> I'm not exactly certain anybody asked for a calculator on their TV, but by this point in time I was now determined to put PCalc anywhere that Apple let me, just out of stubbornness.
I always have a program open that's called pcalc as well, but it's something homebrew and short for Python calculator. It's just a Python repl (enhanced by python-prompt-toolkit) with a bunch of my own pre-defined functions: https://github.com/orlp/dotfiles/blob/master/pcalc/pcalc
I use it all the time, it's so incredibly useful. My most used pre-defined functions are... copy(s) and paste(). E.g. doubling a comma-separated list of numbers is one simple
I would be interested to see how the codebase itself changed over 30 years, using something like [1] "gource", but it predates a lot of version control systems :)
Also, I prefer the earlier UI designs that are more skeuomorphic and look like an actual calculator... but that's just me, they all look great.
I did use the very first PCalc. I even wrote something that would translate a formula to PCalc's programming language. After switching to OSX, I never looked at it again, though. It may sound petty, but it was because of the price. I still think $10 is too much for a (fairly simple) calculator app.
Apple’s calculator that comes with OS X/Mac OS is simply terrible, at least in RPN mode to almost everyone who has used HP calculators.
Two that bite me all the time:
First, “2 Pi” on HP pushes 2 and Pi onto the stack, and the Pi is complete so if you then press “3” you end up with 2, Pi, and 3 on the stack and you are still in digit entry mode.
On Mac “2 Pi” first puts 2 on the stack, then overwrites that 2 with Pi, then when you hit 3 it overwrites Pi with 3.
If you do “2 ln Pi” on both HP and Mac you end up with ln(2) and Pi. On the Mac though if you then start entering a number it will overwrite Pi, whereas HP will lift the stack to make room for the new number.
Basically on HP the Pi key completes any in-progress number entry, pushes Pi, and enabled automatic stack lift so that you can start entering another number without needing to press ENTER. On Mac it lifts the stack if and only if pressing a digit would lift the stack, then overwrites the bottom of the stack with Pi, and leaves automatic stack lift off so that if you then start entering another number (or press Pi again) the Pi is overwritten.
Second, on HP memory recall behaves the same as Pi. On Mac memory recall overwrites whatever is on the bottom. For example suppose 7 is in memory. Then “1 MR” results in 7 on the stack. As with Pi it is still in input mode so “1 MR 5” results in 5 on the stack.
Unlike with Pi it does this even if automatic stack lift would be expected. “2 ln MR” results in 7 on the stack, with the 7 overwriting the ln(2). On HP it would result in ln(2) and 7 on the stack.
I used Mathpad for longer calculations, a bit of a niche application by Mark Widholm, which exists since System 7 or 8. It's a bit like Matlab, and also a bit like those apps where you write small steps and can track the computations (Soulver is one of those), but simpler and more direct than the former and much more capable than the latter.
For simple calculations, I used the standard calculator in "normal" mode. I've also always got one or two calculators on my desk (currently a NumWorks and an HP Prime). And since a few months, I also have Free42 on laptop and iPhone. If you like RPN, check it out: https://github.com/thomasokken/free42
What does it say about me that the brushed-metal look is the most pleasing to my eyes? Maybe this is just an oldie talking, but I do wish interfaces would move away from today's flatness.
Not so much "excited" as "interested": it's 30 years old, and just ask yourself how many OSs and CPU architectures and development environments it has been ported across while retaining some degree of continuity.
I realize most folks on HN trend young and are always interested in the latest new hotness, but here's a tool I remember from when I first began using Macs (in the early 90s) and it's still with us and under active development, unlike the overwhelming majority of hobbyist software projects. There are lessons in here.
Exactly! I respect software that lasts a long time, like LaTeX or XFCE (hardly changed in over a decade). It means you can learn it and not have to learn something new two years later.
In addition to the sibling comments, I find it interesting to compare to a cross-platform Electron-based app.
By building for Mac, the developer was able to target every Apple platform, sometimes with minimal effort, and leverage platform tools like AppleScript for integration.
The Mac has a long history of high quality native apps, and it’s nice to see one survive, even thrive.
They've even extended that motif with Shortcuts, which is surprisingly robust, I just don't know if it has the same level of integration possibilities that AppleScript does.
Even 30 years ago people took calculator apps for granted. This one is a very nicely polished one - the perfect example of what not taking things for granted looks like - and his story is also kind of a dream for a lot of programmers: finding an audience for something well done, and making a decent amount of money building something you like. His hobby lead him to meet and receive compliments from people he cared about, which beats working on ERP forms.
The way I look at it is that 30 years ago this guy created an app. He's kept it updated and it has provided a living. Not bad and not something most of us have achieved over that extended timeframe.
> I'm not exactly certain anybody asked for a calculator on their TV, but by this point in time I was now determined to put PCalc anywhere that Apple let me, just out of stubbornness.