a request for the 8.1.7 and 8.2.2 tarballs and the good reasons why - Mailing list pgsql-general

[ if this belongs on another list, such as -advocacy, i apologize. ]

dear sir or madam ---

permit me to submit a plan which, imho,
   will transform the recent lemon into lemonade.

we all know the famous motto,
   "there is no such thing as a total failure.
   it can always serve as a negative example."

on monday, starting with the oldest first,
   i secured the 7.3.18, 7.4.16 and 8.0.11 tarballs before being interrupted
   [ i have since downloaded 8.0.12, 8.1.8 and 8.2.3 ].
returning late yesterday,
   i found my plan to obtain the remaining two tarballs thwarted.
this didn't bother me at the time,
   as they wouldn't have been useful to me in the usual sense.
however,
   overnight,
   i thought of a way in which they --can-- be useful, in an atypical way,
   and,
   in the process, make the entire organization look good - really, really good.

consider two markets:  students and capitalists.

this summer,
   i am launching an education entity for the purpose of
   teaching ambitious, precocious, motivated teen_agers
   real programming
   based on real computer science [ a subset of the ieee_cs & acm curriculum ]
   using a real language [ c, oop comes later ]
   on a real os [ freebsd, perhaps openbsd later ]
   culminating in the integration of
   real open_source subassemblies [ e. g., postgresql ]
   into their applications.
[ i am also going to teach them money_management, among other things. ]
according to mckusick,
   in his "design and implementation of freebsd",
   at the end of chapter one,
   this is the age range at which typical freebsd committers started.
if i can show my students that
   real, seriously major, people,
   working on a real, seriously major, project,
   that is used by real, seriously major, organizations,
   can make a mistake,
   it will do wonders for their self_confidence.

on the other hand,
   there are some chamber_of_commerce people i know who like to ask,
   "how good can it be if
   it's a bunch of volunteers who
   give it away for free?".
i think that
   being able to show that errors, when discovered, are fixed promptly
   is an excellent marketing point,
   considering the price charged for the service.
i can write up a quick_&_dirty very_simple application to demonstrate the point.
i can quote from some of the emails i read today to
   show the professionalism of the people who addressed the issue
   [ if intellectual property is an issue,
       i can point people to the archives instead
   ].

the common point is this:
   we are all human, so mistakes will happen.
when they do, admit it, apologize and fix it.

in this spirit, i propose two plans:

   a]  i think that the 8.1.7 and 8.2.2 directories
         should be restored to /pub/source.
       i would add a brief, one_off README [ perhaps named README_FIRST ? ]
         that describes the bug and
         warns not to use the version in a non_demonstration situation.
       i think this is the proper action because,
         when new versions are released, typically for security reasons,
         we are advised to upgrade asap,
         yet the prior versions remain available, none_the_less.
       in addition, the time_stamps will demonstrate the rapidity of repair.
       lastly, there is the argument of historical completeness.

   b]  if, for some reason, this is not possible,
         perhaps a reader of this post has copies on an ftp server
         to which i may log_in anonymously or some such.
       while beggars can't be choosers,
         i seek the complete 8.1.7 and 8.2.2 bzip2 versions, if possible;
         of course, gzip also works.
       oh, yes; the md5's would be nice, also.
       all i can promise at this time is to
         name you publicly [ with your permission ] and
         say very nice things about you,
         as i perform the above_described demonstration.
       i am subscribed to -general.

kudos to everyone involved.
you may not think you did anything special because, to you, this is normal.
however,
   so many people just don't expect anything good to happen anymore
   because there are far too many people in this world who just don't give a damn.

thank you.

rob spellberg
woodstock, illinois