Read and understand code faster with programming ligatures in Fira Code font

I had every one of those concerns too. In spades. And then I actually tried using it for real, and none of the concerns turned out to cause any problem in practice.

(Why would I try using something I was confident would make me less productive? Because I felt was in a rut, and was trying out disruptive changes to my workflow. For example the same day, I switched all my calculator tools to RPN. That's not gone badly either.)

>In fact , our visual system has been trained to instantly recognize a!=b in code

Citation needed.

>confusing = with == will lead to lots of errors etc

This is more about NOT confusing == with = -- and several similar characters than the opposite.

Sounds like you have thought of all kinds of arguments to not actually try it and see whether it's any good in practice -- which is how anything should be judged (unless it involves some huge risk). In other words, a priori rationalization.

>If I get used to this font on my machine, i will have a real problem when parsing code in another terminal/editor.

The above argument advices to forever stick to a lowest common denominator coding toolset, lest one has to ever try to work in other environment.

How about maximizing the performance one gets from their own, and most frequently used, environment, instead of "what-ifing" for foreign systems they might have to use 1/10 of the time or less?

Plus, having a tool at our disposal and getting used to it doesn't mean we can't also do without if needed. Especially if it's a small improvement, like the one proposed by the ligatures here, and not some totally different notion that we can't ever go back to how it was without it...

If you give it a try, you may find your brain more adaptable than you think. Mine seems able to parse the same code in my editor (with this font) and in web-based repository viewers (with some other monospaced font) without problem.

>I don't want != changed to ≠ because ≠ isn't a C++ operator.

A colored != is not a C++ operator either. It's about getting our minds to differentiate symbols faster, not about whether they exist as is in some language's syntax.

Yes, it's a gimmick. The question whether it's a useful gimmick is an open one. I'd try it out probably, but probably it would end more confusing than helpful after the novelty wears off.

It is no more a gimmick than the standard mode of rendering. However, it does have an adaptation cost. A comparable example: I went to the trouble of creating a custom key binding for my old laptop that was a standard US layout, except the numeral row was symbols by default (i.e. Press shift to get numbers), because I only used the numpad for numerals. It worked, but relearning was too painful. If I'd learned it that way from the beginning, it would just be normal.

Doesn't have to be the limiting factor to still be an improvement.

It takes some time to get used to and it's annoying to place the cursor correctly to backspace from != to =.

But it looks sexy and confuses the co-workers XD. Joke aside: I actually like making my PC mine and like the look. I don't think it makes me a faster or better programmer in any way...

There are lots of others - my personal favourite is Monoid. http://larsenwork.com/monoid/

I don't even think they add much to readability. If i could choose, the only ligatures i 'd like would be >= and !=

In that case, I would refer you to Monoid font. I love that one. You can even select from options for e.g. ampersand, dollar, zero...

At least with using ligatures in IntelliJ IDEA, I find that it makes zero difference as to how the monospaced font is aligned.

Usually the ligatures in these fonts are designed to fill the same width as the original combination of separate characters in monospace use. That does preserve the alignment of code to either side of the ligature, but unfortunately at the expense of either making some of the symbols much wider than normal or adding a lot of extra whitespace to pad them out. For programming, the ligatured version seems quite risky if it means the only visual difference between single and double equals or underscore characters is a bit of width and/or horizontal padding.

What I would like to try is a language where = actually meant equality rather than assignment and a few other common operators like !=, ===, !==, <= and >= all appeared using the standard mathematical symbols at the same width as =, < and >. That way, some related cases would naturally line up and most lines with comparisons would become shorter and less cluttered, which seems like a modest incremental improvement in language syntax. Alas, you’d also need software and keyboard input to support that in a way that made those alternative symbols effortless and entirely predicable to use. This feels like a job for a standardised programmer’s keyboard and AltGr, except that you’d need a standardised programmer’s keyboard for each of the different languages that already uses AltGr for other purposes, and at some point it all becomes more trouble than it’s worth.

it doesn't unless you use a different font for strings. But it seems that would be fine in your example

I don't think Fira Code recognizes context. It just parses the characters and renders multi-character glyphs into a single one.

So the inequality inside the string will be afforded the same treatment as an inequality anywhere else in the code.

Most of Lua was nicely obvious and intuitive, but ~= was always one of the weird sharp edges. I could maybe see the argument for using it if ~ was already a negation operator of some sort (like C's bitwise complement) but Lua doesn't even have that. Instead you get something needlessly non-intuitive and subject to misinterpretation by anyone who isn't a Lua native, as ~= as an ASCIIfication of ≃ is pretty common.