[illumos-Developer] OpenIndiana and illumos, part 2

[Alasdair Lumsden]

On 18 Nov 2010, at 20:42, Garrett D'Amore wrote:

> Thanks for helping to make my goals clear, Albert.
> 
> Let me elucidate a bit more.
> 
> I'm specifically not interested in distro that attempts to rehash what
> Oracle is doing, or one who's sole benefit is a more liberal license
> than what Oracle provides.  For a variety of technical reasons, I don't
> even think such a distribution is *possible* -- at least not without the
> complete source code (and the right to use it!) for all of Oracle's
> Solaris code.

Okay. Please keep in mind things have changed significantly since both OI and Illumos were launched.

When we first started talking, Illumos (from my perspective) was about:

1. Providing independence from Oracle, and safety from uncertainty for downstream distributions such as Nexenta...
2. ... whilst still working with Oracle for the greater good of all; merging ONNV commits (which were still flowing freely into the public domain) into Illumos, whilst making Illumos changes available for Oracle to integrate if they felt like it
3. Replacing closed source components with open equivalents
4. Innovating and extending on top

Oracle at this point had discontinued the binary builds of OpenSolaris, so OpenIndiana was founded on the principle of building the open source components Oracle released, whilst extending on top, and providing end-users with a nice frequently updated development branch, as well as a very useful and sorely lacking stable branch with security and bug fixes.

Hopefully that is an accurate picture of how things started.

However, since that date, Oracle closed the ONNV gate, decided to make all future Solaris releases a paid-for operating system, and Illumos is now a fully fledged fork. The goals of Illumos seem to now be:

1. Total independance from Oracle, Illumos to now diverge completely from the ONNV code base
2. Assume no future ONNV code drops will happen
3. Innovate and extend
4. Rapid development with major subsystem reworks

OpenIndiana was never meant to completely follow Solaris, the intention from the beginning was to extend things - provide newer versions of packages, provide far far wider package availability (from the fundamentals such as postfix, postgresql, exim etc all the way through to the exotic, such as ffmpeg, vlc, etc), rework subsystems as we saw fit. But the initial goals were fairly modest; build what Oracle provides, extend on top. Why? Because the project started out with a handful of people and it was a realistic, achievable goal.

Keep in mind that there are 2500+ packages comprising OpenIndiana, all of which have to be kept up to date. It's not feasible for us to do this without 30+ people. Heck, it's not even feasible for Nexenta to do it, hence why you use the Ubuntu/Debian packages. So, I hope you'll agree with me, it makes sense that OI continue to leverage the source code drops from SFW, JDS, pkg, etc made by Oracle. Do you at least agree with this part?

Where the misunderstanding seems to come in is that I think you're assuming leveraging Oracle's continuing code drops means that OpenIndiana is not going to innovate, or that we're going to be closed to the idea of reworking major subsystems, or that we have no intentions of gaining independence from Oracle. These assumptions are wrong. Just to be sure we are on the same page, let me clarify what the goals are:

1. Leverage what Oracle chucks over the fence to save us a huge amount of really tedious work keeping FOSS software up to date which would suck up so much time nobody would be able to do anything else. Stopping doing this for a ideological or emotional reasons would be stupid.

2. Despite 1, most certainly have a goal of total independence from Oracle. This involves building a large software packaging community. Guido has been working hard on this goal but communities don't spring up over night. Getting one going requires a phenomenal amount of work; diplomacy, politeness, socialising, fostering an enjoyable, fun and exciting environment. Or, find a sponsor willing to pay people to do it. As the developer community grows, we can start maintaining more and more software ourselves and eventually start to retire Oracle built consolidations completely.

3. Rework, innovate or modify the OS as we see fit, in line with what our userbase most likely wants. For example, providing KDE as a first class citizen in the OS, or replacing the completely retarded 1980s wizard that asks you if you're on a subnet when you zlogin -C into a zone for the first time, or whatever. We are completely open to innovation and change, assuming the changes are beneficial/useful and are not, for example, top-down dictats nobody else finds useful for ideological/personal reasons.

> I *do* need a distribution that is more closely aligned with the goals
> of illumos, which includes significant innovation.  Today there is
> nothing filling that gap properly.  I'm happy to have OpenIndiana be
> that distribution, but some changes in the way OI runs itself (and more
> specifically in what OI drives for) are required to get there.  If I see
> that OI is interested in filling that role and willing to make the
> necessary changes, then I'm willing to invest resources to help.

I don't see that as a problem, in principle. However you haven't really clarified the goals of Illumos. For example, what if you want to rip IPS out of the distro and make it SVR4 based. That for example, would not fly. However I'm pretty sure most of the changes you're after are going to be reasonable, but it would be helpful if you could lay them out for all to see.

I know you're doing things like ripping perl 5.8.4 out - all I can say is yippiee, about time! This is a beneficial change and OpenIndiana will do whatever work is necessary to integrate the change into the OS so that end-users don't get a nasty surprise when they upgrade.

If there are changes you'd like made in OI to help Illumos, then RFE them in the bug tracker, and pick an appropriate severity/urgency level, and we'll see what we can do. If nobody is able to work on it for some reason, we'd welcome the changes being provided by people primarily working on Illumos, and we'd integrate them.

> However, if OI wants to remain truly independent of illumos, and still
> wants to aim for compatibility with Oracle Solaris, or OpenSolaris, as a
> primary goal, even at the cost of innovation in illumos, then I don't
> think OI can fill the needs of illumos.  In which case I need to invest
> elsewhere, even if it first seems like a duplication of effort.

We absolutely do not want to prevent Illumos innovating. Or more positively, OpenIndiana wants to innovate, and be the best Solaris-kernel based distribution available to use. We want a far bigger installed base than Oracle Solaris, and we want to challenge Linux in the mass market. We see Illumos as a key driver to help OI achieve our goals, and hopefully our goals tie in with getting Illumos out to the widest audience possible.

> Let me spell this out a bit more clearly.  There are large financial
> motivators behind illumos, and a number of us have staked not just our
> spare time but our entire careers and in some cases our businesses on
> this bet.  So, we're going to see our goals through, one way or the
> other.  The question is whether OI is going to come along for the ride,
> or be left behind.

We understand, and a lot of enterprises want OpenIndiana to succeed too. I have no doubt that there are significant commercial opportunities open to us all should Illumos+OpenIndiana succeed. It would be refreshing to challenge Oracle and RHEL at their own game by providing a superior product with more responsive and better support.

So count us in.

> If that sounds a bit melodramatic, then I'm sorry, but I do think its
> important the folks involved in deciding the future of OI understand
> what's at stake from illumos' perspective.  This is far more than a part
> time hobby thing.

Again we absolutely understand and we very much want to work with you. I'm sorry communication hasn't been great, but just reaching the current state of play has required a large effort, and unfortunately a lot of the people involved have full time jobs and wives/children/mortgages/etc.

Progress on OpenIndiana has been slow, and this needs to change for the project to thrive and for it to suit the needs of Illumos. I hope that with your help we can start to attract the necessary people to gain some much needed momentum. I know Nexenta have hired Albert Lee and I am hoping this will be a beneficial move for both projects, and help us to all work together.

OpenIndiana is an open project and we are open for anyone to get involved, including commercial backers. We aim to set up developer and admin councils similar to Illumos when we have sufficient numbers to do so (I tried a little while ago but there simply wasn't the numbers to make it a sensible thing to do). OpenIndiana intends to be a democracy, not a top-down dictatorship, meaning your vote will count equally to mine.

I have mentioned that my business has also hired an OpenIndiana developer, this is Andrzej Szeszo. To begin with his primary duties will be related to EveryCity's cloud infrastructure, but he will spend at least some of his time working on OpenIndiana release engineering, and automating the OI build systems. Once we get regular builds going automatically, development speed will pick up significantly. My business is also growing and I fully intend to continue to commit resources to OpenIndiana as I view it as a long term strategic investment, in a similar way Nexenta views Illumos as a strategic investment.

I honestly don't think our goals are that dissimilar; I think perhaps the biggest source of frustration for you is the lack of communication and the glacially slow pace OpenIndiana has been moving at over the past month. These can both be fixed.

Alasdair.