Subject: Re: source access vs dynamism
From: Erik Naggum <>
Date: 1999/08/27
Newsgroups: comp.lang.lisp
Message-ID: <>

* Paul Wallich
| If a hundred people want source enough to ask you, and it takes you five
| minutes to make a decision, that's a day's work.  If 25,000 people want
| source enough to ask you, that's an entire working year.

  and here I thought we were programmers, but instead you argue that people
  should do all kinds of things _manually_?  something wrong, here.

| Obviously you may be able to reject a lot more of the 25,000 out of hand,
| but at some point even reading the email will clobber you.  For any
| project (like an operating system) that has a potentially enormous base
| of interested programmers, personal communication with everyone who wants
| to look at the source code and has a superficially good reason to do so
| is going to clobber the person who acts as a choke point.

  if that person can be trusted to manage something as complex as an
  operating system, I sure hope he's smart enough to realize what a silly
  problem this is before it hits him.  otherwise, who knows what kinds of
  silly things the operating system will do.

  there's a reason companies hire more people when the work-load increases:
  most people who want something done and want to make money doing it have
  figured out that it is beneficial if they can actually train other people
  to do certain tasks and not have to do everything themselves.  given the
  wondrous society in which we live, several people come pre-trained or, lo
  and behold, from other, similar, jobs with a directly useful skill set.

  here's a fairly simple idea: write a program.  publish it.  earn money
  doing this.  support your customers.  include automatic means to get
  upgrades and patches.  include _some_ source, the stuff you'd like people
  to use for innocuous customization and generally to understand your
  program better.  also include a description of what it takes to get more
  or all source, such as printing a file, adorning it with a signature, and
  sending it by ground-to-ground mail.  then do the natural thing in our
  advanced economic society: charge applicants whatever it costs to process
  their application so you have money to employ people doing just that.  or
  write a web thingamajig that deals with the boring administrative stuff.
  it's like the _rage_ among managers and marketing people these days, so
  it's a little odd that programmers don't think about it, isn't it?  ;)

  I think more programmers should have business training or at least some
  _exposure_ to what it takes to start and run a business.  it seems it
  might surprise a great many people, but you don't _have_ to work alone
  and do everything yourself.  basements and garages do _not_ beat a corner
  office and an efficient secretary.  you actually do get a lot more done
  if you hire people who are smarter than yourself at whatever they are
  doing than you would be yourself.  rewarding competence is the best way
  to ensure that the team's competence increases, but it's sadly out of
  vogue in a world of programming where it matters more that people can be
  replaced than that they do outstanding work, because they will leave and
  need to be replaced, and the next guy won't be able to figure it out.

  if, out of 25,000 people who write you with a desire to learn more about
  your software, you don't get 250 "hi, I want to work for you" and manage
  to take proper care of those people, you're doing something _very_ wrong.

  however, I wouldn't hire people who only see problems and refuse to check
  whether the rest of the world perhaps would have to change somewhat if
  you changed one particular factor.  hell, even the free software/open
  source change has a whole lot of ramifications, not all of them equally
  apparent, but I guess I'm used to thinking in terms of cascading changes
  and see that there's no way we can avoid serious scaling problems if a
  lot of people get access to an insurmountable heap of inaccessible source
  as the answer to their _real_ need: software that should fade away into
  oblivion (i.e., not stand out and demand attention) and just do whatever
  it is intended to do, seemlessly and according to how people find most
  beneficial and productive on their own terms.  this kind of software will
  not happen if a whole lot of people value access to source code above all
  and want their mark on software that stands out and demands attention
  like a laser beam right into your eye.  we need to work on something much
  bigger than one person's individual egoboost.  it's _incredibly_ hard to
  do that reliably without forming a loyalty that lasts beyond the feeling
  you get from seeing your name in a ChangeLog entry.  and worse, you don't
  _want_ to work with people who aren't loyal to the goals you have set for
  your project.  if you can't get rid of destructive people, you will have
  very little time available to keep going in the right direction.  this is
  also something you learn PDQ if you try to run a business with employees.

  let me put it this way: I dread the situation where software is written
  by people who are satisfied with name recognition and status among their
  peers -- we'll just get MS-DOS all over again.  granted that we live in a
  culture that adores youth and reveres immaturity as a deity, but if
  everyone who succeeds in any way loses their position to someone younger
  than they were when they were recognized, it isn't just a whole lot of
  disillusioned people we have to deal with: those who aren't wiz kids in
  time won't even have a brilliant flash of youth to look back at.

  I'll do a giant leap to something entirely different: I think a whole lot
  of the issues that plague the world today is based squarely in a rampant
  fear that the world will end _very_ close to 2000-01-01.  Y2K is nothing
  more than fin-de-siècle all over again, as far as the societal response
  is concerned -- technically nothing important will go wrong.  reverence
  for youth is a pretty good sign people don't think they'll get old.  what
  better way to go than when listening to Abba revived by some jail bait?
  I think when the world wakes up with a huge hangover near 2000-01-05 and
  start to realize that the only thing that really ended was the _hope_
  that the world would end and we wouldn't have to take care of things for
  the next 50 to 80 years of our lives, a whole lot of people will start to
  work and value things very differently from what they do now.  when the
  world doesn't end and we aren't plunged back to the dark ages because the
  entire world electricity system didn't fail, after all, I predict that
  all the crap we're doing now with a three-month horizon at most will take
  on much longer horizons, again, like 50 years.  there are some signs that
  some people think like this already: a publisher in Norway has decided to
  revamp their renouned 16-volume encyclopædia of world history and publish
  a special hand-made leather-bound edition in only 2000 copies to those
  who think it's important to maintain excellent craftmanship and some of
  the traditions of the millennium past.  they used to say that nostalgia
  was better in the old days, but I think it'll get better and better in
  the coming years...  but, anyway, let's get this millennium nonsense over
  with so we can get back on track.  we have work to do, damnit.

  save the children: just say NO to sex with pro-lifers