
mw at moni
Feb 26, 1997, 7:15 AM
Post #1 of 3
(220 views)
Permalink
|
I really wonder what If you want to distribute modified versions of qmail (e.g., different packaging formats, porting changes, precompiled binaries) you'll have to get my approval. Note that this means approval of the version, not approval of your distribution method. means? Does this answer the following questions clearly 1) If a Math department has 37 Unix machines, does the system administrator have to compile qmail on each separately (implying that each machine has to have a compiler even if the owner will never use anything else except E-mail and TeX). Or can she compile it on one Linux/gnu, one sparc, one iris, and then install from these machines to the rest ? Is this considered a distribution of the binaries? How about sending the binaries to other departments on campus? How about doing the same with a home made src.rpm/deb? 2) How long does it take to get an approval? How does Dan check a binary package? Does one actually send the binaries to him, or just the description of how they was compiled? Who is going to read and understand all these descriptions, patches? Does he have the capabilities to learn and check every conceivable packaging system? Does one have to send a current version of the packaging system as well? Is he willing to bother with the compilation of these packaging systems, or one should send a binary version? 3) Does the license apply to 1.00 only, or to all previous versions as well? In particular, do we have to remove the .96.rpm from our university's ftp site? Why did the license appear only now? 4) What if a bright grad student at Berkeley decides to write the perfect MTA, and she wants to incorporate ideas from qmail; how much freedom does she have to experiment? Mate Mate Wierdl mw [at] moni University of Memphis
|