Worst ideas/features in programming languages?

Ola Fosheim Grøstad ola.fosheim.grostad at gmail.com
Mon Oct 25 23:34:50 UTC 2021


On Monday, 25 October 2021 at 22:24:40 UTC, Ola Fosheim Grøstad 
wrote:
> On Monday, 25 October 2021 at 22:05:06 UTC, ClapTrap wrote:
>> Well probably they had their own notation on actual pen and 
>> paper. But the fact that when coming to type setting they 
>> preferred an upside down 'E' over writing a word makes it look 
>> like there is something appealing in writing the expressions 
>> symbolically that you just don't seem to grasp.
>
> Why personal attacks? If you read old math writings you will 
> see that they did actually write out math with plain words. 
> Compact short forms came later and mathematicians still don't 
> agree on one common syntax. So there is indeed a personal 
> aspect to writing mathematics. This become obvious if you read 
> proofs, there are many different styles. Clearly strong 
> personal preferences at display.

So, to bring this discussion to an end: I am not saying that 
people should not use symbols and formalisms.  I am saying that 
using "&&" over "and" is not giving a usability advantage, and I 
believe that can be both argued and in theory measured (although 
difficult).

If you read mathematical proofs you'll see that proofs that only 
use formalisms can be very difficult to follow for a human being 
(although easy to verify for a computer). So written proofs tend 
to use words where emphasis and understanding is important.

When language designers talk about their languages they often 
emphasis aesthetic dimensions. So language design is just as much 
art as it is engineering. Which ties to culture, identity and in 
essence philosophy. Mathematicians talk about beauty in relation 
to proofs, programmers talk about beauty in relation to coding. 
Again, influenced by culture and identity.

In essence this is a good thing. If language design was all about 
engineering then we wouldn't have so many options to choose from! 
:-D





More information about the Digitalmars-d mailing list