Geometry app three suggestions

[eried]

- A way to input a coordinate:
I know tap and cursor keys work to adjust the position, but if you zoom you sometimes can't longer get some positions like 0,0

- Better touch control:
Zoom gesture
Tap and drag should just pan the area, not change the cursor position

- More visual cues, example middle of segment, tangent when tapping (you don't really want to input a specific position just by tapping, so it makes sense that object have "magnetic" cues)

I really wanted to check this app once the prime was announced because I always was impressed by the TI one, but in the current stage the app is, it is barely usable for anything simple (and the flickering is so noticeable in this one, its like watching an old crt thru a camcorder viewfinder).

[Tim Wessman]

Just three?

I can think of quite a few more... you will be pleased.

[eried]

(04-08-2014 09:41 PM)Tim Wessman Wrote:  Just three?

I can think of quite a few more... you will be pleased.

Correct, there are several more if you compare to any other geometry app but these three I believe are the required to at least be able to use this app for something.

I really have tried to use this app but it's impossible. 'Notes' is more effective to take geometry data than the geometry app.

[orcinus]

Dumb question: does the zoom gesture work anywhere in the current firmware version?

[eried]

(04-08-2014 11:12 PM)orcinus Wrote:  Dumb question: does the zoom gesture work anywhere in the current firmware version?

In the spreadsheet

[parisse]

(04-08-2014 08:23 PM)eried Wrote:  I really wanted to check this app once the prime was announced because I always was impressed by the TI one, but in the current stage the app is, it is barely usable for anything simple (and the flickering is so noticeable in this one, its like watching an old crt thru a camcorder viewfinder).

To the contrary, you can do much more than on the TI, but you must sometimes enter your input in the SYMB view. For example point(1,2) if you want to enter exact coordinates for a point. If all inputs are exact, then all computations will be done exactly, something you can not do on the TI (the geometry app is a CAS application). You can make a CAS program with geometric and normal arguments returning a geometric object and use it in the geometry app.

[Angus]

You can't be serious... The apps are designed to have easy to use interfaces for doing calculations. The backbone of the new calculator, the innovation if you want. Targeted at people like students in school who even don't want a manual.
Pointing out that there might be a potentially higher usability than in TI's implementation (which implies that you are almost happy with the current state) just if you know non-linear operation and workarounds to achieve your goal, in my eyes, company blinded. You soon fall into the modes-trap und non-linearities then, too.

That makes me fear, that the people from HP really can't follow the many suggestions and just don't understand the complaints from the users, because they just don't get their points. I hope that I am wrong otherwise my expectations for the near future need a serious revision... (I'm not talking about Tim's announced entry modifications for this specific app. Take it more generalized)

[eried]

(04-09-2014 05:53 AM)parisse Wrote:  To the contrary, you can do much more than on the TI, but you must sometimes enter your input in the SYMB view. For example point(1,2) if you want to enter exact coordinates for a point. If all inputs are exact, then all computations will be done exactly, something you can not do on the TI (the geometry app is a CAS application). You can make a CAS program with geometric and normal arguments returning a geometric object and use it in the geometry app.

True, but that defeats the complete purpose of having this calculator format. Have you noticed how in the TI app you can type directly the coordinate or just tap (having a customizable grid to snap the points), two small details that makes that app useful and quick to learn & use.

I have really tried to make some use of the geometry app, but it's hopeless. As I said, the notes app is better to input geometry than the geometry app.

[parisse]

(04-09-2014 06:42 AM)Angus Wrote:  You can't be serious... The apps are designed to have easy to use interfaces for doing calculations. The backbone of the new calculator, the innovation if you want. Targeted at people like students in school who even don't want a manual.
Pointing out that there might be a potentially higher usability than in TI's implementation (which implies that you are almost happy with the current state)

I consider myself that the current geometry app is perfectly usable as it is, other might think differently of course, especially if they are used to a complete different way of doing things. The philosophy of the current geometry app is that you choose one view to input: if it's the PLOT view then it means you don't really care of the coordinates of the point, otherwise enter a commandline in the SYMB view with the precise coordinates. If you define an object from other objects, from PLOT select the command, then select the other objects by typing their initials on the keyboard and enter or comma between the arguments of the function you have selected. In other words, it is a commandline-oriented geometry application, with a helper to enter the commandline from the PLOT view.
If you think that easy to use is contradictory with entering/modifying commandlines then you won't like the current geometry app. But you will miss the unique features of this app: you can do analytic geometric *proofs* using exact computation (unlike conjectures with other dynamic geometry calc applications), and it is fully integrated in the CAS system: you can do any kind of computation in the CAS history with the variables defined in the geometry app.
Moreover from a pedagogic point of view, I really believe it is much more interesting to describe a geometric construction by a list of commandlines (the SYMB view) than to explain which menus you must tap to get your construction, because the list of commandlines is much more universal than something that works only with one software (and you can't get the construction displayed by the software anyway).

[Angus]

Quote:Moreover from a pedagogic point of view, I really believe it is much more interesting to describe a geometric construction by a list of commandlines (the SYMB view) than to explain which menus you must tap to get your construction, because the list of commandlines is much more universal than something that works only with one software (and you can't get the construction displayed by the software anyway).

I second that. Totally.
It is just that I got under the impression that HP prefers the tap here, click there for this kind of task approach in the prime. Wasn't that implied by HP employees?
If you go the way you described that has (again) to be consistant throughout the calculator and it has to be explained und communicated to less versatile people. Which I would consider myself in terms of the prime.
Well - not that discussion again in an other thread :-)

[parisse]

I can't speak for HP of course, only for me.
I think there are 2 approachs to using the calc or any scientific software, one of them is commandline-centric and has a small learning curve and the other one is tap- or mouse-centric. This second way will give easier access to designed computations but it will be harder or impossible to make something else and may be hard to adapt to another similar software. This is what you get in many Apps on the Prime, where it is easy to make a parametric plot but not easy to mix the graph there with a graph from the function plot (you can do that with the geometry App or the advanced grapher on the prime) or a statistical plot (you can only do that in the geometry App on the prime).
I think myself that it is more efficient to learn the commandline-centric way, because once you understand it, you will be able to use it to make things that the designers of the UI did not necessarily think about, and you will also be able to learn any other scientific software quickly, something that is certainly important for people buying a Prime: we speak of an expensive calculator here, twice the price of recommended calcs in highschool - at least in France.
(By the way, the geometry app could be part of the process of learning how to write commandlines, create an object in the PLOT view, then look at the translation in the SYMB view.)
As for the PLOT interface, nobody pretends it is perfect, it can certainly and will perhaps be enhanced, but I don't consider myself that it is the essential point to consider when comparing it to other calc geometry applications.

[Tim Wessman]

(04-09-2014 10:15 AM)Angus Wrote:  It is just that I got under the impression that HP prefers the tap here, click there for this kind of task approach in the prime. Wasn't that implied by HP employees?

I would like you to explain that comment a bit more. I don't think I am following you.

[eried]

I am not really against the command line, the suggestions only add up functionality that in my opinion is needed to be able to use this app (that's why it is only three things); after all this is a geometry app, not a geometry library.

BTW parisse, I think you will love openscad , that's the perfect analog. You can create any complex shape in minutes in a rich GUI based app like Inventor, or you can learn a lot and create something similar in openscad in few days without ever touching your mouse. Both might produce the exact same result but each one has a different purpose. In my case, command line is only an option for batch, automated operations.

[parisse]

eried, my main point is that you will never be able to do the same "with the mouse only" than with a combination of commandlines. For example with the prime you can not illustrate the Moivre-Laplace theorem (http://en.wikipedia.org/wiki/De_Moivre%E...ce_theorem) except with commandlines in the geometry app, despite the fact that it's an important topic in the highschool math curriculum (in France at least). With a mouse only approach you will only be able to do what the designers have designed to do.
I'm not against easy to use helpers to enter commandlines, and that's exactly what the PLOT view of the geometry app provides, but commandlines are not restricted to batch use (and by the way a dynamic geometric construction *is* a batch script), they are an essential component to learn when using a CAS as well as a first step to learn programming, for me it should be mandatory to teach commandlines in any scientific curriculum. And it would have nice side-effects for maths, because it would make students think a little more to things like implicit multiplication, operators priority, function/value of function, chains of implications, ...

[eried]

If that is your example, I think I was wrong about the "geometry app" concept. The other 3 geometry apps I have used weren't designed for that example either, so I guess "geometry app" is perceived a bit different inside france.

[Angus]

hello Tim,

thanks for asking.
When the discussion came to the lack of well written documentation and about conceptional decisions it was said, that the prime is simply wanted to be different from, say, the 48 series. I think it was you who clarified that many things are a result of being targeted at students who want a device to play with without even looking into a manual, without written examples and so on. That is a design goal you can accept. Who don't even go into programming and thus won't face unconsitant typing issues and command behaviour. (Just to point out some aspects I hope my intention is clearer now)

And now there are coming up more and more usage hints, almost workarounds. You learn what command can be used in a program and which one doesn't work, what to be capitalize and so on. You gather such information but the deeper understanding of the device and why things behave as they do is an other matter.
HP still didn't come up with some sort of application note for usage beond the standard manual but such information seems to be vital for advanced usage. That is contrary to the initial concept of a painless to use calc - and thus should not be counted on the plus side when comparing two devices. It is just not self explanatory nor is it explained somewhere.
The official HP statements still say that -playing- with the device is a way to get a mathematical feeling. It is the whole justification for the explorer apps for example.
That is a polarizing approach maybe generated from scholastic theorists.

Did I understand you wrong when characterizing the prime? Simple and intuitive to use i.e. school centric with a cost- and time consciousness development process in mind?

I hope my thoughts got clearer and thank you for taking the time.

[cyrille de brébisson]

Hello,

Well, first, we do what we can, we are only human afterall!

No-one can claim that Prime does not have documentation, there is a LOT of it, I mean, the User Guide is not a small book!

Prime is designed to have a very low entry bar*. It is designed so that basic users can do basic to mid cimplexity stuff (what they will find at school) without having to open the manual and without too much risk of shooting themselves in the foot. Not an easy feat...

Prime is also designed so that advanced users can do whatever they want and have great fun. But with great fun comes great responsability! And yes, not every advanced/corner case use is fully documented. Most of the time because such use mix/match multiple items comming from different parts of the calculator, and these were never tested together before. As a results, some 'interesting' things might happen, and the specific interraction might not be documented.

Prime is also an agregate of various items which were put together, and such integrations never go 100% smoothly...

So, I hope that this helps...

Cyrille



*is bar the right spelling or did I just say that the calculator could offer you an easy drink?

[parisse]

(04-09-2014 07:42 PM)eried Wrote:  If that is your example, I think I was wrong about the "geometry app" concept. The other 3 geometry apps I have used weren't designed for that example either, so I guess "geometry app" is perceived a bit different inside france.

As I already said, at least in France, the Prime price means that the users are certain they will learn science after highschool (and most certainly "hard science"), they will get the Prime when they are 16 or 17, not before. Which means they need an analytic geometry app where you can mix pure geometric objects and other plots and where you can do all kind of analytic geometry computations outside the figure.

[Alberto Candel]

This semester I am teaching a course on computational geometry. I did recommend my students the HP Prime, but couldn't make much use of it. Maybe my shortcomings. I ended up using Mathematica. I do feel that the Prime has a great potential nevertheless, but it requires lots of learning and hands-on programming experience that my students did not have. Of course, I could have focused on actual programming with the Prime, but that was not my original plan.

Anyway, here are a couple of suggestions:

* Make the Geometry App more like the Computational Geometry Package in Mathematica. ConvexHull, Voronoi Diagram, Delaunay Triangulation should be readily available. There is a convex hull function in CAS, but it requires some extra programming to go from the list of points to the picture. I could not find Voronoi or Delaunay.

* It should be possible to save a picture obtained on the Geometry App of the HP Prime as a PDF or EPS or something like that which then could be transfered to a PC or smart phone via Bluetooth (yes, the connectivity kit does not work in Linux). This capability would be great for preparing LaTeX docs for the course.

That is it for now.

[eried]

(04-11-2014 10:48 PM)Alberto Candel Wrote:  This semester I am teaching a course on computational geometry. I did recommend my students the HP Prime, but couldn't make much use of it. Maybe my shortcomings. I ended up using Mathematica. I do feel that the Prime has a great potential nevertheless, but it requires lots of learning and hands-on programming experience that my students did not have. Of course, I could have focused on actual programming with the Prime, but that was not my original plan.

Anyway, here are a couple of suggestions:

* Make the Geometry App more like the Computational Geometry Package in Mathematica. ConvexHull, Voronoi Diagram, Delaunay Triangulation should be readily available. There is a convex hull function in CAS, but it requires some extra programming to go from the list of points to the picture. I could not find Voronoi or Delaunay.

* It should be possible to save a picture obtained on the Geometry App of the HP Prime as a PDF or EPS or something like that which then could be transfered to a PC or smart phone via Bluetooth (yes, the connectivity kit does not work in Linux). This capability would be great for preparing LaTeX docs for the course.

That is it for now.

Hey Alberto thanks for sharing, can you elaborate what things made you use another geometry tool? I never used anything for geometry in school/university so I am a bit away from the field, but I thought these geometry tools were made for interacting "graphically" in the path of finding a numerical solution (like for example I always design small pieces in 3d to 3d-print and I always have to solve hidden angles of the objects, usually on a paper)

Does your suggestions replace the ones I suggested in the first post in "priority"? (from your view point)