pinard@IRO.UMontreal.CA (Francois Pinard) (10/31/90)
[This was posted Fri, 26 Oct 90 12:00:05 EDT in gnu.misc.discuss. I received some suggestions to cross-post to comp.software-eng as well.] I would like to share on the public place some feelings about my very own bug reporting experiences, and GNU. Maybe some other people would recognize their own story in this :-). Control Data and Cray were releasing their operating system in complete source form -- exactly this freedom the FSF is seeking, save for the non-redistribution. On the other end, Digital Equipment, IBM and Sun did binary only releases, they still do so far that I know. I worked several years as a systems analyst in a few big sites using machines from these companies. I worked hard and conscienciously to improve all the habitacles in which I have been. I always tried having the systems more dependable and more useful. However, all companies received the numerous bug reports I filed, even source code corrections, with some kind of reluctance, despite the fact that we were buying high prices for full hardware and software support. This reluctance was seemingly a general rule among the companies, for all their users. All the User's Group made for interacting specifically with these companies (except IBM: I did not participate) had as one of their endless purpose: getting the companies to accept and process bug reports more conveniently. Too often, it was a real challenge getting these companies to simply *recognize* bugs as such. The GNU project, on the other end, has a totally different approach. Not only bug reports are accepted, but also processed faster than anything I ever saw from the big manufacturers. Moreover, instead of inducing reluctance, the bug reports are welcomed as *contributions* to the GNU project. I do understand this attitude is not unique to GNU by the time being, it is spreading around, hopefully should I say. But the quality of support, conjuguated with the very high quality of GNU subproducts themselves, the goals of the project, the high skills of the people involved, and the strong and willingful direction, let me perceive the GNU project as a true blessing. I have no words to express the true content and relief this induces in me, so please do believe me when I thank the GNU developers for all this. I speak mainly for myself, but also feel that I speak for several of my friends who live around computers, and certainly for the whole community in the long view. -- Franc,ois Pinard ``Vivement GNU!'' pinard@iro.umontreal.ca (514) 588-4656 cp 886 L'Epiphanie (Qc) J0K 1J0 ...!uunet!iros1!pinard
edp@jareth.enet.dec.com (Eric Postpischil (Always mount a scratch monkey.)) (10/31/90)
In article <PINARD.90Oct30153737@kovic.IRO.UMontreal.CA>, pinard@IRO.UMontreal.CA (Francois Pinard) writes: >Control Data and Cray were releasing their operating system in >complete source form -- exactly this freedom the FSF is seeking, save >for the non-redistribution. On the other end, Digital Equipment, IBM >and Sun did binary only releases, they still do so far that I know. Not so. I know Digital's RSX system is released in source form; it is assembled during system generation (unless specific tailoring is not needed). VMS has been available at least on microfiche; I'm not sure about other forms. -- edp
jjacobs@well.sf.ca.us (Jeffrey Jacobs) (11/06/90)
Back in the golden days of computing, fixing the operating system was one of the chief duties of "systems programmers", and every vendor supplied source code to the operating system, including the IBM mainframes and DEC 10s. Jeffrey M. Jacobs ConsArt Systems Inc, Technology & Management Consulting P.O. Box 3016, Manhattan Beach, CA 90266 voice: (213)376-3802, E-Mail: 76702.456@COMPUSERVE.COM