While doing various kinds of marketing research around Mozilla development, I’ve noticed a disturbing trend, which is probably well-known to most of you: Mozillians Earning A Living Somehow (MEALS) often seem to resort to code forks. In the mailnews area, we have Spicebird and Postbox. I’m less familiar with the browser area, but Flock is a similar example. This post from lilmatt describes some of the issues for Flock, also discussed by Daniel Glazman. I was particularly intrigued by lilmatt’s comment:
“If, as an example, Flock were to be implemented as an extension and attempted to say, overwrite the affiliate tags for the search box in the chrome with it’s own to redirect revenue, I think they’d be vilified and perhaps even blocked”
This reminds me of an issue that affects microenterprise loans in developing countries, as pioneered by the Grameen Bank. Sustainability is the holy grail of any microenterprise fund – but the interest rates that must be charged for sustainability are shocking to many people, usually near 40% per year.
Sometimes a well-meaning organization will go into an area, and setup a non-sustainable microenterpise program that charges “reasonable” rates, say 10% per year, and is understanding and forgiving when their poor borrowers can’t repay the loans. That program operates for awhile, but inevitably fails. After that happens, the microenterprise community has learned that it is impossible for sustainable microenterprise operations to function in the same region, because they are “vilified” for charging “exorbitant” rates for their loans. Or the borrowers have come to believe, due to the forgiving nature of the well-intentioned organization, that loans are really gifts. This cultural pollution that was inflicted on the region by that well-intentioned organization persists for an entire generation, during which time no real, sustainable loan program can be established, and the local region is deprived of the benefits of microenterprise loan programs. (OK, this is a simplification, but I hope you get my drift.)
I’m afraid we have this kind of cultural pollution in FOSS . As I look at vertical markets, often there are virtually identical applications in both the open-source and commercial spaces. The cultural expectation is that the addon for the FOSS application (like Thunderbird) is free-as-in-beer, but people are perfectly accepting of $10 – $40 charges in the Apple or Microsoft space. You get the sense that some people would be horrified if a Thunderbird addon carried a charge.
But what that means is that the Thunderbird addons are virtually impossible to make sustainable as businesses. So MEALS (Mozzilians Earning A Living Somehow) resort to forks, and we lose the benefit of their future efforts. In the fork, they can try to absorb the revenue streams that Mozilla relies on, or reset the cultural environment to be more in alignment with a commercial model.
What a pity. If we could figure out a way to all work together, then the net effect would be so much more powerful. If there was money to be made, we’d have a much more powerful product to present – and the total pie to share would be so much bigger. Yet I fear that the cultural pollution has already occurred, and any attempt to change the mindset will just fail.
KNIFE anyone? (Kent Nicely Introduces Free Extensions, thereby stabbing himself and the rest of you in the back. But please try my extensions!)
rkent
P. S. As an attempt to put my money where my mouth is, I paid $1.50 at istockphoto for the image above. The process took me at least an hour to figure out, after reading all of the legal language I feel like I must be a criminal, and I’m still not sure that I didn’t do something wrong that won’t cost me my entire retirement savings in liability. There has got to be a friendlier way to get micropayments!
Or maybe it’s simpler than that; upstreaming things is _hard_. We* don’t directly compete with Mozilla, we go after completely different revenue sources, and so at this point financially we don’t really care about holding back our XR patches. Actually, in general the managers are happy to see us upstreaming things. Unfortunately, the schedules (MoCo/us) never match, so patches and reviews linger in bugzilla and ignored. In general only patches Firefox would care about get attention.
As far as I understand it, the current policy at MoCo (who hire so many of the core people they’re having problems with getting people to do the whole granting commit privileges thing) is that the focus should be Firefox and Firefox only. They might be right – I don’t know, I’m not smart/experienced enough to tell. But the obvious effect is that people doing weird crap they don’t care about suffer.
Basically, getting patches in? Hard. Have a fork? Easy. Firefox has a fork of its own, it’s just the canonical one
Oh, and also, sometimes we’re too dumb to figure out patching things in an upstreamable way.
(I’m thinking of one case I’ve done in particular – _I_ wouldn’t want to see it upstream, but I can’t figure out how to make the change cleanly…)
I’ve always wondered though – what happens to Thunderbird when it needs a patch to the core parts of a branch Firefox already released on? Do you get relaxed checkin privileges, or is it security-fixes-only as the policy? Or is it usually synced to Firefox enough to not matter? (i.e. difference between the two is never greater than a few months)
*Sb, though obviously (I hope!) I’m not speaking for my employer / whole organization / guy who sits next to me at work
“what happens to Thunderbird when it needs a patch to the core parts of a branch Firefox already released on”
As I understand it, TB relies mostly on security patches for the core code. So in a typical cycle, FF releases a new version, and a little later TB will release with the same security patch. But there is enough of a delay that it scares the TB people, and so I think that partly played a role in the decision to disable javascript in email HTML. The core TB code in the mail and mailnews directories is separate enough from FF that the vast majority of TB development is unlinked from what happens in FF. That isn’t true for SeaMonkey however, so I think they really struggle to coordinate the timing of their releases with both the TB and FF teams.
But TB of course has a little bit of clout in getting core patches pushed through. I can see how a separate application, that was not considered part of MoCO, would really struggle to get things in that only they needed.
I don’t know whether this is a coincidence but I was contacted by an employee of a search engine company today – with the proposal to add a post-install checkbox that will add their search engine to the list and change the homepage (they are willing to pay of course). I guess that Mozilla is too strict with their policy of only doing things that are in user’s best interest (or maybe just too expensive) and they decided to try it with extension authors. Of course I responded that I don’t want to screw my users – but will other extension authors do the same? Or will we soon have a similar situation as with the freeware applications that will all install you some trashy toolbar unless you are careful enough to opt out? I find this scary.
Re: “I guess that Mozilla is too strict with their policy of only doing things that are in user’s best interest”
I think it is really important to learn to view the world from multiple perspectives – and be comfortable with the tensions of that. If you’re a religious man, we have the quote “Be as wise as serpents and innocent as lambs.”
What does that mean here? Mozilla (as with the vast majority of organizations) can be viewed with either lamb-colored glasses, or serpent-colored glasses. We need to do both. They are both an organization that works for the public interest, as well as a power structure that needs to preserve their own prestige and perks for the benefit of the insiders. That’s just reality, not a criticism, it can’t be any other way.
But I think that your “user’s best interest” comment is a little too rosy when you are talking about the money flows. Whenever you start messing with someone’s rice bowl, you can expect to see more of the serpent and less of the lamb. Yet the main challenge of being a society is understanding that people who try to do it alone, following only their self-interest, will be beaten by groups of people who learn to band together in common interest – be that organization a modern NGO, or a marauding army. So I am not suggesting that more moral purity is what we all need, but instead figuring out ways to share the rice bowl so that our overall movement is strengthened.
I’ve been installing, modifying, and pre-configuring Firefox and Thunderbird for small business customers for a year now. I’m also part of a small group of founders that is trying to address the feeling of hopelessness that many small businesses have in regard to technology. They simply don’t know who to trust.
To me, that is an open door invitation to do something better. The Mozilla platform is perfect for delivering a better user experience in this space using a rich client architechture.
Although I don’t think this model really fits with Mozilla, I do think that we can use this strategy without stabbing the Mozilla community in the back. Instead of selling a shrink wrapped product, we’re more like a consulting firm that also builds custom applications to suit the clients. We charge a monthly per user fee that includes remote data management, email, support, and other services. The delivery mechanism is free as in beer, but the real value to a mid-adopter to late-adopter user (most small business users) is what they call “user-friendly” and what we call user experience design. (and with that I’m including training and support)
Please make a killfile extension for Thunderbird! I don’t want to use filters to killfile asshats anymore! Just one button
[...] team that supports Thunderbird. (That is, I hope to answer “no” to my question “Do MEALS need a fork?“) At what point should I be adding people to this team? A one-person development team can be [...]