2019-10-15
859
#typescript#what's new
Paul Cowan
8357
Oct 15, 2019 ⋅ 3 min read

New in TypeScript 3.7

Paul Cowan Contract software developer.

Recent posts:

Does splitting work across AI agents actually save time? I tested it.

Within roughly the same six-month window, Anthropic shipped Agent Teams for Claude Code, OpenAI published Swarm and the production-ready Agents […]

Ikeh Akinyemi
Mar 13, 2026 ⋅ 6 min read
ai dev tool power rankings

AI dev tool power rankings & comparison [March 2026]

Compare the top AI development tools and models of March 2026. View updated rankings, feature breakdowns, and find the best fit for you.

Chizaram Ken
Mar 12, 2026 ⋅ 10 min read
the replay march 11

The Replay (3/11/26): Eng knowledge gaps, OpenClaw, and more

Discover what’s new in The Replay, LogRocket’s newsletter for dev and engineering leaders, in the March 11th issue.

Matt MacCormack
Mar 11, 2026 ⋅ 26 sec read
ai training alexandra spalato

Your engineering team’s AI training is probably failing: How to fix it

Buying AI tools isn’t enough. Engineering teams need AI literacy programs to unlock real productivity gains and avoid uneven adoption.

Alexandra Spalato
Mar 11, 2026 ⋅ 4 min read
View all posts

8 Replies to "New in TypeScript 3.7"

  1. So typescript is going to have pretty much every feature C# has. (what a coincidence, they’re designed by the same person)

  2. The optional chaining operator just tells me that I am going to show up at client sites where every single dot in the application is preceded by a question mark, the same way that async permeates through the code.

    It tells me that nobody is actually going to know anything about the data that runs through the system, and it’s going to hide bugs that are ridiculous to solve, because they could have happened at any `?.` upstream from where your data was supposed to be.
    Moreover, this affects all downstream systems, as well.

    I’m not saying that 10,000 null checks is good, either. What I am saying is that we are programming backwards. What I am saying is that we should have systems that are hermetically sealed, and operate only on types that are known, not types where every member is optional/nullable. If you can’t build an instance of a type with the pieces you have been given (plus any reasonable defaults), then you should take a different path in code, rather than passing in a tree that is garnished with nulls.

    The invention/inclusion of null in programming languages was Tony Hoare’s billion dollar mistake, and instead of thinking about how we can program without it, we are making it easier to make everything null, at all times, without crashing.

    However, the null coalescing operator is a good tool for building those known and trusted data types from untrusted data sources (server responses, database fields, localStorage, user input), *if* the field you are settling has a valid default/empty/identity state. Optional chaining would be fine to use in the case where you were grabbing an expected value from an unknown data source, and using the coalescing operator to set the default, when building a closed system of well-defined types… but we all know that isn’t how it’s going to be used.

  3. Personally I think the worst thing about their announcement blog post is the dubious formatting of ternary operators. I mean, everybody knows that you’re supposed to format them
    “`
    var result = condition
    ? first_expression()
    : second_expression()
    “`

Leave a Reply

Your email address will not be published. Required fields are marked *

Hey there, want to help make our blog better?

Join LogRocket’s Content Advisory Board. You’ll help inform the type of content we create and get access to exclusive meetups, social accreditation, and swag.

Sign up now