I enjoyed it.

GPU is an interesting topic. Is there a way to use it from Maple?

As far as I understand the current situation, with code in Maple language being 500-1000 slower than, say, in C, it doesn't have much sense to use parallel programming for Maple code other than for some worksheet effects - such as displaying the time spent while a procedure is running, or returning some partial results while a procedure is continuing - other than those that could be done using print. It would be interesting to see an example of that.

Alec

?implicitplot3d gives a link to implicitplot. It seems as if adding digits shouldn't create any problems.

Alec

I agree.

Also, ln should be in roman, not italics, and the fractions should have something like 6*lambda^2 in the denominator instead of being displayed as sqrt(2)/6/lambda^2.

Alec

In practise (not in the CS), programmers often call the necessity of declaring types of variables (like int, double etc. in C) being static typing, and using variables without type declaration (as in Python) dynamic.

In this sense, entering proc(a) is a dynamic typing and entering proc(a::integer[4]) is a static typing example - and it could be actually static if the procedure is (auto or not) compiled.

Stronger vs. weaker types usually mean that the strong types describe exact placing the variable in memory, and a weak type may be a group of strong(er) types. For example, rtable is weaker than Array or Matrix, and Array with C_order and datatype=float[8] is stronger than just an Array.

Alec

Java is Java, it can't be improved.

With the main page in French - no, I had it opened (accidentally) by clicking on the http://www.maplesoft.com in the scroll bar menu in the location field at the top of the browser, instead of clicking on the link to this site that I meant. I'm pretty sure that there were no previously set cookie to the french version. I don't recall the previous page though, from which I tried to go to this site. It could be one of sites located in France (but in English, I'm pretty sure about that.)

That didn't last long - the next day, after repeating the same procedure, just being curious whether it was still in French, it was in normal (well, a kind of) English.

Alec

Will,

Thank you! Yes, I meant to use <maple>, just accidentally entered <math> instead. I used it rather rarely earlier because its behavior was unpredictable, and even if it worked normally first time, it could display something weird after a simple editing of other part of the post etc. Also, if I used prewiev to see how it looks before posting, it was broken after posting, with </maple> tag moving right after <maple> tag and leaving the expression entered outside.

Alec

Will,

Thank you! Yes, I meant to use <maple>, just accidentally entered <math> instead. I used it rather rarely earlier because its behavior was unpredictable, and even if it worked normally first time, it could display something weird after a simple editing of other part of the post etc. Also, if I used prewiev to see how it looks before posting, it was broken after posting, with </maple> tag moving right after <maple> tag and leaving the expression entered outside.

Alec

I posted that after trying to see the worksheet here. At that time it showed the usual page that MapleNet can not display worksheet, or something like that (I, probably, can find it in the cache if the proof is necessary.) It has been opened normally now.

Alec

Yes, for recursive procedures using the remember table, that's different. In this particular example, with autocompile, the remember table doesn't seem to be used. Also, the procedure compiles during its first execution - that also reflects on the timing, and can be prevented by executing that procedure before timing,

XXX(1);

                                 100.

time(seq(XXX(1000000+i),i=1..100));

                                0.546

time(seq(XXX(1000000+i),i=1..100));

                                0.530

op(4,op(XXX));

Still, too long for adding just 10^8 numbers (as anybody writing at least one program in assembler would say). 3Ghz processor speed means 3*10^9 clocks per second, and addition (not packed), is just 1 clock, so adding 10^8 numbers (especially constant), properly done, should take 1/30=0.03 sec instead of 0.53. With 2 processors (2+Ghz each) that should be even faster, and with packing it should be twice as fast. With a GPU used for that instead of a CPU, that would be much-much faster. But this is C (and Watcom) problems, not Maple.

By the way, why does Maple distribute Watcom 1.3 (and not even complete)? It is open source and the latest version is 1.8. It would have sense to upgrade at least includes to make them compatible with official C specification (complex datatype etc.). 

Alec

At this moment, positive changes are noticeable.

Several Maplesoft developers replied in threads related to their work.

Thank you, Stephanie.

Alec

PS Does your picture mean that you have 4 personalities? Just curious -Alec

If crashing was the goal, it wouldn't take me more than 5-10 minutes to crash it (as it usually happens with a new Maple version.) I have destroyed one thread just accidentally, without even trying.

I guess your team has a lot to learn (in the crashing direction.) I wouldn't call that a positive sign :(

And being familiar with other your team products (such as Oracles site, and the previous version of this site), make such negative feelings even stronger :(

What am I talking about actually? - when I went a few days ago (accidentally) to the main Maplesoft site, it was all in French (at least not Japanese, and I speak French, but still) - what sign is this?

Replacing Tim with Stephanie was definitely a good decision though. So there is some hope.

Alec

MapleNet is still not working (not that I care, but still.)

Alec

How to make a link to the ?sum,details or ?sum/details help page?

If I leave it like that, that gives a link to ?sum. If I enter the link directly, either in the source editor, or using the link button, it is ignored and still only a link to ?sum is provided.

Alec

Now, there is no collobarative editing yet (which is good), but I also can't edit my posts after replies (which is bad). No sights of Flash. So these things are postponed to the next update?

Alec

OK, this actually makes sense. It is just not what is usually called "solve a system".

I didn't check the answer, but the following might work,

sol:=[solve(identity(sys[2],t))];

  sol := [{B = 0, a = 0}, {B = 0, a = 1/2}, {B = 1, a = 0},

        {B = -1/2, a = 1/2}]

sol1:=[seq(map2(`union`,i,[solve(identity(eval(sys[1],i),t),{g})])[],i=sol)];

  sol1 := [{B = 0, a = 0, g = 0}, {B = 0, a = 0, g = 1},

        {B = 0, a = 1/2, g = 0}, {B = 0, a = 1/2, g = -1/2},

        {B = 1, a = 0, g = 0}, {B = 1, a = 0, g = 1},

        {B = -1/2, a = 1/2, g = 0}, {B = -1/2, a = 1/2, g = -1/2}]

sol2:=[seq(map2(`union`,i,[solve(identity(eval(sys[3],i),t),{b,k,w})])[],i=sol1)];

  sol2 := [{B = 0, a = 0, b = b, g = 0, k = 0, w = w},

        {B = 0, a = 0, b = b, g = 1, k = 0, w = w},

                                            1/2
                                           3
        {B = 0, a = 1/2, b = b, g = 0, k = ----, w = 1/3},
                                            2

                                              1/2
                                             3
        {B = 0, a = 1/2, b = b, g = 0, k = - ----, w = 1/3},
                                              2

                                               1/2
                                              6
        {B = 0, a = 1/2, b = b, g = -1/2, k = ----, w = 1/3},
                                               2

                                                 1/2
                                                6
        {B = 0, a = 1/2, b = b, g = -1/2, k = - ----, w = 1/3},
                                                 2

        {B = 1, a = 0, b = b, g = 0, k = 0, w = w},

                                                1/2
        {B = 1, a = 0, b = b, g = 1, k = 1/2 I 2   , w = w},

                                                 1/2
        {B = 1, a = 0, b = b, g = 1, k = -1/2 I 2   , w = w},

        {B = -1/2, a = 1/2, b = 1, g = 0, k = 0, w = w}, {B = -1/2,

                          2                             2
                       4 k                           4 k
        a = 1/2, b = - ---- + 1, g = 0, k = k, w = - ---- + 2/3}]
                        3                             9

Now, is this the analytical solution or numerical?

Alec