Live data from Hacker News

Show HN: All desktop software calculators are wrong, so I had to build my own

news.ycombinator.com

81–90 of 169 posts

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#81

I definitely agree with a lot of the complaints. I truly don't understand why so many computer calculator interfaces are so awful. I've been using speedcrunch for a couple of years now and it scratches most of the itches for what I want out of a calculator. It looks fairly similar to Chalk's design as well. Very nice interface!

I really like speedcrunch also, another really cool calculator that is more of a language is Frink https://frinklang.org/

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#82
> Some much needed features are difficult to implement and require a large amount of work. If these features are not available, this is because nobody really cares, and there are few pre-existing parts to build upon.

Some things resemble "nobody really cares", but are distinct. Being unaware (eg "I didn't realize how bad it was until it stopped"). Being impoverished (eg "Any cost, including attention, is an insurmountable cost barrier", or as in education, a chain-of-care disaster triage). And so on.

The "few [...] parts to build upon" is indeed a mechanism for "large amount of work". Yay patents./s But there are others, and subtleties. For example, one can lack a part to build upon, because there are lots and lots of similar parts available, but all are variously not fit for purpose, or are collectively hiding the existence of a part that is - a prohibitive discovery cost (also effort dissipation). Or recursively, a part might not have been built, because its multiple potential uses have not been gathered to incentivize it, or a part it in turn needs - an economic communication failure. All our usual dysfunction.

I emphasize this because the "nobody really cares" model can be very attractive, but also very misleading. It can be a wonderful fit for observations, and predictive of further observations of the undisturbed system, while also being a very poor model for planning and predicting interventions.

For example, from kindergarten to undergraduate, introductory astronomy content tells students the simply wrong color for the Sun. In some sense, "nobody really cares" is an excellent description. Not teachers, professors, authors, publishers, reviewers, parents or students. But... they could care. Individually, it's not too hard to flip their state. And it's possible to imagine interventions with broader impacts. But we collectively don't get to it, for all the diverse reasons things aren't gotten to. Much of science education dysfunction is like that. Because "nobody really cares" that it succeeds... sort of, kind of, but also no - that's very not a root cause analysis.

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#83
The one thing I don't get about all those calculator applications is the existence of the keypad. Like, the users already have the keypad right in front them called keyboard (the names literally share the "key" part). Why would anyone instead choose to use mouse to click those tiny buttons? I feel most calc designers still stuck at the skeuomorphism and never jump out of the box to think about what user really needs.

The only calculators I know that makes keypad optional/hideable are SpeedCrunch and Qalculate!, and now this.

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#84

The one thing I don't get about all those calculator applications is the existence of the keypad. Like, the users already have the keypad right in front them called keyboard (the names literally share the "key" part). Why would anyone instead choose to use mouse to click those tiny buttons? I feel most calc designers still stuck at the skeuomorphism and never jump out of the box to think about what user really needs.…

We have calculators open on our machine label computers. Those computers don't have a keyboard only a touch screen. Without the keypad these computers could not use the windows calculator. That said this is an unusual use case.

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#85

The one thing I don't get about all those calculator applications is the existence of the keypad. Like, the users already have the keypad right in front them called keyboard (the names literally share the "key" part). Why would anyone instead choose to use mouse to click those tiny buttons? I feel most calc designers still stuck at the skeuomorphism and never jump out of the box to think about what user really needs.…

“skeuomorphism” isn’t even a real UX concept; it’s just been carefully designed to look like one.

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#87

Earlier quoted context omitted.

Why not just save those words, and other commonly used words to a startup file? Then, you can just do export PYTHONSTARTUP=/home/savant_penguin/.python/startup.py and those things will be automatically loaded for you. (Edit: add that line to your .bashrc if you want to use it all or most of the time!) If you don't want them always loaded, but rather just for a "calculator mode," you can do something like alias pycalc…

Thank you for this!!! I use the repl all the time and had no idea I could do this.

You're quite welcome. After 7+ years of using Python professionally, I find I'm still discovering things I didn't know. And, I'm not referring to just new stuff that gets added, either. It's cool to pick up on these little things every once in a while.

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#88
post #76

I love this, great idea! I don't need the majority of the fancy features but I can see the use case for them. I built my own calculator app a while ago [0]. I've used an HP-41C as my main calculator for many years and I have not yet found a good RPN calculator app available on ALL platforms. I wanted a simple calculator that felt like an HP-41C that I can use on my phone + all desktop operating systems and get the sa…

I like it, except that the Enter symbol confuses me. I associate it with a windowing button or on/off switch. I probably would use ⏎ (U+23CE) instead.

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#89
post #38

I still use the python repl in a terminal for my desktop calculator. Its a good choice since usually im already in the terminal anyways.

Reminds me of this: https://krasjet.com/rnd.wlk/julia/

Discussed here a couple of years ago: https://news.ycombinator.com/item?id=23414872

Re: Show HN: All desktop software calculators are wrong, so I had to build my own

#90
post #51

I'm surprised no one has mentioned Soulver yet ( https://soulver.app/ ). I have used it for years -- it looks like it has pretty similar functionality to this but (IMO) with a more native-feeling UI. I love it. I don't think it supports as wide a range of math, but it does seem to pass the accuracy test given in the article (sin(exp(37))). It also lets you assign variables which I find handy. On the other hand, the p…

I haven't seen Soulver, but it looks very similiar to the one I use on Mac, called Numi. On Android, I use an app called CalcNote which works in a similiar way. Neither does graphing, but you can do common formulas and functions like cos/sin/rad. I use numi for simple calculations, and find it fantastic when I'm trying to figure out project stuff. A recent use is entering wall dimensions to figure out square footage and compare different building materials. I often leave most of my old calculations in numi like a long-lived numerical memory.

https://numi.app/

https://play.google.com/store/apps/details?id=com.burton999....

Post reply on HN