Created attachment 83169 [details]
screenshot of kpartsplugin in the browser window

I apologize that this is in the wrong category. This is for kpartsplugin, but it does not show up in the list.

I'm using Firefox, but the behavior is the same in all installed browsers that support kpartsplugin.

When I click on a download link, the page rendered by kpartsplugin opens. The attachment is the screenshot of the results. Please look at the picture before responding.

I changed the action in Firefox Preferences so kpartsplugin is not used for archived and compressed files. Unfortunately, many websites do not provide a direct link to the file. Instead, they use a script page that automates the download. In some cases, this appears to be a page that has a mimetype, like php, that kparts tries to open. In cases like this, a right-click and save as does not download the desired file.

I also note that html, xhtml, and xml files are listed as using kpartsplugin. If I try to change this to Firefox, it opens a new window. The option to handle them internally is gone.

In some browsers, the preferences to change the behavior of kpartsplugin is simply not available. Additionally, there preference to disable the plugin is not available. I believe that kpartsplugin is listed as capable of rendering way too many mimetypes. This is an undesirable behavior and is a bug when it takes over downloading files.

One may say that the file still gets downloaded, but kpartsplugin slows down the process. It actually takes longer to download a file and the window takes extra time to render while using extra computer resources. I basically have to wait for the window to render before I can do anything else in my browser. It also creates a cached copy of the downloaded file. The cached copy must then be copied to the desired location. The file does not show in the download history, so a download cannot be restarted or continued as necessary.

Additionally, sometimes the file is not downloaded, so I have to go back in my web history. Clicking back the first time appears to do nothing. Then the download page sometimes must be bypassed. The whole process greatly annoys me.

The solution is to limit the list of mimetypes to multimedia and formatted documents (like office documents) not normally supported natively by web browsers. Any browser can render HTML, so that mimetype has no reason to exist. Compressed and archived files are almost always downloaded then opened later with separate program. In my opinion, images not normally rendered natively by web browsers should also not be included. This basically leaves sound, video, pdf, and Word documents.