Let's steal some issues from https://github.com/davidedc/Algebrite! >:3

[Happypig375]

Tests for our future integral implementation

• 13: 1/x^3*1/sqrt(x^2-4)
• 15: arccos(x)
• 16: arctanh, arcsinh, arccosh [Requires hyperbolic functions and their inverses]
• 17: 1/x [Requires abs function and flag for signifying real variables]
• 18: x^2/(a+b*x)
• 19: x*sqrt((-9)-x^2)
• 22: t^(-1/2)*exp(-t^(-2))

Tests for our future simplification

• 35: incorrect simplification of: sqrt(x*y) , sqrt(1/x) , sqrt(x^y)
• 108: Error in simplify() with division and powers: 1 / (( y ^ 2 ) ^ ( 1 / 2 ) )

New functions

• 20: summation [Should include product as well]
• 33: How to compare two numbers? [Requires boolean logic]
• 80: Is it possible to find the domain of a function over the reals?
• 87: Solving equations with modular arithmetic

Bugs

• 29: incorrect ordering of function calls
• 99: Simplify call significantly slower for "x^(-6)+1" than for "x^(-6)" [AngouriMath literally hangs for this]

Tests to add

• 38: simplify (x^2-xy)/(xy-y^2)

[WhiteBlackGoose]

incorrect ordering of function calls

Huh?

[Rafael-SOWNet]

Triage, 2026-08-11. Recommended for closure, and labelled Opinions wanted rather than closed because declining a proposal is a maintainer's call.

"Steal issues from Algebrite" is not actionable as written — it names no behaviour to add and no test that would pass. Anything specific taken from it would be a better issue than this one.

This is one of thirteen dormant proposals swept in the same pass. If you disagree with any of them the label is the place to say so; if you agree, closing is the action. Nothing here is a defect report, so nothing is lost by closing except the record of the idea, which this comment preserves.

[Happypig375]

At least make sure each of the subissues in the original issue are done with reference to their original issues at the original repo

[Rafael-SOWNet]

At least make sure each of the subissues in the original issue are done with reference to their original issues at the original repo

Done. Measured on master at a45a7256, .NET 10, 20-second cut-off per case.

Integrals

Algebrite #	integrand	today
13	1/x^3 * 1/sqrt(x^2-4)	timeout > 20 s
15	arccos(x)	done — x*arccos(x) - sqrt(1 - x^2) + C
16	inverse hyperbolics	partly — see below
17	1/x	done, including the part the item asks for
18	x^2/(a + b*x)	timeout > 20 s
19	x*sqrt(-9 - x^2)	done — C - (-9 - x^2)^(3/2)/3
22	t^(-1/2) * exp(-t^(-2))	timeout > 20 s

Item 17 is worth calling out, because its bracketed requirement was "requires abs function and flag for signifying real variables" — that flag now exists and is consulted:

"1/x".Integrate("x")                                      // C + ln(x)
using var _ = MathS.Settings.Codomain.Set(Domain.Real);
"1/x".Integrate("x")                                      // C + ln(abs(x))

That is #946, merged as #952. So item 17 is complete as originally specified, not merely "it integrates".

Item 16. The functions exist, but under area names rather than arc ones — arsinh, artanh, arcosh, not arcsinh/arctanh/arccosh, which is deliberate (they are areas, not arcs). They parse into their logarithmic forms. Integration is uneven: artanh(x) integrates, arsinh(x) and arcosh(x) come back unevaluated, and tanh(x) times out. So the item is half done and the remaining half is a real gap.

Simplification

Item 35 (sqrt(x*y), sqrt(1/x), sqrt(x^y)): all three are left unsimplified. The Algebrite issue was about incorrect simplification of these — splitting sqrt(x*y) into sqrt(x)*sqrt(y) is wrong off the positive reals — and declining is the correct answer rather than a missing feature. I would call this resolved by not doing it, and it is worth saying so explicitly since a future contributor may otherwise "fix" it.

Item 108 and the others already ticked check out.

New functions

Item 20 (summation, product): no syntax for either — sum(i, i, 1, 10) and product(...) are parse errors. Not done.

Summary

3 of 7 integrals done, 1 partly, 3 time out; item 17 fully complete including its side condition; item 35 correctly declined; item 20 outstanding. So the checklist is roughly half retired, and what remains is three integration gaps, the inverse-hyperbolic integrals, and summation.

If you would rather track those as their own issues than keep this one open as a shopping list, the three timeouts are the concrete ones and I can file them with these measurements.

[Happypig375]

You can split this into individual issues as you see fit