cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Firefox 98 - PDF Dowload and view

roxpec
Making moves

Managing PDF downloads and viewing has become a big problem with version 89. I am a programmer who has been working on healthcare applications for a few years and there are scenarios where it is convenient to download and in others where it is convenient to preview the document. These scenarios vary from application to application, but also within the same application. So these types of behaviors should be left to whoever designs the application, not to the browser. Today I have a lot of pissed off customers because it is no longer possible to download a PDF file without previewing.

88 REPLIES 88

PCB
Making moves

I have a similar issue.  Much of my job consists of OPENING files from backend applications via the web.  Now I have to download them BEFORE I can view the contents.  Not everything needs to be saved but EVERYTHING I do must be viewed.  So far I have wasted an hour just trying to find out what I was doing wrong only to find out that some DA at Mozilla thought they knew what was best for EVERYONE!!!!  Not a good idea and if not corrected I WILL be switching browsers.

Any files you download will be immediately saved on your disk. Depending on the current configuration, they’ll be saved in your preferred download folder, or you’ll be asked to select a location for each download.

Previously files were placed in the temp folder - is the issue you're having that the temp folder was preventing specific issues with your set-up?

in previous versions you could open a pdf file DIRECTLY on a browser tab to view it through the firefox plugin. Now the file just downloads to the system "downloads" folder or whichever folder the user has configured FF to save files at, without giving you the option to view it while it LITTERS your downloads folder with files you don't need saved.
Genius..... NOT!

I feel the same. I work with a lot of pdf files and most of them I don't need to save in my computer. While it were saved on temp folder, I could deleted all files cleaning this folder. Now I have 2 problems: open all files and after delete them from my Download folder or deleting each them just after open it and close it. Because the download list on browser is limited to 5 files. Or I have to open the complete download list and remove each one from my computer. Personally, I dislike this feature. I take longer with files I don't need to save.
What I need is a way that I can open the file with the program I choose or on browser without saving it on my computer automatically. On new update the programmers could create a way that we can stop this download automatically for those who don't want it.
I made a downgrade on the browser to use it the way it was before the 98 update and faced a new problem: create a new profile and even though I had synchronized all my data with the server before the downgrade, the new profile keeps getting empty. The new feature in 98 update is not good and I lost time fixing it!


The new feature in 98 update is not good and I lost time fixing it!

Yeah, that's why you should never d/l an update immediatelly. Just wait a few days to 1-2 weeks before getting the latest version. By then you 'll know if the new features have been a mistake and have had workarrounds developed or if they re here to be shamelessly shoved down your throat.

Firefox updates are another mystery. When firefox struggles to start, it is surely updating itself. I have tried several times to disavow them but there is no way. ERS builds sometimes become regular versions. The 97 update hit a lot of customers. I'd say the updates are out of control.

Yep, I spend that last portion of my day deleting files from my Downloads folder. And don't get me started with an issue with zip files.

Yeah, I don't have time for that, and it's begging to accidentally delete files I wanted to keep. Hopefully I can stop my work PC from updating on monday morning, otherwise I'm going to lose my sh*t


@Rm4g3dD0v wrote:

in previous versions you could open a pdf file DIRECTLY on a browser tab to view it through the firefox plugin. Now the file just downloads to the system "downloads" folder or whichever folder the user has configured FF to save files at, without giving you the option to view it


The default action for PDFs still should be "Open in Firefox" (formerly known as "Preview in Firefox"), and if you select that on the Settings page (support article), then PDFs should be saved in the web content cache and displayed at their URLs instead of downloaded and displayed from a file: URL.

If that option is not present, you could start a support thread at https://support.mozilla.org/questions/new/desktop/form.

Every single time Firefox pushes out an update they switch everyone back to use Firefox as default PDF viewer, even though we set the default for what WE WANT!!!

What is going through the mind of Firefox developers, where they think pissing off their users is the best choice????

I work at a small company, where after every update, I'm forced to go around and fix everyone's Firefox preferences!

Now this downloading temporary files to download, or last used directory, is another idiotic move!

Windows gives us temp directories (complete with user's TEMP environment variable)  for the specific purpose of TEMPORARILY storing files we don't want to keep!

I finally stumbled onto the about:config pdfjs.disabled boolean to stop Firefox shoving their pdf viewer down our throat, I just wish there was a use temp directory for temp files variable

Hell, I wish every new "feature" came with it off by default, with a prompt  with instructions on how to enable it!!


@Ray wrote:

Every single time Firefox pushes out an update they switch everyone back to use Firefox as default PDF viewer, even though we set the default for what WE WANT!!!


Hi Ray, I can only judge from what I see on forums, but I don't think that is happening in normal updates. It sounds as though your users' handlers.json files might be getting deleted for some reason, restoring the default download handling settings.


I work at a small company, where after every update, I'm forced to go around and fix everyone's Firefox preferences!


You may want to take a look at using group policy to manage settings:

You can use the Firefox for Enterprise support forum for guidance on using Policy, and you can create new Issues on that Github repository if the policies do not work as documented or you need additional policy options.

Custom content/action settings for other file types remain across updates, the only thing changed is "always ask" for PDF's gets changed to use Firefox.

As for forums,

https://connect.mozilla.org/t5/discussions/don-t-change-my-behaviour-for-pdf-treatment-to-auto-open-...

 

Further, Active Directory is not a viable option.

Most small companies don't have a budget for for that, never mind the dedicated staff needed to manage it.

I was finally able to fix it with the  about:config pdfjs.disabled boolean , but it, like every other "feature" should not be enabled by default, unless there is specific, easy access disabling instructions.

I agree. I want to open a .pdf to view it, or download it with an option to where it is put, NOT download it if I just want to look at it or print it. Like it was before these last two versions. Otherwise, I need to find the unwanted file and delete it. No time for that. Firefox assumes that I want the file without giving me the choice. I'm using Windows 10, always updated. This is very annoying.


@EPM33 wrote:

I want to open a .pdf to view it, or download it with an option to where it is put, NOT download it if I just want to look at it or print it. Like it was before these last two versions. Otherwise, I need to find the unwanted file and delete it. No time for that.


As you noticed, Firefox 98 changed from saving downloads in the Windows Temp folder (or Linux /tmp folder) for "Open with [relevant application]" to saving them in your designated downloads folder (the one next to "Save Files to" on the Settings page). It's not possible to pass a file to an application without saving it somewhere.

This change addressed problems with data loss for users who could not re-download files they forgot to save after they were removed from the Temp folder. However, now users who preferred those files to be "out of sight, out of mind" are faced with the task of cleaning them up manually.

In Firefox 102, the developers added a hidden option to use the Windows Temp (or Linux /tmp) folder for those downloads, similar to Firefox 97 and earlier. This is hidden away, so here's how to find it:

(A) In a new tab, type or paste about:config in the address bar and press Enter/Return. Click the button accepting the risk.

More info on about:config: Configuration Editor for Firefox. Please keep in mind that changes made through this back door aren't fully supported and aren't guaranteed to continue working in the future.

(B) In the search box in the page, type or paste browser.download.start_downloads_in_tmp_dir and pause while the list is filtered

(C) Double-click the preference to switch the value from false to true (or click the toggle button toward the right end of the row)

If you notice any problems with it, please speak up. I don't think many people are using it yet, so feedback is needed.

 

"This change addressed problems with data loss for users who could not re-download files they forgot to save after they were removed from the Temp folder. However, now users who preferred those files to be "out of sight, out of mind" are faced with the task of cleaning them up manually."

Surely it is an acceptable behavior for some users. But I think a user who forgets that he has downloaded a file (or loses a file) does first to re-download it than to find it inside a folder full of files.

 


@EPM33 wrote:

I want to open a .pdf to view it, or download it with an option to where it is put, NOT download it if I just want to look at it or print it. Like it was before these last two versions. Otherwise, I need to find the unwanted file and delete it.


You might have received a notification of another reply, but I'll duplicate the info here just in case.

Firefox 98 changed from saving downloads in the Windows Temp folder (or Linux /tmp folder) for "Open with [relevant application]" to saving them in your designated downloads folder (the one next to "Save Files to" on the Settings page). It's not possible to pass a file to an application without saving it somewhere.

In Firefox 102, the developers added a hidden option to use the Windows Temp (or Linux /tmp) folder for those downloads, similar to Firefox 97 and earlier. This is hidden away, so here's how to find it:

(A) In a new tab, type or paste about:config in the address bar and press Enter/Return. Click the button accepting the risk.

More info on about:config: Configuration Editor for Firefox. Please keep in mind that changes made through this back door aren't fully supported and aren't guaranteed to continue working in the future.

(B) In the search box in the page, type or paste browser.download.start_downloads_in_tmp_dir and pause while the list is filtered

(C) Double-click the preference to switch the value from false to true (or click the toggle button toward the right end of the row)

If you notice any problems with it, please speak up. I don't think many people are using it yet, so feedback is needed.

 

 

Hi rtestard,

Indeed this change of behaviour seems to impact quite a lot of people who seem to have a different workflow as the Mozilla Devs thought of.

There are indeed plenty of files that you access on the web (note the word "access", not "download") that you just want to "view" and then forget. Examples include French social security website (60 million users), many utility website (70million+ users), banks, etc. where to view an invoice or a statement, you have to access a PDF or other types of files. In the immense majority of cases you do not want to "save" the file, but just view and forget. And no, I do not want to view the PDF in Firefox, I want to open then with Adobe Reader, I have my reasons.

From a user standpoint, you have 2 actions: "View" and "Save": "View" is when you open the file with a 3rd party app (or Firefox itself) with no intention of keeping it. "Save" is when you want to download it and keep it, i.e. in your "Downloads" folder.

Most people won't understand that Firefox used to download the files in the "View" case (yes it has to, to open the file with 3rd party app). Of course it was hidden from them in the "temp internet files" folder. But this behaviour had the benefit of segregating the downloads into "View" and "Save".

And as some other users mention, the very situation you were trying to avoid (loosing the integrity of a file from the temp folder) is now morphing into a risk of deleting files you did not want to from your "Downloads" folder because it is littered with garbage you never intended to "save".

My 2 pence recommendation would be to add an option to set a specific destination folder of the "open" file action. It could be prefilled with Temp Internet Files which has the benefit of being included in the Windows disk cleanup utility (yes I know, this is for Windows users only). This way, the "Downloads" folder can remain for the files that we intentionally save.

Take care and thanks for all the hard work!

rtestard
Employee
Employee

Hello, can you please help clarify if this issue is specific to pdfs with content disposition:attachment headers?

In any case users can right click on links that point to pdf documents and select "Save Link As..." to download the pdf document directly without viewing it.

My application has a texButton with the following code:

var url = mysite.it/DowloadServerlet?myfile.pdf

window.open (url, '_ self', '');  

The DowloadServerlet add this header

response.addHeader ("content-disposition", "attachment");


@roxpec wrote:

The DowloadServerlet add this header

response.addHeader ("content-disposition", "attachment");


Yes, Firefox now ignores that header for PDFs. In theory, you can force a save by using a non-standard Content-Type header, for example:

response.addHeader ("content-type", "application/x-download-pdf");

But if the user prefers to have the file open directly in Acrobat, they need to take an extra step: open the download list, right-click the file, then click "Always Open Similar Files". That item isn't always available, so fingers crossed you can find a content-type that works.

 

Why don't you put it back the way it was this is a pain for most anyone?

Hi,

Thanks for the suggestion, I did a test and it works in FF v98.0.2 and Chrome. I'm hoping that using a non-standard contet-type comes back at me in the form of another problem. The fact is that this new download feature does not allow the user to choose the file name, I hope my users will get over it.

I suggest you however to investigate the problem, putting the correct content-type my GWT application made logOut as if the window.open rewrote all the content of the page with the pdf.

var url = mysite.it/DowloadServerlet?myfile.pdf
window.open(url, '_ self', '');

 


@roxpec wrote:

Hi,

Thanks for the suggestion, I did a test and it works in FF v98.0.2 and Chrome. I'm hoping that using a non-standard contet-type comes back at me in the form of another problem. The fact is that this new download feature does not allow the user to choose the file name, I hope my users will get over it.


There is a potential problem with future uploads to websites that filter the allow uploads based on content-type if Firefox changes from identifying PDF files as application/pdf to a fictitious content-type. This was an old bug and maybe it was fixed by now.

If users want to choose a name for a download, they need to switch from "Save files to" to "Always ask you where to save files". Is that a change from how it worked before?

 


I suggest you however to investigate the problem, putting the correct content-type my GWT application made logOut as if the window.open rewrote all the content of the page with the pdf.

var url = mysite.it/DowloadServerlet?myfile.pdf
window.open(url, '_ self', '');


The problem is caused by ignoring content-disposition: attachment, combined with your script directing the PDF to open in _self. I assume it was written that way instead of _blank to avoid leaving a blank tab. I wonder whether the blank tab bug was fixed so you can direct the PDF to _blank again?

 


@jscher2000 wrote:

@roxpec wrote:

Hi,

Thanks for the suggestion, I did a test and it works in FF v98.0.2 and Chrome. I'm hoping that using a non-standard contet-type comes back at me in the form of another problem. The fact is that this new download feature does not allow the user to choose the file name, I hope my users will get over it.


There is a potential problem with future uploads to websites that filter the allow uploads based on content-type if Firefox changes from identifying PDF files as application/pdf to a fictitious content-type. This was an old bug and maybe it was fixed by now.

If users want to choose a name for a download, they need to switch from "Save files to" to "Always ask you where to save files". Is that a change from how it worked before?


It was changed only by setting this parameter browser.download.improvements_to_download_panel = false the filename is asked otherwise the name suggested in the url is used.


I suggest you however to investigate the problem, putting the correct content-type my GWT application made logOut as if the window.open rewrote all the content of the page with the pdf.

var url = mysite.it/DowloadServerlet?myfile.pdf
window.open(url, '_ self', '');


The problem is caused by ignoring content-disposition: attachment, combined with your script directing the PDF to open in _self. I assume it was written that way instead of _blank to avoid leaving a blank tab. I wonder whether the blank tab bug was fixed so you can direct the PDF to _blank again?


I did a test and now it also works with:

var url = mysite.it/DowloadServerlet?myfile.pdf
window.open(url, '_blank', '');
response.addHeader ("content-disposition", "attachment");
response.addHeader ("content-type", "application/x-download-pdf");

You cannot assume that in order to perform a download the user has to do "Save link as". do you think all world applications are made as you say? If there is a specific w3c I might as well accept it. But suppose that this management done on the pdf you do it on the images ....

Emmanuel_RA
Making moves

I found a solution in a forum. First, go to about:config look for "browser.download.improvements_to_download_panel" and finally set that flag to false.

It worked for me, hope the same for you 

THANK YOU!!!!  This worked perfectly for me!  What a stupid update that was! 

Many thanks Emmanuel_RA. Now it works as before. The new automatic download feature was very annoying.

This worked for me. We hope that it will be handled appropriately in the future

Thanks!! It worked perfectly here. I searched the internet but found nothing. You helped a lot! Your post should be pinned to the top for those who have the same problem.

roxpec
Making moves

Hi,

Are there any updates on this issue? will a patch be released? or will we have to empty the download folder with all the files unknowingly thinking that the website was not working?

Well, the best fix was supplied by Emmanuel_RA above:

 

First, go to about:config look for "browser.download.improvements_to_download_panel" and finally set that flag to false.

 

I guess you could call it a "setting". I've been running this for several weeks, and problem solved.

There is no guarantee that this setting will be maintained in the future, unless Mozilla acknowledges the need for a different behaviour.

About config false no longer works with the later fire fox updates.


@Marcus wrote:

About config false no longer works with the later fire fox updates.


What version are you running now?

Which parts of the reversion aren't working?

(What should happen in Firefox 98-100 when you toggle the preference to false is: (1) Download dialog pops up for unknown content types instead of saving automatically, and (2) Choosing "Open with [relevant application]" uses the Windows Temp folder rather than your default download folder.)

I had been using version 98, and made the mistake of going to 100. I have my about config set to false but no download prompt, still automatic. I selected all downloads to temp folder, but I should have the prompt with download location option, option to cancel or option to simply open without saving. I was fine with just moving to another browser, but first wanted to let people know that the about config set up, which is hidden from the users anyway, no longer worked. However, I'm also told a future version of fire fox, hopefully between 100.XX and 101 will be bringing back the download prompt right from the settings tab, so I guess I will give Firefox yet another chance when that happens.

I will go to Opera


@Marcus wrote:

I'm also told a future version of fire fox, hopefully between 100.XX and 101 will be bringing back the download prompt right from the settings tab, so I guess I will give Firefox yet another chance when that happens.


Firefox 101 will allow switching from "Save file" as the default action for unknown content types to "Always ask". This is how it looks in the beta:

Fx101-settings-applications-other-files-DE.png

I just noticed the "Save files to" / "Always ask you where to save files" above that changed from two radio buttons to one checkbox. I wonder why?

What about extreme use of ram and processor?