The Old New Thing

2 min read Original article ↗

Latest posts

Windows stack limit checking retrospective: MIPS

Mar 13, 2026

Post comments count 1

Post likes count 2

Windows stack limit checking retrospective: MIPS

Raymond Chen

Raymond Chen

Optimizing out the unnecessary probes comes with its own complexity.

Windows stack limit checking retrospective: x86-32, also known as i386

Mar 12, 2026

Post comments count 8

Post likes count 2

Windows stack limit checking retrospective: x86-32, also known as i386

Raymond Chen

Raymond Chen

One of the weirdest calling conventions you'll see.

How do compilers ensure that large stack allocations do not skip over the guard page?

Mar 11, 2026

Post comments count 6

Post likes count 5

How do compilers ensure that large stack allocations do not skip over the guard page?

Raymond Chen

Raymond Chen

Don't take steps that are too large.

A snappy answer when asked about dressing casually at IBM

Mar 10, 2026

Post comments count 3

Post likes count 3

A snappy answer when asked about dressing casually at IBM

Raymond Chen

Raymond Chen

Oh, this old thing?

The fine print giveth and the bold print taketh away: The countdown timer

Mar 9, 2026

Post comments count 8

Post likes count 1

The fine print giveth and the bold print taketh away: The countdown timer

Raymond Chen

Raymond Chen

Think fast, no pressure.

Learning to read C++ compiler errors: Ambiguous overloaded operator

Mar 9, 2026

Post comments count 1

Post likes count 1

Learning to read C++ compiler errors: Ambiguous overloaded operator

Raymond Chen

Raymond Chen

Look for the conflicting definitions to see where they are coming from.

When <CODE>Read­Directory­ChangesW</CODE> reports that a deletion occurred, how can I learn more about the deleted thing?

Mar 6, 2026

Post comments count 7

Post likes count 2

When Read­Directory­ChangesW reports that a deletion occurred, how can I learn more about the deleted thing?

Raymond Chen

Raymond Chen

It's already gone. If you need more information, you should have been remembering it.

The mystery of the posted message that was dispatched before reaching the main message loop

Mar 5, 2026

Post comments count 2

Post likes count 2

The mystery of the posted message that was dispatched before reaching the main message loop

Raymond Chen

Raymond Chen

Perhaps it's because you dispatched it.

Aha, I found a counterexample to the documentation that says that <CODE>Query­Performance­Counter</CODE> never fails

Mar 4, 2026

Post comments count 6

Post likes count 2

Aha, I found a counterexample to the documentation that says that Query­Performance­Counter never fails

Raymond Chen

Raymond Chen

Of course, anything can happen if you break the rules.

Load more posts