EVERHART%ARISIA.decnet@crdgw1.ge.com (08/29/89)
I'm afraid I must reply to this: >This happens now and then, like in the case of SZ and RZ (the ZMODEM programs). >The C source is completely freely distributable. It was paid for by Telenet, I >think. The problem with this is 1) there would be no games. 2) in general this >only happens when a company needs a unique application, which isn't of much use >to most people. I.E., when someone writes a database application for a >business. Costs the business a bundle, but it's only of use to them. For >programs which are of general interest, why should one business pay a fortune >for it and everyone else get it free? They shouldn't. So all the people who >need it share the cost. >-- >James A. Treworgy >jtreworgy@eagle.wesleyan.edu >jtreworgy%eagle@WESLEYAN.BITNET In fact, writing and sharing software occurs in some quarters in major ways and WORKS. Consider: 1. There have been dozens, possibly hundreds of games produced PD on various engines. I've seen them by the score on various software swap tapes - and they represent an insiginficant fraction of the software shared, legally, by computer professionals. 2. Selling software is not cheap. In fact, to a big company, it is VERY costly. RCA used to sell an IC design system, but sold it for $1000K or so a copy. The reasons were things like: * A big company must be very careful of liability lawsuits. Any bugs in their code are likely targets for these, and because of their size, large companies tend to get hit with larger damage claims than a guy in a garage (who in reality will probably never get sued at all because a lawyer will realize there are probably not assets enough to pay HIS bill, let alone damages.) * A staff of employees is needed full time for support, because of the above and to avoid damage to the corporate reputation in other lines. Benefits to a large company are NOT cheap. * Advertising is costly. This hits everyone but must be remembered. Several thou for an inch square in a magazine is normal. More for bigger ads that someone might actually see. * Marketing people have to also be hired to sell software. It's incremental cost, but it's COST. 3. Internal development of software is not enough. I've seen and heard of hundreds of software tools developed internally at various companies which had considerable value and utility. Many would have been useful in outside organizations and indeed in many cases commercial software later appeared doing the same things...sometimes not as well. Of these, only a VERY few survived more than a year or two, and most became nonfunctional with newer releases of OSs and the like. The ones that survived were generally the ones that were released to the world. The others died mainly due to their authors' moving on to other things and sources never becoming available for maintenance, even where others in the organization were interested in DOING maintenance. The value of outside help in maintaining software cannot be overstated. Where a piece of software cannot be sold by a company, it makes great sense to release it freely. This has the effect (and yes, I've SEEN this effect in practice for years) of encouraging other companies to do the same and greatly multiplying available software. To the author, it means his code is likely to be useful to him in 5 years without his having to do ALL the maintenance; it also means some fame. To the company, it avoids some mainenance expenses, provides goodwill among computing professionals, and maintains the competitive advantage that the expertise to use the tool exists internally. I edit the DECUS VAX SIG tapes, which lately have been running to around 150 megabytes of freely redistributable software, much of it contributed in this way, every 6 months. 300MB per year is NOT "now and then". I should add that the total amount of software so exchanged is far higher, even within DECUS, since the DECUS Library gets a lot of material not on the sig tapes, and there are several other SIGs besides the VAX SIG that do this. The RSX SIG Tapes go back to 1977 and are the oldest in the current form. All this material is indexed and almost all of it is in source. I therefore claim the assertions "there would be no games" and that "this only happens when a company needs a unique application" are false and even harmful. (By the way, the Spring 1989 VAX tapes contain a dBase III approximate clone, in source and executable; several other DBMSs have appeard in the past. A spreadsheet, project planner, and many more tools are also on that collection.) One can share costs by sharing what one does, and obtaining the fruits of others' doing the same. Lots of us do. We wind up FAR richer, without hassles, than we could ever have imagined being otherwise. Some of the commercial market survives because people are not aware what's available in the freely redistributable area. Since ads are very expensive, symposium tapes (for example) don't get advertised there. However, those reading this are on the metanetwork and have the essential tools available for becoming informed. You can join DECUS (free; phone 508 480 3418 and ask for forms) and find about what's in that library (several GB of PD code; lots of it for Amiga as well as other systems) if you wish, too. You SHOULD be aware of at least what's on the Fish disks, if not more, for the Amiga public domain. In other words, it takes a little work to find what the real alternatives are. It's worth it. Also don't knock the value of knowing how to use a bit of code. That is a major advantage to have and it remains yours even when you give the code out. Still, I've found you can get some significant help from others you don't even know by releasing your stuff. One final note: it takes only one really fine piece of work made freely available, sometimes, to make life tough on commercial people. The DECUS C compiler has made life miserable for years for PDP11 C compiler vendors, and Gnu C is doing so for other C compiler vendors now. It isn't necessary to be able to supply 40 or 50 word processors in the freely redistributable market to mess up the commercial market; one, if well done, can be enough. The intellectual power available by the occasional help of unorganized general benefactors is enormous and can make the freely available software the best in its market. It has done so on many occasions in the past. Glenn Everhart Everhart%Arisia.decnet@crd.ge.com