Re: Docker

From: Paul Pathiakis <pathiaki2_at_yahoo.com>
Date: Mon, 17 Apr 2023 01:51:34 UTC
 I'm sorry.... I almost ROFL when I read that FreeBSD is used 'less' in the industry....  You 'like' both Linux and FreeBSD...However, you don't seem to understand the philosophical differences in the development of the two projects.  

List of products based on FreeBSD - Wikipedia

| 
| 
|  | 
List of products based on FreeBSD - Wikipedia


 |

 |

 |


The LIST OF PRODUCTS....  That's just stuff that is commercial and based on FreeBSD....  Yes, I'd like to think between iOS, OSX, etc. Apple is a massive shipper of what works.  I've worked on both iPhones and Android as personal products.  Don't get me started on iPhone - just works and Android's Linux has been less than entertaining.
There's a large list of Open-Source products as well.

If you looks at WHO is using FreeBSD... it's pretty much the biggest commercial companies in the world.  Why is that?  The philosophy of proper engineering.  Not the moronic "Cathedral vs Bazaar" mentality that everyone said you can just keep throwing tons of non-engineering types at problems and they'll write the works of Shakespeare.  Maybe writing a piece of literature is possible.  Writing a bulletproof, high performance OS that adheres to STANDARDS like POSIX for interoperability with other POSIX standards is a 'novel' idea but it is the only one.
Also, stripping/non-adhering to standards?  You get MS's no longer relevant Web Server.  Apache (uses FreeBSD too) still dominates the world and will continue to do so.  Why?  A lot of the same FreeBSD philosophy.  Why was MS web server faster than Apache in certain benchmarks?  Not doing all the proper checking of the web standards.  Two cars....  One beats the other 2 out of 3 times, however, it gets hijacked 5x more than the other.  Do you really want the car that has no security like locks built into it?

When someone gets a degree in a technical discipline, it changes their mentality in how they approach problem solving.  Computer Science, Computer Engineer, Physics, Systems Analysis, etc.  If someone puts a piece of 'spaghetti' code in front of a person like that, you get what is now emerging here....
What problem are you trying to solve?  So far, you've claimed there's a problem that there's a large pool of people that want to see docker on FreeBSD.  Again, doubtful.  If there is, again, you've been repeatedly told to start studying the Linux structures in its kernel and the corresponding components in FreeBSD and make an attempt to understand that Linux is NOT standards compliant.
So, as someone that sees this more and more as trolling, please stop asking people to solve a non-problem.
FreeBSD doesn't care about what the 'latest, greatest and kewl thing' that someone thought of.  FreeBSD cares about building structured, ordered, QA'd, release engineered product that isn't part of a crappy, latest and greatest software development paradigm.  Anyone care for Waterfall, devops, AGILE, and everything else that falls by the wayside every 5-10 years where a bunch of companies pop up to consult about the 'BEST WAY' to do software engineering instead of the classic examples/methods that have worked for ages?  How about 'fake deadlines'?  They are all fake when someone in management doesn't understand what they are trying to accomplish.  They set an artificial deadline.  (It's due to the fact they don't understand proper engineering they just think that if they don't put a deadline, the engineers will just screw around. - VERY DOUBTFUL)  Yet, I watch FreeBSD push out at least 3 full releases a year.  They have a 'timeframe' that they'd like to see.  However, it has to pass QA and release engineering via RCs (release clients).  I've seen them 'predict' a norm of 3 RCs.  What happens if they find bugs?  In the commercial world, with AGILE, the attitude is document the bug, and keep moving....  That bug can actually make it into the release.  In FreeBSD, NO!... there will be a revisiting of the implementation, the correction, the QA process again, another RC and more testing.
The day I read what AGILE was proposing was the day my classic computer science degree churned and almost puked.  Sorry, that's my feeling towards artificial deadlines, and SCRUMS, and constant churn of buggy code that should be fixed waaaayyy upstream versus being allowed to linger until the next Sprint or release.
Paul


    On Saturday, April 15, 2023 at 07:09:48 PM EDT, Mario Marietto <marietto2008@gmail.com> wrote:  
 
 With this general attitude, full of firm convictions and systematic stigmatization of other 
people's ideas (when they aren't canonical), there is not enough space for the emergence of 
divergent ideas,ideas that can create innovation because they are able to break thought 
patterns became too static and historicised. It is a pity because in order to develop, 
progress needs to be nourished above all by unpopular and divergent ideas. 
I'm not talking specifically about porting docker to FreeBSD, but I'm talking about the way 
you express your ideas, which, in my opinion, aren't dynamic and open to changes and aren't
ready to welcome innovations that come from other systems such as Linux, which always have
something that's wrong. Some of you demonize Linux too much and you don't see the great 
amount of innovations it has brought. I like both Linux and FreeBSD and I am sad to read
that the latter is always used less in the industry. There may be reasons, but few of you 
seem to realize that some of them are inherent to their own views about how FreeBSD 
should be.
On Sat, Apr 15, 2023 at 9:32 PM Ralf Mardorf <ralf-mardorf@riseup.net> wrote:

There are countless container approaches available under Linux. None of
these approaches are mature and almost all, if not all, will be phased
out before they are mature.

I guess nobody wants to experience a Canonical and/or Google universe.

Somewhere in the Internet I found "While Docker is a container runtime,
Kubernetes is a platform for running and managing containers from many
container runtimes".

At next somebody will introduce a platform that does manage platforms
for platforms. An energy turnaround looks way different.




-- 
Mario.