It's still bloated and slow
I switched to Lynx a couple of years ago, and I'm never going back to firefox.
Mozilla has released an 11th Firefox beta, adding the "Do Not Track" http header it hopes will become an internet standard. We’ve continued our work to improve performance and stability, while also implementing a 'Do Not Track' privacy feature to provide more control over online behavioral tracking," the open source outfit …
Lynx is unusable for serious web browsing - I know, because I use it and elinks regularly (including right now).
The lack of graphics isn't necessary a huge problem, but the different font sizes and lacklustre (non existant : lynx, flawed : elinks) javascript support, plus the continual prompting when loading SSL pages and the need to mess around with the config file to make it half way viable.
Then there's the over zealous cacheing - lynx simply does not refresh pages when it should do. There's probably a config setting to fix this, but frankly I'm tired of fixing something that works in proper web browsers.
It is, however, ideal for emergency/casual use in constrained environments. The OpenBSD developers wisely ship it in the base system, and it should be included in almost every standard Linux distribution as without X or another system, obtaining files can be troublesome.
I'm much more worried about the fact that the latest stable release on Windows lets infected websites deliver drive-by malware exes straight onto my desktop. Has happened twice in the last week, and I'm running anti-virus. Switched to Chrome yesterday - would rather have my privacy invaded than my desktop.
Mozilla, this DNT idea isn't going to work. You're relying on trackers being ethical which, given the amount of money Google has proved there is in that activity, isn't a realistic expectation.
What you really need to do is expose the mechanisms to the user in a meaningful way. For example, bring DOM storage out from under about:config so that it can be easily disabled and, more importantly, the user knows what it is for. A mechanism to restrict the session to the parent domain would also help keep web beacons and rogue scripts under control. Just a couple of examples of useful controls that could be applied instead of this false sense of security. Oh, and get rid of the "safe" browsing traitorware, please.
If you're really serious about this, you need to assist the user in safeguarding her privacy by bringing the mechanisms and methods of such tracking under her control. Since you are mostly funded by Google who have a vested interest in keeping the user tracked, I really doubt that this will happen. Toss as many bones as you like in this direction but until a substantial increase in user control over privacy matters happens, it'll mean nothing. In fact, you're making matters worse: It won't be so very long before I hear "I didn't bother with AdBlock and NoScript because this do not track thingy protects me."
...but it's only a start.
I think the main point of this is that it provides a starting point for a possible legal case.
In this situation, a person has turned on the DNT header, then (somehow) discovers that he is still being tracked.
In every transaction with every server, he is specifically telling the server "I do not give permission for you to track me". This is clear and direct. If the server ignores this, it is like a company ignoring the "don't use my details for marketing" tick-box on an application form. Therefore, a legal case may be able to be brought, which in turn would allow a precedent to be set (one way or the other).
The danger is that people may assume this header is being used by everyone straight away and give a false sense of security.
If they are still adding features, then it is NOT beta. It is not even Alpha. It is still in Development stage.
What the hell is it with developers these days that they slap "beta" on everything from pre-production concept to full-production-but-can't-be-arsed-to-support-it (I'm looking at YOU Google)?
*sheesh*
The plugins sit there watching all the time and, as soon as they see the links, get excited and load themselves fully in the browser. I make sure QuickTime is *not* installed on my own machines for that very reason - because I know how it likes to sink its teeth into everything and ignore other applications' settings.
Perhaps this is a flaw in how NPAPI lets plugins work, I don't know, but if it is, fixing it now could break well-behaved plugins that may not see update. Though, I do admit sometimes you do have to break away from the past and stop legacy support. (Unfortunately, this doesn't happen often enough, or else SMTP would've been replaced by now with something you can't easily spoof.)
In any case, mini-rant aside, Mozilla COULD probably change how file associations are handled and make user settings prevent the loading of plugins that lust for those formats, but that's not necessarily a simple task. Since the matter is easily fixable by disabling those plugins in the Addons Manager, I'd rather they work on higher priority bugs first.
Finally, for those complaining about a feature freeze, just how much work do you think it really is to add one setting that simply modifies a single string it sends to a server, and how many bugs in Firefox do you think that that would actually introduce?
As a regular traveller to a country where English is not the first language, it annoys me intensely that most websites, including Google, Yahoo, Amazon and IMDB to name a few, completely ignore browser headers that would give them a great indication of what language to present me with. This is an existing useful header which would allow them to redirect me to a language based on my system preferences, rather than my current location determined by IP address. They choose not to use it.
Given that they ignore existing useful headers, I don't see what about this new header would make major websites choose to use it header unless it's made mandatory in some way (like that will ever happen).
.deb
packages