zlacker

> With jupyter and spyder I have been bitten by lingering obsolete state from re-running cells or code blocks various time. I find the program much easier to debug if it runs on a clean slate and builds all the state from scratch.

Hm, reading this just made my brain connect the reasons I don't like Jupyter with the reason I am baffled by people who like Smalltalk. An image that you can't fully restart just... doesn't seem like a pleasant development environment.

>>blep_+(OP)
Exactly. How are you supposed to code review something that lives in RAM?

>>blep_+(OP)
Jupyter would really benefit from having the output and state of notebooks stored to separate files. That way you could easily version the code itself in git, while putting the output and state file types in .gitignore.

>>codefl+D
By committing the code to a git repo and having a code review like every other language out there.

I'm guessing you have never tried these things but image based Smalltalk implementations have supported VCS for decades now, literally. In Pharo this is with git using Iceberg:

https://github.com/pharo-vcs/iceberg

They even wrote a tutorial to make it easier: https://github.com/pharo-vcs/iceberg/wiki/Tutorial

It's not magic, it's not even a problem, because the problem you're imagining doesn't actually exist. So long as the user of the system has at least half a brain (and maybe less) they will be capable of distributing their code with git these days.

>>Jtsumm+sN
I’m not imagining anything, I’m arguing from logic. If the code is distributed via Git, then you have a normal programming language (with a particularly cool IDE/REPL). Which isn’t in any way a bad thing, but I hope you will agree that this is not how Smalltalk is usually advocated. You can’t have it both ways.

>>codefl+r71
> I hope you will agree that this is not how Smalltalk is usually advocated.

I will not agree, and I hope you will understand why after I give some evidence. Here's what Pharo advocates typically point people towards:

https://mooc.pharo.org/#week1 - I've linked to the week1 anchor so you can see that Iceberg makes it in pretty early. I hope you understand that despite your hopes for me I cannot agree. Sorry to crush your hopes, I hope you can forgive me.

>>Jtsumm+9a1
> I hope you understand that despite your hopes for me I cannot agree. Sorry to crush your hopes, I hope you can forgive me.

Honestly, I don't have the slightest idea what's going on with you.

>>codefl+8l1
> Honestly, I don't have the slightest idea what's going on with you.

I was having fun with the word "hope" which I find to be a weird sentiment because it expresses a lack of agency or control over a situation. But you didn't lack agency or control, you had an opportunity to persuade. An opportunity to convince me of your position and to try and bring me into agreement, but you didn't use that opportunity. It would have made more sense to me to express a hope for agreement if you had at least taken the chance to provide evidence for your claim.

>> I hope you will agree that this is not how Smalltalk is usually advocated.

You offered no evidence, you just hoped (asserted a lack of control in persuading) that I would agree with a bald assertion. I offered evidence to the contrary, the Pharo MOOC, which turns up in most discussions I've seen, here and elsewhere, about Pharo in particular and Smalltalk in general. Interestingly, to me, you didn't even engage with my evidence, which is unsatisfying.

You could have responded to it. You could have provided your own counter-evidence. But you didn't.