News for hackers. Stuff that matters.
Re:Smalltalk's Browser: Unbeatable, yet Not Enough by brazzy on Thursday March 05, 2026 @07:22AM (#47260794)
Do Smalltalk IDEs really not have the concept of different "views" of the code? The 4-pane hiearchical view is clearly valuable, but why would it need to be "surpassed" rather than complemented by other views that are available when needed and can be switched to or even shown alongside the traditional view?

If that kind of thing doesn't exist (and the article sure sounds like that), then yes, it appears the smalltalk ecosystem really has fallen decades behind the state of the art in the IDE area.

Re:Smalltalk's Browser: Unbeatable, yet Not Enough by projektfu on Thursday March 05, 2026 @09:50AM (#47262176)
They have many views, the trouble is the proliferation of these views as you go through the code. You browse to a class and method, then you want to see the senders, and so you open that, and then see the implementors of a method a sender calls, and so on.

https://blog.lorenzano.eu/content/images/2026/01/Pasted-imag...

Re:Smalltalk's Browser: Unbeatable, yet Not Enough by brazzy on Friday March 06, 2026 @02:04AM (#47271880)
Then it sounds less like a specific problem with the 4-pane browser, or SMalltalk IDEs, or Smalltalk, and instead simply an overcomplicated way to state the obvious: large codebases are complex and difficult to navigate. No shit.
Re:Smalltalk's Browser: Unbeatable, yet Not Enough by projektfu on Friday March 06, 2026 @04:31PM (#47281346)
Smalltalk makes it a little harder than some languages because a method has no real idea who will be receiving the messages it sends. People have done work on implementing types (Strongtalk) to, among other things, limit the search space. But you're right that a lot of moving parts means that eventually you have to have a lot of open context to view it statically.

Another way is to trace things in the debugger, and then you get the dynamic picture.

Re:Smalltalk's Browser: Unbeatable, yet Not Enough by neilv on Thursday March 05, 2026 @11:29AM (#47263691)
Historically, Smalltalk has many browsers (views). This System Browser is one of many browsers, and the most busy-looking.

You can browse within it, and also spawn off other kinds of browsers from it.

And these browsers are extensible with others. As someone new to Smalltalk, I was pretty easily able to add a visual class hierarchy browser into this environment:

https://www.neilvandyke.org/smalltalk-chg/

Half the things we know or think about in HCI, the people at PARC figured out before we were born, and sometimes before the hardware to test it existed.

https://worrydream.com/EarlyHistoryOfSmalltalk/

Re:Smalltalk's Browser: Unbeatable, yet Not Enough by pjmlp on Saturday March 07, 2026 @06:40AM (#47286723)
Including how to build a full workstation with memory safe systems languages, besides Smalltalk,

Interlisp-D, now recovered and playable in the browser,

https://interlisp.org/

Mesa with XDE (think the safety Zig is selling today, but in 1985)

http://bitsavers.informatik.uni-stuttgart.de/pdf//xerox/xde/...

Cedar, a full graphics workstation in the evolution of Mesa, now with proper mix of reference counting + cycle collector, as full GC implementation,

Here the demo done by Eric Bier for the Computer History Museum,

https://www.youtube.com/watch?v=z_dt7NG38V4

Re:Smalltalk's Browser: Unbeatable, yet Not Enough by Frolo on Thursday March 05, 2026 @01:01PM (#47264962)
smalltalk invented the concept of views