The 1970s calculator race | Hacker News Reader

The 1970s calculator race

twitter.com

169 pointslxm60 comments

The article failed to mention possibly the most influential one of them all, the Busicom. While it was an obscure Japanese company, they indirectly kicked off the microprocessor revolution -- via another obscure company called Intel.

"Busicom asked Intel to design a set of integrated circuits for a new line of programmable electronic calculators in 1969.[1][2] In doing this, they spurred the invention of Intel's first microprocessor to be commercialized,[3] the Intel 4004"

https://en.wikipedia.org/wiki/Busicom

Intel shipped to its customer Busicom the first 4004 chips in 1971. But because that year Busicom was in financial trouble they accepted that Intel could resell the chips to others in exchange of lower prices. Commercial success of the 4004 isn't present during the first few months and they didn't patented their chip yet. Then it got better.

last entry it's just a photo of a busicom calculator

Sinclair's 1974 Sinclair Scientific calculator was amazing: http://files.righto.com/calculator/sinclair_scientific_simul... https://news.ycombinator.com/item?id=6302364

They sold a scientific calculator with 12 functions using a chip designed for 4 function calculators with just 320 instructions of ROM!

I remember testing various calculators with the sequence

  25 sin cos tan atan acos asin

to see how close they came back to 25 degrees.

The Sinclair showed 42 point something.

    >>> degrees(asin(acos(atan(tan(cos(sin(radians(25))))))))
    24.999999999999996

This is just one ULP off from the correct result; the next larger double is 25.0.

If I round each result to 6 significant figures and then use that as the input for the next stage, my bog standard scientific calculator gives me 9.32631 for the final answer.

(chained using full accuracy as sibling comment does give 25 of course).

Thinking about the curves of the functions near zero in the middle steps and a few terms of the taylor series around those values suggests lots of accuracy lost in rounding. We are spoiled with 16 sig fig these days...

I think you entered a number wrong, since your calculator otherwise got the right answer. None of these functions are going into particularly dangerous areas for precision. When I round every step to 6 figures on my calculator, I still get 24.9999

Heck, rounding every step to 2 figures still gets me .41 radians which is 23.5 degrees.

I think he means rounding to 6 decimal places. If I do that I get 9.326312.

I...how? That's the same on most of the numbers, but even more precision on one of them.

Here's my calculator, with exactly three decimal places:

25->rad: .436

sin(.436): .422

cos(.422): .912

tan(.912): 1.292

atan(1.292): .912

acos(.912): .422

asin(.422): .435

.435->deg: 24.924

Where does your calculator start to diverge, that you end up with an answer like 9?

And I got the same result as Aardwolf that the Sinclair emulator returns .434, only a tenth of a degree off.

Edit: Oh wait, you're using a calculator set to degrees. That's a totally different problem, because it makes the 0-1 output from sin and cos become an extremely small fraction of a circle. It works as a type of precision test, but it's not at all the same thing you'd try on the Sinclair that only does radians. It still doesn't explain your original memory.

Side note: Even in degrees mode, the answer for 6 digits is kind of an outlier. If you use 5 or fewer digits the calculation crashes. 7 digits gets you 24.9. It boils down to just acos(cos( [about half a degree] )), and the other operations don't really matter.

...but it's not at all the same thing you'd try on the Sinclair that only does radians.

Right. Apparently I wasn’t aware of that at the time. Nice to find out here on HN some fourty years later :-)

It works in radians, did you convert the 25 degrees to radians first?

That being said, I cannot even get it to compute the sine of 2 correct. If I type: 2, UP, + (sin is under the "up" part of the plus button), it returns 1e-3. I think I'm entering it correctly, what could be wrong?

EDIT: found it, I think it only supports values between -pi/2 and pi/2 radians. It computes the sine of 1 radian correctly.

EDIT2: after doing the sin/cos/tan/atan/acos/asin sequence on 0.43633 (25 degrees in radians), it gives 0.43400. Not that bad! I mean, not great, but at least 2 digits.

My droid48 (running an HP48 ROM) gives me 25.000000576

Wow, this is brilliant. Thanks for the super interesting read.

Vintage calculators are my hobby! It's a shame I only saw this thread when it's 12 hours old. If anyone has any questions about them or would like to see photos I'm happy to share.

The Twitter thread, while educational and cool, is a bit inaccurate/misleading about a handful of things. A lot of the photos are of later models or something unrelated to the tweet, and there were a lot more innovations not talked about. Progress also did not happen as early as the tweets would indicate.

Personally, one thing I love about old calculators is the craftsmanship. Each one of the early ones was assembled by hand, and it shows - some of them are true works of art.

Then you should post a blog or site to correct these mistakes and share your hobby with the world.

There's already tons of sites about this stuff, probably hard to add much that hasn't already been written about at this point. I did a bit of collecting myself for a while but gave up when I both no longer needed calculators very much anymore and the prices started getting ridiculously inflated.

http://www.vintagecalculators.com/html/calculator_time-line.... - similar timeline but goes back farther

http://rskey.org/ - programmable calculators

http://datamath.org/ - TI calculators

http://www.hpmuseum.org/ - HP calculators

https://web.archive.org/web/20070528063100/http://www.tasweg... - Soviet calculators, original site sadly went down

http://thimet.de/CalcCollection/index.html - huge collection that includes accuracy and performance testing (for programmable models)

I wouldn't start a blog just to correct those mistakes, I think the twitter thread is cool and got people excited about vintage calculators which is great.

The problem is just that I don't have enough time/energy to dedicate to making a site with substantial content about vintage calculators. A HN comment is orders of magnitude easier than making a proper site with photos, good content, researched details, etc.

Maybe it's a factor of my age, but I love these kind of 'walk down memory lane' articles.

After my Dad passed away, I kept only a literal handful of things of his as mementos. Of these was his TI-59 (from when he was studying electronic engineering). It's the first 'computer' I ever used and wrote my very first program on (a number guessing game). Still have a few mag-cards with 'lunar lander' (turn-based, enter the thrust per iteration and see the resulting height, velocity and fuel ... if I remember correctly). Haven't fired it up in years. Might do so this afternoon :-)

Addendum: the Casio credit-card sized calculator? I soooo wanted one of those!

I love them because it's a reminder of how I became interested in computers and programming in the first place. 1980 -> '84 I lived in a country town in South Australia (then the second-largest city, but still only around 30k people); after school I waited for my parents to finish work in the library of the government education centre, and quickly discovered copies of Byte magazine. They had all the issues from the first in the 70s' to the then 'current day', and I can remember seeing all the adverts for very cool (or so I thought) calculators and computers and hardware bits and pieces. My interest became an obsession, and I cajoled my parents into buying a TRS80 Color Computer in 1982, started typing BASIC programs into that machine, and haven't stopped programming since.

These kind of articles restore, at least a little bit, the wonder and excitement I felt at having interaction with these new amazing machines, and make it a bit easier to deal with the next JS problem, or developing that new Web API for a client.

> the Casio credit-card sized calculator

EEVBlog did a teardown of that one if you want to see how it was put together (there's actually more going on in there than one single blob and some flat-flex!) https://www.youtube.com/watch?v=vwo7sSrSSW4

I love this nostalgia thing with the older calculators that's been going on for a while now. Getting more current, does anybody else like to stroll down the aisles checking out the current lineups of calculators at their local office supply stores? I have a small collection now and really enjoy checking out the latest Casio or TI models, and sometimes even the eye-catching store-brand scientifics in various unique color schemes, and so on.

Looking for things to do with my calculators was one factor in making the decision to become a licensed ham radio operator. I was kind of disappointed in the end that the exams didn't end up requiring much calculator-math, but anyway I'm kind of going back to school these days so to speak, getting a bit more involved with mathematics as an avocation and definitely enjoying my various calculators again after a long hiatus post-formal-education.

Haha, I recently bought the HP-12C and the HP-50g mainly just to own them.

The iphone has an app for an HP 41CV which is very cool

If you enjoy that nostalgia, I highly recommend a visit to the Computer History Museum in Mountain View; they have a good selection of calculators (including some mechanical and electro-mechanical ones!) in their main exhibit.

> Getting more current, does anybody else like to stroll down the aisles checking out the current lineups of calculators at their local office supply stores?

Yes!!!!! Still love my Casio fx-4500PA and fx-50F. It's funny. I am a computer scientist working in big data etc. and you can find me standing open-mouthed in front of the shelf with the $20 scientific calculators in the store.

About 10 years ago I found the Casio 9860 slim (clamshell) which can do RPN with an addon, and integrated it to my desktop.

It is faster to grab it than firing off any piece of software, and easy to take with me when having coffee and doing back of the enveloppe calculations

> Casio 9860 slim (clamshell)

Wow, I really like the looks of that one! Nice. I agree on the convenience. I usually have something around my desk like an fx-260 and it's really nice to have handy.

A pleasure to use. Runs on AAA batteries, backlit, and has a USB port to transfer things back and forth

I do like to see what calculators they have whenever I'm in a pound shop. Last time I was in one, they had scientific calculators. I just admire the ability to get the price so low, to just £1 for something that a few decades ago was so cutting edge.

For those of us in science/engineering path, the HP-35 was as lust inducing as pretty coeds. To actually own one was to be both envied and hated. Not being able to afford the $200 or so it took to buy one, I bought the much cheaper TI-52. Turned out to be a great choice that carried me through my BSEE and well into my first job until it was replaced with the programmable TI-59. Unfortunately, being unable to envision there ever being a collectors market for these things, I got rid of both of them.

HP-15c is a work of art. I made it through college using the Voyage-200(souped up TI-89).

For those that want a video overview; LGR has done an episode on this.

https://www.youtube.com/watch?annotation_id=annotation_99646...

And Techmoan (great channel) did a review of the HP calculator watch: https://www.youtube.com/watch?v=gGQrsSEaZkI

The fact that this entire article was written on twitter is more interesting and remarkable to me than the actual content. Although they're both pretty interesting.

"Begin Japanology" has an episode on calculators. It contains a segment about calculator product wars among Japanese manufacturers. https://youtu.be/1_GVkR0SITo?t=1050

Amazing fact: I have never owned a pocket calculator. I learned to use slide rule in school and universities and stepped right into computer age around 1973. Hence I have had embarrassing moments, especially with those with Forth-like operation.

Not a year goes by I don’t kick myself for getting rid of my HP 28s. It wasn’t from the same era as those in the article but damn, what a trusty work horse it was for me in college.

Free 42S is my favorite phone app. I've the original 42S manual PDF -- not easy to find -- their `charity' pay-wall.

~35MB Download: original HP 42S 358pp manual: e3pi.org

I've many HP rpn's, Free 42S is better at everything, except keys haptic satisfaction.

If you're voyager/landscape 10C 11C 12C 15C 16C centric, rotate phone: rot90:HP-42S <--> HP-42C

yes, ballistically-black ruggedized Otterbox-ed, cost $0.00.

when I was a kid at school late 70s early 80s, calculators were the awesome thing, this series of game calcs were the "THE" thing to have, especially 8 attack and boxing... http://homepages.ihug.co.nz/~pinwhiz/cascalc.htm

A proficient abacus user can be faster than a proficient calculator user in tasks involving addition and subtraction.

This is a motivation for abacus/calculator hybrids, http://retrocalculators.com/digicus.htm

I cherished my SR-52 in high school. My parents got it for me for Christmas my Junior year. I replaced it with a TI-59 (with magnetic cards!) in college. And the top of the line for me was the 'landscape mode' TI-92+.

I sold the -52 and -59 at a garage sale one year but still have the 92+.

Non-mobile link: https://twitter.com/pulplibrarian/status/1030830856720527360

Peak calculator for me was that Casio MG-880 with the space invaders game. Many an hour of math homework was consumed with frantic tapping.

This is pure gold. Calculators were what got me to choose Computer Science as a career. I will be sure to save all the photos.

from the mid 70s to the mid 80s my parents had a "compact" calculator on their desk at home (for the bills etc) that plugged into power -- no batteries. It was about the size of a book, and sat on the desk next to the (wired, of course) telephone. What a modern household!

Ah the Little Professor.

I don't think I ever used one growing up that wasn't a little bit sticky.

Too hard to read this is Twitter format

Twitter is the WRONG platform for this type of information.

100% agree. Came here to say exactly this. Great content which would a great blog or medium post rather than a real-time 280character stream.

I dunno. As weird as this format is, I'd prefer this to the typical article these days with autoplay video, janky slideshows, ads moving the content around, a pop-up begging you to subscribe to a newsletter, Outbrain/Taboola ads showing cracked toenails and butt implants, etc etc.

That’s the other end of the extreme nastiness scale. True.

I would certainly hope there is still a middle ground somewhere!

At some point in time, there was a middle ground between those two extremes: Wordpress, Blogspot and so on.

https://threadreaderapp.com/thread/1030830856720527360.html

The best platform is the one where you have an audience.

Twitter has the benefit of less friction - if it's on your feed you can read and get hooked on the story with no clicks. The reader funnel has one less step.

It certainly is. I can't possibly fathom what goes on people's minds that leads them to think Twitter is a great platform for information sharing.

It makes me sad that humanity gave up blogging for tweeting. I can't think of a more unsuitable platform for making this kind of posts: it limits the number of characters, the UI is awful for continuity of posts, the webpage is really heavy since it loads much more content than the post itself and the platform treats unregistered users as if they didn't deserve to read the content.

People use Twitter to write long blog posts because of one simple fact: the presence of people to read your writing is more valuable than any possibible combination of typesetting, CSS, and noninvasive ad placement.

It would work just as fine writing a blog post on an appropriate platform and then sharing the link through Twitter.

I came into possession of a couple of TI-74 basicalcs a little while ago with a printer attachement and a couple of ROM cards with some kind of tax software on them.

They're pretty cool little devices. Kind of frustrating to program on though. I wouldn't have wanted to have to do that for a living. But still pretty cool to have something with that kind of power you can carry in a large pocket in 1985.

TI-74 specs:

TMS70C46 CPU[1] (C70009, another chip from TMS 7000 family also reported)[2]

31 5×7 character LCD

32+4 KB ROM

8 KB RAM

RAM/ROM memory expansion port

Hexbus port

80 characters per line (31 visible)

Reply on news.ycombinator.com