The curse of the gifted programmer (ESR's email to Torvalds)
lwn.net
lwn.net
1) Ignores the content of the submission 2) Ignores the other discussion comments 3) Unsupported personal insult 4) Upvoted to 'best comment' position.
Look, pretend it's about someone else - Ray Monderic or someone - doesn't the idea that early formalising your development procedures pays benefits for other people and eases the transition from a system small enough that the leader can manage it to a system big enough that they can't have some merit?
Your parent is an unsupported personal insult, you're not wrong about that, but if it discourages others from advertising esr essays it's legitimate and meaningful.
It's bizarre.
Linus Torvalds is a gifted programmer. Eric Raymond is an insufferable blowhard. From the very first sentence of his mail, he's making it apparent that he is part of the senior cadre that makes the Internet works (to borrow a turn of phrase) and that he is condescending to explain to Linus how this software development stuff really works.
I think my comment does a rather fair job of cutting to the quick of the issue, but I respect your take on it as well.
Your unsupported assertion about Linus Torvalds being a gifted programmer is one I have no problem with since I think "Linux kernel, git, lots of people respect him, yes I can accept he's a gifted programmer".
That is, I don't question it because I already have some basis for agreeing with it and none for disagreeing. If you claimed he was rubbish, I'd ask for supporting evidence.
But when you say ESR is useless, I can't name projects he's worked on like git, but that doesn't mean they don't exist - for all I know he might have worked on something I've never heard of, or something big I don't associate with his name.
So when you throw your dislike for ESR around, that cuts to the quick of the issue for you, because you already know why you dislike him. "Oh don't read it, it's ESR" is enough for you. To cut to the quick for me, you would need to say something like:
"ESR worked on Projects X and Y which both failed, his results aren't as good as his talk, beware".
I hope all the people who upvoted you did so because they know why they dislike ESR and agree with you. I fear that the people who upvoted you did so because they know other people they respect dislike ESR and want to demonstrate their group membership credentials by supporting the popular view.
I don't hate ESR (as shouted in my earlier post) and I also don't like his patronising village elder tone, but I do like and can fit into experiences in my life the idea that running things from one person's head is the business equivalent of a code smell, and that the pain for them of changing it soon will result in smoother operations for everyone else sooner and smoother operations for them too later.
> for instance, note the timestamps on the comments
OK.
While the email was undoubtedly pretentious, I've got to say that ESR brings to light some really important points... at some point, systems that "mere mortals" use to keep track of a large code base become a necessity even for people like Linus.
I don't like how comments like these (even if they're from a user who normally has insightful/useful comments, albiet generally snarky) get voted so high up. It's time like these I wish I could vote down a response.
EDIT: Interesting to note for future readers: my comment was voted down at least once within 5 seconds (or less) of posting this.
Comments without content are timeless, they don't suddenly gain usefulness through age. Are you suggesting that if only someone had let your comment reach maturity, they would have been obligated to upvote it?
> As Linux grows, there will come a time when your raw talent is not enough.
is the key bit.
...and I think it was proven true several years back, when Linus found it increasing hard to keep on top of dealing with the inflow of bug fixes, patches, features, etc.
Raw talent is never enough for the really big problems. Proper use of routines and procedures are a brain amplifying technology.
At the time the email was written, Linus was 30.
> Linus is quite patronizing in practically all of his e-mails
Maybe so, but -- for me at least -- Linus has earned more of a right to be patronising than ESR has.
Linus earned his stripes; 30, 20 or 10 years old, the man delivered a world-class piece of software.
Patronizing him is just pulling non-existent rank so ESR doesn't feel like a fish out of water. Remember, while Linus oversaw the hottest piece of Unix software to come in decades, Eric Raymond was making a living off of "seniority" and "lore".
Eric Raymond is the spokesperson of an "in-crowd" whose membership he didn't earn.
I don't see what your point is. I clearly didn't mean that he was literally a child.
His point, BTW, was valid since well before Linux was born. Or Linus, or ESR.
I am curious whether splitting drivers early is a symptom of this.
I've personally seen too much code sharing as a more symptom of hubris than too little code sharing.
why not the hundreds of mailing list archive where you can read the prequel, or the followup?
Have a look yourself: http://lkml.indiana.edu/hypermail/linux/kernel/0008.2/0240.h...
The more interesting "follow-up" is historical, when kernel development started using BitKeeper and then git, instead of Linus' email inbox.
Linus can come off as smart but an asshole in person. But despite this, it seems like his smart extends somewhat into smart insights about human nature.
Linus has at least two to rub together.
Why do you quote ESR as though he's some demi-god? He's a freak who takes credit for other people's work without so much as a citation.
Also, show us why ESR doesn't have a clue.
Here is Linus' letter on which ESR responded: http://lkml.indiana.edu/hypermail/linux/kernel/0008.2/0201.h...
And here are Linus' points, in short: http://lkml.org/lkml/2000/8/22/12