It’s a nice little device but it falls a bit short in one key area: it lacks a CAS (Computer Algebra System) for symbolic calculations.
That functionality was available in an earlier release of the software (Epsilon) but it was at best rather rudimentary.
I’ve flashed Omega (a fork of Epsilon that preserves the rudimentary symbolic capabilities of earlier Epsilon) which also allows running third-party apps (including Xcas). You can check it out here: https://getomega.dev
I’m a strong believer in calculators and I understand the restriction on CAS systems in order to facilitate assessment in exam environments, but almost all calculators (including NumWorks) have Exam Mode settings anyway so… I don’t really see why not. It’s a critically important tool.
I don't think it primarily competes with CAS calculators, but rather with the TI84: graphing, statistics, table of function values, plus elementary programming in Python, but faster, and at a lower price.
CAS is the cherry on the cake, which you can enable by flashing another version of the OS. And as you probably know, it's not harder than upgrading the OS, so I'd say it's a pretty decent compromise.
Pretty much this. This calculator is meant to break the hegemony of TI in high schools by providing a (cheaper) alternative. You'll end up having to disable CAS in school anyway, sometimes you're not even allowed to have it, even if you can disable it.
What kind of calculator do you need for high schools anyway? Is there a reason why Casio isn't as popular? It's cheap, more responsive than TI (from my experience with TI-83).
That would probably depend on the school and the country (perhaps even the teacher). In my case they didn't even mention that there was an alternative, the TI-83 Plus was pretty much presented as a must have.
I’m somewhat aware that Texas Instruments enjoys some kind of quasi-monopoly on high school graphical calculators in the USA. However, when I was taking IB Maths Higher way back in 1999, I was using a Casio. Right after that I bought a HP 49G when I went to university. More recently I’ve been using an HP Prime (which I adore, but it’s a bit of an overkill in many ways) and the NumWorks with Epsilon (which is underwhelming in other ways, principally the CAS and lack thereof/need to use a separate and rather clunky interface for Xcas).
Indeed… throughout my whole school career (ending with my graduation in 1999, albeit from an international school featuring the International Baccalaureate) nobody ever gave a fraction of a dam about what calculator we had, as long as it didn’t sport ‘banned’ functionality or features (the latter including a blanket ban on any QWERTY-layout keyboard, which was very odd).
It also still treats RPN as a second-class citizen. You can import a front-end, flash a custom firmware, yada yada... In ten years maybe it'll be right.
But realistically, adult human beings who still want a graphing calculator want either a CAS and/or RPN.
It's ridiculous that buying a vintage HP-48GX is still the best way to get solid RPN on a calculator. The software environment was incredible.
You can accept the worse keyboard of the HP-50G for the increased speed and memory, sure, but there was a product category, it was almost perfect, and Carly Fiorina killed it. We have chips with 100,000 times the capabilities and yet we can't get a calculator with solid software and decent keyswitches at any price.
This is one of the typically recommended calculators in high school in France. The schools usually get a reasonable deal to promote it to pupils.
It is a good calculator, but the Python part (that was supposed to be a strong asset) is not that well managed. I just tried a few things on my son's calculator but gave up. i was expecting some kind of live interface between the REPL or an editor and the calculator but it really looks like an Arduino that you have to update with code all the time.
Interestingly the license in the software and hardware is the: "Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
Public License"
> The license must not restrict anyone from making use of the program in a specific field of endeavor. For example, it may not restrict the program from being used in a business, or from being used for genetic research.
> Q: Can I say "You must not charge [much] money for distributing the program"?
> A: This is non-free. We want Debian to be distributed by for-profit CD vendors and improved by corporate resellers. We can only include programs whose authors allow this.
> For many users buying Debian on disks is more convenient (and cheaper) than downloading. For-profit distribution is the most reliable and convenient way to ensure that Debian disks are easily available everywhere where there is a demand. Because everyone can download the CD images and start producing their own disks, competition ensures that nobody will make an undeserved fortune distributing Debian.
> Q: Can I say "You must not use the program for commercial purposes"?
> A: This is non-free. We want businesses to be able to use Debian for their computing needs. A business should be able to use any program in Debian without checking its license.
It also goes against the Creative Commons people's guidance on the matter:
> Can I apply a Creative Commons license to software?
> We recommend against using Creative Commons licenses for software. Instead, we strongly encourage you to use one of the very good software licenses which are already available. We recommend considering licenses listed as free by the Free Software Foundation and listed as “open source” by the Open Source Initiative.
Debian and NumWorks obviously have different perspectives. The points raised in your comment are more ideological than practical.
Personally I am happy that a company that cares about source accessibility is making decent calculators, and slightly amused that Debian and the OSI seem to find that worse than copyleft.
>The points raised in your comment are more ideological than practical.
The primary point is that this is neither Open Source nor Free Software, by OSI and FSF definitions of these terms, which are the generally accepted ones.
When a claim of "open source" is wrongly made in the interest of marketing, the expectation should be for this misrepresentation to be called for.
> When a claim of "open source" is wrongly made in the interest of marketing, the expectation should be for this misrepresentation to be called for.
On the other hand, nobody has a monopoly on the “open source” concept, or even on putting the “open” qualifier in front of things. Broadly speaking, source code ranges from completely closed to completely open, and it is difficult to argue that code under a CC licence is not more open than, say, Windows’. Or indeed the usual calculator operating systems.
It is like the free software flame wars all over again, this time with the OSI in place of the FSF.
>They really aren’t. These are colloquial terms, not trademarks.
They are terms with strong definitions that are well accepted in the community, including the industry.
>It’s not a misrepresentation. “Open source” is a generic term.
It isn't. See above.
>Pretending it has a legal meaning is just not a reflection of reality.
Nobody said a thing (yet) about legal meaning.
However, there are licenses (legal territory) which are Open Source, and there are licenses (such as the ones used by NumWorks) which are not. And, to top it off, they are not Open Hardware either.
And I'd suggest great care on avoiding the misuse of "Open Source", as the OSI might invoke their right to protect their trademarks[0].
Looks like this single codebase is used to build for the calculator device, and all targets for the simulator: Android, iOS, macOS, Windows, and the web.
> NumWorks Epsilon is released under a CC BY-NC-SA License.
I had no need for one but got it anyway. The OS is really easy to build and install. Amazing work. Try booting it up next to your smartphone. There is no perceptible boot time -- and its arm cortex processor is probably much weaker than your phone's. It reminded me what computing could be like if you're willing to cut out the layers and build from scratch.
The Nokia "brick" would boot up really quickly too. But a modern smart phone is such an "enabling" device, that I don't think you could realistically build a competing platform without wholesale incorporating something like Linux, with all its startup processes. It would take ages to write a minimal OS with the functionality and compatibility that modern apps (and thus users) expect.
BTW, I have an N0100 too. I completely agree there's something pleasing about the immediacy of a calculator.
Despite claims in the title: This is not Open Source, nor Free Software, due to the choice of a license that does meet neither OSI nor FSF definitions, as it has restrictions on commercial use.
In case anyone sees this as an exercise in a technological "false ceiling" (as in, why make a physical calculator with limited RAM and display instead of a more capable calculator app for a modern cell phone) consider this page your answer:
Looks neat. Will support it (buy them as gifts). However, I started with the original HP41 RPN calculator and just can't make myself use an algebraic calculator unless I have absolutely no other option. I probably own a dozen HP calculators of different vintages. I bought my kids HP calculators as well, which they love.
I would love something like this to become standard, but I recall needing a specific calculator in HS. Our math classes were based on the TI-83, and people with an 84 or 86(?) would struggle because the interface isn't exactly the same. Maybe there is more variety now and classes can better handle other calculators.
My wife is still using my old TI-89 about 20 years later, but if that ever dies I will probably pick up something like this for general use. I would also much rather my 6-year-old learn python instead of spending her time making goofy games in TI-BASIC like I did.
> I would also much rather my 6-year-old learn python instead of spending her time making goofy games in TI-BASIC like I did.
There is something about programming in very restrictive environments though (granted it's for fun). Trying to see what you can make a small device do if you're creative enough can be a lot of fun.
Btw, the newest TI-84s and Nspires also have Python.
I find it wild that US HS classes seem to demand specific calculators at all, let alone graphing calculators which can do a bit too much of the work for you IMO (for education purposes obviously). In HS we used the class set of TI-83s for like 1 week to easily play around graphing some functions something that could probably be done on most student's phones with a free app these days.
My HS classes let you use a scientific calculator but I think honestly that's a bit too much too for most purposes. In university we were almost never allowed a calculator for the exam on tests and questions were just designed to be doable without one which is a much better test of how well you know concepts than how well you know your TI.
> let alone graphing calculators which can do a bit too much of the work for you IMO
I can no longer do calculus because I got a TI-89. Granted I don't have to do much calculus day to day, but if someone walked up to me and demanded I tell them the derivative of 2xe^2x I would have no idea what to do.
The TI 68k devices are great. Or more actually, they were great for their time 20 years ago, and it is a damned shame that no really better device has emerged since. The document centric approach of the Nspire makes them a pain to use and the CAS does not seem to be any better.
That said, you can debate rechargeable batteries vs AAAs (both have pros and cons), but a backlit screen is a really good thing.
> Documents (or being able to have/save separate "calculation sheets") are actually progress vs not having them.
Yes, documents are good. But the Nspire architecture that wants you to create a document before choosing what to do in it and the weird tab idea is not great. As you say, a TI-89 is quick and efficient in comparison, which is why I still use it despite having got a Nspire. If I need to mess around before I can get something useful, I might as well start Mathematica.
> The impression I get from newer calculators is that some corporate decision maker decided that "cellphones are popular and we must copy them".
That is definitely a problem. They are not getting any more ergonomic.
> If you need to use the calculator in a dark room, there's worse problems with your setup than lack of backlight in the calculator.
My eyesight is not perfect and anything that helps readability is an improvement. Sometimes I have to work in the evenings as well.
Having backlit screen is probably the reason why it needs lithium battery in the first place, as a simple pack of 4 AA batteries wouldn't last as long.
Lithium battery has a life cycle of around 1000-2000 charges, and may deteriorate even faster when exposed to high temperatures. Depending on the capacity though, you can expect a cycle to last more for a calculator compared to, for example, a smartphone, so it might not be that much of a problem (but then again you have the backlight, a major drain). When the battery dies, you can't simply go to a nearby store to get a replacement. You usually have to send the device back to the manufacturer to get it replaced.
I can't speak for the NumWorks but battery replacement on the Nspire is trivial and their availability seems to be just fine. Of course the ease of replacement and availability of AA/AAA batteries is better than the lithium battery, but it would not be a reason for me not to get the calculator.
If you want algebra, use a computer. I’m talking about doing calculations and I think it is hard to beat the efficiency when entering them into an HP–48-like device
1. That’s not a calculation. It’s a special case you either need to know or have to be clever to compute. A calculation might be, say, compute zeta(2). But I don’t really want to argue about semantics.
2. This can evaluate this correctly on the (slightly later) HP-49G+
So given this new information, will you change your mind or try to find some other silly example?
"arithmetic also includes more advanced operations, such as [...] exponentiation, logarithmic functions, and even trigonometric functions, in the same vein as logarithms (prosthaphaeresis)."
I own hp50g and TI-89, and I keep picking the latter. It is simply that much more slick.
When using hp50g, I find its input<>reaction latency makes me want to throw it at a wall. I used to tolerate it and rely on the input buffer, but that simply isn't good UX. I'm glad I grabbed a used TI-89 at some point.
Still keeping the hp50g for nostalgia reasons. Is what I used back at university; We went a long way together.
Intriguingly, log is computed by first doing a prime factorization. Perhaps that gives slightly more precision, but I always thought that prime factorization is way harder than computing log (say via an approximation of the infinite series).
Actually, the only page in the site to claim "open source" with regards to the calculator is a blog article[1] by a non-technical staff. Elsewhere, it is only referred to as an "open" graphing calculator. The HN submission is titled incorrectly.
They should also inform their non-technical personel during on-boarding about the significance of the "open-source" term in terms of licensing.
There are organizations and corporations that focus on such terms during on-boarding either because that's their development model or they are aware that using such a term while not adhering to it could draw valid criticism.
I believe such market domination in the technological stack that powers most servers isn't evident but I would appreciate a citation. In addition, market domination in any market domain is determined by several complex factors.
Those servers use that stack as a component in closed systems, exactly the way Apple and Google use open source as part of their closed systems.
Frankly the argument that Linux powers servers is a red herring.
Look around at the devices people use.
Yes, some hackers like us use Linux boxes. Yes, there are some possibilities opening with things like the raspberry pi.
No, open systems are not succeeding in displacing closed ones.
Recommending people buy a closed system instead of one which is open source to punish the company for not complying with an ideology does not help the growth of open source.
It’s a nice little device but it falls a bit short in one key area: it lacks a CAS (Computer Algebra System) for symbolic calculations.
That functionality was available in an earlier release of the software (Epsilon) but it was at best rather rudimentary.
I’ve flashed Omega (a fork of Epsilon that preserves the rudimentary symbolic capabilities of earlier Epsilon) which also allows running third-party apps (including Xcas). You can check it out here: https://getomega.dev
I’m a strong believer in calculators and I understand the restriction on CAS systems in order to facilitate assessment in exam environments, but almost all calculators (including NumWorks) have Exam Mode settings anyway so… I don’t really see why not. It’s a critically important tool.