[GTALUG] Linux Unicorns

[D. Hugh Redelmeier]

| From: Evan Leibovitch via talk <talk at gtalug.org>

| On Fri, Nov 24, 2023 at 12:19 PM D. Hugh Redelmeier via talk <
| talk at gtalug.org> wrote:
| 
| It seems to me that Linux has won and been commoditized on the server side.
| > No high-level discussion needs to address this.  Several distros are good
| > enough and almost interchangeable in abilities.
| >
| 
| Agreed.
| 
| The whole CentOS/Rocky/Alma/Oracle/SUSE/RHEL dustup has brought some
| instability to the realm (and for those of us old enough, bears remarkable
| resemblance of the old SysV/OSF1 wars). Ditto some of the jerk moves
| Canonical seems to be doing to leverage the Ubuntu base. But I think that
| these instabilities will settle down, even if they only evolve into
| stalemates.
| 
| On the PC side, if you don't wish to demand a dedicated machine, you
| > have to fit into what the market has.
| > 1. Windows of various vintages
| > 2. MacOS.
| > 99. Linux.
| >
| 
| People want virtual dedicated desktops, for something other than UI
| testing? Uh, OK.
| Where do Chromebooks fit into this?

I was talking about an application publisher.  If they want to run on the 
user's existing PC...

| Old Unicorns came from different times.  25 years ago, it wasn't obvious
| > that Linux would be where it is (and isn't).
| >
| 
| Speak for yourself. 🙂
| 
| I was kinda certain that Linux would overtake Unix early in its life; that
| feeling was confirmed when the first Beowulf clusters came online and DEC
| imploded in the late 90s, and fully cemented when Oracle bought Sun less
| than a decade later. The increasing commoditization of PC components
| combined with the neverending balkanization of Unix hardware increasingly
| made the progression to dominance inevitable, the main issue was how long
| it would take.

25 years ago, it was clear Unix was not much of a contender.  And Windows 
NT looked like a winner.  You could choose to build a Unicorn on Windows 
in the cloud.  I think you'd be nuts to use Windows now.

| Netflix uses NetBSD for at least some stuff.  Making a decision like that
| > today would take serious conviction.
| >
| 
| BSD is still a good choice for some, so long as they're OK with
| self-support. It failed to get Linux's mindshare for two unrelated reasons:

So you are agreeing, right?

|    - They got caught up in their own little petty Unix wars in the 90s,
|    just when the world was looking for some platform stability (that Microsoft
|    and Apple were providing on desktops). Going with one Linux distro over
|    another, especially on the server side, was far less of a critical decision
|    than choosing Open- or Net- or FreeBSD.
| 
|    - Also, the GPL enabled big vendors to contribute code without a
|    competitor being able to take it and re-lock it proprietary, as Apple
|    famously did to BSD-licensed code in Project Darwin. The GPL enabled the
|    old Ray Noorda philosophy of "co-opetition" which BSD could never match
| 
| 
| 
| > | Does anyone on the list actually know (as in personal experience)
| > | whether/which (or even if) most tech unicorns are Linux based or with a
| > | Linux origin story?
| >
| > That may be interesting history but I doubt that it is worth consideration
| > for a new business.
| >
| 
| Agreed. Linux is infrastructure, not a factor in whether an effort gets
| funded or expands.

Or even work.  In other words, it isn't an innovation any longer.

| More interesting is to figure out how to use the cloud without being tied
| > to a particular vendor.
| >
| 
| Both the choice of OS and where hosted is infrastructure, operational
| issues, and rarely part of a startup's pitch. If there is a privacy or
| security component to the proposal's added value the hosting may be a
| factor (ie, not hosted in a Five  Eyes country) but I think even that is
| less of a factor with revelations that much claimed protection is an
| illusion
| <https://www.forbes.com/sites/thomasbrewster/2023/08/08/protonmail-fbi-search-led-to-a-suspect-threatening-a-2020-election-official/>
| .

Yeah, I was grasping at straws.

There is room for innovation in infrastructure but it is hard to
become a unicorn there since the area is mostly occupied by giants:

- Amazon

- Microsoft

- Google

- Oracle...

Being cloud platform agnostic is smart.  How you do it isn't something
I know much about.

I vaguely remember that Netflix can switch platforms.  (The few tech
people that I know of there are fairly impressive.)  That's probably
very useful when negotiating with their cloud providers.
-------------- next part --------------
---
Post to this mailing list talk at gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk