From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:0.9.9+) Gecko/20020403 BuildID: 2002040314 On my site (see url), I have some rar archive files which can be downloaded. But instead, Mozilla opens them like a text file ! Another archive format, .ace is downloaded without problem. Is this related to bug 126782 ? Or it is a server problem ? Reproducible: Always Steps to Reproduce: 1.Go to that url 2.Chose a link in one of the tables. Actual Results: mozilla opens it like a text file. Expected Results: mozilla must start a donwload instead.

free.fr servers the .rar files as text/plain so Mozilla displays them in the browser I guess they are serving unknown files with a text/plain mimetype (it is the default in Apache). You have to contact free.fr so they change the mimetype of .rar files to application/octet-stream (I don't think there is an registered mimetype for rar files).

Thanks for this useful info !

Nothing seems to have changed. On another site, a rar file cannot be downloaded without using "save link as" function.

*** Bug 147678 has been marked as a duplicate of this bug. ***

I see this problem when I left-click on .rar files, for example any of the ones mentioned in duplicate Bug 147678, at this page: http://www.infonet.civ.pl/download/ but when I left-click and select "Save Link Target As" I only have a choice of saving the file as text, and the .txt extension is automagically added. Now, if I delete the .txt extension and save it anyway, I do in the end download the file correctly, but this workaround DOES seem quite confusing.

Following comment #5, it seems to be related to bug 120327 :-/

Well, I think this behavior IS resulting from Bug 126782, but... Since because of that bug I could not download a .rar file by left-clicking it, I tried to right-click->Save Link Target As, but that was confused by Bug 120327 which saved the file as a TEXT FILE and tried to add the .txt extension. Now, if I had had the choice of saving the target as ALL FILES (*.*)instead of Text File ".txt" (bug 116951) this workaround would have been a lot less confusing. Anyway, that's my daft newbie's understanding of it.

see http://bugzilla.mozilla.org/show_bug.cgi?id=124114#c2 for contact info. in the past, Free.fr has been quite reponsive to fix their incorrect mime types.

I sent a mail to webmaster@free.fr to ask them for a fix.

anyone have a .rar file on free.fr so that we can test the issue ?

just did a quick test, they are still sending .rar files as text/plain. same problem for .tar.bz2 files.

Contact : webmaster@free.fr

this bug happens to many for mozilla unknown filetypes .rar .vl2 .iso file becomes auto extension .iso.exe and mutch more .... this bug existing since the first mozilla release on all os types :(

collection of related problems will come as dups (mostly other mime types)

*** Bug 134701 has been marked as a duplicate of this bug. ***

*** Bug 155194 has been marked as a duplicate of this bug. ***

*** Bug 133915 has been marked as a duplicate of this bug. ***

moving.

*** Bug 213981 has been marked as a duplicate of this bug. ***

Hi all, I just had to comment on this bug and the related bugs 173564, 218372. I must "respectfully" and whole heartedly disagree with your resolution to fix it as a "Evangelism" bug. I think this should be evangelism towards the Mozilla developers. This is not just limited to .rar files but almost any file type. I have encountered this in a number of different situations. When IE, Opera, Konkeror, you name it, correctly downloads these as files yet Mozilla and friends do not, that is a problem. This is not an isolated occurance with only a certain type. It could also happen on a GET, POST. Sometimes the user has no choice in the matter. I agree in a perfect world your solution is good enough. But this isn't a perfect world. Server admins make mistakes, forget things. Sure it may be a problem with Apache. Maybe Apache should fix it. If I need to get a certain file, I know I for one am not going to take the time to search there web site, or do a whois search for the web admin email address, send them an email, wait days or weeks, or months for them to get around to make this one stupid little change, and then finally use Mozilla to download it. NO. I am going to just use something that works the way I want it to, like IE or Opera. I don't care about what W3C standard was broken for this to work the way I wanted it to, it just works. But my point is this. You have taken the choice away from the users and given it to the web server. You have taken the users choice away from them to use Mozilla. You have forced them to use an alternative. Ok rant off. Why not give back the users the choice? Make a preference setting, menu option, something. (Right click may not always work i.e. POST) Isn't that what Mozilla is all about. Choice. Please reconsider this. See I said please. Again this is just my two cents take it or leave it.

The standards-violating MIME type second-guessing that MSIE does (and now, unfortunately, Opera too) is infuriating to me... I can't even create a Perl script which outputs plain text results (something I do a lot for system-maintenance CGI scripts) without it likely failing in MSIE (while working fine in Mozilla). Usually, MSIE prompts to download the results as a binary file or some stupidity like that. Try this: http://webtips.dan.info/cgi-bin/plaintext.pl Or, for more fun: http://webtips.dan.info/cgi-bin/plaintext.pl/test.exe http://webtips.dan.info/cgi-bin/plaintext.pl/test.zip http://webtips.dan.info/cgi-bin/plaintext.pl/test.gif http://webtips.dan.info/cgi-bin/plaintext.pl/test.html A standards-compliant browser will display all of these as plain text.

Is this bug now resolved with the mime type sensing that was added to Fx 0.8? - rmjb

I think so. could someone mark it fixed?

Simple question : does the lizard have this type "sniffing" too or not ? As reporter of this bug (and not using mozilla right now), I wanted to know if mozilla is as smart as firebi...fox now :) If so, this bug can be closed, or am I wrong ?

This bug is still here. With .wmz files as well as .rar files that are sent as text/plain I see the hex code onscreen. Even more frustrating are the pages with right-click disabled, and the code in an inner frame so that the file can't even be saved with a 'save page as'.

Sorry, that last was supposed to be .wmv files

crassius@hotmail.com, the bug has really been fixed. What version of FF d you have, and please give an example link of where your problem occurred.

my comments above were for mozilla 1.6 rather than for firefox 0.8

I just tested downloading RAR files from free.fr. It works fine now. Who is able to close this bug? It seems resolved now, at least when using Mozilla 1.7 RC2. If this bug still exist with other websites, this bug must be closed anyway as it is related to free.fr only, as mentioned by the reporter.

Can you gave me an url for a rar file hosted on free ? The old url is not avaiable anymore. Removing it.

Frederic, have a look at: http://ferdi67.free.fr/ The first link on this page points to a RAR file. Some other RAR files are available at: http://www.infonet.civ.pl/download/ In both cases, I have no problem downloading these files.

Thanks for the new link. Works ok in both mozilla 1.7rc2 (Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7) Gecko/20040514) and in my yesterday homemade based AVIARY firefox build (Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7) Gecko/20040517 Firefox/0.8.0+ (MozJF/Pentium4)) So closing bug as fixed.