I personally use Numi, which is very similar. I like the ability to assign variables and reuse them (and change the value of the variable on the fly, instantly redoing all other calculations).
I use tydig[1] on my iPad which is really nice. Especially being able to drag around results with links pointing back to them and then graphing things as needed.
Seconded (though I'm tentatively switching to dc[0], reverse polish notation ftw), can anyone who has used both Numi and Soulver point out the differences between the two? They look super similar to me.
For me the currency conversion, easy ‘totalling’, syntax highlighting, and mixing Markdown text inline (which gets ignored) is the key feature for these notepad/numi apps.
These still could be done in Vim but it’s more than just a unixy calculator or text-editor spreadsheet.
To each his own I guess. As far as I'm concerned, I'm mainly interested in the one-column spreadsheet, tweaking a few input rows and checking how it impacts some other rows. I still use instacalc from time to time, but now I often reach for vim instead.
I started hacking something up to invoke dc more easily…
$ . calc.sh
$ * 4 5
20
$ / 21 3
7
It’s a WIP. I’ll probably add floating point and a stack.
Which will mean moving away from `dc` to awk/python/something else running in the background, using sh co-processes.
The main difference is that you can easily drag n drop line result anywhere inside the document with Soulver for more complex computations. With Numi you are forced to make up variable names, which is silly.
I have written to Numi author about this a few months ago and he replied he wants to add this feature to Numi but I haven’t heard any news about that since. I am not following Numi closely, though.
Sadly, I am Android-only. Is there any difference in functionality, or it a pure clone (edit: or, is Numi the clone? Sound like Soulver has been around a while)?
I’ve never used Numi. Soulver is 14 years old and is from my favorite era of indie macOS development. He’s really responsive on twitter or to bug reports/emails.
Numi is also available in Setapp [1], which I pay for and really love (it’s a great value for me), but I must admit I’ve never used it. I’ll check it out later today, but don’t know if I can offer a comprehensive pro/con.
I prefer Calca. I just find variables easier to deal with. I used it most recently when I was working the food table at my kid’s school fete, and I didn’t have a point of sale. Just do all my calculations with variables so it was easier to spot errors. (Kids spending their own pocket money are ruthless at challenging your math).
I've been using Soulver for ages, I have Soulver 2 from the Mac App Store. It's a terrific little scratch pad, I love the ability to naturally interleave notes/language with arithmetic. I was unaware that there was a Soulver 3, however, as apparently they are no longer using the MAS as a distribution channel. Just a little top level note for anybody else like me that came here off the vague title alone and wasn't aware of Soulver 3.
> Why would anyone buy [any app] outside the Mac App Store?
I almost never buy an app from the Mac App Store when it's also available directly. I'd rather that the developer gets to keep the extra 30%.
Free apps I typically get from the store when available for the convenience of the updating system (though sparkle has made non-app store app updating pretty painless).
I don't find v3 a compelling upgrade for me but will upgrade anyway once syncing to iOS works again (important feature for me) just to support the dev.
> I almost never buy an app from the Mac App Store when it's also available directly.
Unpopular opinion, but I always try to buy Apps from the Mac App Store. It is easy to purchase and with buyer protection. You can easily install the app on another Mac and it updates automatically and all my apps are in the same place with just one account. But most importantly, Mac App Store Apps must run in a sandbox which is much more secure, than just downloading an app from the internet which might run with root privileges on your Mac.
I did purchase Soulver 3 though because I wanted to support the developer as I have been using Soulver 2 almost daily for 10 years. But sure enough, the in-app purchase experience did not work and crashed when going to PayPal. Ended up purchasing through the web-site. The downside now, is that my Soulver 3 App has megabytes of Paddle API frameworks embedded in it that is not used, and Little Snitch reports that Soulver 3 tries to connect to paddle 2 times every freaking time I start the app. Very annoying and unnecessary.
Soulver 3 has actually been a bit of a disappointment. They rewrote the parser and math engine in Swift and a lot of functionality has been lost in the process. I'm sure this will be addressed in further releases, but for now, it has no iCloud sharing and you cannot share documents with the iOS app. Bit flicking operators does not work anymore, which was so useful for programming. It has gained support for date time calculation which is very nice, like 'today + 30 days', but Numi does this better and also has a 'now' variable for current time you can do calculations with. The new Soulver 3 UI could be better and I especially find it annoying that the window has a rather large fixed min-width and I also don't like the different color for the right column. Numi is a a copy of Soulver, but has bypassed the original. Unfortunately, I found Numi after I purchased Soulver 3, but I have ditched Soulver 3 for Numi. Numi has more functionality, looks much better and has fewer bugs. Edit: I don't like that Numi is not available on the App Store either. Numi is created by a developer from Belarus, which is a country only slightly better than DPRK and I would very much like to have Numi in a sandbox.
Ha-ha. ))) C' mon man! So superficial about Belarus. )) I just don't want to bother with MAS, so it's much easier to automate all stuff. Btw all recent updates are notarized.
> I just don't want to bother with MAS, so it's much easier to automate all stuff.
That might sound good to you, but not to your users. Please put it in a sandbox and on the MAS. There is no reason for a calculator to _not_ run in a sandbox. Discoverability and sales will probably improve as well on the App Store. I for one, would purchase in an instance, whereas now the only purchase option is a paddle (sic) preorder button.
It is the newer version, there definitely appear to be some nice new features[0] and a UI re-fresh (although as a Soulver 2 user, I think I'm pretty happy as is)
I'm pretty sure that's a price reduction. I think Soulver used to be either $19.99 or $24.99; Soulver 3 is a price increase (although not unreasonable).
However, it doesn't sync with Soulver for iOS yet. I suspect Soulver 3 will probably come to the Mac App Store once the iOS app is updated to use the new file format.
Are you sure? Looking at my Soulver macOS receipt I did purchase it for $10 in 2012. I can't tell if it was a sale price but would be coincidental to have the same price cut 7-years apart.
The problem I have with this type of app is that I am never sure what will work and what will not. It's "natural language" only after you learn which subset of language the app actually understands.
I still can't find anything that beats my HP 50g (either physical or running in one of the emulators) for just about any kind of computation involving several pieces of data. For more data, a spreadsheet is really hard to beat.
I mean, I use PCalc[1] all the time too, but I love Soulver b/c unlike Excel —- which is great but can be a lot —- it’s super easy to just jot stuff down and have it calculated, just like a spreadsheet.
You’re not wrong on the syntax stuff, but as long as the math operations are there, it’ll do what you want. The special stuff for currency or other conversions isn’t very hard to get the hang of once you start using an app.
Really, the frustrating thing is when I’m on a machine that’s not mine and I hit my Soulver system-wide shortcut and can’t open up my scratchpad. But that’s my problem.
I know what you mean, but the grammar feels intuitive and the GUI is really clear, so it's pretty easy to extrapolate once you get a few things. For example, you want to calculate a price for three items at $425 in British pounds:
$425 * 3 in £
Once you've got "in £" you know it's available anywhere. I use it a lot for unit conversion too:
3 inches in mm
The other thing I use a lot is referencing other lines. So this will output £4:
Yeah I agree. Instead of learning a new language subset, I just use what's already there.
For me, the go-to for multi-line calculations is Octave. At that point I can annotate to my hearts desire, plus I've always preferred Octave plotting over spreadsheet functions.
The ability to share your calculations nearly universally, though, is the major advantage of spreadsheets.
I've been using v2 for years, and unless 3 changed significantly, the syntax is truly minimal. More importantly it only is introduced where it makes sense, ie not forced. It's also very succinctly documented in a well organized help menu item.
Eg "what % of 198 is 22"
You can still calculate this the normal way, but this syntax sugar is natural enough and easier to remember.
Or the other big example somebody already posted is currency conversion. "1 USD to GBP"
Even if you don't use any of that, the ability to have a readable notebook, variables, and other features make it far better than a traditional calculator.
I could never go back.
I have similar concerns as you, but my current approach of copy/pasting from notepad (so I can label things) to a calculator is very prone to error or being outdated. I think the tight feedback loop in an app like this makes things discoverable and builds confidence in the output.
edit: when I'm "deep in it" this is similar to what I use Excel/GDocs for, but they're heavier than I'd like for more casual use. I spend half the time I'm using a spreadsheet in Google looking at documentation for Excel functions.
There's an old version available, but for no good reason. I should make this public - email me at gabi [at] purcaru [dot] com if you want to contribute
I’ve been using Soulver for like 12 years and it’s one of my favorite and most-used apps.
There was a huge amount of time between Soulver 2 and 3 so I gladly paid for version 3 (even though it doesn’t sync with iOS which isn’t a big deal, and some features haven’t made their way over so I still keep 2 installed).
There might be decent alternatives but at this point, I’m so used to Soulver I don’t want anything else. Plus, the new Alfred workflow is really good.
Me too, but don't you find the way Soulver 3 manages documents a bit weird? I much preferred the old way where you could save document wherever you wanted, whereas version 3 seems to want to control all of that for you. Unlike the previous version I couldn't wrap my head around how it was supposed to work quickly, and went back to version 2. I was happy to pay for a new version just to make sure Soulver isn't abandoned because my working life would be noticeably more difficult without it.
Yeah it took me a bit of time and I admit I still use 2 a lot but I’m trying to force myself to use 3. I think part of it is the 9 years or whatever that I used 2.
My editor JOE has a built-in calculator equivalent to a Casio scientific: for example it can do statistics (sum, count, average, standard deviation) on a block of numbers you select in the edit buffer. I use this all the time.
In another project, I wrote a preprocessor for TeX- somewhat like Markdown, so that you can make TeX documents with an easier syntax. One of its features is a built-in calculator/spreadsheet. The idea is that \{1+2+3} is expanded with 6. You can also assign and reference variables \{a=3}. If these appear in a table, you can reference other cells: \{RRD+7} "Get first expression from cell two to the Right Right one Down and add 7 to it".
As Marco Arment once said, if Soulver isn’t open he’s not working. Rings true for me as well. I almost always have it running to help me through some calculation or more complex modeling. Highly recommended.
I reimplemented a simplified version of it 7 years ago (wow, time flies) and open sourced it, if people are interested. https://github.com/vjeux/soulver.js
This. Excel as a notepad would be great. Don't know enough about windows desktop development to implement my own solution, but I'm surprised that nothing currently exists.
What about Numbers given where talking MacOS and syncs across iOS and it’s basically free ? Seems like it would fit the bill for this kind of scenario.
I've used Soulver for several years and just upgraded so I guess I like it. I have to confess I originally bought it simply because I was delighted someone was reimagining the calculator interface and I wanted to support that. But I do use it every day.
It's not my automatic go-to calculator (that's either dc or RPN pcalc). But if I want to do any slightly complicated calculation but don't need the overhead of Excel/Numbers or Mathmatica (or am on my phone!) it's really nice to lay out some variables (cost/m2, length width, blah blah) and be able to tweak them. And I can fiddle on my computer and then see the model on my phone where it's also easy to edit.
Not a tool for everyone, but surprisingly good as both a programmer and general purpose calculator.
I really like Calca and still use it a bunch, but it has a few annoyances (some of which cause incorrect answers) and seems to get updated about once every 4 years.
Putting the answers inline rather than over in a separate column (ala Soulver and Numi) makes more sense to me.
Calca is absolutely incredible and an invaluable part of my workflow. I only wish that praeclarum would open source it (with a prominent option to donate) so that it can live on forever, since it seems a little undermaintained at the moment. As far as I know, there's nothing exactly like it out there (with respect to typing calculations in plain text, then immediately seeing results inline).
I've been a Soulver user since 2012, and it's been an integral part of my works, even acting as a scratch pad for ideas with numbers.
The UX is the plus here - I am aware that there have been several other apps that have offered a polished UX since then, but the convenience and my learned affinity to Soulver makes me keep it.
Insect is an open source, high-precision scientific calculator with support for physical units, which I wrote some time ago. There is a version on the web and one for the terminal.
I didn’t continue working on it after the class finished, though – I lost all interest after I realized that the Frink programming language, despite being free, is not open source.
I never knew there were so many people into notebook/repl style calculators, for lack of an accepted term. How did these never take over from the skeuomorphic style? There clearly seems to be a market below the spreadsheet/Jupyter notebook use case.
Back in the early days of the Newton, there was an app that let you manipulate algebraic expressions by dragging and dropping individual elements of the expression. Think factoring out a common factor represented by the expression sprouting parentheses with the factor out front.
There was a later effort called MathDrag'n but that seemed to run out of steam.
Does anyone know if a similar thing exists and is maintained now?
What's new between Soulver 2 and 3? I can't seem to find out any info.
This type of app seems like it would be useful, and I know I'm basically asking for free candy, but I don't find 30 day trials that useful. I'll toy around with it now, but 35 days later think about it for serious work and ever really get to evaluate it. Or, I'll drop $30 on it now and never think about it again.
You can evaluate emacs lisp everywhere in the editor with C-x C-e. C-u C-x C-e pastes into the buffer. IDK if that fits your use case or not but I hope it's helpful.
eval-last-sexp, which I believe is , e s in Spacemacs. Also, most Emacs defaults like that are carried over to Spacemacs so you can try SPC h d k C-x C-e and get info that way.
This looks neat, but I'm struggling with the landing page.
There are a bunch of animated gifs that show off how Soulver works - great! But just as they finish the exercise and show the result, they immediately blank and restart. I don't get any time to figure out what just happened. It's incredibly frustrating.
It’s also annoying that each example on the landing page only fades in when it is scrolled almost to the top of the window. I have enough vertical space to see two examples at once, but it always hides the bottom one to look “mysterious”. This is especially disorienting when scrolling up or down a page at a time.
Apple’s marketing pages do this type of fade-in better – the fade-in happens quickly as soon as the page is scrolled to a certain point, and after that, the faded-in item doesn’t fade out again unless you refresh the page.
Soulver is great, but there's one feature that seriously makes me hate using it.
I often open up "scratchpad" text documents. I Paste some text, edit it, Select All, then Cut. I press CMD-W to close the window. Because the document is blank, I'm not prompted to save the unsaved File.
If I try to do the same thing with Soulver, when I close the (blank) window the app pops up a modal dialog asking me where I want to save the (blank, never-saved, scratchpad-only) File. So annoying.
It seriously prevents me from quickly opening up or using Soulver just because I don't want to deal with clicking through that modal once I'm done. I'll just use a Text document instead or put the math equation into the Chrome omnibox.
In the save confirmation dialog of any Mac app, the keyboard shortcut Command-Delete (⌘⌫) chooses Delete for you. I found such dialogs much less annoying after I learned that.
So when you want to close a window without saving:
There is a setting for “Suppress save warning for unsaved documents”. Would this address your need, or do you only want to suppress the modal when the document is blank?
I have the setting left on because I am also annoyed by this, and I rarely need to save my Soulver documents.
A tangent, but there is an unspoken need for natural language processing/interpretation in EHR charting for nurses.
Very often, you enter values manually into a flow sheet around that status of a patient (ie vitals, PQRST for pain, assessments) but you may also enter a supporting note around the circumstances of your charting.
It would be beyond helpful to have an interpreter to pull in data within that note to place it within the appropriate data fields. It would eliminate a significant portion of duplicate charting.
I am putting it out there, because this is beyond me to solve this issue from a technical standpoint.
Soulver looks like a tiny step in that direction/way of thinking.
Is this really a good solution? My biggest concern is more psychological than technical. If the error rate of interpretation is too high, then it obviously won't reduce the effort. However, if the error rate is low, then the errors are much more likely to be overlooked.
The enemy of good may be perfect, but it seems that the closer to (without being) perfect the interpreter, the worse the outcome for the patient.
Like the concept, but I'm on Windows so I'll try out some of the alternatives mentioned in these comments.
My solution for Windows has been the NppCalc plugin for Notepad++. It is not quite as advanced, but does have a lot of functions that the alternatives don't (like some common encryption algorithms which came in useful recently)
I've used this sometimes, and still use it sometimes (a few times a year, probably). But I find it not very intuitive (beyond the basics of typing some names, numbers and operators in a list), and I couldn't figure out how to easily do subtotals in a long list (I don't want to write a long formula adding every other item above). Better documentation and examples could help.
I'm using a thing called microMath+ on Android (available on F-Droid) but it's kind of clumsy. I haven't tried hacking on it yet to improve the UI, but I'm wondering if there might be something else out there that's already better. (Not proprietary software, obviously.)
That's right. I'm sorry I didn't post the correct name in the first place; I don't have https://github.com/kragen/xcompose set up on my hand computer (because it doesn't run X).
I don’t use Soulver that much anymore other than while travelling, but it’s great for tracking my travel budget on my phone: able to clearly list what I’ve bought, for how much, what the grand total is in my home currency and leave comments on things.
I feel that with non-standard syntax like $100 + 25% tip it become very confusing whether the answer would be 100.25 or 125. I’ll end up spending a lot of time trying to verify the answer.
Type annotated calculator. If the author had a Windows version he would probably sell many more licenses, but understandably that's probably not high on their priority list.
I love Speedcrunch and it's an integral part of my Windows workflow, but not as flexible as Numi or Soulver. I would incorporate either of those two into my Windows workflow if they were available as such with the same responsiveness and footprint.
Has anyone found anything remotely comparable for Android? I'd love to have a Soulver or Numi analog on my phone. I don't even need it to sync with anything.
Since getting a Mac I've been finding that there's a whole market of mac only apps and I get the sense that it's just choosing your demographic: people who buy macs are more likely to pay for software.
Spreadsheets force you to pick cells, and the formulas are hidden. They are non-linear, this is linear. They also take a lot of screen space, this is tiny.
Well, since you can literally build game engines in free spreadsheet apps, probably none, but, as you noticed yourself, the selling point is the UX/UI, which seems to be excellent.
If you're only concerned about functionality and not UI, you can always write your calculations in hexadecimal machine code, at least if you don't have the misfortune to be on iOS or some similar game console platform where users aren't allowed to write code. But I think it's more practical to use a better UI.
I haven't used it, but not having to type your data points into separate columns seems to be the benefit here. Instead of forcing rigid entry you can type human-like numeric notes and it will perform the calculations on the fly.
A paid, Mac only app which could fairly easily be implemented as a plugin to any popular text editor... I'm not sure I understand the appeal - what am I missing here, please? :)
But to answer your question, yes, it turned a profit:
> The maker of cloud-based file-sharing software reported profit, excluding some costs, of 10 cents a share, above the 6 cent average estimate of analysts polled by Bloomberg.
It is funny - but to add a little nuance here, it sounds like they're reinvesting tons of cash back into R&D. If they were not reinvesting as heavily, they would be cash positive.
> Dropbox has demonstrated impressive growth since 2015, roughly doubling the number of paying users and revenue. Dropbox became cash-flow positive in 2016, but recorded a loss of $111 million that year after pouring money back into its product development efforts.
Yes! "It is better to have 100 functions operate on one data structure than 10 functions on 10 data structures."
I don't care if it's literally Emacs, but I want all my data in one place. Half my programming career was spent taking one dumb format and converting it into another dumb format because the systems were incompatible, and for no good reason.
People complain about needing "dongles" (i.e., adapters) with Apple's new laptops, but there's a simple and obvious fix (buy one), and we can see the endgame (everything USB-C) where even that won't be required.
Programmers spend all day building custom software "dongles", with no end in sight. Why aren't they angry about that?
Yes, the answer is yes it should do it if people want it and are willing to work on it. Let people decide what they want to build and use within the software they use.
I think you're missing the gap between "could fairly easily be" and "has been".
As a programmer, the REPL and/or notebook interfaces for my favorite languages provide what Soulver offers already, so it isn't as valuable to me. To a non-programmer, Soulver seems like a very pleasant, user-friendly tool.
I'm a programmer, at least occasionally, and Soulver is a pleasant, user-friendly tool to me. When I open Soulver, it's because I want to do some unstructured multistep math work and it's faster and easier to do it in Soulver than a spreadsheet program. Jupyter Notebook could undoubtedly do similar work, but it doesn't read to me like it would be the best tool for the job.
https://numi.io/