Packages and Trademarks – An Observation

Better to have tried and failed than not to have tried at all, right?

Generations of fathers would say to their kids “of course”. But it seems not in every case.

A while back, Mozilla and Debian had several goes at coming to an agreement for the use of the Firefox name and logo. In then end, we collectively failed[0], which is something I consider a great shame. Sadly, this failure seems to have left a significant number of free software people with a residual unhappiness and antipathy towards Mozilla. Which is, in some ways, an even bigger failure.

On the other hand, no-one seems to have even suggested that there’s a possibility that the new kid on the block will allow their name to be used when Debian packages their software. So Debian are happily and peacefully packaging it under the alternative name. I haven’t noticed anyone being upset about this.

Why is that?

[0] And I’m not interested in reopening the debate about who was right. That’s not the point of this post.

38 thoughts on “Packages and Trademarks – An Observation

  1. “Chromium” is both similar in name to “Chrome”, and the actual name of the open-source project. A clever idea, but I wonder whether it hurts Google’s claim to the “Chrome” name, or risks customer confusion when a malicious “Chromium” appears.

  2. Dan: I don’t understand your point. Firefox has an official name, Firefox, and an unofficial name, Minefield. (Actually, it has a few more, named after parks, as well.) Chrome has an official name, Chrome, and an unofficial name, Chromium. None of these names are “made specifically for Debian”.

    Debian invented the name IceWeasel themselves; it wasn’t forced on them by us. They could have used Minefield if they wanted. (Or Namoroka, or Shiretoko, or Bon Echo).

  3. > I don’t understand your point. Firefox has an official name, Firefox, and
    > an unofficial name, Minefield. (Actually, it has a few more, named after
    > parks, as well.)

    Sorry, Gerv, you’re wrong.

    Firefox doesn’t have unofficial name. Those names are codenames for each release. Just like Chrome has code names for releases (Eclair for instance) You can’t have a product name that changes between releases.

    This is a problem. A big problem as far as I am concerned. You can read my opinion here:

    http://kaply.com/weblog/2008/05/28/the-browser-with-no-name/

    Chrome has handled this much better with Chromium

    And if Firefox had a non trademarked name, I’m sure it would have been used instead of Iceweasel.

  4. Mike: Well, OK, sort of. Would it be better if we always called it “Minefield” until we were ready to ship it as Firefox (i.e. had codenames for each release, but didn’t brand the betas with them)?

    I don’t think that would make people happy, because Minefield isn’t the sort of name you want for your browser. And it was chosen for exactly that reason.

    So is your point that Debian were annoyed with us because we didn’t have a standard alternate name they could use? That didn’t come through in the discussions, if so.

    If people want to band together around a common name for all their different non-Firefox variants of the codebase, there’s nothing stopping them. It could be IceWeasel, it could be something else. Why do you think that hasn’t happened?

    Nit: Eclair is an Android code name.

  5. My point was Debian had nothing else they can use. It’s Firefox or “make up your own name.”

    So they picked “make up your own name”

    A nice third option would be unencumbered by trademark/unbranded Firefox.

    I don’t think people need to band together to solve this, I think Mozilla needs to solve it.

    And the reason I think people haven’t done it is because up to now, people are able to use Firefox as they need.

    But as Mozilla cracks down any and all uses of Firefox, I think it will become more necessary.

  6. Another observation: I think I would have preferred Ubuntu to name it IceWeasel as well. Whenever people report issues on Linux I have to ask: “Is that the official Firefox build you are using or the Ubuntu one?” And quite often the issues evaporate immediately when these people download the official build.

  7. “If people want to band together around a common name for all their different non-Firefox variants of the codebase, there’s nothing stopping them. It could be IceWeasel, it could be something else. Why do you think that hasn’t happened?”

    Well, that didn’t happen with Chromium either – Google had things all set up with a somewhat similar name before it became an issue. Maybe either or both Debian or Google learnt from the previous experience?

    I guess maybe the annoyance was because it wasn’t clear that the name couldn’t be used, and then the work of finding an alternative name and adding it to the browser wasn’t handled by Mozilla.

    You also seem to be assuming that everyone on both sides was acting logically, which is pretty much always a bad assumption when it comes to humans – once people start arguing, they will generally carry on defending their position even if it’s not logical. Which is also the reason that finding some nice solution now without stirring up all the previous arguments is going to be pretty hard…

  8. There is one big difference, which is that Chrome itself isn’t open source. For Debian (or anyone else) to take the content of the Chromium repository and label it “Chrome” would be like building a stock OpenOffice.org and calling it Lotus Symphony, not taking into account the proprietary branding and extending that differentiates them. As a concrete example, Chrome has proprietary media codecs built in. The situation with Firefox is somewhat different, in that the Debian builds did at least come from a source repository used to build the complete Firefox application.

  9. But that sounds like saying “Debian isn’t upset with Google because, hey, they aren’t properly open source. Debian only gets upset with real open source organizations”. That seems precisely backwards… They are supposed to be supporters of software freedom!

  10. I think the argument over the use of the “Firefox” trademark and logo – both with Debian and elsewhere – is really a proxy for something larger. Most free software projects have source code tarball releases as their main deliverable. There is an expectation that someone else is doing the compiling; it might be the end user, or a “distribution” organization. That someone will naturally make small changes to suit their own needs, and with rare exceptions (there was a huge argument over name-change requirements for LaTeX packages, IIRC) that’s just fine by the “upstream” developers.

    Mozilla isn’t like that. Within the core development team, the “official” Firefox binaries from mozilla.com are considered the main deliverable. The source is thought to be of interest only to people who are making major modifications, or building something entirely new on top of the XPCOM and Gecko infrastructure; in neither case would they want to call the result “Firefox”. Thus, our perspective is that anyone who wants to redistribute something called “Firefox” ought to be redistributing the official binaries, or else they’re probably up to no good.

    I claim that this is the key point of disconnect between Mozilla and Debian, and more generally, between Mozilla and people who want to rebuild Firefox with small modifications – most prominently the various other Linux distributions, with their integration constraints.

  11. @Gerv But that sounds like saying “Debian isn’t upset with Google because, hey, they aren’t properly open source. Debian only gets upset with real open source organizations”. That seems precisely backwards… They are supposed to be supporters of software freedom!

    I think it’s generally true that redistributors of free software tend to get more upset with organizations that are producing stuff that’s almost in keeping with their rules for free software, than organizations that are distributing stuff that isn’t even remotely close to conforming. There’s more hope of persuading the almost-in-compliance people to change what they’re doing, so there’s more of a point to the argument.

    Clarity of messaging from Day One surely also helps.

  12. >But that sounds like saying “Debian isn’t upset with Google because, hey, they aren’t properly open source.

    Chrome isn’t FOSS, but Chromium is, and always has been. See also: Safari, WebKit (ok, one is an engine, the other is a full browser, but still similar).

    Minefield is close to that type of name, but there is never a final release of anything called Minefield. And as you said, the fact that the Minefield name is meant to scare away people that shouldn’t be testing it is why it would be a rather poor name for an actual product.

    >So is your point that Debian were annoyed with us because we didn’t have a standard alternate name they could use? That didn’t come through in the discussions, if so.

    I don’t know if that situation and that solution had come up before. Other trademarked brands (Gnome, KDE, Apache, OO.o, etc) were fine with the distros using the name, even if some changes had been made.

  13. Hi,

    Interesting question. My view is that

    Google did elegantly by 1) immediately having separate “projects” and 2) explaining what they were for. In my mind Chrome is the inferior binary blob aimed at Windows users – chromium is the cool upstream project that linux distros package.

    I think mozilla could still improve the firefox situation by de-emphasizing the importance of firefox branding on linux, and creating a similar “upstream” project that could cater to linux distributions.

    Best
    Anders

  14. > Thus, our perspective is that anyone who wants to redistribute something called
    > “Firefox” ought to be redistributing the official binaries, or else they’re
    > probably up to no good.

    I don’t think that’s true at all. I think we have a very clear understanding that Linux distributions do not ship the official binaries and that they are not “up to no good”.

    Michael Kaply and others are suggesting that “Chromium” is not trademarked by Google. But it certainly is:
    https://chrome.google.com/extensions/intl/en/branding.html

    I wonder if Google gave Debian permission, or maybe no-one has thought about this yet.

  15. Another interesting question is whether Debian or other distros will be able to use the Chrome icon.

    It’s actually a pretty big deal to have users moving to Linux see the same app icon(s) that they saw on Windows.

  16. Robert: Chromium has a different icon to Chrome (no idea which one Debian uses). The killer point about the Firefox icon is it’s non-free, not that it’s trademarked.

  17. 1) The Minefield name is deliberately chosen to turn people away. Chromium isn’t.
    2) Chromium isn’t release-specific. On Jaunty, installing Firefox 3.5 as a backport gives you a broser named Shiretoko (what’s Shiretoko? I asked for Firefox.) and an ugly icon.
    3) The ugliness delta from the Chrome icon to the Chromium icon is smaller than the delta from Firefox icon to Minefield icon.
    4) The Chromium icon and name are real pieces of identity. The Minefield icon and name aren’t.
    5) The Chrome/Chromium distinction was announced on day 1.
    6) Google says that Chromium is Open Source but Chrome isn’t. Firefox is said to be Open Source even though the release builds come with a non-OSI EULA.

    I think Firefox would benefit from a name that doesn’t imply danger (Minefield, isn’t obscure (Shiretoko) and isn’t meant to mock Mozilla (Iceweasel) and polished icon for a fully Free Software version. (E.g. Icefox with the same icon as Firefox but only with a white fox. Disclaimer: IANAL, I don’t know this would protect the Firefox icon sufficiently for the purposes of trademark law.)

  18. James: Firefox’s icon is not non-free. The icon files, like everything else shipped with the official build of Firefox, are available under the MPL. And have been for quite a while now.

  19. Gerv: other-licenses/branding/firefox/LICENSE begs to differ. And that file hasn’t changed since its creation in 2004.

  20. I am not a Debian Developer, but here is the core of the issue, as I understand it:

    http://mxr.mozilla.org/mozilla-central/source/other-licenses/branding/firefox/LICENSE :

    You are not granted rights or licenses to the trademarks of the Mozilla Foundation or any party, including without limitation the Firefox name or logo.

    For more information, see: http://www.mozilla.org/foundation/licensing.html

    http://www.mozilla.org/foundation/trademarks/policy.html :

    Modifications

    If you’re taking full advantage of the open-source nature of Mozilla’s products and making significant functional changes, you may not redistribute the fruits of your labor under any Mozilla trademark, without Mozilla’s prior written consent. For example, if the product you’ve modified is Firefox, you may not use Mozilla or Firefox, in whole or in part, in its name.
    […]
    Again, any modification to the Mozilla product, including adding to, modifying in any way, or deleting content from the files included with an installer, file location changes, added code, modification of any source files including additions and deletions, etc., will require our permission if you want to use the Mozilla Marks

    This is a problem. And I believe Mozilla did offer permission to Debian (just as they did for Ubuntu (and others?)), but the DFSG (#8) says:

    License Must Not Be Specific to Debian

    The rights attached to the program must not depend on the program’s being part of a Debian system. If the program is extracted from Debian and used or distributed without Debian but otherwise within the terms of the program’s license, all parties to whom the program is redistributed should have the same rights as those that are granted in conjunction with the Debian system.

    So while Ubuntu could accept the offer, Debian could not.

    Meanwhile, Chromium simply has the 3 clause BSD license, which is fine:

    Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

    * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
    * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
    * Neither the name of the Google Inc. nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.

    No special stuff regarding the icons or the trademark.

  21. From what I understood reading the tedious debate when it appeared, Debian wanted to be able to stick the Firefox name and logo over *anything at all without restriction* and Mozilla vetoed it, emphasizing the distinction between an (open-source) copyright and a trademark (which is by essence “proprietary” to the trademark owner). At the end of that reading I came out feeling for Mozilla, but that’s just my bystander’s opinion, and IANAL, nor do I play one on TV.

    OTOH I’m using openSUSE Linux, which AFAICT is just like the good old SuSE of before being bought by Novell (with some enhancements such as the possibility of downloading a DVD image over the net then using it to install the next release without even burning a physical DVD), and if someone wants to boycott SuSE because Novell made a patent agreement with Microsoft it’s not me: nowhere in SuSE do I see Bill Gates’s hand. I feel SuSE (and RedHat/Fedora) more to my liking than Debian (and Caldera, and Ubuntu), so maybe I was prejudiced.

  22. Gerv: Your hint that official builds weren’t always available under the MPL reminds me of the EULA fiasco, which was another source of antipathy, even from Ubuntu. Even now about:rights says “Mozilla does not grant you any rights to the Mozilla and Firefox trademarks or logos.”

  23. James: other-licenses/branding/firefox/LICENSE says “You are not granted rights or licenses to the trademarks of the Mozilla Foundation or any party, including without limitation the Firefox name or logo.”. I agree, it should probably say “_trademark_ rights or licenses” for clarity, although I feel that the fact that it says “to the trademarks of” means that this is how a court would interpret it anyway. That’s certainly what it means. Feel free to file a bug if you think the wording needs fixing.

    Kelly: My view of that is that the “License specific to Debian” clause of the DFSG should be about copyright and patent law, but not trademark law. As RMS keeps telling us, the three are very separate and should be handled separately. Allowing someone to use your trademark is like saying “I trust you”. Is Debian really saying that “if you trust us and want to show it, you have to trust everyone else in the world”?

    Logos should not be a functional part of the product – removing the logo should have no effect on how well the product works. If that’s true, then I think having a trademark licence specific to Debian is not disadvantaging or disenfranchising downstream Debian users (which is what the clause is trying to avoid).

    James (2nd comment): The fact that official builds weren’t under the MPL was a problem, and we fixed it. Hope that’s OK with you :-) about:rights is perhaps ambiguous, like the LICENSE file mentioned above; it should say “doesn’t grant you any trademark-use rights” or something like that. The logo file ones and zeroes are free software.

  24. >Is Debian really saying that “if you trust us and want to show it, you have to trust everyone else in the world”?

    Yes, that is what the DFSG says. To leave out trademark rights, Debian would have to pass an amendment.

    IMHO, trusting others is one of the points of FOSS. After all, licenses that say things like “don’t use this code for evil” or “don’t use this code in the military” aren’t considered FOSS. The are times when trademark should not be looked at the same way as copyright, but I feel that this isn’t one of them.

    I still don’t understand why Mozilla feels the need to defend their trademarks more heavily than anyone else in the open source world.

    @Tony Mechelynck: Debian had some patches they were carrying for FF (particularly security patches backported to the FF version that was in stable), but that is nothing unusual. Here is the current patch info for iceweasel: http://patch-tracker.debian.org/package/iceweasel/3.5.6-1

    A larger patch set than some, but certainly not the largest.


    I’m done, I don’t think this discussion can usefully go much farther.

    Cheers,
    Kelly

  25. Gerv: Yes, fixing it was great, but a problem fixed still leaves more antipathy than not having had the problem in the first place. Is it unfair that Chromium has learned from your mistakes without paying the price?

    I still don’t accept that the bits are under the MPL – even http://www.mozilla.org/MPL/license-policy.html says “Mercurial hg.mozilla.org All Mozilla Code except other-licenses/branding is MPL/GPL/LGPL.”

    Filing a bug asking to change what is clearly a thoroughly considered policy would be pretty pointless IMHO, it’d just be closed because bugs aren’t for general discussion. Do I think there needs to be a better balance at the intersection of trademark and copyright law? Yes. Am I directly able to affect this balance in the case of Mozilla? No, as is clear in this and other cases like UbuntuOne and OpenSolaris, the FLOSS community is unable to affect the actions of single-vendor trademark holders. But we’ll still complain about it if they get it wrong.

    Projects which are clearly components like Linux, GNOME or X don’t have this trademark problem. Firefox and Thunderbird do, because uniquely on the free software desktop they think they’re a standalone product that should not be altered even in good faith, not a component of a greater whole. http://mjg59.livejournal.com/68112.html?thread=344336#t344336

    Google Chrome doesn’t pretend to be part of the free software desktop so no-one complains that it’s non-free. RHEL is also clearly a non-free compilation and service, so no-one complains much. Firefox is trying to be free and non-free at the same time, so people complain.

    Cf the license text on http://en.wikipedia.org/wiki/File:Gnomelogo.svg http://en.wikipedia.org/wiki/File:Chromium_Logo.svg http://en.wikipedia.org/wiki/File:Chrome_Logo.svg and http://en.wikipedia.org/wiki/File:Firefox-logo.svg

  26. Further on component/product – source is a component, a binary package is a product. Binary tarballs don’t count – people want packages. Since Mozilla doesn’t provide an official Linux package for any distribution, you have to rely on the distributions with all that entails. Set up an official apt repository for Debian and Ubuntu with official branding and the latest version and you’ll get plenty of users switching to it. Those who are happy to let their distro handle things will keep using iceweasel. This is how Chrome/Chromium works on Linux.

  27. Kelly: I still think most sane Debian developers would not claim that trust works that way – in Debian or anywhere else. As for why Firefox needs to defend its trademark more than other free software projects, I’ve written about that several times. Here is one example. Summary: we have 300,000,000 non-technical users. “Mozilla Firefox (Steal-Your-Data Edition)” is not something we want to see. And we have successfully used trademark law against bad actors in the past.

  28. I still don’t accept that the bits are under the MPL – even http://www.mozilla.org/MPL/license-policy.html says “Mercurial hg.mozilla.org All Mozilla Code except other-licenses/branding is MPL/GPL/LGPL.”

    Fixed – the fix should show up in an hour or so. That’s a document I can change easily :-)

    Filing a bug asking to change what is clearly a thoroughly considered policy would be pretty pointless IMHO,

    The policy’s already been changed. Download any current Firefox and visit about:rights. It says that the binary build is available under the MPL. No exceptions. The bug would be about clarifying the wording to better reflect the true policy.

    As for Wikipedia, I think their “Mozilla logo copyright” exemption text is out of date. (Seems like we didn’t do a particularly good job publicising the change when we made the logos free software. I thought we had, but clearly not.) The Firefox logo files nearly fit the definition of a Free Cultural Work. The only exception (which is not the one cited by Wikipedia) is that the underlying source data is not freely available. AIUI, we don’t make the vector version of the logo available under a free licence.

    Assuming the Chromium logo has the same usage rights as the Chromium name, then the Chromium logo and the Firefox logo have (to my understanding) exactly the same level of freeness.

  29. Can you point to where you did announce the change to the logo licensing? As in, specifically the logo, not just the binary distribution. I can’t goo^Wfind anything with a Googleā„¢ search, and as you may have guessed this is a subject I have interest in, so I’m very surprised this is the first time I’ve heard about it.

    BTW, you should put a notice at the top of http://www.mozilla.com/en-US/legal/eula/firefox-en.html saying it’s no longer used.

  30. I wouldn’t worry *too* much about the impact that the whole Iceweasel debacle has on Mozilla’s reputation. Quite frankly, it reflects more on Debian. I say this as a Debian user.

    Everyone in the open-source community knows that the Debian legal people are very exacting in precisely what they will allow and what they won’t. They don’t consider a vanilla Linux from kernel.org to be something they can redistribute either. They’re… picky… and they’re known for it. Some people agree with their standards. Some disagree. Most of us just don’t want to bother thinking about it too hard, because we have other things to do with our time. (Personally I mostly just want my software to do exactly what I ask of it, without a lot of fuss.) But all of us (who know anything about Debian) know that this is something the Debian legal team is known for. We’ve come to expect it from them.

  31. > IMHO, trusting others is one of the points of FOSS.

    We *know* there are malicious actors who use the Firefox name and logo to p0wn our users. If software freedom requires us to trust *them*, then software freedom is broken. (I don’t think it does.)

  32. Robert: You’re not being asked to trust malicious actors, only those acting in good faith. I can’t comment on DFSG vs Debian-specific trademark licenses, it’s not something I cared about.

    Gerv: filed as https://bugzilla.mozilla.org/show_bug.cgi?id=541761 There’s also quite a few EULA bugs that could be closed now the EULA isn’t used.

    http://weblogs.mozillazine.org/roc/archives/2010/01/video_freedom_a.html#c2674042 is another example of how Mozilla hasn’t communicated the change in the logo’s copyright licensing at all.

  33. James: if Debian want to be able to pass the trademark rights they get on to any downstream person, then we are being asked to trust anyone, including potentially malicious actors.

    Thanks for your work on getting the logo licensing change communicated more effectively. Keep pushing on that :-)

  34. Gerv: Well, that gets into DFSG vs trademark licensing, which as I said is not something I tried to understand. The issue of the non-free logo short-circuited the discussion before any meaningful discussions about that could be had.

  35. From our lawyer: The extensions page roc points to above is in error. We are going to fix this. There will be an update here shortly. Apologies for the confusion.