[layer8]

The full stop at the end of your last sentence isn’t strictly needed. It isn’t even strictly needed in the preceding sentences (except for the first one with the question mark), because the capitalization already indicates the beginning of the next sentence. We still use full stops because redundancy and consistency help preventing errors and ambiguities. Reducing the tolerances to zero increases the risk of mistakes and misinterpretations.

Ideally, adding/removing/changing a single character in valid source code would always render the code invalid instead of “silently” changing its meaning.

[lisper]

> The full stop at the end of your last sentence isn’t strictly needed.

yes, that's true | but this redundancy is not necessary | it's there for historical reasons | there are other ways to designate the separation between sentences | some of those alternatives might even make more sense than the usual convention

[layer8]

The point was that the full stop is currently a redundant element in most cases, yet we would not want to omit it just for the reason of being redundant.

The spaces in your “ | ” punctuation are also not strictly needed, yet one would want to keep then for readability and for risk of otherwise mistaking an “|” for an “l” or similar.

Again, something not being strictly needed isn’t a sufficient argument to do without it. There are trade-offs.

[lisper]

> The spaces in your “ | ” punctuation are also not strictly needed.

yes.that.is.true|spaces.are.not.strictly.needed.at.all|there.are.alternatives.and.there.are.situations.where.using.those.alternatives.actually.makes.sense|however.the.use.of.whitespace.is.so.deeply.ingrained.that.if.you.dont.do.it.the.rendering.of.your.text.will.generally.be.very.annoying.on.contemporary.systems

The Right Answer is to separate the underlying representation from the rendering. We already do this to some extent in modern systems. For example, the meaning of text generally doesn't change if you change the font. This is not always true. The font can matter in math, for example. And some font modifications can carry semantic information -- using italics can provide emphasis, for example.

The Right Way to design a programming language is to have an underlying unambiguous non-redundant canonical representation, and then multiple possible renderings that can be tailored to different requirements. Again, we kinda-sorta do that in modern systems with, for example, syntax coloring. But that's just a half-assed hack layered on top of deeply broken language designs.

[vacuity]

Considering all the "tabs or spaces" flamewars and standardized formatting as with gofmt for Go code, I think this would get restricted at most professional codebases to some person's favored style. Not sure that's a good reason, but it's worth considering. For projects that are solo or along those lines, feel free.

[vacuity]

You're being disingenuous. Your suggestion is more like if you wrote

yes, that's true but this redundancy is not necessary it's there for historical reasons...

without any breaks. That might be exaggerating compared to your actual position, but surely you can see that "unnecessary in this situation" doesn't imply "unnecessary overall". "Not necessary" if we're cherrypicking, sure.

If my program now has no semicolons and then I write something else that behaves differently than expected, I'm going to be sad. My mental model for programming fares better when semicolons are used, so I will favor writing programs with semicolons. To me, the cost is trivial and the benefit, while minimal, outweights the cost. I consider it separate from actual boilerplate. You can disagree and use other languages, but then we're probably being moreso opinionated than divided into better or worse camps.

[lisper]

> That might be exaggerating compared to your actual position

To the point of being a straw man.

There was actually a time when neither white space nor punctuation was used andallwordswerejustruntogetherlikethis. Note that it's still possible to decipher that text, it just takes a bit more effort. Natural language is inherently redundant to a certain extent. It's mathematically impossible to remove all redundancy (that would be tantamount to achieving optimal compression, which is uncomputable).

The spaces around the vertical bars in my example were redundant because they always appeared before and after. That is a sort of trivial redundancy and yes, you can remove it without loss of information. It just makes the typography look a little less aesthetically appealing (IMHO). But having something to indicate the boundaries between words and sentences has actual value and reduces cognitive load.

---

[1] https://en.wikipedia.org/wiki/Kolmogorov_complexity#Uncomput...

[vacuity]

I think you forgot the analogy. Why is it bad to have semicolons in programs then?

> You become keenly aware of the fact that 90% of your mental effort is going not towards actually solving the problem at hand, but appeasing the compiler or conforming to some stupid syntax rule that exists for no reason other than that someone at some time in the dim and distant past thought it might be a good idea, and were almost certainly wrong.

You said this originally. I definitely agree for something like parentheses in if conditions in Java, but I think semicolons are a great example of how

> having something to indicate the boundaries between words and sentences has actual value and reduces cognitive load.

[lisper]

> Why is it bad to have semicolons in programs then?

It's not bad to have them, it's bad to require them when they aren't necessary. It's bad to make their absence be a fatal syntax error when they aren't necessary. (Some times they are necessary, but that's the exception in contemporary languages.)

Also, I know I'm the one who brought them up, but semicolons are just one small example of a much bigger and more widespread problem. It's a mistake to fixate on semicolons.