OGB Election: Bio & Positions

Posted on March 10, 2008

Here we are again, its election time in the OpenSolaris community. I’ve been nominated again and the playing field is larger this go around, so I’m going to provide a bio and some positions here so that voters know what their voting for. Historically I’ve not been a big campaigner, I don’t ask anyone for their vote, the OGB is a representative body, and as such you should vote for those persons that you feel most adequately reflect your own position, that may be me, that may not, choose wisely. That said, after this post I’m going to stop writing about politics so much and instead try to rekindle the technical roots of this blog, because thats generally a lot more interesting.

I’m currently the Director of Systems at Joyent, Inc, the largest customer facing production Nevada installation in the world. I currently manage several hundred systems all running Nevada, as such all our “support” is done through community process not a contract with Sun. Previously I’ve worked for Homestead, Inc where I managed their Sun Midrange database systems and the various storage systems, I worked for MCI Systemhouse in a call center for a while doing corporate UNIX support, and I spent several years at Taos Mountain Inc here in the Silicon Valley where I worked for a long list of enterprise customers including Fujitsu, Clarify, and Sun in 3 different locations, including the iPlanet migration from Netscape and MPK17 to Santa Clara, I worked as an admin on “The Ranch” in Sunnyvale (E4500/A5200 heaven!!!) prior to the move into the new datacenter, and I managed 3 of the QA labs on the 3rd floor of MPK17. I’ve been around, spending several years in the enterprise space and then the last 1.5 years at Joyent in the very different world of startups.

I’ve been a part of OpenSolaris since the beginning, round Sept/Oct of 2004 when the pilot phase was spinning up. I came on board after introducing myself to Jim Grisanzio and begging a little, which didn’t turn out to be required, I then proceeded to bring other administrators in from the #solaris IRC group and other places that Solaris admins congregated. I was nominated for the initial CAB during the pilot and was 3rd in line behind Rich and Al, nominated again for the OGB but came in just short again, I’m currently serving on the board because of the resignation of James Carlson. During the CAB days I worked with the board on the drafts of the Constitution and established the Genunix Wiki for that initial purpose. I’ve participated in a wide variety of aspects with regard to OpenSolaris, have attended every major conference possible since 2005 to represent community involvement in OpenSolaris, etc, etc. I’ve been around a while, I’ve known the players, and I thin I’ve done my best to be part of the history of OpenSolaris to date. Blah blah blah.

As for my positions… there is a long history of my positions on things in this blog, in the press, and on various lists. I’m not really timid or quiet and its probly fair to say I’m the most vocal non-Sun participant in OpenSolaris. This extends back, actually, to matters regarding Solaris and Sun prior to OpenSolaris but I didn’t have a microphone quite as large.

A good place to start is with my recent Governance talk at the Silicon Valley OpenSolaris Users Group. Its a long presentation and largely background but will give you a better idea of where things are at with the board.

I’ll list out my positions on some controversial topics…

  • GPLv3: This issue kinda came and went, but some other candidates are mentioning it, so I’ll follow suit. I do not oppose GPLv3 outright. I do, however, oppose GPLv3 dual licensing just to make a splash or get some coverage. OpenSolaris is a very big rock and throwing it into the GPLv3 world would make some very big waves and there are many potentially excellent outcomes. I have opposed GPLv2 out of fears that large chunks of code would be raped out of Nevada, but it should be said that the reality of “stealing” ZFS or DTrace for Linux sounds a lot simpler than it is and doing so would certainly not take Sun developers out of the loop. If the decision to go GPLv3 is done slowly and cautiously with full community support it could be a good thing to do.
  • The Sun Contributor Agreement (SCA): I largely support the use of the SCA. Contributor Agreements are not uncommon (I discuss this as part of my SVOSUG talk) and ours is pretty tame. I do however support alternatives or limits on the SCA, that is, I don’t think that a every type of contribution should call for an SCA. Managing contributions with a valid SCA Registration Number is excessive and beaurocratic overhead that should be limited whenever possible, nevertheless, there is a valid legal reason for Sun to require it and I don’t challenge its use.
  • The Constitution: I support the Constitution and our ASF heritage. I am frustrated in our inability to fully implement the Constitution currently and believe that the Constitution should succeed or fail based on its own merit after being fully and properly implemented, which to this day has not occurred. During my time on the board I’ve started to address these issues but simply ran out of time with only about 2 months to really get it done. We currently are debating our future but we haven’t fully accepted our present and to change without really putting an effort into our current form is simply wrong.
  • Parliamentary Order (PO): I’ve been criticized for being overly formal and at times the “bureaucrat” shirt has almost been handed to me. During the drafting of the Constitution I spent a lot of time learning and researching PO and come to believe that when properly utilized it ensures the rights of individuals and expedites balanced debate moving issues to resolution quickly and fairly. Especially in a Con-Call environment, its hard to ensure that everyone has a fair opportunity to share and contribute. I have little sympathy for people who disregard PO simply because they feel silly saying “Mr. Chairman”. And, besides, its mandated by the Constitution regardless of what anyone thinks. I hope that in the next year we provide the board with a very strong and skilled Chairman.
  • Representation: I believe that the OGB represents the community and should act based on what “the people” want. I put forth a motion to make the “OGB-Private” mailing list publicly viewable, not necessarily because I think a private list is a bad thing but because I know there are those in the community who don’t like it, and I wanted to make their voices heard. The most important role of any board member is to make the community feel heard.
  • Open Development: I’ve tried over the last year to make a distinction between “Open Source” and “Open Development”. Efforts to continually open development are on going and we can not thank individuals like Mike Kupfer, Dr. Stephen Hahn, and many others, for working so diligently behind the scenes to make this happen. But some times we focus too much on ON, we need to look at some really great success stories like IPS or the Documentation Community Group, efforts which are truly inclusive, transparent and stand as models to the rest of our community. ON is the crown jewel, but its also the most complex element to really open up development on widely, lets not put too much energy into that front that we ignore the rest of our community.
  • Scope: Times are changing. Our community has grown beyond its origins, now including Open HA Cluster (“Sun Cluster”), elements of SAMFS/QFS, the StorageTek Availability Suite (AVS), etc, etc. The board much take a scope much larger than just ON or Indiana or whatever… we need to look to a growing community which addresses a wide range of interests, needs, and diverse persons. One size doesn’t fit all.
  • Industry Involvement: I believe that we need to provide more established and transparent lines of communication between industry community elements and Sun as it pertains to OpenSolaris. I advocate the creation of an Advisory Board in which industry players, commercial and non-, can collaborate. This includes entities like Nexenta, Intel, Joyent, Dell, Blastwave, 4Front, etc. Many organizations invest time, money and effort into OpenSolaris and we need to foster, encourage, and reward that commitment to our community.
  • Quality Control & the ARC Model: I do not advocate moving the core code repository of “Solaris” outside of the Sun firewall, at least not based on our current trajectory (GPLv3 could substantially change that). The ARC model is tried and true and to rip it apart would be costly and dangerous for Sun. However, it is far to complex and tedious for the average developer and the existing Sponsor model just sucks. The sponsor model has been described as a “stop gap” or “bridge” and I agree with that position. Average development (bugfixes, minor features, minor improvements, etc) should not be restrained by these types of concerns, if Sun wants to do that dance its their problem, not the developers. With new SCM’s such as Mercurial, however, we have some new and existing possibilities, such as the ability to provide a central hub for developers to share Hg changesets and WebRevs, encourage branching and experimentation, etc.
  • Inclusion: I believe strongly in inclusion, bringing as many people in from as many places as possible. But this requires a strong structure or things just turn into chaos. By properly implementing the Constitutions CG/Project structure, and restructuring the existing CG’s and Projects to fit the appropriate model, that we can make life smoother for everyone and become much more scalable. The OGB should never discuss whether or not we will allow something to happen… rather, the OGB should direct it to the appropriate place. We should always encourage new work, no matter how crazy or contrary, and encourage it to grow and success. For instance, there is a small effort to adopt rPath’s Conary build/packaging system. This is opposed by some as pointless given that Indiana has already decided on IPS. I believe that despite this, those advocating Conary should be given a Project to work on it, grow it, and lets just see where it goes. The minute we say “No” to community work we start down a very dangerous and destructive path.
  • The OGB: Here’s the big one, please see my presentation noted above for a full explanation. I do not believe that the current OGB is appropriate. Historically the CAB/OGB has struggled to understand its purpose, this is solidified by the inability to fully adopt and implement the Constitution. The most pressing need has been to restructure the existing CG’s and Projects… but thats so controversial its inevitably tabled and forgotten. The OGB doesn’t have money, control over development, control over the trademarks or IP… there is nothing to govern! Furthermore, the OGB currently has as its primary task communication with Sun, however Sun has recently made decisions without the community (such as the Trademark debacle). Now… we can fight this, but we don’t have a lot of ammo, so thats a fight we’re just gonna loose. Or, we can instead try to be constructive and seek a solution in which Sun proper plays a much greater roll in OpenSolaris governance in general, so that Sun and the community are working together rather than sending silly responses back and forth. Why not change the constancy of the OGB to include Simon Phipps, Ian Murdock, and Bill Franklin? It would certainly be more constructive and productive that what we have now. I think that with a model like this we could be even stronger than ever and far more unified that we currently are.

I could, and usually do, go on much more, but you get the idea. There is a lot for the community to do and a lot of opportunity before us. We’ve had some tremendous success and we need to foster and encourage that, while building toward something even better. Lets not worry about where we’ll be in 2 months or 6 months, but rather where we’ll be in 1 or 2 years. As Sun puts weight behind OpenSolaris via Indiana we’ll see OpenSolaris become more and more integral to Sun’s success and we need to prepare for that new reality when the commercial product and the Open Source community aren’t so distant.

In closing, as I said before, the OGB should represent you! When you look at candidates I want you to ask yourself “Who would be better in a discussion with Jonathan Schwartz… them or me?” Being a board member is extremely time consuming and frustrating, and thats effort put forth with the hopes that you, the developer, don’t have to! So choose wisely in people that you trust, believe in, and believe that represent you. Engage board members, hold them accountable, and let your voice be heard.