First Warren - thanks for the refresh.   Both Dennis and Ken's Turing lectures should be required (re)reading to remind us all on some of the important lessons and gifts the system gave us, besides the technologies themselves.

On Thu, Sep 10, 2020 at 10:30 AM Nelson H. F. Beebe <beebe@math.utah.edu> wrote:

In his paper, Dennis Ritchie referred to another UNIX article that I
did manage to track down and record in unix.bib:

Thanks.  I remember that paper and the argument at the time (also a fun re-read).

I remember that from those days and I do admit that when I too first encountered UNIX in the early/mid 70s, like Prof. Norman describes, UNIX did seem a little 'different' from my then comfort zones of the PDP-8 and PDP-10 worlds, much less the IBM/360 and TSS.  In my first encounters, I certainly felt some of the things Prof. Norman and others in those days would describe.  But what I find fascinating about this paper is that the complaints about UNIX described there by Norman and by others at the time also, now are used to describe Linux - that is, the observations (complaints) are unchanged in 50 years. 

So I realize that you either "get it" or you don't.   You can be educated and overcome bias by keeping an open mind if you come from some other system (somewhere else), or you don't find it strange if you are new to computers - i.e. accept it as is [like Lesk describes in the side bar].   Put another way, as I used to say in those days to people that were seeing UNIX for the first time, the learning curve was different and could be longer >>if<< you come in with expectations from some other system.  But if you had never seen or used a computer before it was not that difficult. What is 'normal' behavior for you -- such as the case-folding or C:mumble in filename conventions?  

I think that Norman's complaint is really the 'baby duck' syndrome at its highest level (vi vs emacs, LISP vs anything else, etc.).  Once a person working with a new system learns to use and appreciate a feature (like typing only a few characters for as 'ls' instead of dir<tab>, having to have case independence, or not having to device the storage device in a filename) UNIX or the like is not so strange, becomes comfortable,  if not desirable.

In fact, just yesterday, I was trying to reconfigure a used Cisco switch I had picked up to use at home.   I would have loved to have been able to type: 'dir' much less, 'ls' instead of: show file information flash:? which I suspect some IT person that uses Cisco gear does not find strange.

I also think Dennis's comments about Xerox's GUIs are interesting BTW.  The BLiT that Rob, Bart, and friends were working, was just making the scene around the time of this paper and certainly, GUI's were becoming all the range and would make their mark as Dennis suggests.  But, I also think it's interesting that 36 years after his paper GUIs did not wholesale replace CLI's.   Not because people are stubborn, as much as people discovered what each gives you (and thus I use both -- I run the Apple GUI for simple things, but there are always 3-5 'iterm2' windows open with a (shutter) C-Shell prompt in each [the later cause the ROMs in my fingers are burned to the old UNIX maxim: 'Bourne to Program, but Type with Joy.' 

Anyway, thank you both for a refreshing reread and reminder during these bizarre times that some things in our world remain constant.  People like what they like, and as Paul Simon reminded us years ago: "One man's ceiling, is another man's floor."

Clem