The future of GhostBSD

[ericbsd]

For those who don't know, I am now involved in TrueOS, and I participate in all the meeting, TrueOS Core is slowly becoming a fork of FreeBSD. Most of the people know ASX started to dislike and I slowly going that direction too. With a lot of discussion with ASX and Joe, my manager, I decided that GhostBSD 11.1 is the last release under FreeBSD.

The first official release of TrueOS is coming sone as TrueOS 17.2 for 2017 second release the year. I have been running on TrueOS unstable for the last months for work Joe did add all our port in their repository, and I run a spin-off GhostBSD on TrueOS CORE and other than the TrueOS update manager it is stable. I did not have any problem with stability.

After two months of working on TrueOS Core, I decided to go forward to replace FreeBSD with TrueOS CORE as the base system.

Why abandoning FreeBSD?

FreeBSD is lagging behind with drivers, the new feature we see in Linux and the lack of support for the desktop and more. FreeBSD is not well tested for stability with the system, ports build, and pkg build, also ports are not well maintained. I can add more and more, but it is not my point.

Why going with TrueOS?

TrueOS focus on Desktop first and Server second. It uses OpenRC and ports are almost all ready for OpenRC. Build of Ports are tested, pkg build are tested, system build is tested, iso boot is tested, and they are tested both by human and automation. I have been hired for doing automation testing for everything iXsystem produce. There is no code that goes into TrueOS Core base that is not functional or buggy, Yeah sure it can happen error happen, but less likely than FreeBSD. I am learning to do GUI test automation for QT and GTK, for GTK it is not well documented, but for QT it is.

Ports under TrueOS are a bit better maintained, but still, follow FreeBSD ports and try to push change upstream.

Now, what that change means For GhostBSD?

Since I work for iXsystem and now contributing on TrueOS Core, I can voice my concern about everything and get a voice in the direction of the base system. I have the support of Ken, Kris, Joe, and Matt. Joe will contribute to GhostBSD whenever he has time since he is a manager he is busy with meetings it is normal. The base is battle tested the ports build is battle test and on and on.

What is the benefit to GhostBSD?

• OpenRC
• more stable ports
• latest stable drivers
• latest stable abi
• latest stable linux abi
• latest stable linux supports
• latest stable kernal change
• kenel modul not in FreeBSD

There is more but you get the picture.

Now, what is the short-term plan?

Release GhostBSD 11.1
Merge the work of Joe on the ghostbsd-build has done most of the heavy work for us.
Update one server to build pkg from TrueOS ports.
Build iso for everyone to test.

Now, what is the long-term plan?

The adoption Ravenports
The adoption of SysADM but I don't like how TrueOS does updates I will leave that part out.
building GTK GUI for SysADM

There is probably more to add for short and long term. But it is the basic.

I start this discussion to give everyone the opportunity to voice there opinion, concern, and subjection.

GhostBSD will still function the same and going the same direction, but with a different system base.

[ASX]

Admittely I have been very concerned about a move like that, mainly because TrueOS is based on FreeBSD-12-current and that is not stable.

However, talking with Eric, he explained me that they had put in place a lot of test and checking for Quality Assurance, so I decided to give them a chance and at the same time to try to make GhostBSD reborn, after all I'm still interested in a "Good Desktop OS".

It is also my understanding that FreeBSD management is splitted ... the "old guard" from one side and the "new guard" on the other side ...

GhostBSD and TrueOS are the only two BSDs pushing hard on the "Desktop" (you can consider OpenBSD and MidnightBSD too, I know you can build a desktop with them ... but ...).

They will remain independent projects, that's my understanding.

FreeBSD is a militarized zone, think of it like of sort of DMZ (labeled "de-militarized zone", labeled "open to community contributions") which in practice means it is the most militarized one. (just like the North Korea DMZ). My opinion., of course, and that is not the point of the discussion, it is just to explain my position.

ravenports, I'm not yet sure ravenports is a solution, but I'm confident that it is a step forward.

Anyway, I accept the challenge, I want to see GhostBSD succeed, and I'm glad something is changing, because I know that when there are no changes there is no life.

What will be the future will depend from us, we can talk.

PS: noticed that ericbsd didn't mentioned one thing we discussed: in addition to i386 also support for 10.x is already dropped.

[kraileth]

Hooray for the great news!

I've been passionate about TrueOS in the past and tried to like Lumina, but that has killed the experience for me. I'm a GTK person and for whatever reason I cannot stand the Qt looks. GhostBSD on a TrueOS base (with OpenRC goodness and the much better driver support) sounds... well, just excellent! I was hoping that we could switch to -STABLE after 11.1 was released, but this is even better.

Two thumbs up for the plan of bringing SysAdm to GTK+, too! I'm very much looking forward to this. And dropping 10.4 is something that I also support. It would have been unnecessary work IMO.

And about Ravenports... Wow! Just wow. I would not have dared dreaming of this! I was the first external committer to Ravenports after John started it and am the POC for a good handful of ports there. This mainly happened because I offered help over at the FreeBSD ports but wasn't exactly happy about their adoption (or rather: lack of) of my work. ASX is not the only one who kind of has a set on the ports team. Here's my story:

After finding a broken port that I was interested in I decided to mail the maintainer. I told him that it was broken, that I was willing to help and asked if I should open a PR. Two weeks later he answered that opening a PR would be great and so I did that. I'm a patient person, but five weeks later I was slowly getting unhappy with the situation. My first attempt at submitting a new port to Ravenports was answered by John five minutes (!) later. It has never been more than a couple of hours and I had an answer from him. And he's basically running all of Ravenports all by himself... Now it's way over two month and despite asking again, the FreeBSD port is still not even marked broken. This is a completely unsatisfactory situation for me and I'm determined to help to push Ravenports forward.

Sounds quite a bit like we're a bunch of FreeBSD rebels. I can live with that. FreeBSD is for servers, there's probably nothing you can do about that. Bringing the BSD desktop forward may require a bold move like this. I'm very much intrigued about how this will work out. So let's get this baby going!

Are there any plans regarding a replacement DM so we can get rid of SLiM? Eric: What were the problems with LightDM that you encountered which forced you to back out and switch back to SLiM for 11.1? I'm using LightDM and in general it does work.

[ericbsd]

kraileth about the Slim me and John will work on making lightdm to work with the live CD.

[ASX]

follow up: viewtopic.php?f=97&t=933