:

https://mixitconf.org/en/blog/messaging---social-architectur...

I did a talk a few years ago on the history of layout in CSS, and those newsgroups were an invaluable source of information for why decisions were made.

> A lot of discussion of early web technologies, like MOSAIC and HTML, took place in newsgroups that are still recorded and displayed online.

Which is why I'm devastated how much new conversation about how things are put together is generally not hosted by the organizations involved, but is stored places like issues on Google Code (all gone), SourceForge (increasingly difficult and likely will begone), GitHub (which people don't seem to like to believe will one day delete their data); imagine if some of these early standardization efforts had done all of their communication on Internet protocols using the contemporary equivalent of Yahoo Groups or Google+--maybe using CompuServe or AOL?--and all of that work was now only around due to haphazard archiving efforts. (Google+ really burns me, as there was so much discussion of how things in the Android ecosystem works--particularly among all of the third-party forks, such as CyanogenMod--which is now just deleted and gone :/.)

And Google Groups, which had a truly wonderful searchable Usenet archive (inherited from DejaNews), but several redesigns later is nearly useless.

Though given how many people have old Usenet postings they'd like to forget, maybe it's not that bad.

I'm only vaguely aware of this myself, though there are a few places I've learned to look.

A history of the Brooklyn Bridge and its builder notes, apparently as a form of admiration, that he designed every detail down to the nuts and bolts. Thing is, that's because he had to. There were no standard fasteners. (The book doesn't detail this, it's just a passing reference I happened to catch.) David McCullough's The Great Bridge: https://www.worldcat.org/title/great-bridge-the-epic-story-o...

Herbert Hoover, as US Secretary of Commerce, instituted the National Bureau of Standards (now the NTIS, I think), which among other things standardised weights, measures, and yes, fasteners. Rather hagiographically addressed here: http://lcweb2.loc.gov:8081/ammem/amrlhtml/inhoover.html

Joanne Yates has written on such thrilling topics as the evolution of the interoffice memo and standarisation of business communications (and though I'm writing tongue in cheek, I really do find the work fascinating), and has been working on a book on standards and their evolution. I've not read it but it may be of interest. Engineering rules : global standard setting since 1880 https://www.worldcat.org/title/engineering-rules-global-stan...

Simon Winchester has a book on machining and measurement, also not read, though based on his previous works it should be excellent. The Perfectionists: https://www.worldcat.org/title/perfectionists-how-precision-...

There was a fairly popular book on containerised freight shipping that came out within the past five years or so called The Box That Changed the World which looks at the standardisation of containerisation: https://www.worldcat.org/title/box-that-changed-the-world-fi...

Somewhere someone commented that one of the computer containerisation standards was simply a bunch of people coming to an agreement as to how to organise and and do things. My history with Linux has taught me that various packaging formats and systems (and the infrastructure surrounding them) have profound consequences and impacts, for users, adminstrators, and software producers.

I wish I could point to a single source and say "read this", but I'm really not aware of one presently.

You might also enjoy https://www.cl.cam.ac.uk/~mgk25/iso-paper.html about standard paper sizes.

Somewhat familiar with that.

The details of printing signatures and folios, and having the pages come out the way you want them to when bound, is another minor fascination. As are canons of page design.

The key lesson here is not about a specific image format or company, but the fact that even a very light assertion or even hint of proprietary restrictions can kill or handicap a nascent technology or protocol in the bud.

Is it? Gifs have been wildly successful. Today, many web deverlopers prefer newer formats like png for both technical and philosophical reasons, but gifs continue to make up a huge fraction of internet pictures. The average 12 year old knows what a gif is.

> The average 12 year old knows what a gif is.

Depends on what you mean by "know". If you mean "moving picture that is not a video", sure they do. But if you showed them an APNG they will probably call it a GIF as well.

Gif owned lossless image formats before Unisys asserted their patent. In the aftermath, PNG was created, even animations shifted elsewhere -- mp4 and, though also proprietary, Flash. SVG popped up later, but fills some rolls previously held by gif.

And simply using or supporting gifs became an issue -- it was political. That's not something you want to have happen to your standard.

Also: because gifs are so predominantly used for animations and not static images, I've taken to using a client-side CSS style that renders all gifs as transparent, unless specifically hovered over.

This very rarely affects anything other than animations. People simply aren't using gifs for images as a matter of course today. For photos, it's largely jpeg (vastly higher compression, though lossy), for text-based images, png, though SVG is making inroads, particularly for data visualisations.

Reply on news.ycombinator.com