[OpenIndiana-discuss] Merging OI + OmniOS? (And OpenZFS vs ZFS)

Wed Dec 30 18:10:14 UTC 2020

 I have a problem with my mail list settings, so I write one single reply, adressing three different persons. I am a newbie to the OpenSolarish scene, so I dont know much about it, or who the people we should talk to, though. :)

@Aurelien,Yes, I was quite certain I was not alone in seeing the disadvantages of having separate small teams instead of one single larger team. It is good that you people already had this discussion. Maybe it is time to revive your old discussion again?
It was not diplomatic to say that OI was inferior, but now some additional water has passed under the bridge and you all have had some time to work and deliver your own product. I think that if people were more diplomatic it would benefit OpenSolarish, and the world as a whole. We dont want a cursing and yelling culture as in some other OSes no? 
I dont understand your statement of no vanilla illumos anymore. Is OmniOS not illumos, but OI is? In that case, that is sad, yes. But in the greater scheme of things, it would be better for OpenSolarish community even if we have to give up vanilla illumos. If we merge OI and OmniOS, it is not irrecoverable, it will be possible later to try to tweak our new distro to a more vanilla illumos distro. Or, if someone in the future, did the additional minor work to tweak the new merged distro to vanilla illumos - that would be much less work than creating OI today, I guess?

I bet many more than I, would be very excited in one unified illumos distro where we merge OmniOS and OI. I understand there is lot of prestige involved, but if we dont do want OpenSolaris to diminish and die to Linux, we all maybe should put aside the prestige? :)

@Andy,Regarding who will do all the work. Realistically, it will be you and the other OI devs who will do all the work. I bet several other people would love to help, but not all of us can do what you can do. I want to help though. Maybe help to coordinate, and help discuss with the OmniOS team? And write manuals? Of course, if we can merge OI and OmniOS there will be lot of work for you guys in the beginning. But in the long run your work will be less as there are more developers involved. And as you guys already have discussed this earlier, maybe it is time to rethink this old train of thought of yours, again? You know the process better now and have a shipped product, so you do stand on your own legs. A large team of talented developers is better than a small team? This would also cause some attention in the FOSS world, and the OpenSolarish community. More people will be exposed to your work, and benefit from it. There are already lot of work done by OmniOS, you and the OI devs could leverage their work? What is your thought of working with a new set of talented developers creating something new together that we all could benefit from? Is there too much prestige involved for this to be realistic? 

@John Groenveld,
I had this problem that Solaris 11.3 could not import an Linux OpenZFS zpool. I created the zpool from Solaris 11.3 using zpool v28 and ZFS v5. Then I imported that zpool into Ubuntu 2020.10 with OpenZFS v0.8.4, and I did OpenZFS send and receive data to the zpool. And then Solaris v11.3 can not import the zpool again, as it is UNAVAIL. Have you tried this, and could import a zpool between Linux and Solaris? I guess MacOS also has this same problem, as it also renames the disks? And Windows OpenZFS version renames the disks too?

    On Wednesday, December 30, 2020, 03:05:55 PM GMT+1, Aurélien Larcher <aurelien.larcher at gmail.com> wrote:  

 This is something that was discussed in the past when OmniOSCE was created but back then there was a focus on providing continuity of releases and too few people: the goal for to have a true reference distribution.
Back then Adam and I had suggested to merge the distributions given the size of our respective communities: use OmniOS as a core OS with their release schedule and use oi-userland to provide extra packages but there was no interest.
Tobias suggested to me that OI being an inferior distribution I should better use my time and provide X11 packages to OmniOS instead. Obviously his comment did not motivate me much :)
I think merging both makes sense to gain momentum and stopping reinventing the wheel given the number of developers (us pulling from OmniOS and Dominik packaging stuff that we already have).
But then it means that there is no illumos distribution actually running vanilla illumos which is a bit sad.
We've had good interactions on compilers and pkg so that's at least a great start.
Also this may be an overhead for Andy and he should have a good idea if it is doable, if this is something that would be positive for him and what would be the milestones. 

On Wed, Dec 30, 2020 at 12:39 PM Kalle Anka via openindiana-discuss <openindiana-discuss at openindiana.org> wrote:

A question. I am sure this has been discussed before, but maybe it is time to update the discussion? :)
As I have understood it, OmniOS is a robust server OS with no good desktop environment. It is possible to install X11, though:https://geekblood.wordpress.com/2017/10/26/installing-x11-and-a-desktop-environment-on-omnios/
Would it be possible to merge OI and OmniOS? Say, try to transform OI into a desktop environment ontop OmniOS? Both OSes share lot of common ground, so technically it would not be impossible to for instance port MATE to OmniOS? 
OpenSolarish is a bit fragmented as of now. There are several developers working on their own stuff, with different IPS repos, etc. If we could unite some of the work, OpenSolarish would benefit as a whole. 
OmniOS would get new developers, a good desktop environment, and also lot of users coming from OI. 
OI would benefit from all the work the OmniOS developers are doing in getting a stable and robust server OS, so it would free up OI developers to more quickly advance desktop environment. Lot of synergy effects and win-win. OmniOS does the server backend, and OI do the desktop frontend. This lessens the burden for the developers. The more developers, the better.
The situation is similar to two competing camps developing ReactOS - the open sourced WindowsXP clone. Why not unite the two teams instead? It would be much less work for the OI developers if we could build on the excellent work from OmniOS team, and vice versa. Remember, we all love OpenSolarish and it would be better to have one single strong distro, than several competing distros. FreeBSD has only one distro and that is better than the Linux fragmentation mess today. There are only so many OpenSolarish developers, and our community would benefit from one single strong distro.
Or, have I missed something? Like, for instance, do the illumos community also suffer from "Not Invented Here" syndrome like some other OSes? The only truly innovative OS is Solaris, and lets keep it that way by consolidating and uniting our teams so we can make faster progress? I am interested in trying to help the community too. Because Solaris is the best OS! :)

BTW. Earlier I reported that OpenZFS renders ZFS disks unusable: If you import a zpool v28 and ZFS v5, into Linux using OpenZFS, then Solaris 11.3 cannot import the zpool again. Someone suggested that Linux OpenZFS renames the disks as "/dev/sda" instead of "/dev/c0t0d0" in the zpool. So this problem could maybe be fixed by importing and exporting the zpool by a illumos distro which would automatically rename the "/dev/sda" entries to "/dev/c0t0d0" back which means Solaris v11.3 could import the zpool. I have not tested this yet, but it sounds plausible?
_______________________________________________
openindiana-discuss mailing list
openindiana-discuss at openindiana.org
https://openindiana.org/mailman/listinfo/openindiana-discuss

-- 
---
Praise the Caffeine embeddings