On Wed, Jul 29, 2020 at 9:42 AM John Gilmore <gnu@toad.com> wrote:
There was another chapter to the "tar wars" after UNIX and after POSIX.
Ah ..  indeed - I left out the Gnu Tar story as I was not 100% sure of how it came about as I had not taken part in it.  And pretty much for the purposes of how we go to where we are today, other than it exists, works, is a popular implementation and can read/write things when called upon ... I did not think it would add to the (already) long story. The null vs space filling is an interesting point which I had left out.  Thank you for that detail - I do remember it.

If I missspoke/was confusing (I hope not) about the UIDs I thought I had said the way you did.  The key was that USTAR added the names in ASCII which was not there before in Ken's original version.

Again, thanks for the friendly addition/update.


After I left Sun in about 1985, I worked on a project with GNU and the
BSD folks, to find or write freely available implementations of many
popular UNIX commands. 
Yep, I do remember all that...

 
Since we didn't find a free "tar" program, I wrote one from scratch, based on the SunOS man page and on running the
tar binary from SunOS 3.3.
I always found that strange the folks that wrote that that tar implementation (i.e. you and your mates) had not found the pax code, as the USENIX version had been previously posted/was in the wild by then.  Keith certainly knew about it (he could have even been part of the finding a student to write it, but I don't remember), but he also might have been off at BSDi by that time.  I think by then that the USENIX FOSS implementation even knew how to behave like cpio, tar, or pax depending on its name.

I'm fairly sure, that Apple and HP had picked it up soon after it's release.  DEC had a different set of tar switches, so pax was put in the Ultrix contributed library, and they left theirs alone.  That said, the USENIX version did have an MIT/UCB/CMU style license, not the gpl, which our common 'friend' in Cambridge often (??always??) found suspicious.  So, I had always >>suspected<< the licensing style was driver for yet another version, and have always been a little curious.

But to me it was like C compilers, as long as they all worked, I didn't care.  As you know, I have never been super religious about the different license flavors as long as I could use it.  Probably a good beer discussion/story behind it all when I see you next at a future conference post CV-19.



Clem