InformAction Forums

Greetings -

For a few weeks now I've been having rendering problems when using Firefox/NoScript to view some local pages via the file:// protocol. Some items on the page are misplaced vertically.

This mis-rendering issue only happens when the pages first load. If I click again on the bookmark for the page or if I refresh the page it displays correctly.

If I view highly similar pages using the http:// protocol the issue does not occur.

The pages in question comprise only HTML and CSS, there's nothing "complicated" (like any scripts) involved. The HTML and CSS have both been checked OK against the W3C validators. In any case these pages have all been working OK for years, up until a few weeks ago when things suddenly started displaying incorrectly.

At first I thought that an update to Firefox was causing the problems. However upon investigation I discovered that if I disabled NoScript it fixed the issue.

Of course I don't want to disable NoScript altogether, so for the time being I've reverted to NoScript version 13.4, and everything is displaying perfectly.

Immediately prior to this I was running NoScript 13.5.5 (though I suspect the trouble may have started prior to 13.5.5 - I was very busy during December and could not investigate at the time). NoScript was set to auto-update so I didn't really notice when the new versions went in.

Now that I've worked out how to install a particular version of NoScript, it should be possible to ascertain which was the first version to cause the problem. If I can do this then I will post back with further info.

best wishes, Peter.

Linux Mint 21.3 (Mate desktop)
Firefox 146.0.1 mint001 - 1.0

Could you please provide at least one of these malfunctioning pages? Thanks!

petergb wrote: ↑Sat Jan 03, 2026 5:38 pm Now that I've worked out how to install a particular version of NoScript, it should be possible to ascertain which was the first version to cause the problem. If I can do this then I will post back with further info.

Narrowing it down to a specific old development build can provide more insight than knowing only which release first introduced the issue. You might have already found this, but in case not, old NoScript dev builds are available here: https://noscript.net/feed?quantum&c=150&t=a

Hello again...

Thank you both for your replies.

I got rather side-tracked yesterday after I discovered that the problem was not actually 100% reproducible as I had previously thought. This made me suspect that it might perhaps be something to do with some garbage stuck in Firefox cache. However deleting all history and cache etc. on that machine was very inconvenient , so I did the next best thing which was to get hold of another computer with a similar software setup and copied the relevant files across. I discovered that yes, I could reproduce the issue on the other machine - just not every time.

Despite quite a bit of playing around I still don't know what it depends upon. The browser seems to go through phases when it goes wrong almost every time, and then at other times it's hard to get it to go wrong at all?! ...and this "moody" behaviour seems to be the same on both machines.

After that I discovered your posts and set about preparing a sample page to demonstrate the issue. There was a lot of personal information etc. on the original pages. After pruning that out and cutting things down to a minimum I've zipped up a set of files which should work.

In the first instance I will try to send the link for this via PM as the forum rules don't seem to keen on including links in forum posts!

I also recorded a short video clip showing the effect on my screen here, and will send the link for that too.

best wishes, Peter.

Please check latest development build, thank you:

v 13.5.5.902
============================================================
x [nscl] Generalize DTD serialization (file:// quirk mode
compat, thanks peterbg for reporting)

Thanks!

Unfortunately the link is currently giving me "404 Not Found"

page: https://noscript.net/getit/#devel
link text (under RC for Firefox): "Install NoScript 13.5.5.902 on Firefox Desktop by simply clicking here."
link target: https://secure.informaction.com/downloa ... .5.902.xpi

I shall try it again in a while.

best wishes, Peter.

Sorry, it's been a temporary network glitch preventing the proper downloads synchronizations.

Please try again now, thanks.

Thanks, I have it now.

Firefox grumbled a bit and didn't want to let me install it at first (boo, hiss) then I sussed the "open it in a different page" trick and it went in happily after that. My backup of NoScript settings has imported successfully.

I've tested the new version (13.5.5.902) about a dozen times so far, using the same method as on the video clip I sent you (launch a new blank tab, then go to the bookmark) - and it's looking good: No errors seen yet! Thank You!!

First observations:
- no rendering errors seen as yet;
- page seems to load more "smoothly" again (more like it used to do with v13.4)
- no brief appearance of the top RH graphic* flashing far too large and in the wrong place as the page loads**;

* I had to redact the original top RH graphic, so I replaced it with the picture of a cat
** this would only happen about 10% of the time, but it was highly noticeable and rather off-putting when it did!

However (remembering that I discovered the problem wasn't 100% reproducible) I think the next thing I should do is to "hammer" it a bit to see if I can persuade it to go wrong; so I will put it v13.5.5.902 on the other "test" machine as well, and then give them both some exercise.

So far so good though! ...and the way the pages are now loading smoothly (without "spluttering") is giving me a positive feeling. I will report back with more results a bit later on.

Thanks once again for this commendably fast response
...very best wishes, Peter.

(edited for clarity)

Hello again

I've now given v13.5.5.902 quite a lot of testing on both the machines mentioned above, and also on an old laptop (which although i5 based goes back to before processors had "Generation" numbers, it just says M520@2.4GHz) which I thought might be interesting as its graphics acceleration is probably very different to that of more modern machines.

I've now run over 70 tests and have not observed any rendering errors at all.

I use these pages many times every day, and if anything does turn up in the future I will come back and let you know. However as far as I can tell at the moment v13.5.5.902 seems to have fully fixed the issues which I had been experiencing. Thank you!

best wishes, Peter.