Hacker News Reader

“ Gopher Protocol was created at the University of Minnesota for searching and retrieving documents. Its original design featured green text on a black background.”

Yeah. All you need to know about this article.

Gosh I could have sworn I remember it being amber text on a black background.

The VTs I've used on VMS were amber.

Probably a VT220. IIRC those had amber screen phosphor. The VT100 had white.

Yeah, I think the point that @setpatchaddress was trying to make was that the author is confusing the color of the monitor itself with the color prescribed by the software. Thus, take the accuracy of the entire article with a big grain of salt. That's how I understood the comment....

Monochrome monitors were often either green or amber or white. If the monitor used P1 Phosphor, then it was green, P3 Phosphor was amber, and P4 Phosphor was white. So the color of the text on the screen was a reflection of the hardware spec/design, not of the software.

Indeed, apologies for being terse. I suspect @nsxwolf understood that, though.

Re>> "I suspect @nsxwolf understood that, though."

You're probably right. I still struggle with tone sometimes. I'm still getting used to this whole "Internet" thing... ;-)

So what color was P2 Phosphor?

Blue-green.

Check standard phosphor types section here: https://en.m.wikipedia.org/wiki/Phosphor

My first response was ¯\_(ツ)_/¯ -- but I dug some more and found out that P2 is a Blue-Green color used in oscilloscopes, but it doesn't look like monochrome computer monitors were ever made for it though [1].

[1] https://en.wikipedia.org/wiki/Phosphor#Standard_phosphor_typ...

P2 was a long persistence phosphor. It was used in analog oscilloscopes because those couldn't be refreshed -- the waveform was drawn onto the display as it was received, and the phosphor was responsible for making that visible.

You wouldn't want that in a computer monitor. The computer is perfectly capable of refreshing the monitor, and the long duration of the phosphor would make anything rapidly changing (like a scrolling display) impossible to read because all of the previous contents would interfere with the current image.

Storage tube monitors did use long persistence phosphors. They were a way to simplify the refresh and memory requirements on early hardware.

Blue-green. It was used mostly only in oscilloscopes [1] because the persistence time was really long - good for instruments where the signal might be changing faster than a human can perceive not so much for low latency inputs.

[1] https://www.wellenkino.de/565/565-1.jpg

And Asteroids arcade machines, I believe.

There were also monochrome flat-panel displays, either LCD or plasma. My first¹ computer was a 25Mhz 386 lunchbox luggable with a plasma (amber) display. It was portable-ish and ran on AC power only.

---

1. Well really my second. The first was a Spectravideo computer I got for writing demo programs for the midwest distributor when I was in high school. It wasn't very useful though.

How did you end up with that for a first computer?

Something similar was the first computer I used too. My answer would.be: My dad bought it. This was a common form factor for personal computers back then, and they were sold in stores to to the general public.

At the time I periodically travelled to give LaTeX workshops. I wanted something that I could set up in my hotel room to be able to work on course prep. Laptops were much more expensive and this came out to be competitive, if not cheaper than buying a desktop plus monitor.

And while we're on the topic.

Why the hell were our monochrome displays not all white. What's the purpose of adding a color.

Green phosphor coating was cheaper and had a longer "glow" time so it worked better on lower refresh rate equipment. Why they weren't all white later, once that all mattered less, is a good question. I do know that HP charged us more for the white ones...that never changed.

I wonder..."once that all mattered less".... I wonder if, by the time that white CRTs were cheap enough, if color CRTs were also cheap enough (albeit, more expensive than white monochrome), that most consumers just said "f-it, for an extra X, I can have full color instead of _just_ white?"

I'm sure that was true for PCs, but for dumb terminals, there weren't that many that supported anything other than monochrome.

I was "that guy" during those days and insisted on buying a monochrome CRT (which at that time was just a CRT, since your only choice WAS which "chrome" you chose for your monochrome) whose text was kind of pastel white. It was glorious.

I also found noting the Linux kernel's color schemes as rather odd.

Same here, considering that there were lots of prior art for white text on black background.

Yes, now that you mention it. MS-DOS 5.0 (which predated the first Linux release a few months in 1991) had a nice grey text on blue background for EDIT.COM (or was it exe?) and QBASIC.EXE. and NC.EXE (Norton Commander). Netscape and Mosaic inverted this color scheme with a grey background and blue hyperlinked text. COINCIDENCE? PERHAPS NOT!

I grew up with MS-DOS and 4M of RAM, perhaps I should make a tiny VM with Windows 3.11 and Netscape to revisit these days and try to get my Jekyll blog running on it (served from CERN httpd or the popular webserver at the time). Not sure if VirtualBox virtual network adapter can be recognized by Windows 3.11. Perhaps someone can backport a Gemini browser to Windows 3.11.. perhaps someone can backport a modern WWW browser to Windows 3.11!

edit: someone provided VM's already! Just 33M for DOS622+WIN311 (but still a lot of floppies). I disabled the default USB support and the startup to GUI is ~5s. EDIT.COM has a "hypertext" links in it consisting of 2 little green arrows, 1 triangle rotated left on the left side and 1 right on the right side. The grey on black text is not underlined. http://virtualdiskimages.weebly.com/virtualbox.html

I think the article is more of a musing than a mathematical proof…

I mean who has the time to deep dive into 100% historical accuracy of these things.

I thought it was a neat look at a historical context of something mundane like a hyperlink.

The whole thing felt like "intern from high school searches for 'old computer screenshots' with Google Image Search and comment on what they found".

Hence the musing… when you muse about something you want to keep it loose exploration no?

No one has time to bother caring about the past, and we should be grateful that someone half-assed it (at best)?

I strongly disagree with this premise, there are plenty of people out there taking the past seriously and "full-assing" it.

I’m not trivializing history just calling out all the pundits who armchair complain about people’s good-natured blog posts.

Someone purportedly writing authoritatively for an audience. You don't get to just invent history out of whole cloth because you "don't have time." If you don't have time to write a factual article, you don't have time to write an article.

So now this piece of crap is out there on the internet for the rest of time, permanently memorializing completely nonfactual statements. Fake news. Literal definition of. Wonderful.

They're also pretty seriously wrong about HyperCard:

> Apple brought color to its HyperCards, but notably, the text links were still black and not blue.

HyperCard never natively supported any form of "text links". You could make a button with a text label, or a transparent button hovering over text, but there was no way to attach a behavior to a span of styled text without a lot of custom scripting.

(And, for what it's worth, the color XCMD for HyperCard never really caught on. It was a late addition, and never felt entirely like a native part of the application. Even when it was available, most users kept on authoring stacks in black and white.)

> However, some UI elements did have blue accents when interacted upon

I have no idea what the author is referring to here. Possibly the blue tint in the system UI (like window titlebars), which has nothing to do with HyperCard and didn't apply to its in-app UI?

> I have no idea what the author is referring to here. Possibly the blue tint in the system UI (like window titlebars)

I get the impression that the author was trying to answer two questions, the second being: where did the color blue come from?

I was also under the impression that the color XCMD was not a part of HyperCard and was created by a third party, but I could be wrong there since it has been over 20 years since I've used HyperCard.

There were a couple of different color solutions for HyperCard, but the official one (Color Tools) was released with HyperCard 2.3.

The window tint in System 7 was a user preference. The purplish blue seen in the screenshot was the default color, but there were about a dozen other options. In any event, it seems a stretch to assume that Apple's choice of this color influenced Mosaic in choosing a different blue color for a different purpose.

This article is poorly written, I was expecting an interesting investigation but stopped reading halfway through. Here's a bunch of underdescribed things with non-blue hyperlinks, and some completely unrelated ones that didn't even have hyperlinks. The answer to the headline could have been made in one sentence and there would have been no loss of value.

> The answer to the headline could have been made in one sentence and there would have been no loss of value.

That sentence would be "I have literally no idea, it seems completely arbitrary." There really isn't any value worth preserving, it's just misinformed commentary on a bunch of screenshots taken while the author was probably in diapers.

Well, it does follow the Best Practices for Getting the Right Answer on the Internet protocol.

(a/k/a Cunningham's Law)

For all their talking about early Windows versions, they missed that Windows 3 introduced a hypertext help system. It used green links.

This is what it looked like: http://toastytech.com/guis/win30help.png

That's a nice looking color.

I guffawed at "I do not believe that this is the first instance of the blue hyperlink since this color is cyan, and not dark blue."

No Blue Scotsman!

Tim Berners-Lee told Ben Shneiderman at the time that he was influenced by the design of the HyperTIES-based "Hypertext on Hypertext" project from the 1987 HyperText conference that the ACM published, which had light blue links.

https://news.ycombinator.com/item?id=28317104

>My students conducted more than a dozen experiments (unpublished) on different ways of highlighting and selection using current screens, e.g. green screens only permitted, bold, underscore, blinking, and I think italic(???). When we had a color screen we tried different color highlighted links. While red made the links easier to spot, user comprehension and recollection of the content declined. We chose the light blue, which Tim adopted.

>His systems with embedded menus (or hot spots), where a significant user interface improvement over early systems such as Gopher. But Tim told me at the time that he was influenced by our design as he saw it in the Hypertext on Hypertext project that we used Hyperties to build for the July 1988 CACM that held the articles from the July 1987 Hypertext conference at the University of North Carolina. The ACM sold 4000 copies of our Hypertext on Hypertext disks.

Ah that's probably why the color green is also in the EDIT.COM hypertext arrow links in the MSDOS 5.0 Help menu in 1991.

Windows help functionality has been declining ever since. Those old help files were so clean and snappy, well-organized and searchable. Now, I dread accidently pressing F1 in any major windows app.

I'm quite sure that students, even grad students, at UIUC, had lots of access to windows based computers in 1993. Just because they weren't doing their work on them didn't mean they didn't have PCs running DOS or Windows (and maybe dual booting Linux).

They were most certainly not dual booting Linux in 1993.

I'm pretty sure we were dual booting linux in 1993, why were we certainly not?

I think you could barely fit one OS in your 64MB hard drive at the time…

Windows 3.11 (1993) was ~15MB once installed.

Computers came in towers back then which could contain many hard drives

250MB was common by then and Linux distros only took up a dozenish MB. Lilo existed by then and everyone dual booted to DOS/Win3. If you didn't want to modify the MBR you could boot Linux from DOS with loadlin.exe.

1993 is roughly when I bought my first computer and it came with 120 MB HD. I dual booted Linux (Slackware) on it in 1995. Still on Win 3.11 though, I believe I bought a bigger hard drive before upgrading to Windows 95, but still dual booting with Linux on another partition.

You could fit a lot more OS on an HD in 1993 than you can now!

Windows 3.11 required 14 MB HD space.

https://www.technologytips.com/windows-system-requirements/

At 64 MB, you wouldn't have a whole lot of space for data and applications / utilities, but you could certainly cram a couple of OSes on it.

The first "user-installable" linux was arguably SLS, which came on 24 1.5 MB floppies. I'm unsure what the installed size was, but likely in the 24 -- 48 MB range.

https://opensource.com/article/18/8/first-linux-install

Not in my university, i had a used toshiba 386sx25Mhz as a laptop with Windows 3.11. The fastest public desktop computer was a 486DX2 running at 66Mhz, around 6 of them on the 4th floor from the physics faculty, that was early 1994. It was running Mosaic and Netscape 0.9 or so with Win 3.11, no dual boot. When I got my own room and built myself a 486DX5@133 Mhz I got RedHat linux on it, dual booting, I think it was '95. My roommate had a pentium with Windows98SE running Moria, he was quite addicted to it. Perhaps some very adventurous types were trying out Linux on their own machines in 1993 but distributions were all pretty new, I am not sure it was written about in Byte or C'T magazine in 1993. Perhaps it was discussed on mailinglists or usenet groups, but it was not really findable on the www I think. I frequented a site called 'on the bleeding edge' featuring a picture of a knife (or so I remember) and a list of kernel version numbers and patches. Not sure what date, but might have been '95.

I remember downloading and compiling kernel 1.1 or 1.2 - which were released in 1994. In 1995 I bought the offical RedHat CD's, I might have tried Slackware or Debian floppy disks earlier. A friend was running Linux with X-Windows, Emacs and Mosaic on his Pentium with WindowMaker or AfterStep. No dual-boot. He was on usenet (with the built-in netscape usenet reader) and played with Lisp, he had studied physics (or was it philosophy) at the Utrecht University. Heaps of Byte and Astronomy magazines and SF books littered in his room where he enjoyed a glass of wine and a cigar. I occasionally dropped by for a chat about linux, SF, astronomy and the meaning of life and my own troubles. I think he had dropped out of college for some reason years ago but I never dared asking why. He was brilliant and stayed in the student dorm as a senior. Unfortunately at 45 or so he died in his sleep I heard later when I wasn't living there anymore. He had a good job at the municipality and was well respected for his knowledge, seriousness, dry and dark humor and wit and was still living in the student dorm.

my next door neighbor was running a dual boot DOS and slackware 0.8 kernel in 1992. I still have an account on a machine descended from it.

> They were most certainly not dual booting Linux in 1993.

This just isn't true.

I don't know about UIUC, but I was at university in the UK in 1993.

We had "labs" with about 30 Windows/DOS PCs in them that were for general use (meaning word-processing). Probably 1 in 100 students had their own computer (and if so it was more likely to be an Amiga or Atari ST).

My course wasn't computer-related, but I took a few computer-based modules (graphics, vision, AI, intro to programming). All of these apart from intro to programming where done on university SunOS/Solaris or Irix machines (and Intro to Programming was in Turbo Pascal for DOS on those Windows machines).

The SGI Irix machines (made famous in the original Jurassic Park) were really nice.

> Windows 1.0 in 1985 did not have...overlapping windows.

Actually Windows 1.0 did have overlapping windows.

They just weren't the default style for application windows.

But there were popup windows that overlapped other windows on the screen. These were typically used for dialog and message boxes, for example the End Session message box midway through that filfre.net article.

There was nothing stopping anyone from using a popup-style window for their application, and adding a titlebar so you could move it around on the screen. It just wasn't the custom, and people would think your app was weird if it did that. And on a typical system of the day (no GPU!), dragging your window around on the screen would perform rather poorly.

Great catch!

I had read that Digital Antiquarian piece a few months back and had remembered about the "sub windows" that an app could have. But I had thought they were scoped to just be on top of the window for the app that spawned it, and couldn't leave that "tile."

However, you are totally right. This image right here clearly shows it overlapping another application's tiled window.

https://www.filfre.net/2018/07/doing-windows-part-3-a-pair-o...

This is the best kind of correct: technically correct!

The second screenshot is from windows 2, which was released in 1987

Ahhh! Good catchup. I thought those "minimize" and "maximize" icons looked odd. Updated

The part about project xanadu is also shady. They list it as 1964, but first paper was published 1965, while their idea of hypertext was born 1960, with the word choosen 1963. Not sure where the 1964 comes from. Not to mention that this wasn't even an actual implementation. It's disputable whether xanadu really should be the first one to mention here. The ideas of referencing documents is older and has previous implementations in analog world. It they wanna go just about hypertext, they should at least started with memex.

>> WorldWideWeb was not created in 1987

WorldWideWeb was created by (mostly Republican) computer engineers who broke from IBM in Silicon Valley in the 1980s, forming little democratic circles of twenty to forty people with their laptops in each other's garages.

This is a reference to a bizarre anachronism in David Graeber's writing. The conflict over this passage once carried over to the comments here on HN while Graeber was still alive, and he and a longtime foe quarreled about it to a fairly unsatisfactory conclusion.

I also disagreed with,

> Here Microsoft uses the “hyperlink blue” for active states when a user clicks on different drives, folders and icons.

Hyperlink blue was a much brighter blue (pure blue, on the interfaces I used, but I was a litter later in the timeline) than Window's blue, which was a (noticeably) darker blue.

Yeah, if I were to 'pick' a reason it would be that it worked in the EGA colorspace. You did not exactly get a huge range of colors there. While in 1993 256 (small screen res) or 'truecolor' (very expensive vid card) was not unheard of but it was decently uncommon on low end hardware.