Allen Wirfs-Brock

Blog Post: All I want is a Personal Digital Habitat wirfs-brock.com/allen/posts/10…

@aboodman @etrepum You're right, it about cleaning up nested generator activations. See #sec-iterator-interface" target="_blank" rel="nofollow noopener">tc39.es/ecma262/#sec-i… particularly Note 2. The `throw ` method is called by the `yield*` operator. See the yield* semantics at #sec-generator-function-definitions-runtime-semantics-evaluation" target="_blank" rel="nofollow noopener">tc39.es/ecma262/#sec-g…

No, I mean the `throw` method of the `Iterator` protocol: #throwexception" target="_blank" rel="nofollow noopener">developer.mozilla.org/en-US/docs/Web…. I get that the idea is to throw an exception that a catch statement around yield will receive. But the runtime doesn't ever call this method, so it seems like it is only useful if you can guarantee your callers will manually iterate your `Iterator`. Which seems like it makes it pretty non-useful since everything else encourages you so strongly to use generators with for-of.

Said it before and will again. I think TC39 has done a remarkably good job evolving JS. So many of the additions over the last ~15 years feel like so nice – like natural obvious additions to the language that just fit in with everything else, without adding new dimensions of complexity. The way promises + await + `try/catch` snap together is an old example I have always loved. But the way `Iterable`, `Iterator`, `Generator` and `for of` all fit together is another really pretty one. So is `class` + `prototype`. I know there was prior art to look at for a lot of this, but I still think that fitting this all together with the world's largest ecosystem of existing code and developers must have been quite a challenge. Props and gratitude @TC39.

@aboodman If you aren't already familiar with "JavaScript: The first 20 years" you will probably enjoy it. It only touches upon the design history of iterators, but their story is pretty similar to other ES6 features that are covered in greater detail. dl.acm.org/doi/abs/10.114…

@jonathoda @andy_matuschak Not necessarily true for corporate data processing (EDP). Some places overnight processing kept the machines very busy.

@andy_matuschak We worked late at night when you could get a turnaround in minutes

When the compiler’s slow, or the design tool’s too cumbersome, the joy gets sucked out of the work for me. Feels like death. But: man, in the punchcard / mainframe days, people had day-long feedback loops and couldn’t tear themselves away. What gives? Novelty wearing off?

@beffjezos Sorry, can't talk now.

We're so back. ☢️

The Onward! Essays 2024 submission deadline of April 25 is rapidly approaching. If you have an essay relating to software, programming, or programming languages which has not yet been formally published please consider submitting it. #Call-for-Essays" target="_blank" rel="nofollow noopener">2024.splashcon.org/track/splash-2…

Call for Onward! Essays Submissions due 25 Apr 2024 #Call-for-Essays" target="_blank" rel="nofollow noopener">2024.splashcon.org/track/splash-2… Onward! is more radical, more visionary and more open than other conferences. Onward! Essays invites software developers&academics to submit essays on the state & possible futures of software development

To: the crazy ones, the misfits, the rebels, the troublemakers, the ones who see things differently Subject: you should write an Onward! paper alarmingdevelopment.org/?p=1708

@bradleymeck @dillon_mulroy dl.acm.org/doi/pdf/10.114…

@dillon_mulroy @awbjs has a fantastic book on stuff like this

I found out today that we almost had types in JavaScript in 1999 AND JSX (in the form of E4X). Have you ever wondered why we went from ES3 to ES5? Check out this article on the skipped ecmascript spec that could have changed the web two decades ago. evertpot.com/ecmascript-4-t…

@HernanWilkinson Here's the thing, everything we deal with in CS and software development is an abstraction.  Nothing is “real”, the concepts we use to design and create software are all in our heads.  Programmer are masters at creating useful things out of “smoke and mirrors”.

So, how is right? Jim Highsmith (screenshot from Wild west to Agile), Bob Taylor (youtube link), neither or both?? Same joke, different ways to tell it, same idea: The software does not weight😂😂😂 youtu.be/Y0MsrrTo8jY?t=…

Another post in my series on architecture as a metaphor for software development. Inspired by staring at the ceiling in Centre Pompidou after @ncraftsConf tomasp.net/blog/2023/pomp…

@BrendanEich @TypeScriptBob @aaronfrost The full (189 page) HOPL JavaScript history paper is at dl.acm.org/doi/abs/10.114… NOT paywalled!

@TypeScriptBob @AkiraMakes brendaneich.com/2017/12/my-dot… web.archive.org/web/2020060608… (cc: @aaronfrost in case he can ping Charles of whoever might restore the transcript archived thanks to the Wayback but now lost from the served version). Also of course twitter.com/awbjs/status/1… by @awbjs & me from HOPL-4 (2020).

wow, found an official guide from Apple on how to structure your GUI app to support chat commands using LLMs with function calling 🤯 imagine saying to chatgpt "set that chart to be a 3d pie" and that command can get routed to a Part of your Compound Document

@getifyX @FITC Remind me of an invited talk I gave at the OOPSLA conference in 1996. wirfs-brock.com/allen/talks/oo…

People have been asking for my slides from "Imperative vs Declarative: Weathering the storm"... here they are: speakerdeck.com/getify/imperat… #slides #webunleashed @FITC

@msimoni @gasnerpants yup, see x.com/awbjs/status/1…

@awbjs @gasnerpants So long since I used Smalltalk. Do they have resumable exceptions, too?

everyone arguing about checked exceptions (without knowing that's what they're called) without understanding the ergonomic issues

@msimoni @gasnerpants Don't forget the designers of what ultimately became Smalltalk's standard exception handling system.

@gasnerpants My whole shtick about resumable exceptions is in fact anti-continuations. Common Lisp demonstrates that you don't need first-class continuations to resume after an error. But I'm afraid, other than the writers of the CL standard, I'm one of a handful of people who understand that

Ecma TC53 workshop in Cupertino, on 6 Dec 2023 "...participants will discuss proposals and consider options for embedded JavaScript standardization including new APIs, interoperability, packaging, and Hardened JavaScript (aka Secure ECMAScript)" ecma-international.org/news/invitatio…

@codexeditor @atlantis__labs Tek 401x terminals were based upon Direct View Storage Tube CRT technology. The image is actually "stored" on the phosphorus at the front of the CRT. No refresh was required. The bright "cursor" is the electron beam that is "writing" on the phosphorus.

@atlantis__labs It's absolutely key to the charm ... but the limitations of monochromatic display and resolution also contribute some creative constraints to the use of space and 𝘭𝘪𝘨𝘯𝘦 𝘤𝘭𝘢𝘪𝘳𝘦.

9 minutes of the greatest Dark Mode graphics ever devised ... youtu.be/IztxeoHhoyM?si…

@ljharb @AdamRackis I vaguely recall that meeting. My sense is that there was too my divergent views toake any progress. One of my points was that cjs modules had always been syntactic extensions to ES3&5.

@awbjs @AdamRackis we spent hours in one plenary and the room was convinced there was no way for node to do it and remain compliant. If we were wrong it would have been great to figure that out in 2017 :-p

Wait this was possible all along? What the fuck.

@ljharb @AdamRackis Clause 17 says "Except as restricted in 17.1, a host or implementation may extend Script syntax, Module syntax,...". Noting in 17.1 restricts adding import statement syntax to Script. Impl would have to define its semantics. A justifiably good idea ???

@AdamRackis lol i mean, if the person who edited/wrote the spec doesn’t like what it says, but we can’t change it because browsers already shipped it, the only option left is just to do what it says