Hacker News Reader

Comment by JadeNB | Hacker News Reader

I like the author's careful descriptions of what is good about the document, but it seems like common-sense pragmatism. Surely there are other, equally pragmatic and sensible, documents? (I admit that I don't have any examples, but I am not often exposed to good or bad procedures documents.) Why does this one deserve particularly to be singled out?

EDIT: As dragonwriter (https://news.ycombinator.com/item?id=9682195) correctly points out, it would be wrong to demand that everything written about be the unique best object in its class, and I do not mean to do so. I mean instead that the author literally singles out this document as "one of the finest written processes of all time", which seems to be more than the analysis demonstrates.

Also, because I am that guy, this:

> there's thousands of permeations of things that can go wrong

should be 'permutations' (or, more likely, probably 'combinations'), not 'permeations'.

View on HN

Surely there are other, equally pragmatic and sensible, documents?

I was reminded of the Four Rules for Firearms Handling. I think you're selling this sort of thing short, however. It's not as if some junior officer just sat down one afternoon and wrote these rules. Rather, they are the end result of a long process of trial, error, disaster, and rewriting.

> I think you're selling this sort of thing short, however.

I don't mean to do so!—although I see how "common-sense pragmatism" can sound derisive. I agree that it is much harder than it might seem for a document not to get in its own way; sort of the procedural version of Twain's "I would have written a shorter letter if I had had more time." I didn't in any way mean to question the value of this document, only its unicity.

Sebastian Marshall writes in a fairly hyperbolic style. If I were to apply a de-hyperbolizing filter, I think the post title would be something more like "this document is a pretty good example of how processes can evolve and become perfected over time."

Not sure if anyone is reading this thread any more, but...

> If I were to apply a de-hyperbolizing filter, I think the post title would be something more like "this document is a pretty good example of how processes can evolve and become perfected over time."

Actually, I wasn't using hyperbole at all: I've studied a lot of operations documents, and this is (literally) my favorite one of all time. It operates on many levels, effectively changes normal human behavior for a number of different actors with very different motivations and communication styles, and it's critically important.

It's also very well debugged. It's really a very fine document, again, my #1 in operations that I've read. (#2 is probably Dalio's Principles at Bridgewater.) If anyone has anything equally good as a candidate though, shoot me an email -- I love studying, corresponding about, and discussing this stuff.

> Surely there are other, equally pragmatic and sensible, documents? (I admit that I don't have any examples, but I am not often exposed to good or bad procedures documents.) Why does this one deserve particularly to be singled out?

Because its good enough to serve as an example, and the author was familiar with it? If there are many equally good examples, and there is value in an example, then there doesn't need to be anything above the other alternatives (and, if they are equally good, there won't be, by definition) to select a particular one -- doing so at random, or by aesthetic preference unrelated to the function is fine.

I didn't mean "why is it worth writing about?", though I see how my comment could come across this way. I meant rather that the introductory sentence:

> "General Orders for Sentries" is one of the finest written processes of all-time.

seems to indicate an almost uniquely special document, whereas what I saw was a well and carefully written document that nonetheless seems not to be unique (though I have no examples to prove my contention).

It is permutations if the order of events is relevant to the outcome. Theoretically, every single step you take may look "correct", but still end up badly if performed in disorder.

I agree, and did not mean to suggest that 'permutations' was absolutely the wrong word, but I think that it is not really what the author meant.

To be sure, I agree that the correct outcome is likely to result only from one (or only from a few) of the possible permutations of the correct steps, and "doing the right thing at the wrong time" might as well be "doing the wrong thing"; but, I think, the same is not true of things going wrong (which is of what the author was speaking), where it is unlikely that permuting them will cause a good outcome.

> but, I think, the same is not true of things going wrong (which is of what the author was speaking), where it is unlikely that permuting them will cause a good outcome.

No, but there are still different permutation of things that can go wrong, and different permutations (not just combinations) present different challenges, even if none of the permutations converts "gone wrong" into "good outcome".

"there's"?

Reply on news.ycombinator.com