FreeBSD has a politics problem

Rozhuk Ivan at
Sun Mar 4 02:57:28 UTC 2018

On Sat, 03 Mar 2018 17:30:14 -0500
John Darrah <timmcgrawfan at> wrote:

> FreeBSD recently introduced an updated Code of Conduct that
> developers and members must adhere to. There has been much backlash
> online about it and about introducing identity politics into a
> technical OS project in general. The Code of Conduct was adopted from
> the "Geek Feminism" wiki's version, which claims (among other things)
> that racism against whites doesn't exist, sexism against men doesn't
> exist, and that certain protected classes of people should not be
> criticised.

All people on the Internet are equal, there is nothing to discuss.
Any rules/CoC asserting the opposite are absurd.

The more you learn about the relationships and decisions in FreeBSD, the more doubts it is to continue to deal with this.

- Take the commit bit from an Iranian developer. (for me it's rumors, I can be wrong)

- History with John Marino a year ago:

- Now this strange code of conduct, when quite neutral could be taken:
and do not spend money.

As for me personally, I do not understand where the logic is in many technical solutions, from those that I see.
Perhaps these are differences of mentality and language.

1. The system does not still have the O_EVTONLY and O_NOATIME flags for open ().
CoreTeam believes that FreeBSD should only work on servers?

2. 3+ years in the glib broken GFileMonitor and it was completely off then the applications crashed and eerily braked.
While 1+ year there were two solutions to the problem:
a. go to libinotify and do not use the clumsy kqueue () code from glib
b. use my alternative kqueue () code:

As a result, wait clumsy patches of glib with which everything is still slow.
I got the feeling that no one here at all does not care about FreeBSD users on workstations.

3. For 4 years there is still no support for modern LTE / 4g modems.
I rolled up the patch, checked it, then I also made a couple of significant, I think, improvements in logic, as a result, my first version of the patch, slightly modified in the design, is added to the system.

4. I do not understand why I had to add an extra driver to the system
"amdsmn - Family 17h System Management Network" which is only needed in amdtemp.

5. Constant begging in the mailing list to pay attention to the patches in the bugzilla and added to the ports / system.
Otherwise, patches in bugzilla hang for months and years

6. Imposing the use of poudriere port committers.

7. TCP stack which at considerable RTT and losses gives in times less speed than in Linux.
The minimum set of CC algorithms in the base.

In my experience CoreTeam (or whatever it is) is engaged in politics and code writing, and takes little care of both project and community management.
I see how people who do much go away, nobody takes their place, nobody does their work. This is bad for the project and for users.
Everything described above quite strongly demotivates me personally from any activity in the project.

More information about the freebsd-advocacy mailing list