MJ Ray wrote: > Arnoud Engelfriet <[EMAIL PROTECTED]> wrote: > > I guess then I don't understand why. What I saw was a package > > called 'firefox' that Depends: on Iceweasel. So that means > > if I type ``apt-get install firefox'', apt-get will see the > > dependency and install Iceweasel. That's where I got worried. > > I wonder if there's any good way for the firefox transition to > silently transit users of the old firefox package but bale out > if someone requests it as a new install.
The FreeBSD ports system has a "BROKEN" flag. If you try to install such a port, the system will tell you the port/package is broken and cannot be installed. It's just a Makefile flag: BROKEN= this port is unsupported on FreeBSD 5.x which effectively does this: root:/usr/ports/example# make install ===> foobar-0.1 is marked as broken: this port is unsupported on \ FreeBSD 5.x. Is there something similar with dpkg? Then the "broken" message could explain the situation, and people will then apt-get install iceweasel (or install the real Firefox another way perhaps). > > It's the fact that the package Depends: on Iceweasel and > > so the act of running ``apt-get install firefox'' results in > > Iceweasel being installed and not the Firefox browser. > > I just checked and apt-get displays what it is going to install and > asks ``Do you want to continue? [Y/n]'' unless the user has explicitly > configured their system not to do so. How is that significantly different > to 'coke' at the pub being a transition package that asks whether you > want to continue with Pepsi? I wasn't aware of that, from what I had understood dependencies install automatically (sorry about that). This is a very big step in the right direction. Is there any way to make apt-get say "Iceweasel is a replacement for Firefox that will be installed instead of Firefox if you say yes below"? > > Good point. It's always about context. I would say that > > commercial vendors of Debian (-based) software are trading > > among other things this package. They could be held liable for > > a false designation of origin or false label (15 USC 1125). > > http://www.bitlaw.com/source/15usc/1125.html > > I can see that they could, if they claim that the software includes > Firefox in connection with the trade, or they put that on the label > of debian software. If the package is silently included, it seems a > mere functional thing, similar to lists of unix for DOSsers 'equivalent' > programs. Are there cases where something like the lists of equivalents > have caused a successful trademark prosecution? A list of equivalents would be something like "Carbonated drinks we stock: Coca-Cola, Pepsi Cola, River Cola". Absolutely nothing wrong with that. There can be no confusion there. I could imagine a case where a very exclusive brand was put with shady low-quality brands, but that'd be more a tarnishment than a confusion issue. Trademark confusion is usually about people trying to pass off their product as someone else's brand, or associating their stuff with that brand, or substituting their alternative for the brand. In Beijing I once saw a shop that had a big logo for Coca-Cola; once you got close you saw the rest: "we don't sell <BIG>Coca-Cola</BIG> but we do have brand X". If vendors don't mention Firefox at all, and a customer by himself tries typing "apt-get install firefox", the chance of confusion will be much lower. Especially when apt-get first identifies Iceweasel and then asks for confirmation. I agree with you it's a niche. And I hope that that little extra clarification in the apt-get procedure is possible, because it'd be a very simple fix that clearly solves the problem. Arnoud -- Arnoud Engelfriet, Dutch & European patent attorney - Speaking only for myself Patents, copyright and IPR explained for techies: http://www.iusmentis.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
