Dreaming the future of software; Project Editor ECMAScript 2015; Reformed Smalltalker JavaScript historian dl.acm.org/doi/abs/10.114… @allenwb@mastodon.social
Sherwood Oregon USA
- Tweets 8,520
- Following 218
- Followers 6,374
- Likes 436
Allen Wirfs-Brock retweeted
Also, if you are representing a company on TC39, when performance review time comes arounds its easier to score points for having advanced a new feature then for slowing things down.
Particularly true for younger companies that don’t have a lot of experience with standards.
1
2
6
Standards committees like TC39 are intentional bottlenecks. That’s a good thing.
Do you agree JS has some strange quirks arising from its 10-day origin sprint? Now imagine what future JS would be like if every “good idea” that came along was immediately added to the language.
7
27
143
TIL: After 25 years of being involved with writing programming language standards I just discovered that there is an ISO meta-standard on writing programming language standards.
However, I don’t think I’m going to pay $158 to read it.
iso.org/standard/37765.html?…
8
22
141
The brittleness of chained data accesses was one of the original motivators of “Responsibility Driven Design”
en.wikipedia.org/wiki/Respon…
Valuing code reusability is mostly about making sure the code you write today will continue to work tomorrow.
cc/@rebeccawb
Hot take: Optional chaining encourages writing brittle code.
Example (currently a stage 1 JS feature):
const data = myObj?.firstProp?.secondProp?.actualData
Instead, normalize object structures at the time of creation (in JS apps, that often means in API call response handler)
1
2
9
I'll take:
Number.isFinite(x) ?
x.toFixed(2) : 'N/A'
Has fp fad destroyed all appreciation of readability?
5
2
24
Allen Wirfs-Brock retweeted
In my experience, consistancy is highly rated by individual language designers and given lip service by committees. But once a committee process is involved, actual agreement on consistancy issues and trade-offs is rare and consistancy loses to compromise.
1
2
Allen Wirfs-Brock retweeted
Stanford uploads 111 lectures by Donald Knuth. There go your next five years of spare time…
youtube.com/playlist?list=PL…
4
642
1,540
Allen Wirfs-Brock retweeted
True experts would not have been at Netscape in that crazy year, or else would have noped out. This is true based on all I know and learned the hard way!
3
13
You might think differently if you read through the papers from the 3 HOPL conferences. The next HOPL is in June 2020. Consider attending.
en.m.wikipedia.org/wiki/Hist…
1
8
8
Allen Wirfs-Brock retweeted
I don't share my opinions because I want to remind others how important I think I am, or in the vain delusion that it will create some wider movement.
I share my opinions because I'm hoping that one or two or a half dozen folks will say, "Huh, I hadn't considered that before."
4
2
37
"Shipping an imperfect project is always better than not shipping at all" 😧
Useless the imperfection kills people, or wastes millions of dollars, or destroys your business, or ...
Correctly deciding whether to slip, ship, or abandon is hard and always situational.
It's nice that 'thought leaders' share the best way to do stuff and if you have the time and space to practice this: by all means do it.
But don't let their opinion prevent you from building stuff.
Shipping an imperfect project is always better than not shipping at all.
5
15
68
Great Talk by Richard Gabriel about this Inkwell system.
How RPG combined old-school symbolic AI and ML to turn a misinformation bot into a provocative poet.
youtu.be/6Vax07bD_Lw
4
21
Great respect for @waltmossberg but I don’t think he “coined” that term. I first blogged using it in Jan. 2011 & pretty sure I wasn’t the first. It’s typical for ideas and terms to circulate in tech circles many years before reaching popular press.
wirfs-brock.com/allen/posts/…
Amazon wants to master 'ambient computing,' a term coined by my pal @waltmossberg theverge.com/2019/1/4/181685… via @Verge
7
Allen Wirfs-Brock retweeted
Replacement of a 20 year old system doesn’t mean that system was unsuccessful. It just means that time has marched on and tech and requirements have changed.
1
1
3
The referenced article is great. A masterful introduction to how modern JavaScript is dynamically optimized and how trivial optimizer bugs can enable security exploits. (Warning: you probably need to know at least a little bit about optimizing compilers)
🙈 It feels somewhat funny to start your new year reading a detailed description of how to turn a seemingly trivial bug of yours truly into a serious security exploit in @v8js.
h/t @anbiondo
abiondo.me/2019/01/02/exploi…
2
10
It is in reference to the Aug 1978 Byte cover which shows Smalltalk in an ivory tower on an isolated island. The balloon on the Aug 1981 Byte cover announces Smalltalk's escape from that ivory tower.
archive.org/details/byte-mag…
I mean ... what the actual hell was the rainbow hot-air balloon about?
That's not "open in Corel DRAW", is it?
I can't sleep 'till I know!
2
6
The last time I saw any numbers there were high hundreds (perhaps low thousands) of still deployed mission critical legacy enterprise Smalltalk apps. With perhaps a handful of new ones each year (mostly at shops with other legacy ST apps).
Such apps have long lives. 1/
AFAIK OOCL is still driven by Gemstone smalltalk, no? And then there is JP Morgan Chase that build their risk management with smalltalk. And more. I implanted smalltalk in 2018 to the biggest german rail company. It is not over ✊🏻
1
5
Many of you are probably aware that for awhile Smalltalk was a serious IT programming language before Java and the web emerged. This 1995 article by Dave Thomas is a good summary from the high point of the Smalltalk era:
web.archive.org/web/20130612…
8
59
137