Handling of incorrect MIME types

[Dunderklumpen]

I was reading something yesterday and it says the W3 say that "if a content header is not present the browser may attempt to guess"

That´s right, at least from waht I have understood. Wich means that Internet Explorer actually does it wrong (no surprise). Bad thing is that it solves the "problem".

I'd agree with thumperward though - it is very easy to fix this problem on a server and surely any host who cared about their customers would comply with requests of this sort as it's very simple and only needs done once.

I think no one is arguing about that. The miss-configuration is easily fixed and any good host should fix it also. But - that requires that the user acutally knows this and that´s where I mean that the problem is. This is causing average users to blaim Firebird = bad reputation = bad reviews and since the most important marketmethod is "mouth-to-mouth", well this is a problem.

And there are already some non-standard functions in the Mozilla code - blinking text and marquee support. They got in there due to the Chinese market (I think) so how about the rest of the market then? .

[jrobbio]

I want to post a message on my hosts forum about this. What is the best way of getting across my message?

[iamnotniles]

I was reading something yesterday and it says the W3 say that "if a content header is not present the browser may attempt to guess"

That´s right, at least from waht I have understood. Wich means that Internet Explorer actually does it wrong (no surprise). Bad thing is that it solves the "problem".

of course it causes other "problems" by doing that.

[Paradox52525]

Not necessarily, no one is suggesting full mimetype guessing like internet explorer uses. In fact all you would really have to do is is look at the file extension of any link server as text/plain, and if it's not a known text file extension (or if it contains invalid/binary characters) change the memtype to application/x-[file's extension]

[My Yute]

I'm seriously considering it. It's hard enough finding good sites do download anime from. When a good bittorrent site doesn't use correct MIME types

let the link load in firebird, then hit ctrl-S or File -> Save Page As... and double-click the saved file.

Or if you really want to get into heaven, email the site with this:

"Dear admins of-Torrentz-R-Us.com,

Mozilla users have to right-click your links in order to download files. This is incredibly easy to fix. Just set your server to present .torrent links with the MIME type 'application/x-bittorrent'".

- Chris

Right-click will not do it on many sites since they have a second page handling the sending of the file. Right-click will get you to save the page - not the file.

The average users does not understand that there is a server confuguration error. The average users blaims the browser and since things works in Internet Explorer "the case" is lost.

Also - a supportdepartment will finally be rather fed up having to anser this kind of question/problem over and over again and since most supportdepartments have other, more important things to attend to - in the end that "case" is also lost.

I have seen it happen more than once, I´m sorry to say. This is a matter of reality and in reality people expect things to work - if they don´t they switch back to something that does. Simple question - should Firebird solve peoples problems or create new ones?

Are you talking about bitorrent sites specifically? Because this is the way I handle my bittorrent files and have never had a problem. When you click save as it allows you to save the torrent file which you then double click on and that starts the download.

[Dunderklumpen]

Are you talking about bitorrent sites specifically? Because this is the way I handle my bittorrent files and have never had a problem. When you click save as it allows you to save the torrent file which you then double click on and that starts the download.

No, at least not me and as I said - right-click will not let you save the actual file on many systems since you are redirected to a second page that handles the download. This is common on sites handled by databases such as MySQL and others together with php (just an example).

[chrisgeleven]

There are cases where workarounds to perform tasks are acceptable and there are cases where it is not. This is a case where a workaround is not acceptable in the long-term. While I agree servers should be setup properly, name me a mother who isn't a computer geek that is going to e-mail some web server admin and ask them to fix their MIME-type settings. Then see if you can find anyone else that is going to be willing to right-click most links just to download them. Ain't going to happen folks. They will just go right back to IE (like my mom and dad have despite my reeducation effort).

Right now right-clicking and e-mailing web site admins might work. But in the future it will not be feasable for an end-user product especially when more and more sites redirect you when you click on a download link before downloading.

[alexkreuz]

why is it so hard to let windows manage the filetypes? if its an html file you clicked on and the default is firebird, then it'll open in firebird. if it's a winamp skin file, then windows will launch winamp. if it's a bittorrent file, then windows will launch bittorrent. is that really such a hard thing to do? the rest of the windows system works like that, why do the browsers have to be so anti-windows and follow those geeks at w3c?

[Dunderklumpen]

why is it so hard to let windows manage the filetypes? if its an html file you clicked on and the default is firebird, then it'll open in firebird. if it's a winamp skin file, then windows will launch winamp. if it's a bittorrent file, then windows will launch bittorrent. is that really such a hard thing to do? the rest of the windows system works like that, why do the browsers have to be so anti-windows and follow those geeks at w3c?

I think standards are very important and I can go along with the goal for the project to create a browser that does. The standards from W3 should be implemented and I would not like to see anything else. Creating a solution for badly configured servers is therefore not standard and that´s why I suggest it to be an extension.

The problem here apperantly lies in the fact that a lot of us users would like to see this extension but so far I have´nt really seen anyone willing to create it (not whining).

Anyone up for the job?

[Ted Mielczarek]

why is it so hard to let windows manage the filetypes? if its an html file you clicked on and the default is firebird, then it'll open in firebird. if it's a winamp skin file, then windows will launch winamp. if it's a bittorrent file, then windows will launch bittorrent. is that really such a hard thing to do? the rest of the windows system works like that, why do the browsers have to be so anti-windows and follow those geeks at w3c?

Because

1. Not everyone uses windows, so file extensions don't mean the same thing
2. Lots of CGI pages produce content of varying types that has nothing to do with their file extension
3. If it wasn't for those geeks at the w3c, most of the web wouldn't work together the way it does

That being said, I think this problem does need to be addressed. However, I don't think we ought to "solve" it the way IE does. I would like to see something like <a href="http://httpd.apache.org/docs/mod/mod_mime_magic.html" title="Apache module mod_mime_magic">mod_mime_magic</a> from apache used to sniff files served as text/plain.

[Dunderklumpen]

That being said, I think this problem does need to be addressed. However, I don't think we ought to "solve" it the way IE does. I would like to see something like <a href="http://httpd.apache.org/docs/mod/mod_mime_magic.html" title="Apache module mod_mime_magic">mod_mime_magic</a> from apache used to sniff files served as text/plain.

Good suggestion, I just hope that someone with enough skills to create such an extension has the time to really do so because in my opinion, this is something we really need. Is there nobody out there that can build this thing or anyone that know anyone that can?

I mean we have had this discussion before and it will crop again from time to time. This "problem" will not go away and as me and many other have said - it drives people away from a really good piece of software.

[Lenders]

The Idea That People Don't Just Automatically Hit Back And Then Right-Click The Link Instead Like They've Done On Every Browser For The Last Eight Years.

I have yet to hear from one single solitary person that they switched back to IE because of this. One.

- Chris

Some sites just don't allow you to do this. Example: www.deskmod.com
If you right click and select save link to, it saves you the php script, not the file you want to download.

This IS a bug issue and users don't bother which is responsibile of this, the browser or the server, they only notice that with IE it works so they use IE.
What does the Mozilla community want?
Do you want people to finally abandon IE for the rest of their lives and switch to Mozilla?
If this is your goal, issues like this must be solved, otherwise Mozilla Firebird won't be a real competitor of IE.

[MechR]

On this page of a review of Moz 1.0: http://www.arstechnica.com/reviews/02q3 ... moz-3.html

I found this:
"When Gecko loads a page, it looks for a valid "doctype" at the beginning of a web page. Doctypes tell the browser what kind of content it should expect and where to find the "Document Type Definition" that applies to a particular page. If there is no valid doctype, Gecko drops back to "compatibility mode" and attempts to render badly written pages the way older, less-compliant browsers would. Gecko's compatibility mode is pretty effective and completely transparent to the end user. It does a good job of rendering pages that were written to look nice in Netscape 4.x and older versions of IE, but it's not quite perfect."

Is this still true (or was it ever)? It seems like something similar to this would help the MIME-type problem.

[wheerdam]

Is this still true (or was it ever)? It seems like something similar to this would help the MIME-type problem.

I think that will do nothing for the MIME-type problem. Server use MIME-type to tell Firebird what type of file the browser is trying to view, and Firebird will act accordingly.

[dtobias]

One of the things I can't stand about IE is its standards-noncompliant MIME type sniffing, which makes it an exercise in frustration to try to get it to properly display plain text content which happens to be served from a URL that doesn't end in .txt (e.g., via a Perl script ending in .pl). At its ever-changing whim, IE will treat such content as HTML, binary data, or who knows what else... anything but actual plain text despite the "text/plain" MIME type.

If Mozilla ever changes to be bugward compatible with IE on this score, that will be time for me to abandon it and try to find a more standards-compliant browser to switch to.