I have to disagree here, both from a Scalix product point of view as well as philosophically.
debian have - in the scope of the Iceweazel project - modified the Gecko engine with their patches; for this reason, it is not safe to assume that this Gecko engine will behave exactly the same as in Firefox, therefore we don't want to and cannot take support for it. We specifically do not support the Gecko engine, but the exact browser versions we have tested, and for good reason. We have seen so many very subtle AJAX issues, even in dot releases of Mozilla Firefox browsers that we need to be very specific.
Having said that - in most cases this is known to work and can be tweaked around by modifying the UserAgent string.
I might be convinced that it could be a valid enhancement request to introduce an option that would 'relax' browser checks and possibly, in the Mozilla case, go back to the Gecko engine check.
On the general Iceweazel statement.... and the following is strictly personal opinion - I disagree as well. I don't think Mozilla Corp and the foundation is all goodness, but I appreciate what they've done for the 'net in general and the balance they bring in. They are the ones that have made sure that "we" (in the sense of non-Windows users, me personally working on a Mac most of the time) are able to see almost all of the content the Web has to offer by creating a browser platform that is seen so relevant in terms of market share that basically noone dares to ignore it.
That hadn't been the case with any other browser before FF came along. They've done a tremendously important job here and I appreciate that very much.
So why not let others take this further and use it for their own purposes? It's Open Source, after all, with many contributors.
Correct - and that's why the Gecko engine and the source code for Firefox are free and I presume always will be.
However, if everybody was allowed to recompile FF without rebranding, there would suddenly be tens, if not hundreds of different FF versions floating around. Some might just be recompiles, some might have patches selected by someone, some might (and actually would be, I'm sure) malware.
For the purpose of consumer users (and FF has a broad market share amongst those as well), this may not happen. Definitely not. And the only way to protect this from happening is to preserve the integrity of official builds. And not allowing anyone else to use logo and name, IMHO, is, at least for a consumer target audience, the most effective way of doing that. Which, again, ensures a high level of integrity for the official versions, no cloning, no bad things, and a good market share and a stronghold against a pretty boring IE-only world.
I personally would probably even go a couple steps further and even introduce proper code signing into the game, so that the product checks itself and makes sure it's not only a paper limitation.
So as friendly and nice the idea of backporting patches, stable releases, etc., is, at least for a consumer/desktop oriented packaging, the debian people should not do what they do - they should just ship the browser in binary form consisting of the original Mozilla Corp packages, so that everyone can use it; my understanding is that this is what Ubuntu is doing. They can certainly maintain their own copy of it with their own patches, but it should not be the mainstream. If debian in itself wants to become viable for consumer end users and company desktops, they will need to define a more flexible policy (similar to Ubuntu's Universe policy) for some bundled apps....... and that's part of it.
I know I might have started a flame ware here, but I believe I have expressed this opinion before in the Iceweazel debate. For me, as it stands, as long as Iceweazel doesn't gain browser share in it's own right and it makes sense to start testing it, it is unsupported with Scalix and whoever wants to use it will have to tweak it to make that clear....
Florian.