🚀 go-pugleaf

In article <1991Jul6.090850.18499@cs.sfu.ca> miron@cs.sfu.ca (Miron Cuperman) writes:
>Lets try some numbers.  Let us say that a company can develop software
>at a cost of $200K and sell it to 10 customers for $40K each.  Let
>us also say that they can provide support at $15K per year costing
>them $10K per year per customer.  If they "hoard" the software they
>will make $200K plus $50K per year.  If the put it under the GPL
>they will sell one copy and all the other customers will copy that.
>They will lose $160K and will make $50K per year on support.
>If the lifetime of the software is 3 years or less or if the margin
>of profit on support is decreased they will not be able to recover
>the costs.

In article <1991Jul6.222325.13344@cs.umb.edu> betsys@ra.cs.umb.edu (Elizabeth Schwartz) writes:
>  You are right, and it gets even worse if you consider high-volume, low-cost
>software. Take something like dBase for the PC, or even MSDOS. Most users
>of these products don't NEED support, except for a small amount of training
>or installation assistance. WHere is the incentive for the manufacturer to
>continue to make these products easier to use, and where is the guarantee
>to the  customer that these products are extensively tested, if the only 
>money to be made is on support?
>  I don't mind debugging at home. But if I am buying something which I am
>going to use to keep my firm's financial records, or give demos to customers,
>I want to pay money to someone who will guarantee me that their product will
>perform as advertised, not someone who promisis to start hacking if it breaks!

Hold on! Some assumptions have gone into this line of reasoning that bear
examination. These assumptions are of several kinds:

1. A software producer selling GPL'd software will behave the same as one
   selling proprietary software.

2. You have to start from scratch for each programming project you wish
   to make money on.

3. People are schmucks.

4. Only the manufacturer can improve software.

Now, let's examine these in detail:

1. As a producer of GPL'd software, my behavior can be modified from that
   of a producer of proprietary software. I can charge for the product 
   and give free warranty protection to all who buy it from me. Or, I can
   give the product away free and charge only for warranty protection
   (thus bypassing the first company in the above example), but make
   an initial warranty fee required.

   One of these, or any other equivalent scheme, will bind the warranty 
   protection to an initial as well as a yearly fee, and so can be used
   to set up the original cost structure. Or, you can set up any king of
   cost structure that you want.

   The trick here is that if you have to get the customer to buy your
   product, and they want to get certain things for their money: a quality
   product and assurrance that it will work (on an ongoing basis).
   Whether your software is copylefted or not makes no difference on these
   topics (as is being currently demonstrated by Cygnus).

2. I don't have to start from scratch if I can take/buy a GPL'd package
   and modify it into a new product. If I could save a significant portion
   of the development cost, that's several sales sooner that I make a
   profit!

   Such a scenario can only come about with GPL'd software---never with
   proprietary stuff.

   Also, IMHO, this scene is better for society. Resources are not wasted
   reproducing the same functionality in software. If, however, you just
   want to get a lot of programmers employed reproducing functionality,
   they you will absolutely hate the GPL.  I, for one, take the former
   view, however, the latter smacking too much of socialism / communism
   (employment for the sake of employment) for me.

3. Do you think that one guy who bought the software package is such a
   schmuck that s/he will just give it away to all comers? No, s/he won't,
   but something equivalent will probably happen to deprive the original
   seller of profit in this scenario.

   However, it is important to remember that (most) people are not
   schmucks! This can be exploited in setting up cost structures (as in
   item 1 above).

4. How about all the little PC guys improving the software! I know that
   they can do a better job of it than the big company! This is clearly
   demonstrated by all the PD software out there that works great!
   Also, Elizabeth notes that most PC software doesn't NEED improvement, so
   why does it have to cost a lot to get it?

   The problem here is weeding out the crap and organizing the improvements
   so that they can be shared by all in a trusting manner. Also, Elizabeth
   is correct in noting that she wants to buy support, not hang on hackers'
   whims, since she has a job to get done. Proprietary software can do both
   these things.

   But, while PD does well on the former, it does poorly on the latter.
   Also, while proprietary does well on the latter, it does poorly on the
   former.

   However, GPL'd software can promote both of these contrary aims. The GNU
   C compiler is a good example. If I have some modifications for it, I can
   use them myself. I can also send them in, and, if they're ok, the
   distributors can include them in releases to all. Also, I can go buy
   support from Cygnus or equivalent vendors just for the service of
   support, if I want it.

Think about the scenarios in full before passing judgement. The copyleft,
GPL and LGPL, have a lot to offer as alternatives to PD and proprietary
software.

-- 
Kurt Baudendistel ---                                           --- GRA
Georgia Tech, School of Electrical Engineering, Atlanta, GA  30332-0250
internet: baud@eedsp.gatech.edu              uucp: gatech!gt-eedsp!baud