Knuth versus Email (1999) | Hacker News Reader

Knuth versus Email (1999)

cs.stanford.edu

164 pointsjoshuacc128 comments

I had an uncle who was a vice president at a fortune 50 company. The executives at the company were not allowed to have a computer at their desks. They all had secretaries who dealt with email, calendars, spreadsheets, etc. The point was that if they were sitting behind a computer doing time consuming clerical work required to interact with a computer, they were being distracted from much more important tasks. The company was more interested in having people at that level who could delegate and organize than to actually do the work themselves.

The skills and kind of work that makes a person successful at one level often leads to lack of success at another.

Exactly. The C level human I work for types at 10 WPM and their responses looks like Charlies posts from Flowers for Algernon. Not the best use of their time.

They talk at about 2.3 million words per min and a 20 second conversation gets sent back by the exec assistant or one of the reports. Brilliant person, email isn't their forte.

"Is this the most important use of my time" is on their wall. Email isn't it. It works other ways, a voice mail from Hawking isn't the best use of his time.

Charlie at what point in the book?!

While I agree with your premise about different types of work, and while I agree there's some level of clerical work required to interact with a computer, computers can do far more than clerical work, and can provide intellectual leverage when used properly. Banning a high-level executive from using something that gives them intellectual leverage holds the organization back.

You have to think, people - especially senior executives at fortune 50 companies - don't like being arbitrarily held back when they have a job to do. If they want to use a computer, I imagine that these ambitious people would fight the CEO tooth and nail to get access to one. Probably people on the board whom the CEO is politically motivated to please are also banned from using these computers.

People tend not to ban things before they're widely used and cause some sort of problem. I imagine there was a time at his company where his executives used computers. People would have a memory of the time before the computer ban, and compare it every day to the time after the computer ban.

These executives probably had stellar track records, had the motivation to do better, had the previous experience of using computers, had political sway, and still failed to convince the CEO.

And yet, with nothing more than a brief description of the abstract idea, and a firm belief that computers are valuable, a Hacker News commenter has been able to diagnose that it's holding their company back.

That's not to say that CEOs can't make bad decisions, but we don't even know what the company does or what they sell. Probably neither of us have worked there. It seems a little premature to rebel at his decision.

You can delegate the non clerical tasks to your "office of the CXO type staff as well, thereby creating additional leverage.

Read an article last month that was somewhat seriously asking if Trump has ever used a computer: http://gizmodo.com/has-donald-trump-ever-used-a-computer-176...

When was this?

I was reminded of RMS's autoreply:

I am not on vacation, but I am at the end of a long time delay. I am located somewhere on Earth, but as far as responding to email is concerned, I appear to be well outside the solar system.

After your message arrives at gnu.org, I will collect it in my next batch of incoming mail, some time within the following 24 hours. I will spend much of the following day reading that batch of mail and will come across your message at some point. If I write a response immediately, it will go out in the next outgoing batch--typically around 24 hours after I collected your message, but occasionally sooner or later than that. Please expect a minimum delay of between 24 and 48 hours in receiving a response to your mail to me.

If your message is hard to understand or responding takes real work, the response could take longer.

So please wait 48 hours after sending a message before you resend it, remind me about it, or ask if I have received it. If it has been less than 48 hours, the absence of a response from me only means you have not given me time to answer.

If you are having a conversation with me, please keep in mind that each message you receive from me is a response to the mail you sent 24 to 48 hours earlier, and when writing it, I probably had not yet downloaded your later mail.

If you are in big hurry to speak with me, and one day's delay would be a serious problem, you can ask my FSF assistant to phone me. Send mail to <rms-assist@fsf.org> saying what you would like to talk with me about, and giving your telephone number. You can also call the Free Software Foundation office at 617-542-5942 (weekday Boston business hours) and ask them to phone me on your behalf. If it's really important, try both!

An intermediate measure is to email me your phone number and ask me to phone you.

But if there isn't enough hurry to warrant phoning me, please don't bother the FSF people. The mail you already sent me will reach me before any mail they could send me now on your behalf. I will respond as soon as I can.

To contact the Free Software Foundation, use one of the addresses in https://www.fsf.org/about/contact/email.

If you do not wish to receive this message ever again, please send a message to rms-autoreply-control@gnu.org with the subject "OFF". Otherwise, you might receive a reply like this one up to once a month.

Stallman also talks about how he does his computing and how he checks e-mail: https://stallman.org/stallman-computing.html

So he's self-centered enough that he expects everybody to read through all of this essay. Interesting.

Actually, back in the 80s, people normally wrote email of about that length, and since gaming was uncommon, had the leisure to read well-written prose from their peers.

Sometime around 2000, busy people started complaining about this practice, and old timers started to hold their tongues. I knew the era was over when my Millennial co-worker one cube away asked me to not interrupt him and IM him instead, since he had to take off his earphones. About the same time, my boss complained about how long my (senior engineer, explaining matters related to the immanent demise of the company...) emails were -- frankly managers were busy and 1-2 sentence ought to explain about this O-ring problem.

Your projection of self-centeredness on the part of your elders is unbecoming and ill-informed. You are merely young.

rms is just old school.

I am young. RMS has written a very simple email, well laid out and information dense. There is no external referencing and it is easy to skim read.

I find most "short" messages lack key details, requiring me to either search out more information, request a clarification, or abandon the conversation (due to anticipation of the pain continued communication would cause). This is more time and mentally expensive to me than reading a longer more precise message.

To me, this seems to be making two opposite points. One reason to prefer asynchronous interruptions (like IM) to their synchronous counterparts (like out-loud conversation), is so that you have the leisure to deeply read long-form writing (like well-written emails or (hopefully) well-written documentation).

There is no need to insert arbitrary ageism. It just makes you look immature, whatever your age.

I don't see that as self-centred at all. He is politely explaining how email fits into his schedule and offers you alternatives if you need to speak with him urgently. That is a lot better than people who get your email but just ignore it for a day or more until you end up sending a "did you get my email".

Also RMS replies to a lot of the random emails he gets. Much more than a lot of people who are as well known as him to be fair.

Sounds to me like you are just not a big fan of RMS so find a negative in everything he does that isn't quite the norm.

Huh. Am I the only one around here who thinks that it's completely normal to wait a few days for an email reply?

I have an ongoing exchange with a guy that's contributed bug fixes to one of my projects where the average response time (both ways) is on the order of months. It's become a standing joke. I have a mail from him sitting in my inbox right now. It's been there a few months - I'm sure I'll reply to it in a month or three.

Old enough to have done email by way of dialup?

Even big computers (e.g., VAXes running Unix) used to transfer e-mail by way of dialup. In the days before everyone had internet access, some hosts would only have intermittent dial-up connections to other hosts, and they'd transfer e-mail over UUCP. If these machines were in underfunded organizations like university departments, they'd make their calls at night when long-distance calls were cheaper. It could take a day or two just for an e-mail to get across the country, and another day or two for the reply to get back.

For more on this, see: https://en.wikipedia.org/wiki/UUCP#Mail_routing

My first access to the web was via UUCP exchange via a local BBS that exchange mail with their upstream 4 times a day, coupled with the CERN mail<->web gateway...

Seem very similar to Fidonet.

Probably old enough to learn by experience that you have to fight for the control of your own time and attention. If you get people used to quick replies to any small issue, they will later expect it.

I tend to do email once a day in the morning when I get to work. I mostly don't look at it again after that unless there's something particularly critical at issue.

No, he's reasonable enough to explain why and how long it takes to respond to the massive amount of mail he must receive daily.

He also has rather complex rules about inviting him for a talk

https://groups.google.com/a/mysociety.org/forum/#!msg/mysoci...

To people who think like rms, these rules are very helpful, but I think many people will get the impression that he is being very rude by having a long document with his preferences.

Reminds me of the infamous "no brown M&M's" rider...

Those sort of riders are a sentinel to signal that the venue read the contract carefully, not diva behavior, as is commonly assumed. If they caught the brown M&M's thing, they probably got all the stuff about rigging safety, so you don't have to grill them on those things. Otherwise, you'd better double check.

Honestly, I think that such clauses can have opposite effect. I mean: how can you treat the other party seriously if they put such ridiculous, childish demands like 'no brown M&M" in contract?

> ...how can you treat the other party seriously if they put such ridiculous, childish demands like 'no brown M&M" in contract?

If you don't know what you're doing, you won't treat the other party seriously. This is a strong signal to the other party that you don't know what you're doing and SHOULD be avoided.

If -however- you do know what you're doing, then you take the other party seriously even when you come across such demands because (as shawn-furyan said):

> Those sort of riders are a sentinel to signal that the venue read the contract carefully... If they caught the brown M&M's thing, they probably got all the stuff about rigging safety, so you don't have to grill them on those things. Otherwise, you'd better double check.

I'm not sure if this is accurate, but the brown M&M thing started after Van Halen performed at a college gymnasium and their stage setup inflicted tens of thousands of dollars worth of damage to a facility that, if the organizers had read the rider, was never fit to host such a show.

http://ultimateclassicrock.com/david-lee-roth-van-halen-brow...

This applies to RMS -- when he's on the road, he's often at the mercy of the event organizer for transportation and lodging arrangements (either directly or indirectly by them providing information). If they mess up, he's in real trouble.

You do not know why that clause is there. When somebody asks for something and you do not understand why, and decide you don't want to do it, it is the sensible thing to ask for explanation. It is the childish thing to ignore it.

Early in my professional career, I learned something: nobody in the professional world wants to hear what anyone else has to say, unless money is involved (e.g. one of you is the other one's client or customer), and even then they want to hear as little as possible. I had to learn how to professionally communicate with my employers and clients, which means learning to say as little as humanly possible while still mostly getting the point across.

So you optimize for communicating with that sort of person if you need money from them.

Never having met the guy, that doesn't really sound like rms' MO.

His time is important enough, and he is contacted/harassed/solicited by so many people that he expects people who waste it have to do it on their own terms.

If he reaches to YOU, and he still demands that, then that's another story.

You gotta understand, despite what you may think of RMS (and I may share some of your negativity), he is practically WORSHIPPED amongst parts of the *nix community. The man certainly has some condescending opinions about software, but he is not conceited about his own celebrity - he quite certainly would gladly give it up if he could. This is the message of someone who knows he gets more email than he can adequately respond to in a timely basis and still do anything productive at all with his day. He's doing the best he can.

I agree that we can't fault famous people for not being able to answer every email immediately, but

>The man certainly has some condescending opinions about software, but he is not conceited about his own celebrity - he quite certainly would gladly give it up if he could.

Which is why he's happy to let Linus get all the credit?…

It's not like he wants to call it RMS/Linux. Many people have contributed to the GNU project, and one doesn't have to be looking for self-publicity to advocate for its recognition.

It continues: "please do not buy me a parrot..."

top-notch hackers are not ordinary people ;)

I don't really understand why people become so anti-email. I'm a deeply introverted person, so it's not like I don't understand wanting to be left alone, but if I don't want to be interrupted... I close my email viewer. The asynchronous nature of email makes it great for introverted people. I used to get a TON of email at one of my prior jobs, and, like, it was fine. Just wait till it's convenient to open it. If it's actually urgent they'll call or send an IM.

I'm not accusing Knuth of this, but I think some people just use email as a way of humble-bragging. "I'm so important it's overwhelming!" Yeah wow. We're all so impressed.

> I'm a deeply introverted person

That there might be why you haven't learnt to hate Email (or other forms of asynchronous communication). If you want to learn to hate, try running a few mildly successful Open Source projects. Or try participating in other lame company flame email thread.

The fundamental problem with email (and other forms of asynchronous electronic comms including Github pull requests and Facebook messages) is there's little burden on the sender. The entire burden is carried by the recipient. It's too easy for the sender to turn impulse into wasting someone else's time. Snail mail at least costs a stamp and a trip to the letter box.

TLDR: Right now email doesn't scale

Email has been scaling just fine for years. And I've run some reasonably successful open source projects (including SpamAssassin at the height of its success). I really don't see the problem.

The fact that an email is easy to send without knowing the recipient is a huge benefit of email. No other electronic communication system offers this benefit (and unfortunately it enables spam, but that's an entirely different discussion).

The whole point is that it enables spam, broadly interpreted to include unsolicited mail of all kinds.

It really isn't that much of a burden to ignore an email.

Synchronous communication costs just as little, and it's vastly more disruptive. I've done all the things you mention, I guess the fact that I can easily turn off email makes me not care. Sure someone can callously waste my time with an email.. And I can callously delete it. Plus, if I think an email is a waste of time, I can just not respond.

"Email is a wonderful thing for people whose role in life is to be on top of things. But not for me; my role is to be on the bottom of things." Wow!

I think that many of us have jobs in which we have to be both "on top of things" and "on the bottom of things" to varying degrees.

For example, when I'm trying to write a tricky piece of code, it goes much better if I can shut out interruptions like Knuth does and concentrate on the task at hand. If people expected me to reply immediately to their non-urgent e-mails, it would be a lot harder to concentrate.

Also, people who expect me to answer their e-mail or phone call immediately are making the presumption that whatever they have to ask me is more important than whatever I currently happen to be doing. In most cases, it's not.

Previous HN discussion of Knuth's email containing anecdotes of people's experience interacting with him: https://news.ycombinator.com/item?id=2537624

So I'm guessing he's not on Slack?

I know this is just a glib quip - and we ought not be encouraging such things, but it made me laugh.

Hard to tell; Slack doesn't let you know who's online, what they've read etc.

Slack lets you know who's online.

You're right - I had to dig around to find it (on the Android app). It's nowhere near as clear as it is in Facebook Messenger. (And of course there's no `read to here`).

"The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man." --George Bernard Shaw

Wow! That looks like my new favorite quote

There is something that feels disingenuous about taking a conscientious position against email, and then saying that you do actually use email, just by way of a "human spam filter" instead of an automated one.

With that said, I certainly recognize the need for very-deep-focus jobs to be uncoupled from the "right-now" pace expectations of email. So I think Knuth is making a smart decision for the kind of work he does (and one that could be useful to other people in similar lines of work), but he's just presenting/explaining it poorly.

It's sorta like the person who proclaims "I don't have internet access, and I'm better off for it", and then goes on to say that they have a personal assistant whose job is to handle all their online accounting/shopping/research/etc for them. Maybe they personally are better off for their particular workflow, but the really are using the internet, just by proxy, and their implied value statement about themselves compared to people who do use the internet (directly) is not really valid.

He is very specific when he said that the only incoming email he welcome is about bugs and errors in TAOCP. Also he doesn't rant against email, just explained why is not for him and encourage people to use regular mail to communicate with him.

Book "Deep Work" contains some great thoughts on this topic, talking about Knuth too.

I made an app that shows you how often you check your email: http://focuslist.co/escape/ (and other distracting websites)

Looks neat. I tried to download from the MAS, but I use the incompatible 10.10, and I don't want to upgrade my OS. Is there a way to download it regardless?

So you made one more distracting website to check.

I've read this a long time ago and it seems to be still mostly true! A more recent statement of Knuth about email is here: https://www.youtube.com/watch?v=QS8qwMna8_o

Knuth has a point. Although I use email and couldn't work without it, I think we haven't yet realized email's insidious invasion of our world, and its pernicious change of our work habits. Suddenly the Inbox has become a tyrant dictating our daily schedule. Since when do others have priority over our time and attention, expecting immediate replies to their intrusions?

Companies and organizations have not yet grasped the negative impact on productivity that email has wrought. Like drinking from lead goblets in another era, it's going to be a while before we get it. We need email, but it needs to be put in its place.

He was still on email. He just delegated the work to someone else. Sounds like a good arrangement for someone with his level of contribution. I'd rather he didn't spend his limited time reading his junk mail either.

The form ``email'' has been well established in England for several years, so I am amazed to see Americans being overly conservative in this regard. (Of course, ``email'' has been a familiar word in France, Germany, and the Netherlands much longer than in England --- but for an entirely different reason.)"

I sense there's a joke here. Did anyone get it?

It's not a joke -- email (or émail in French) means enamel or glaze in those languages.

Canada too at least the French parts but anywhere here you can buy spray cans of émail in English is enamel.

Before everything was wired to the internet in the late 1980s, UNIX and bboard message apps were as popular as email. You were either there to get the message or it didnt matter. I see mobile message apps enjoy a resurgence in the past decade.

Knuth does read emails and reply them, the only difference from the rest of us is that he may do it on print out of original emails, and usually in batch.

He's old, and he's notable enough that people put up with his eccentricities.

That statement works fine, perhaps even better, without "he's old".

The existing comments are already largely meaningless, so I'll add another meta comment:

Is the use of "versus" in the headline correct in English? There isn't anything adversarial going on here, it is just his email policy.

In other places "versus" even seems to be used as a synonym of "and".

> There isn't anything adversarial going on here

On the contrary, this passage sounds quite hostile:

> If I run across such a message that was misaddressed --- I mean, if the message asks a question instead of reporting an error --- I try not to get angry. I used to just throw all such sheets in the wastebasket. But now I save them for scratch paper, so that I can print test material for The Art of Computer Programming on the blank sides.

Anybody know the year of this? It was posted in 2007 (https://news.ycombinator.com/item?id=81289) and https://news.ycombinator.com/item?id=9997073 has 2003 in the title, but it's likely even older. The delightful Eco quote was from 1995.

When I was around Stanford in the early 80's (second employee at Imagen, a laser printer spin-off from the TeX project), I'm pretty sure Knuth already handled his email via his secretary.

Edit: Oh, right, see the summary elsewhere about his secretary Phyllis Winkler at her retirement.

Another date for you: there's an HTML comment in the <head>

    <!-- Changed by: Ursula N. Owen,  8-Mar-1999 -->

It's possible that it's referring to the HTML skeleton, not the article itself.

Bravo for finding a bit of actual textual evidence. Let's use 1999; it's at least an upper bound.

Actually I heard of him stopping to use email during the late 80's and such

That's consistent with cpr's post. But the article was clearly written later.

After catching dust for 15 years on my bookshelf, I sold the 3-volume edition of The Art of Computer Programming this week through a used book service. No regrets.

Reply on news.ycombinator.com