Mozilla

Thunderbird Proces of Change, Part 2

October 8th, 2007

As I noted in my last post, in late 2006 we started thinking very hard about creating a new organizational home for Thunderbird. A number of us came to the conclusion this was the best plan, including the Foundation Board and the key Thunderbird developers.

In early to mid 2007 the Foundation board designated two members (in addition to the work I was doing) to meet with the key Thunderbird developers to work on developing a plan together. In a series of meetings it was determined that a new Thunderbird organization would need an organizational leader — the developers couldn’t both develop and lead an organization. And as I have mentioned, there was a very strong interest in seeing Thunderbird vision expand and that clearly required additional people. A set of questions were raised but few answers were developed.

These meetings were not public for a couple of reasons. There are some key personal and personnel issues. I wanted to make sure that everyone already involved had a good opportunity to express their thoughts quietly, in a safe setting first. Second, we didn’t yet know what we were likely to do. I didn’t want people to worry we were adrift. I know people worried about this when we did start the public discussion. So either I was wrong to not start the public discussion much earlier, OR I was right to do a lot of ground work first and have an outline of how we would proceed before opening the discussion. Or I was *both* right and wrong, and there is no perfect solution.

We then spent some time thinking about who might be a good organizational leader for Thunderbird. The Thunderbird developers and I also spent time trying to think through how to create a new organization. By June or so it was clear that it was time to begin a public discussion about our goals for Thunderbird and desire to create a new home for it. I started working on this. Around the same time the Thunderbird developers came up with the suggestion that they create an independent company and we move Thunderbird development to it. The company would be interested in promoting the Mozilla mission, but would be outside of the Mozilla umbrella of organizations; a private company owned and run by the developers. We spent some time thinking through the pros and cons of this possibility, how the developers and a new company like this would interact with the source code and with the Thunderbird product. This became the third option included in my post opening the Thunderbird discussion. This option had the advantage of exciting the developers.

When we considered the idea of Thunderbird moving to a private independent company, a number of significant disadvantages emerged. First, this would mean that Mozilla as an organization was leaving the mail/ communications space and hoping that space would be filled by another organization. We weren’t ready to do that. Second, it felt like this would be moving Thunderbird to a more private space. We’re eager to see Thunderbird become an even more public project, with more contributors with greater authority. This concern is not remotely a reflection on the motivations of Scott and David. They have been devoted Mozilla participants for many years. This is a *structural* concern. It reflects the desire that Thunderbird — the product as well as the code — remain dedicated to the public benefit through the Foundation.

As is the Mozilla way, the public discussion of Thunderbird allowed us to move forward. Through this process we realized that David Ascher, long a part of the Mozilla project, was an excellent candidate for a Thunderbird / mail organization. I talk to David periodically about many Mozilla topics, and I think perhaps I realized he might be interested in leading the new organization even before he did 🙂 After this, the first step was to have David come to Mountain View (he lives in Vancouver) and meet the Thunderbird developers, followed by a bunch of other folks. I became convinced that David could lead the organization, and had the personality and technical chops to work with the Thunderbird developers. In particular to navigate between their phenomenal commitment to continuing to serve existing Thunderbird users and the Foundation’s goal of broadening the product vision.

I outlined my view of the future to David and Scott, the Thunderbird developers — David Ascher joins Mozilla to lead the new organizaton, David Bienvenu and Scott McGregor join the new organization, they continue to work on Thunderbird as they have been and participate in the efforts to broaden Thunderbird. David and Scott reiterated their interest in forming a private company but agreed that David seemed a good person to run an organization and a reasonable person to work with. We (meaning the Thunderbird developers and I) then spent time talking through how this arrangement could work.

These discusions involve nitty-gritty details like the status of the build system, ensuring that different versions of a product based on the Thunderbird code could be built. They also involve discussions of how modules owners like Scott and David would work within the new structure. Module ownership isn’t related to employment; it’s related to activity and the ability to lead and draw others to one’s work. So we expect the current developers to remain module owners for as long as they are active, interested, and leading a healthy community. We expect someone from Mozilla (eventually at MailCo) to make the final decision about when that code is ready to have the official Thunderbird name and be released as a product. There’s nothing new in this, but it’s always good to reiterate a shared understanding in a time of change.

In September I announced Mozilla’s plans for a new Thunderbird / mail organization. David Ascher is already at work getting to know the community, find out who is interested and able to help, and working through the details of setting up an effective organization. Scott and David are working on their plans as well. One of the fundamental aspects of Mozilla is that participation is not dependent on employment. Mozilla has phenomenal contributors who have never been employed to work on Mozilla. We have people who change jobs and remain equally involved with Mozilla. We have people who change jobs and later change their involvement with Mozilla, based on their preferences and their ability to lead others. Both Scott and David have stated their plans to continue their involvement with Thunderbird. It’s an unusual setting, and extremely powerful.

In the coming months a large set of Mozilla folks with be working on getting the new Thunderbird / mail organization organized and running, as well as serving Thunderbird users. We are also very eager to see Thunderbird become a broader product vision and to see a community with greater distributed authority. More people with the expertise and ability to authority to work deeply in the code will lead to a better product faster, and will spur the development of new experiments to improve mail. If you are driven to see Internet communications improve and can assist, please get in touch with David.

19 comments for “Thunderbird Proces of Change, Part 2”

  1. 1

    Jilles van Gurp said on October 8th, 2007 at 9:44 am:

    Basically how I interpret recent developments is that Scott and David don’t see their future in the mozilla organization or the new mailco. It’s pretty clear they are basically scaling back their commitment to pursue their careers elsewhere. It’s also pretty clear that they decided to do that in response to the proposed organizational change. I’m sure they’ll stay involved in some way (after all it’s their baby) but it does look a bit odd and is hardly the best start for a new organization if the two people who have been doing the bulk of the work for a very long time choose to be not a part of it.

  2. 2

    Abdulkadir Topal said on October 8th, 2007 at 1:48 pm:

    You know, it would be great to have posts like these *before* rumors are spread all over the news sites and the community is at a loss about what’s going on.

  3. 3

    C said on October 8th, 2007 at 2:31 pm:

    Thanks for this Mitchell. The community were in the dark especially after the laconic announcements from David and Scott. This came two or three days late, but at least, it came, and it’s clearer now, though after this uncertainty about the future of Thunderbird has been /.’ed and widely commented over the globe.

  4. 4

    C said on October 8th, 2007 at 2:31 pm:

    Thanks for this Mitchell. The community were in the dark especially after the laconic announcements from David and Scott. This came two or three days late, but at least, it came, and it’s clearer now, though after this uncertainty about the future of Thunderbird has been /.’ed and widely commented over the globe.

  5. 5

    C said on October 8th, 2007 at 2:31 pm:

    Thanks for this Mitchell. The community were in the dark especially after the laconic announcements from David and Scott. This came two or three days late, but at least, it came, and it’s clearer now, though after this uncertainty about the future of Thunderbird has been /.’ed and widely commented over the globe.

  6. 6

    Mitchell Baker said on October 8th, 2007 at 2:43 pm:

    Yes, I can see that it would have been better if I had gotten this written earlier. There’s so much going on in the Mozilla world that just keeping up is hard; staying ahead of everything is quite a challenge. I see this as a problem of success. Still a problem, of course, and I always hope to do better. But when we reach the point that everything is planned perferctly and there are no surprises then I will be worried for all new reasons 🙂

  7. 7

    Al Billings said on October 8th, 2007 at 4:50 pm:

    Mitchell can’t really speak for David and Scott and it probably wouldn’t have been appropriate (or as appropriate) to post this immediately after their announcements. If their announcements were overly terse, that may be something worth taking up with the two of them as it did not have to be done that way. That has little to do with Mitchell though.

    Personally, I do hope for the best for the two of them in their new company (assuming that they really are doing it since they haven’t said) and I look forward to continuing to see them contribute to Thunderbird as it moves onward.

  8. 8

    Daniel Glazman said on October 8th, 2007 at 11:42 pm:

    There is one point that’s still unresolved, and it’s not a minor one : who will train the new hires ? I hope there are tb contributors available for hire, but they will always be behind scott and david in terms of knowledge of the code. There’s a learning curve concern here.

  9. 9

    Ruediger Hof said on October 9th, 2007 at 3:30 am:

    I hope that Bienvenu und McGregor will fork the program. Bury Thunderbird.

  10. 10

    Martino Jocaio said on October 9th, 2007 at 4:21 am:

    I welcome the departure of Scott and David, as I hold them responsible for the fact that no real community has emerged around Thunderbird in the last 3 years after the Netscape departure. While there is no doubt that both are great developers, there is also no doubt that they aren’t great communicators and that has really been the main differentiator between Firefox and Thunderbird in the past as many Firefox devs have tried to keep the Firefox development process as open as possible and to communicate actively about it.

    IMO that is the main reason why Firefox has attracted such a huge community with so many outside developers, testers and people spreading the word.

    Therefore I for one applaud the decision of the Mozilla Foundation here and I really hope that David Ascher does a good job in selecting adequate developers with at least some communication skills.

  11. 11

    David Ascher said on October 9th, 2007 at 7:52 am:

    Daniel —

    Great question. I agree that there’s a significant learning curve. I’m hoping that Scott and David and Mozilla will figure out a working relationship even if they’re no longer employees, that will make it worthwhile for them to continue to assist as they have for many years. In addition, there are other people who have contributed to the mailnews codebase over the years, several of whom seem interested in helping me (and via me, Thunderbird) out.

  12. 12

    Amsterdammer said on October 9th, 2007 at 8:53 am:

    Martino, this is very unjust against Scott & David to compare the possible lack of communication with the success of Firefox!
    The power of Firefox is based on a colossal marketing machinery, which I missed for Thunderbird (in a smaller design).
    After http://www.spreadfirefox.com/ u see a powerful team with big money, after http://www.spreadthunderbird.com/ u see a measly redirect.

  13. 13

    Martino Jocaio said on October 9th, 2007 at 11:09 am:

    Amsterdammer, you make the same mistake that many other people do by attributing the huge success of Firefox solely to its Marketing. What you and many others seem to forget is, that Firefox (and its predecessors) was already very popular before it was heavily marketed and that active communication is at (least partly) a form of marketing that was totally missing from the Thunderbird devs.

    But you don’t have to look at Firefox. Just take a look at SeaMonkey or the Calendar project. Both were often pronounced dead in the last few years, but both have gained a large supporting community mainly because their front people actively communicate via blogs, forums, newsgroups or conferences.

    That is a major item that needs to change in Thunderbird-land for it to gain more traction in the consumer and enterprise market

  14. 14

    Eyal Rozenberg said on October 9th, 2007 at 12:15 pm:

    You’ve been killing Thunderbird / Mail&News.

    You want to cut off everything that’s not firefox from the googlebucks.

    You make up flimsy excuses about email not being important or the web exploding in 2005/6. The only thing exploding was your coffers maybe.

    And now instead of 2 to 150 or so TB to FF devels we have 0.

    Say hi to Larry and Sergeyi for us.

  15. 15

    wjl (Wolfgang Lonien) said on October 9th, 2007 at 3:55 pm:

    Mitchell,

    I use a free rebranded Thunderbird – known also as Iceweasel – as my main email client since I discovered it. It is not perfect, but the closest thing to perfect I’ve found. It is small enough, yet modular so I can use extensions like Enigmail or even Calendar if I would need that. The best thing about Tunderbird/Iceweasel is that it is no bloatware.

    Please don’t change it. If you want an Outlook killer for the Executives, then develop a server of equal quality like Thunderbird.

    What am I missing? Openness, maybe, in the discussions. You said:

    “So either I was wrong to not start the public discussion much earlier, OR I was right to do a lot of ground work first and have an outline of how we would proceed before opening the discussion. Or I was *both* right and wrong, and there is no perfect solution.”

    In developer space, we say “release early, release often”, because only openness leads to success and to quality. Why didn’t you let us hear the discussions? Where is the mailing list where we could follow your opinions – plus those of your lead developers?

    It’s all about trust in the free software world. There is nothing like “users”; forget about that term. A so-called “user” could be your best QA officer, or have the one idea that will earn you billions later.

    So please tell us: where can we get the thoughts of Scott and David?

    cheers,
    wjl
    http://wolfgang.lonien.de/
    http://blog.thedebianuser.org/

  16. 16

    Peter Lawson said on October 9th, 2007 at 10:48 pm:

    I would like to see some convergence in the project. We currently have SeaMonkey, TB version 1.5, TB version 2.0, and Penelope, all being developed in parallel. Let’s settle for one good, comprehensive product.

  17. 17

    Daniel Glazman said on October 10th, 2007 at 2:05 am:

    @David : uuuuh ? I don’t understand. So you really think that if Scott and David work for another company or start their own shell, they will have time to invest to be hand-in-hand with new hires ? Mentoring only one person on such a complex code will require a lot of time. Hum, to say the least.

    @Martino : it would be laughable if it was not so serious and unfair. We all remember that Firefox did not create a community in the beginning because checkins in mozilla/browser were limited to the team of 4 or 5. We all remember that major features were unexpectedly (JS console) removed without discussion with the community. And as the author of Nvu, I know too well how hard it is to build a community of core contributors on a complex code.
    Thunderbird has a lot of extensions and support. It also has Milimail (milimail.org), contributors to the core.
    Firefox has a wider community because it’s a BROWSER ! Because any help you provide is visible to litteraly a hundred million people. Because the internals of a Mail User Agent are terribly complex (just for the record, I implemented one in a former professionnal life). Because when you’re only 2 or 3 on a product like thunderbird, you unfortunately don’t have time to mentor new people ; happened to me with Nvu exactly.

  18. 18

    Mitchell Baker said on October 10th, 2007 at 1:01 pm:

    Daniel

    Do we hope Scott and David remain involved with Thunderbird? Yes, of course we do. Do we think they will? They have said they want to. Is it their choice? Absolutely. If they want to make a private company and try something new, it is absolutely their choice and I intend to try to be supportive. Are we willing to stand still and leave Thunderbird as it is, to stay with the status quo? No. Do we wish there were more developers who had authority in the Thunderbird code already? Yes, we do, that’s an explicit goal for the future. Do we treat the need to build a strong Thunderbird community with more great people involved as a serious and important task? Absolutely, this is part of how we will measure success.

    David and Scott have always been devoted to the Thunderbird userbase; I have every expectation that this will continue. Does that mean they will help train other people? That’s their decision, it’s not something we can force to happen. I suppose that helping people get up to speed on the code and able to contribute is a part of being a good module owner. That’s a discussion that’s broader than Thunderbird though and probably should be addressed elsewhere.

    In any case, there’s a lot of work to do in the mail space, that’s for sure.

  19. 19

    Old Sarge said on October 11th, 2007 at 7:53 am:

    Tell the Truth. You simply decided to kill Thunderbird because it is a desktop client and it has no place in SaaS model. You have already stated that unless the application is a browser extension it doesn’t fit into the Mozilla vision.

    Call me a skeptic or call me a cynic, I don’t care. The Mozilla goal is to hitch its wagon to Google and join the SaaS movement. The world doesn’t need another greedy M$. I do not need nor will I be convinced that it is in my best interest to allow a vendor of desktop software or web-based applications delivered via a browser to dictate what is best for me.

    You should have seen Thunderbird as the best opportunity to provide the eMail/News module of the best open source desktop suite of applications in the world. How much did M$ pay Mozilla to kill Thunderbird? (I use it because it is safer than Outlook.) How much did Google bribe you to drive a stake into it? (Some people don’t have the latest hardware to make webMail a viable alternative.)

    I wish David and Scott the best. They didn’t desert. they were betrayed.

Skip past the sidebar