Algebra Rizing
in [Lisp]
|
From: Kenneth Tilton on 4 Jul 2010 20:03 Kenneth Tilton wrote: > Mario S. Mommer wrote: >> Hi, >> >> Kenneth Tilton <kentilton(a)gmail.com> writes: >>> Thanks again for the report, but to anyone out there, please relax on >>> the bug reports. I have prominently announced that the thing is known >>> to be broken and incomplete and is being shared only for what does >>> work (which will expand in Algebra tutorial functionality before >>> anything else, except for tree bugs that make the tutorial forest >>> impossible to see). >> >> Kenny, that kind of thing very rarely, if at all, works. If you publish >> an unfinished prototype with a ten megawatt lightsign of a disclaimer, >> people will still complain about how buggy it is. No idea why. > > Human nature. I would do the same! It is impossible to look at something > broken and not point it out. > >> >> Same thing with text. If you send someone an unfinished text, clearly >> saying it is just an early draft, asking to just check if the concept >> has half a chance of going somewhere, you'll get even the spelling >> corrected 95% of the time. >> >> So, good luck with this. But please try not to insult all the nice folks >> that have given you feedback. > > You are confused. I have had a couple of perfectly pleasant exchanges > with people offering bug reports by email and I thanked them and > reassured them they could relax on that, tho there was at least one case > where something I thought worked did not (because I had TeX fonts > installed and did not notice I had cocked up the fallback image fonts). > Other stuff as well, I think. > > The flaming you are seeing is me making life miserable for the js > library haters, who yell about anything they can because things like > qooxdoo do work. We have had fun before in c.l.js when I made the > mistake of coming here looking for help and got attacked because I was > using jsQuery (my first try, on an unrelated project). > > Can you imagine someone coming to c.l.lisp asking for help with ASDF or > Slime and being attacked for using.... oops. > >> >> Now feedback on the algebra app: I'm usually somewhat weary of telling >> kids that maths is easy, because that does not correspond to what their >> experience with it will be. After a while, when they don't get it, they >> not only feel frustrated, but also stupid (because you claimed it was >> /soooo/ easy, remember?). In my experience, an important part of the >> difference between people who do better in math and those who do not do >> that good is the level of tolerance to frustration. > > I agree that forewarned is forearmed, but in what we USians call > "Algebra I" their experience is a function of the process, not the > content. I explained it here: http://www.stuckonalgebra.com , tho it > sounds like you read that. Oh, no you just made it to the first screen. Forgot I had put that there. That is inside joke for c.l.l'ers who right recall one of us responding once to a nooby whining about some challenge with "Lisp is hard! Let's go shopping." which I subsequently learned was first said by a "Talking Barbie" doll about math. Was that as far as your review got? :) Try again: http://teamalgebra.com/ and try this: Solve OK 3x-6=12 Return 3x/3-6/3=12/3 Return (a) wasn't it fun typing that? (b) what do you think the software will say given the tortured approach to the solution? Correct or incorrect? btw, anyone mindlessly banging away at the tab key foaming at the mouth screaming "dysfunctional!" must not understand much about software: the site may be a train wreck, but it presents a one-of-a-kind wysiwyg math editor as well as a one-of-a-kind automated tutor able to assess intermediate steps in a problem. I think a couple of things must be functional. :) kt ps. If I'm a troll, we need more trolls. k -- http://www.stuckonalgebra.com "The best Algebra tutorial program I have seen... in a class by itself." Macworld
From: Stefan Weiss on 5 Jul 2010 00:50 On 05/07/10 02:03, Kenneth Tilton wrote: > Try again: http://teamalgebra.com/ and try this: > > Solve > OK > 3x-6=12 > Return > 3x/3-6/3=12/3 > Return > > (a) wasn't it fun typing that? > (b) what do you think the software will say given the tortured approach > to the solution? Correct or incorrect? Are you genuinely interested in feedback, or are "yes" and "correct" the only answers you'll accept? If you think you can't learn anything from critical feedback, please feel free to ignore the rest of this message. (a) Was it fun to type it? Not really. There was a quite noticeable lag here (25mbps connection, central Europe, reasonably fast computer, FF3.0). I don't have the TeX fonts installed, which is probably also the case for ~99% of your prospective buyers. You may want to test your application without those fonts on your system. "Foaming at the mouth" critic that I am, I also dislike having to relearn how to enter text. Most of the keys I usually use for this purpose don't work in your editor (but still cause HTTP requests), like End, Home, Delete, Shift+Arrow, etc. Copy/paste are right out of the question, of course. It's also not possible to open a new tab with Ctrl-T when the editor has focus. It intercepts the key combination and prints a "t" character instead. Entering "3x/3-x" yields "(3x/3)-x", whereas entering "3x/x-3" gives "3x/(x-3)". Unexpected, but that may have been explained in the typing tutorial section, which I was unable to complete for technical reasons (which you don't want to hear about). Okay. I guess I can live with that, if it's the price for an interactive algebra tutor app. But is it *fun* to type like that? Hell no. On to your second question. (b) what do I think of the solver's reaction to my input? Is its reaction "correct" or not? All the solver had to say was "Hmmmmm....". I don't think a student would consider that very helpful. Playing some more with this example: 3x-6=12 Return 3x-7=11 Return The solver's comment for this development is now "Sweet". I tried the "3x/3 ..." example a few more times with the same input, and it looks like the responses are randomized. Sometimes you get "Whoa" or "Unusual" or "That's different" instead of "Hmmmmm...". Here's a suggestion: these responses may look funny to you, but they're going to annoy people who are struggling with their task. Don't just say "Hmmmmm...", tell them what's wrong or unusual. They're kids, not idiots. Yes, I do work with kids, and I know what I'm talking about. Just in case I haven't landed in your killfile yet, and you've made it to this point without throwing me in your imaginary "library haters" pot, here are a few more observations: You asked Gildas why he was watching HTTP requests. Your application automatically includes Firebug Lite, which just about screams "debug me" to all developers (i.e. most of the people reading c.l.javascript and c.l.lisp). The only way I can stop Firebug Lite from opening and obscuring most of the visible area is by activating the real Firebug extension for your site, and then, yes, I will see the HTTP requests. Which brings me to another point: with Firebug Lite open, the usable vertical space in the content area is 70px on my screen, barely enough to see the huge headline. Even without FB Lite, the area occupied by the equation editor is about the same size as the app's header area on my screen. I'm currently on a laptop with an 800px vertical resolution. I think it's safe to assume that many of your prospective users will also be on laptops or even netbooks. There's no need to load a 512x512px clip art image which serves no purpose at all (it's rendered as 200x200px, but that doesn't exactly improve things). If the header is only used for this image and "Whoa" comments, make it smaller. The right mouse button is disabled on your site, which is something I absolutely, positively loathe. It serves no purpose at all except to annoy your users. Anybody who really wants to copy/view anything will be able to do so anyway. In case you're interested, the application took about 20 seconds to load with a blank cache, and 15 seconds after the external assets were cached. I didn't have any gray screens of death this time, but a "loading" indicator would be a nice touch. Usually, if a page doesn't load in 10 seconds, I'll close the tab, but I do make exceptions for more specialized applications like yours. > btw, anyone mindlessly banging away at the tab key foaming at the mouth > screaming "dysfunctional!" must not understand much about software: the > site may be a train wreck, but it presents a one-of-a-kind wysiwyg math > editor as well as a one-of-a-kind automated tutor able to assess > intermediate steps in a problem. I finally found the message about your application being a work in progress. You have to double click some text (not a link) reading "Help!!!!!". Did you expect everybody to find and read that? Your site may be one of a kind. I didn't check. It may be a valuable product, and I can certainly see how an interactive approach to algebra can help students. But you're posting to a JavaScript newsgroup here (and to c.l.lisp for some reason). If you want to promote your product, this is not the right place; try alt.algebra or alt.algebra.help instead. You posted a self-described "train wreck" prototype to this group, so what the hell did you expect? If you did it on purpose to annoy somebody here, that's classic troll behavior. > ps. If I'm a troll, we need more trolls. k No. Nobody needs a troll. Least of all this group (cljs). We've got our share of difficult characters; there's no need to pour oil into the fire. Judging from what I've seen on your blog and elsewhere, you look like a nice enough person, but your recent posts in this group _really_ irritated me. You apparently throw everyone who doesn't agree with you into your "library haters" group, and completely dismiss any advice you're given. You've managed to antagonize just about everyone who replied to you, and most of those people didn't even mention libraries. When you act like that, the moral high ground you imagine yourself on just disappears. You're getting free beta-testing here. Whether you want it or not, when you post to this group, that's what you get - including all the not-so-flattering comments about your "train wreck" site. I think the least you could do is reply in a civil manner to the people who are actually trying to help you. I don't even know why I bother to write all this. Maybe because I think that you could easily post in a friendlier and more diplomatic way, if you wanted to. All kinds of people read and post here. If you can't get it into your head that we're not a homogenous "anti-library" cabal, you should stop posting in this group. Or you could open your mind a little and listen to what people are actually saying. stefan PS. At least your algebra tutor is already better than some of its more shady online competition. There's a site named algebrahelp.com, which claims to show all the required steps in solving an equation (non-interactive). You can click a "sample equation" button and see for yourself: --- begin quote ----------------------------------------------------- Sample equation: 2v + 22 + -3v = 2 + -1v + 0v Reorder the terms: 22 + 2v + -3v = 2 + -1v + 0v Combine like terms: 2v + -3v = -1v 22 + -1v = 2 + -1v + 0v Anything times zero is zero. 22 + -1v = 2 + -1v + 0v Reorder the terms: 22 + -1v = 2 + 0 + -1v Combine like terms: 2 + 0 = 2 22 + -1v = 2 + -1v Add 'v' to each side of the equation. 22 + -1v + v = 2 + -1v + v Combine like terms: -1v + v = 0 22 + 0 = 2 + -1v + v 22 = 2 + -1v + v Combine like terms: -1v + v = 0 22 = 2 + 0 22 = 2 Solving 22 = 2 Couldn't find a variable to solve for. This equation is invalid, the left and right sides are not equal, therefore there is no solution. --- end quote ------------------------------------------------------- Heh :) I might have a talent for breaking online algebra tools. -- stefan
From: Kenneth Tilton on 5 Jul 2010 09:12 Stefan Weiss wrote: > On 05/07/10 02:03, Kenneth Tilton wrote: >> Try again: http://teamalgebra.com/ and try this: >> >> Solve >> OK >> 3x-6=12 >> Return >> 3x/3-6/3=12/3 >> Return >> >> (a) wasn't it fun typing that? >> (b) what do you think the software will say given the tortured approach >> to the solution? Correct or incorrect? > > Are you genuinely interested in feedback, or are "yes" and "correct" the > only answers you'll accept? If you think you can't learn anything from > critical feedback, please feel free to ignore the rest of this message. > > (a) Was it fun to type it? > > Not really. There was a quite noticeable lag here (25mbps connection, > central Europe, reasonably fast computer, FF3.0). I don't have the TeX > fonts installed, which is probably also the case for ~99% of your > prospective buyers. You may want to test your application without those > fonts on your system. I doubt it's the fonts. One character per keystroke? Bah. I just tried it and it was molasses. One thing I have not bothered with is tossing data structures as sessions go away. I'll try to rememebr to bounce the thing every so often until I get around to GC. And with the next release I'll buy AWS instances in Europe and Cali and leave them up for a day. > > "Foaming at the mouth" critic that I am, I also dislike having to > relearn how to enter text. Oh, you have used a math editor before that worked differently? Hint. > Most of the keys I usually use for this > purpose don't work in your editor (but still cause HTTP requests), like > End, Home, Delete, Shift+Arrow, etc. Copy/paste are right out of the > question, of course. (a) complaining about the HTTP requests suggests you are coming up to speed on this thread way too slowly (b) Copy/paste have indeed been implemented before, and jsMath would make highlighting the selection area easy, but it's pretty far down on the triage list given that I doubt students learning Algebra will be entering work and wanting cut/paste to do it. I might bring back the "ditto" function that let's you recopy the preceding step when you want to make one small change while still showing all the work. > It's also not possible to open a new tab with > Ctrl-T when the editor has focus. It intercepts the key combination and > prints a "t" character instead. I will continue to dissuade folks from being so picayune, but thanks for the "heads up". And don't blame me: FireFox kept jumping into search mode when I typed a "/", so I stopped all event bubbling. > Entering "3x/3-x" yields "(3x/3)-x", > whereas entering "3x/x-3" gives "3x/(x-3)". Unexpected, but that may > have been explained in the typing tutorial section, which I was unable > to complete for technical reasons (which you don't want to hear about). Yeah, that is problematic. What I had before was a so-called "quick-fractions" option which one would turn on only if one was working only with numeric fractions. Otherwise that feature gets in the way. > > Okay. I guess I can live with that, if it's the price for an interactive > algebra tutor app. But is it *fun* to type like that? Hell no. What math editor do you like better? Hint. > > On to your second question. > > (b) what do I think of the solver's reaction to my input? > > Is its reaction "correct" or not? All the solver had to say was > "Hmmmmm....". I don't think a student would consider that very helpful. Kids are smarter than you know. And I am disappointed you did not notice that the math engine was able to analyze intermediate steps, let alone even accept them. Most other "tutorial" math software makes you do the whole problem and type in only the answer. > Playing some more with this example: > > 3x-6=12 > Return > 3x-7=11 > Return > > The solver's comment for this development is now "Sweet". Yeah, I have to err on the side of /not/ expressing doubt: the last thing we want to do is be discouraging about good work. Otoh, you have a good point: "Sweet" is a bit much. I'll add "mute the approval when borderline ok-but-screwy" to the do-list. I am also often tempted to take out the feature altogether. It's one of those deals that seems like a good idea but probably adds little difference to the educational value, adds a bit of complexity to the app, and even risks doing more harm than good. otoh, students /do/ make this kind of mistake and it's worth heading off if possible. we'll see. > > I tried the "3x/3 ..." example a few more times with the same input, and > it looks like the responses are randomized. Sometimes you get "Whoa" or > "Unusual" or "That's different" instead of "Hmmmmm...". > Here's a suggestion: these responses may look funny to you, but they're > going to annoy people who are struggling with their task. Don't just say > "Hmmmmm...", tell them what's wrong or unusual. They're kids, not > idiots. Yes, I do work with kids, and I know what I'm talking about. You may work with kids, but if you think telling what they did wrong will make a good educational product you do not know much about how anyone (kid or not) learns. The learning comes from the head-scratching, and from figuring things out with the absolute minimum of help. An upcoming feature will be to let kids clone a problem to see how the engine solves it, but we will never help them with their problems. In the Practice Room (coming soon) kids can give up on a problem and see a solution, but in Homework or Mastery mode (where they earn certification) never. > > Just in case I haven't landed in your killfile yet, and you've made it > to this point without throwing me in your imaginary "library haters" > pot, here are a few more observations: No, I just have you in the "persistently negative and unable to notice anything positive" pile, but you were in the Humanity pile anyway. > > You asked Gildas why he was watching HTTP requests. Your application > automatically includes Firebug Lite, which just about screams "debug me" > to all developers (i.e. most of the people reading c.l.javascript and > c.l.lisp). The only way I can stop Firebug Lite from opening and > obscuring most of the visible area is by activating the real Firebug > extension for your site, and then, yes, I will see the HTTP requests. Oh, calm down. I left that in the index.html by mistake. The only requests I see sail by are my server application being chatty. > > Which brings me to another point: with Firebug Lite open, the usable > vertical space in the content area is 70px on my screen, barely enough > to see the huge headline. Gee, maybe when I take firebug light out that will stop happening. <sigh> You People have such a great grip on the important. > Even without FB Lite, the area occupied by the > equation editor is about the same size as the app's header area on my > screen. I'm currently on a laptop with an 800px vertical resolution. I > think it's safe to assume that many of your prospective users will also > be on laptops or even netbooks. There's no need to load a 512x512px clip > art image which serves no purpose at all (it's rendered as 200x200px, > but that doesn't exactly improve things). If the header is only used for > this image and "Whoa" comments, make it smaller. Hmmm, even being fully aware that there was no need to pore over the thing with a magnifying glass, wow, that is what you are doing. Cool! > > The right mouse button is disabled on your site, which is something I > absolutely, positively loathe. It serves no purpose at all except to > annoy your users. Someone needs an anger management refresher. > Anybody who really wants to copy/view anything will be > able to do so anyway. Also a bit paranoid. There was no intent behind that. I might be inadvertently overusing the "prevent default" thing. See above on how FireFox made me do it. > > In case you're interested, the application took about 20 seconds to load > with a blank cache, and 15 seconds after the external assets were > cached. I didn't have any gray screens of death this time, but a > "loading" indicator would be a nice touch. Usually, if a page doesn't > load in 10 seconds, I'll close the tab, but I do make exceptions for > more specialized applications like yours. Yeah, I'll put up a dancing bear animated GIF while it loads. I had one in there but it was loading so fast for me it never actually appeared so I yanked it. > >> btw, anyone mindlessly banging away at the tab key foaming at the mouth >> screaming "dysfunctional!" must not understand much about software: the >> site may be a train wreck, but it presents a one-of-a-kind wysiwyg math >> editor as well as a one-of-a-kind automated tutor able to assess >> intermediate steps in a problem. > > I finally found the message about your application being a work in > progress. You have to double click some text (not a link) reading > "Help!!!!!". Did you expect everybody to find and read that? No. Most of the disclaimers are in the emails announcing the site. The one on the typing tab is more prominent, and the welcome page has it prominently. I just arrange for the newest tab to appear first. > > Your site may be one of a kind. I didn't check. Then I can imagine you with the first Mac in 1984 complaining that there was no hard drive. That's OK, I am here talking about qooxdoo and qooxlisp, but surprisingly none of the JS and Lisp programmers want to talk about that. I suppose this has to do with none of you doing anything but surf Usenet looking for parades to rain on. qooxlisp: http://wiki.github.com/kennytilton/qooxlisp/ qooxdoo: http://qooxdoo.org/ > It may be a valuable > product, and I can certainly see how an interactive approach to algebra > can help students. But you're posting to a JavaScript newsgroup here > (and to c.l.lisp for some reason). Dude, it is a Lisp application. Even the JS gets sent over bit by bit. As for the c.l.js purpose: the subject is not really algebra, it is to show by example a great way to deliver RIAs using at least qooxdoo and at best qooxlisp, a way to program RIAs with a serious language (unlike JS). Now I realize programming the web is not something the loudmouths around here do, but I am sure many lurkers would benefit from discovering qooxdoo. btw, I learned about that splendid JS project on c.l.lisp. I'll go back now and yell at the guy for talking about JS in a Lisp group. <sigh> > If you want to promote your product, > this is not the right place; try alt.algebra or alt.algebra.help > instead. Have the moderator of this group tell me that. Hint. Besides, the software is not ready for alt.algebra. All we really have is a successful melding of Lisp and qooxdoo, so these are the only groups I am spamming. > You posted a self-described "train wreck" prototype to this > group, so what the hell did you expect? If you did it on purpose to > annoy somebody here, that's classic troll behavior. No, the purpose is sharing about qooxdoo and qooxlisp. I only say I am doing it to annoy the bullies to annoy the bullies. The only reason we ended up talking about algebra is because one of the bullies dragged my sig into the flamewar, then another tried to say I was just a troll so I decided to point out what I had produced in the way of real work. Thanks for the time you spent on the feedback. kt -- http://www.stuckonalgebra.com "The best Algebra tutorial program I have seen... in a class by itself." Macworld
From: Stefan Weiss on 5 Jul 2010 10:23 On 05/07/10 15:12, Kenneth Tilton wrote: > Stefan Weiss wrote: >> Even without FB Lite, the area occupied by the >> equation editor is about the same size as the app's header area on my >> screen. I'm currently on a laptop with an 800px vertical resolution. I >> think it's safe to assume that many of your prospective users will also >> be on laptops or even netbooks. There's no need to load a 512x512px clip >> art image which serves no purpose at all (it's rendered as 200x200px, >> but that doesn't exactly improve things). If the header is only used for >> this image and "Whoa" comments, make it smaller. > > Hmmm, even being fully aware that there was no need to pore over the > thing with a magnifying glass, wow, that is what you are doing. Cool! I noticed a problem with your layout and suggested an improvement. If you think the best way to handle this is a sarcastic reply, I hope you'll hire somebody else to do customer support. I won't make the mistake of commenting on your application again. >> The right mouse button is disabled on your site, which is something I >> absolutely, positively loathe. It serves no purpose at all except to >> annoy your users. > > Someone needs an anger management refresher. > >> Anybody who really wants to copy/view anything will be >> able to do so anyway. > > Also a bit paranoid. There was no intent behind that. I might be > inadvertently overusing the "prevent default" thing. See above on how > FireFox made me do it. Then it's an irritating bug instead of an intentional nuisance. The global "prevent default thing" you have going at the moment is not the right way to deal with your problem in Firefox. But that has nothing to do with Qooxdoo or Qooxlisp, so you won't want to hear about it, and I'll save myself the time. > I am here talking about qooxdoo and > qooxlisp, but surprisingly none of the JS and Lisp programmers want to > talk about that. I suppose this has to do with none of you doing > anything but surf Usenet looking for parades to rain on. .... > Now I realize programming the web is not something the loudmouths around > here do, but I am sure many lurkers would benefit from discovering qooxdoo. Remember what I said about antagonizing people? </persistently-negative-unexperienced-loudmouth-out> -- stefan
From: Kenneth Tilton on 5 Jul 2010 11:40
Stefan Weiss wrote: > On 05/07/10 15:12, Kenneth Tilton wrote: >> Stefan Weiss wrote: >>> Even without FB Lite, the area occupied by the >>> equation editor is about the same size as the app's header area on my >>> screen. I'm currently on a laptop with an 800px vertical resolution. I >>> think it's safe to assume that many of your prospective users will also >>> be on laptops or even netbooks. There's no need to load a 512x512px clip >>> art image which serves no purpose at all (it's rendered as 200x200px, >>> but that doesn't exactly improve things). If the header is only used for >>> this image and "Whoa" comments, make it smaller. >> Hmmm, even being fully aware that there was no need to pore over the >> thing with a magnifying glass, wow, that is what you are doing. Cool! > > I noticed a problem with your layout and suggested an improvement. And you persist in trying to talk about layout issues! When the app has now been clearly advertised to be a train wreck in progress!! Good for you. Never say die! > If > you think the best way to handle this is a sarcastic reply, I hope > you'll hire somebody else to do customer support. I won't make the > mistake of commenting on your application again. I supported the desktop version back in the 90s, that was a lot of fun. Folks flipped when they realized they were talking to the author. They were calling for help with a finished product, and every call went well, thanks so much for asking. > >>> The right mouse button is disabled on your site, which is something I >>> absolutely, positively loathe. It serves no purpose at all except to >>> annoy your users. >> Someone needs an anger management refresher. >> >>> Anybody who really wants to copy/view anything will be >>> able to do so anyway. >> Also a bit paranoid. There was no intent behind that. I might be >> inadvertently overusing the "prevent default" thing. See above on how >> FireFox made me do it. > > Then it's an irritating bug instead of an intentional nuisance. What?! A bug in a train wreck?! > The > global "prevent default thing" you have going at the moment is not the > right way to deal with your problem in Firefox. Uh, nothing else defeated the beast. I'll just be more selective about what I do not let bubble. > Remember what I said about antagonizing people? Sorry I spoiled the fun for you. kt -- http://www.stuckonalgebra.com "The best Algebra tutorial program I have seen... in a class by itself." Macworld |