Help page for structured types shows examples where they add delayed quotes around things. For example

So I started writing  'specfunc(anything,y)' instead of just  specfunc(anything,y). Only to find this breaks things when used inside a proc.

Here is an example.

Removing the quotes, it now works

Download delayed_vs_not_v2.mw

Download delayed_vs_not.mw

It took me 2 hrs debugging to find this, as when I copied code which was working in my worksheet to my function, it stopped working and did not know why.

So for now, I will not use quotes around specfunc or any others like the help page shows in its examples.

Is there a rule of thumb to follow here? The help page is very inconsistent. It shows some examples with quotes around and some without. Makes it hard to know why and when to add the delayed quotes. It seems arbitrary.

I gave up on this. I know how to find all normal derivatives such as diff(y(x),x) and diff(y(x),x$2) so on in expression. 

But now I want to find all those used for initial conditions, which do not have x in them. This makes it harder.

These have the form   D(y)(0) or (D@@2)(y)(1) and so on.

I tried indets and select and many other things, but can't figure how to tell Maple to find these in expression. Here is an example.

Given

expr:=y(3)+y(8)+5*(D@@2)(y)(4)+Pi+3*exp(2)+77*D(y)(0)+1/(D@@4)(y)(7);

I'd like to get list that has in it only the derivatives and nothing else, anywhere they show, which will be

{(D@@2)(y)(4) , D(y)(0) , (D@@4)(y)(7) }

some of the things I tried are

indets(expr,'satisfies'(s->op(0,s)=(D@@n)(y) and n::integer) );
indets(expr,'specfunc(anything, D)');
indets(expr,'specfunc(anything, (D@@anything)(y))');

And many more. I still struggle with structured types in Maple.

In the above, we can assume the dependent variable is always same which is the 'y' symbol.

I can do the above using patmatch. But I am trying to move away from that in Maple and use types.

For reference, this is what I do using another software. I am trying to do the same in Maple, but using structured types and not patmatch.

Iam sure there is a way to do this in Maple using structured types, but so far not able to figure the syntax needed.

Maple 2025.1

I have more examples like this. Sometimes odetest give false negative result. dsolve solution is actually correct, but odetest does not return all zeros as result of its checking.

Example 1

Download odetest_problem_oct_8_2025.mw

Would you agree that odetest in the above should have returned [0,0] instead? Should this result be considered a bug?

Example 2

More examples where odetest gives false negative. The solution by dsolve is correct but odetest says IC is off but it is not. 

Download odetest_problem_oct_11_2025.mw

Example 3

Download odetest_problem_oct_11_2025_V2.mw

Example 4

Download odetest_problem_oct_11_2025_V3.mw

Example 5

Download odetest_problem_oct_11_2025_V4.mw

Will add more examples as I find them....

Any one could find why dsolve can't solve this first order ode when adding IC,  but Student:-ODEs:-ODESteps can?

Download why_dsolve_cant_solve_ode_oct_4_2025_maple_2025_1.mw

Given an ode with IC. When solution is explicit, Maple resolves the constant of integration as expected and returns solution with no c__1 in it.

But when asked for implicit solution, also with same IC, it now returns solution with c__1 still there.

Is this by design or a bug? Should not constant of integration be resolved using IC in both cases? If unable to solve for c__1 because solution is implicit, should it then not return solution all?

Does this happen in earlier versions of Maple?

Download why_C_still_in_solution_maple_2025_1_oct_4_2025.mw