Hey everyone!
It has been a while. I got busy with life, learning DevOps and such. I always intended to come back. I have a second daughter being born soon (due date is March 27th), so around that time I might go into hiding again. I have some learnings I might post here soon (I have completed them, but now I need to make them into a more digestible way).
I recently saw this article on The Register, and I thought a lot of people here would like that. I found it to be pretty illuminating with how it talked about the development of Unix and its derivatives. Also I think it raises an interesting point - what is Unix or Unix-like? The rules have been broken of a Unix-like operating system, but there is also no reason to think that the people who developed Unix did everything 100% correctly, but I think it is worth it to reexamine how they though an OS should be.
Being old enough to have witnessed most of what that article discusses means that for me it was a fascinating recap.
There are some things about original Unix that were a breakthrough at the time, and are worth preserving.
I did not know that Ritchie and Thompson worked on
academically. We should certainly look at what they did.
There was computer networking before Unix…mainframes with proprietary OS’s were networked… but it was a messy
case-by-case challenge. There were also packet switching
networks before the internet. There were interactive OS’s before Unix. One of Unix’s achievements was to standardize
networking. Another was the shell, which removed the separation between interactive and batch computing. There are other really basic achievements, like everything is a text file, which the article mentions.
My first hands on Unix was BSD4.1 It was a huge improvement over other OS’s at the time. That is what early Unix users remember and would not like to see lost.
That does not mean we cant have some modern features
like GUI and VM.
I think the author of the article was being a bit non-serious when he suggested that Unix-like would mean no GUI. But I think what was serious about it was his suggestion that we go that far back in Unix’s history to see what we have forgotten, which is why he brought it up.
Certainly, while there are some advantages to working with a command line, I am personally too used to a DE to do that 100%. Even if it did break a huge Unix rule to have a DE, I would say that is a rule worth breaking, at least for non-server usage (no reason to have a DE on a server afterall).
I was not around during this time, so its interesting to note what made Unix a pioneer. I remember reading a lot what a big deal Java was in the late 90s early 2000s because suddenly it was easy to make software for everything. Nowadays, we pretty much just have to consider making software for x86 processors and ARM. And ARM is pretty new.
The text thing is important. I have to use Windows 10 for work and I hate how I have to go through a Windows 8-esque menu to get to a Windows 95-esque menu so I can finally change the setting I want. Text files aren’t pretty, but they work and they are easy to find. If people really need it, a graphical wrapper for the text file(s) can be made.
What they make possible is having the output from one
program readable by another program. That was something new and useful. It did not exist before Unix.