undefined | Better HN

0 pointslolive2y ago0 comments

Went to his website. There was an interview of him from 2022. The answer that really stroke me was not about Vim, but about software craftmanship vs professional programming:

« I have been working for a company where quite a few managers, educated in physics and mechanics, thought the software was just the same as what they knew and they could decide how to make it. That company went downhill and was eventually taken over. The same happens in places where decision-makers can get away with failure, such as in government. The people writing the code probably just make sure they get paid and then run away from the crime scene. On the other end of the scale are people who want to write beautiful code, spend lots of time on it, and don’t care if it actually does what it was intended to do or what the budget was. Somewhere in between, there is a balance. »

I am not so sure about the last sentence. But the rest is SO true!

0 comments

bennyschmidt2y ago

> people who want to write beautiful code, spend lots of time on it, and don’t care if it actually does what it was intended to do

There was a funny React framework a while back that has all kinds of cool state management and this and that, but it renders nothing. For the developers who spend more time theorizing, migrating, refactoring, than shipping.

It's funny when people go back and forth about the time complexity of a click handler (that is debounced anyway), and they arrive at the optimal solution, but when you use the app the entire thing is super clunky lol where is the concern for time complexity of me using the product xD

asveikau2y ago

I feel it's much more common to create abstractions without regard for performance cost of the abstraction. So keeping time complexity down, even for a piece that doesn't matter much, is above average for this sort of thing.

Gibbon12y ago

Also creating abstractions without considering the mental overhead of reasoning about what the code is doing.

Not enterprise: event --> action.

Enterprise: event --> abstraction --> abstraction --> abstraction --> abstraction --> abstraction --> abstraction --> abstraction --> abstraction --> action_parta --> abstraction --> abstraction --> abstraction --> action_partb abstraction --> abstraction --> abstraction --> action_finalizer.

1 more reply

hock_ads_ad_hoc2y ago

Creating abstractions without considering performance is probably the correct approach. It’s much better to at least start with a program with enough abstraction to allow reasoning about how things should work. Save performance concerns for when optimization is seen to be necessary.

3 more replies

taneq2y ago

> people go back and forth about the time complexity of a click handler (that is debounced anyway)

This is the modern software world in a nutshell.

twism2y ago

Re-frame?

gemstones2y ago

This was immediately what I thought of, too.

tonyedgecombe2y ago

I'm pretty sure that company was Océ, I worked there at the same time Bram was there (although I don't think I ever met him).

They were a photocopier company that struggled with technology in general, not just software. I suspect the root of their problem was that they had been taken over by sales and marketing people. That's never good for a technology company.

cxr2y ago

What's wrong with the last sentence?

For ease of consumption, the context again:

> On the other end of the scale are people who want to write beautiful code, spend lots of time on it, and don’t care if it actually does what it was intended to do or what the budget was. Somewhere in between, there is a balance.

sillysaurusx2y ago

It’s remarkably uncommon to find an employee who strikes a balance between the two. But I think this doesn’t make the sentence wrong.

myst2y ago

Literally everyone strikes the balance. According to their experience. The older, the more experienced, the more balanced.

2 more replies

loliveOP2y ago

After reading all the comments, I realise I misinterpretated the last two sentences. I thought Bram was concluding that you have to accept the same kind of trade-off you accept in the frame of professional software development, when you deliver an app you are really dedicated to.

Obviously, Bram, or Linus, or Stallman, and many other devs I don’t even know, are not there for delivering half-baked pieces of art. And that’s eventually a GOOD thing. [the counterpart being that the dedication level is MUCH beyond the one in professional environments]

creer2y ago

I'll go with: the sentence takes this as a dimension, this at one end, that at the other. While perhaps what matters is on an entirely different dimension. From a point of view of usefulness to its users, is there a strong correlation with whether the project was "thrown over the fence" or is a work of software artistry?

WalterBright2y ago

I understand wanting to write beautiful code, it's a major motivator for me. But it does have to actually do something useful - or there's no point.

The_Colonel2y ago

The code itself, its elegance and cleverness can be a point in itself. I'd argue it's a stupid point (with a few exceptions ala CCC), but for some people it does seem to be the main driver.

aerique2y ago

But that's the whole point, isn't it?

1letterunixname2y ago

I've worked with enough "professional" programmers to know I don't want to work with them and their sloppy, reckless, uncurious, and inconsiderate ways.

Exoristos2y ago

So what do you do for a living?

1letterunixname2y ago

Well from your low karma, I know you're a troll.

1 more reply

ugh1232y ago

The last sentence was the point. Either end of the spectrum are people who are not caring about the right things (typically about themselves, egos, etc).

j / k navigate · click thread line to collapse

0 comments

bennyschmidt2y ago

> people who want to write beautiful code, spend lots of time on it, and don’t care if it actually does what it was intended to do

asveikau2y ago

Gibbon12y ago

Also creating abstractions without considering the mental overhead of reasoning about what the code is doing.

Not enterprise: event --> action.

1 more reply

hock_ads_ad_hoc2y ago

3 more replies

taneq2y ago

> people go back and forth about the time complexity of a click handler (that is debounced anyway)

This is the modern software world in a nutshell.

twism2y ago

Re-frame?

gemstones2y ago

This was immediately what I thought of, too.

tonyedgecombe2y ago

I'm pretty sure that company was Océ, I worked there at the same time Bram was there (although I don't think I ever met him).

cxr2y ago

What's wrong with the last sentence?

For ease of consumption, the context again:

sillysaurusx2y ago

It’s remarkably uncommon to find an employee who strikes a balance between the two. But I think this doesn’t make the sentence wrong.

myst2y ago

Literally everyone strikes the balance. According to their experience. The older, the more experienced, the more balanced.

2 more replies

loliveOP2y ago

creer2y ago

WalterBright2y ago

I understand wanting to write beautiful code, it's a major motivator for me. But it does have to actually do something useful - or there's no point.

The_Colonel2y ago

The code itself, its elegance and cleverness can be a point in itself. I'd argue it's a stupid point (with a few exceptions ala CCC), but for some people it does seem to be the main driver.

aerique2y ago

But that's the whole point, isn't it?

1letterunixname2y ago

I've worked with enough "professional" programmers to know I don't want to work with them and their sloppy, reckless, uncurious, and inconsiderate ways.

Exoristos2y ago

So what do you do for a living?

1letterunixname2y ago

Well from your low karma, I know you're a troll.

1 more reply

ugh1232y ago

The last sentence was the point. Either end of the spectrum are people who are not caring about the right things (typically about themselves, egos, etc).

j / k navigate · click thread line to collapse