@charlie_fcl Interesting. Thanks.

If I make that my personal initialization file then both worksheets run ok in my Maple 2024.0 for Linux.

Please do not create separate Question threads for this.

Please add followups here, instead.

Please do not create separate Question threads for this.

If you have followup queries on a revised attempt, then please add such in Replies here, instead.

@Bubbleee I have seen fewer reports of corrupted files from Maple 2024.

@Bubbleee Unfortunately, the attachment you've linked consists of only two lines, the first empty and the second containing only the text:

</Worksheet>

The file is only 13 bytes in size. Nothing useful can be recovered from that.

@charlie_fcl

Is there code in the Startup region of the worksheet? Are there any hidden inputs?

If neither, then could you show us the contents of your initialization file (supposing that it doesn't contain secret stuff)?

Your problematic worksheets (saved using Maple 2024) don't appear to contain much more than some stock examples. But none of us can tell all that's really in them without direct access to the acutal .mw worksheet files. Could you provide the actual worksheet files, instead of .pdf exports?

The most helpful way to proceed is to provide code and files, giving the best chance to reproduce and diagnose the problem.

@Ali Hassani 

The type-check against,

    specfunc(identical(x^2), sin)

will be true if it's a function call of sin, and all arguments being of type identical(x^2). That includes the case of no arguments at all.

The type-check against,

    specfunc(sin)

is the same as that against,

   specfunc(anything, sin)

where it will be true if it's a function call of sin and all the arguments are of type anything. Ie, any function call to sin. That includes the case of no arguments at all.

The type identical(...) is a check for an identical match, as the name suggests.

These are all described, close together on the structured types Help-page. There it states:   

@C_R If you're using that plaintext then you'd have to remove the errant `*` instances (which Carl has likely correctly surmised comes from transcription here of faulty 2D Input).

The OP does not provide actual worksheet attachments, which makes helping him unnecessarily awkward. It's not helpful.

@Alexandrakristensen24 I was unable (as yet) to recover that corrupted portion, which contained a faulty image portion.

Your attachment's faulty portion does seem to start at line 274 of the XML. But there are only 286 lines in the file. It's quite possible that your attachment doesn't contain much more data, and that the (nor properly saved) original was actually much larger.

@C_R 

As I'd already mentioned in my original Answer: the custom mechanism is faulty in M2024.0.

I've already reported that functionality regression.

@C_R Regardless of which command you might extend, you'll need a conditional to test for type `=`.

Otherwise you'll have weird errors when applying rhs,lhs accidentally to type algebraic, Matrix, piecewise, etc.

What version are using? Did you see the part of my Answer that explained that the `simplify/ext` mechanism doesn't work properly in Maple 2023.2 and Maple 2024.0?

The general mechanism worked in 2022.2.

Which are you using here?

@C_R It may be due to the special characters in the file name. The OP could try renaming it.

@C_R Experiment and the quest for understanding are great motivations. I was curious, is all. Thanks.

I'd have to study those other examples later.

@mmcdara 

I gave the Compile variant because I didn't know whether you wanted to repeat this operation on many Matrices or column choices.

If anything else has been compiled then part of the overhead is lessened. Also, you might be able to get a faster compilation without inmem=false, but that may depend on version and OS.

In later Maple versions there are also some variants using andseq (possibly with evalhf), which also allows for a quick bailout.