X server issues and ports

[ericbsd]

I will not delay, I did that in 10.3 and it resulted to be released in same time of 11.0, wish does not make sense once I started from alpha I am not putting new feature or drastic change, I have made my choice, I also did not know if you was coming back.

I do not want you to take that the wrong way, but after 11.0 I intended to make some release form 11 stable to start testing the repository. The repository was not in the roadmap of 11.0 there would be no one waiting for this other than us. I would also add that 11.0 was suppose to be release 2 month ago. I did fall behind I might remove one option in the update manager, but I not backing from releasing 11.0 as intended.

[ASX]

I will not delay, I did that in 10.3 and it resulted to be released in same time of 11.0, wish does not make sense once I started from alpha I am not putting new feature or drastic change, I have made my choice, I also did not know if you was coming back.

As you prefer, the final choice is yours.

I do not want you to take that the wrong way, but after 11.0 I intended to make some release form 11 stable to start testing the repository. The repository was not in the roadmap of 11.0 there would be no one waiting for this other than us. I would also add that 11.0 was suppose to be release 2 month ago. I did fall behind I might remove one option in the update manager, but I not backing from releasing 11.0 as intended.

I have difficult to understand your position, i.e. tracking strictly the fbsd releases, even at the cost of a doing things "the wrong way", and knowingly. (and I specifically refer to mixing packages and ports).

I'm unable to see a "fast release" as a plus for itself, while looking at the whole picture.

[ASX]

The ISO are build on quarterly and not on Latest to clarify, every build should be same, but for the part that is build from ports which I always had disagree, but I have let convbsd do this part that is an other story.

Wasn't that just written from you ?

[kraileth]

I will not delay, I did that in 10.3 and it resulted to be released in same time of 11.0, wish does not make sense once I started from alpha I am not putting new feature or drastic change, I have made my choice, I also did not know if you was coming back.

In general I'm also for delaying things because that's usually worth it. In the current situation however I also think that getting out GhostBSD 11.0 quickly might be of higher priority (I've already read people saying that they'd stay away from GhostBSD because it's still based on 10.3). Still I wonder if we couldn't introduce our own packages after the 11.0 release? In the end it's just a configuration file to activate a different repo? Since we're of course not building the whole ports tree - how are we going to do this, anyways? Will we have both the FreeBSD repo and our own with the later having a higher priority so that by default packages come from there but people can install software that we don't provide, too? In that case it's probably just a technical decision which the user shouldn't really even notice (of course we should announce it nevertheless) if we test things properly before activating the repo.

I do not want you to take that the wrong way, but after 11.0 I intended to make some release form 11 stable to start testing the repository. The repository was not in the roadmap of 11.0 there would be no one waiting for this other than us. I would also add that 11.0 was suppose to be release 2 month ago. I did fall behind I might remove one option in the update manager, but I not backing from releasing 11.0 as intended.

That doesn't have to exclude the above mentioned option, does it? I'm all for the stable (whatever name it gets in the end) branch for developing things. But that doesn't mean that we cannot bring useful things over if they are "home grown" by GhostBSD (like the repo) and don't break anything. Or do I miss something here?

[ASX]

Since we're of course not building the whole ports tree - how are we going to do this, anyways? Will we have both the FreeBSD repo and our own with the later having a higher priority so that by default packages come from there but people can install software that we don't provide, too?

"My" idea was to provide a basic repository, that's why I requested users feedback about that, plus some amount of packages ... say a total of 10.000 pkgs, and no FreeBSD repository.

People wanting additional software should have to ask on forum or issue tracker, and we will include it.

EDIT: mixing two repository would be an error, just like mixing port and packages.

Or do I miss something here?

It is a matter of priorities, it is about what we think is more or less important, and clearly we disagree about something ... guess it is normal.

[ericbsd]

The ISO are build on quarterly and not on Latest to clarify, every build should be same, but for the part that is build from ports which I always had disagree, but I have let convbsd do this part that is an other story.

Wasn't that just written from you ?

I did not created GhostBSD ports repository it is convbsd, I did not have mush problem against that I a was in favour of building ports on top of the pkgs, but in that time I did not have choice to follow convbsd, sever and repository was not possible at that time. convbsd implemented that and nearly in the same time you started helping here.

[ericbsd]

I agree with you ASX on

not mixing FreeBSD and GhostBSD repository.

I did that on time and it is a mess.
It is better to start with a small repository the a big one I agree to that too.

When it come to the release of 11.0 loot of people a waiting for it and train is started, but the repository can be done tested during the development of 11.1 i prefer to take time to make the repository for 11.1 then rushing it for 11.0 or release 11.0 late like 10.3.

I know we will all disagree on that.

[ASX]

Eric, I'm not interested about discussing past choices, there were of course good reasons why things went that way.

What I'm interested now is about "to define" a path for the near future, possibly with some agreements, else also without agreements but however fully defined, so that we don't discuss things for itself.

You want to release gbsd 11 as is now, fine, state that clearly. Well actually you already stated that.
Am I allowed to say that I consider this an error ?

The new own repo can be introduced later of course, but I fail to see what improvement will bring gbsd-11:
as I see it gbsd-11 is "consolidating" a "pkg + ports mistake".

Of course we disagree.

i prefer to take time to make the repository for 11.1 then rushing it for 11.0

Again: who is putting that pressure on you ? other than yourself ?
Who said you can't release gbsd-11 a bit later, but for good reasons ?

[ericbsd]

Can it be done in 30 days?

[ASX]

Can it be done in 30 days?

The new repo ? probably yes, what I'm unsure about is listed below:

- build time at 250 pkgs x hour, will require 40 hours to build 10.000 packages
- additional 40 hours for i386 arch

a) network manager is already in fbsd ports (thanks)
b) the other softwares (update-station, ....) need to be included in ports (our ports-tree) and built together the other packages, not sure how much time it will be required: you could esitmated it better I think.
c) there is the need to add the repo to the webserver, guess one day will be enough.
d) some config file need to be tweaked to make use of our repo
e) ghostbsd-build need to be modified to fetch packages from our repo.

Guess that is all is required.