Re: Call to NOT move to github or gitlab

[Dennis Clarke]

On 6/10/22 06:39, Benno Schulenberg wrote:
> Op 09-06-2022 om 09:38 schreef Tomasz Kłoczko:
> > So you are not using pages where JS is used??? Really?🤔😱
>
> I didn't say that I don't visit pages that use Javascript, I said
> that I don't allow my browser to execute it (except for a few
> trusted sites).  Many pages are perfectly usable without Javascript
> (as they should, if they want to be standards compliant).

        * * * WARNING : long rant follows  * * *


    You are speaking a very foreign language for an entire generation.
I have hit the same problem over and over wherein it is beyond some
people ( or nearly impossible ) for them to even enbrace the concept
that one may run a computer without a mouse and a screen.  A serial
terminal into a hulking HP SuperDome ( or similar ) looks entirely dead
to them. They *need* endless graphical clicking to do anything at all.
Somewhere in the last twenty years we lost many people from the computer
industry and that is not a bad thing. Simply put, a cell phone is a
massively capable 64-bit machine with many gigabytes of memory and long
term storage. That works just fine for most people. However I wish you
luck trying to talk to anyone about shifting bits through a UART or even
using memory on big endian servers and why that looks so weird compared
to their Windows x86 machine.  The communication gap is in perception
and the ability to visualize the actual problems.  That seems to have
been lost in the last twenty years.  Mostly.

> > If you need to find something all you need is click on the search input
> > line and enter what you are looking for.
>
> "All you need to do is click..."  I /don't want to click/!  I want
> stuff to be local, where I can use my keyboard to access things.
>
> > Maybe it is not clear how github/gitlab works [...]
>
> Christ.  I /know/ how Gitlab works, ...

    This may collapse into an old school flame war. I hope not. However
it may simply be a long whine and push by the popular Disco Dancers to
demand that they click everywhere and submit everything to some super
massive corporation controlled site from hell. These popular Zoot suit
wearing folks have allowed trivial communications to fail. Here is an
example :

    1) OpenSSL is one of the most important open software projects

    2) There has always been a mail list for OpenSSL users

    3) They made the horror show leap into github

    4) Documentation has been updated to claim that ordinary users
       should report any problems via a "trouble ticket" on GitHub.

       Really?

       Please see :

         https://github.com/openssl/openssl/blob/master/SUPPORT.md


    5) That document claims :

        If you have general questions about using OpenSSL. In this
        case the openssl-users mailing list is the right place for
        you. The list is not only watched by the OpenSSL team members,
        but also by many other OpenSSL users. Here you will most
        likely get the answer to your questions.

       Furthermore :

        Just because something doesn't work the way you expect does
        not mean it is necessarily a bug in OpenSSL. If you are not
        sure, consider searching the mail archives and posting a
        question to the openssl-users mailing list first.

    6) However OpenSSL has jumped into GitHub and the flow of traffic
       to the mailing list dropped considerably. In the past there were
       questions posted and a response the very same day. Sometimes in
       the same hour. Not so anymore.

    7) An ordinary user working with recent OpenSSL may run into a
       problem :


https://mta.openssl.org/pipermail/openssl-users/2022-June/015240.html

    8) Hold on ... the certificates inside the test suite for OpenSSL
       have expired? That information is hidden inside GitHub :

            https://github.com/openssl/openssl/issues/15179

       There is no such discussion in May of 2021 on the mail list:


https://mta.openssl.org/pipermail/openssl-users/2021-May/thread.html

    9) A new certificate was created nine days ago :

            https://github.com/openssl/openssl/pull/18444

   10) However there is no discussion on the mail list anymore for these
       sort of problems. The information and knowledge is being slowly
       transfered over to GitHub entirely and the Disco Dancing Zoot
       suits seem to follow along without even a care or thought to what
       that means long term.

The actual business objective of services like GitHub and GitLab is to
migrate all knowledge into their domains with the long term plan to lock
that out from people forever. All one needs to look at is the history of
Oracle and the OpenSolaris project to see what happens when an open
source project falls under the money hammer. This has nothing to do with
the javascript problem. It is just a long term business plan and only a
fool would be unable to see it.

I was eventually able to get a reply from someone who told me to go get 
a patch at https://github.com/openssl/openssl/pull/18444.  There is no 
patch there. There is a big complex messy page with a pile of clicking 
and searching required to get the new certificate. I then waste my time 
doing the click search dance to eventually find :


https://github.com/openssl/openssl/pull/18444/commits/456de6e73c05fc413aacedcdd551e2a259f93262

That is garbage. The actual file is five more clicks away :


https://raw.githubusercontent.com/t8m/openssl/456de6e73c05fc413aacedcdd551e2a259f93262/test/certs/embeddedSCTs1_issuer.pem

I try to confirm this here :

    https://mta.openssl.org/pipermail/openssl-users/2022-June/015241.html

Yes I use the mail list as described in the documentation.

Eventually, a day later I hear back from someone :

https://mta.openssl.org/pipermail/openssl-users/2022-June/015243.html


    All of this is a waste of time.  Disco is dead and no one wears a 
Zoot suit anymore. Who knows, it may all be popular again someday along 
with the telegraph. However when that happens all knowledge will be 
locked away inside a corporation and you will be forced to pay for it. 
The Microsoft way has always been "embrace, extend, extinguish" and you 
can bet all your future dollars that will happen.[1]

    Regardless, after a pile of work by me I see :

All tests successful.
Files=243, Tests=2874, 9634 wallclock secs (39.16 usr  3.89 sys + 
9446.48 cusr 149.27 csys = 9638.80 CPU)
Result: PASS
make[1]: Leaving directory `/opt/bw/build/openssl-3.0.3_rhel6_amd64.006'


    I could have had that result in one day. Less than six hours of 
work. However not when the transfer of all people and knowledge falls 
inside Microsoft GitHub.


--
Dennis Clarke
RISC-V/SPARC/PPC/ARM/CISC
UNIX and Linux spoken
GreyBeard and suspenders optional

[1] https://drewdevault.com/2020/08/27/Microsoft-plays-their-hand.html