Return, interrupts and error handling

Mon, 01 Jan 0001 00:00:00 +0000

A sequence of commands (a code block) can be interrupted at any point by special interrupt statements. TrustLang uses two kinds of interrupts, which in meaning correspond to their notation:

“positive” interrupt ++ — return of a value/through-flow (analog of return);
“negative” interrupt -- — error (analog of throw).

An interrupt can return a value: ++ 'String' ++ — return a string, -- err -- — “throw” an error (a rough analog of throw err). If no value is specified, the interrupt returns nothing (an empty value).

Reflections on the Turing machine and the causes of errors in programming languages

Sat, 15 Nov 2025 00:00:00 +0000

A couple of years ago I wrote the article Reflections on structured programming, in which I tried to sort out the misconception that Edsger Dijkstra proved that any algorithm can be built from just three constructs (sequence, selection, loop). Whereas in reality:

This theorem was proved (with conditions and restrictions) by the Italian scientists Böhm and Jacopini, and Dijkstra merely referred to it.
The three mentioned basic algorithmic constructs are insufficient for modern programming (for example, exception handling is required).
The famous criticism of the goto statement by Dijkstra (as well as of the assignment statement, which is always forgotten to mention) was not a strict ban, but merely a recommendation for improving code style, so that programs would be easier to analyze.

And now the time has come to write about some problems of the Turing machine — the fundamental basis of all information technology.

Errors on TrustLang

Return, interrupts and error handling

Reflections on the Turing machine and the causes of errors in programming languages