Two Sites that cause fatal Fx problems

[Fatal Sites]

Two sites that, if I instruct NoScript to temporarily allow all, cause Firefox to go into a not-responding mode, will temporarily transfer me to Thunderbird and then back to Fx, cause Fx memory use to climb to well over 1,xxx,xxxK, and force the use of the Task Manager to close Fx.

These are links to RSS news feeds, both of which showed this problem today:

http://abcnews.go.com/International/vid ... l-22802999

http://atwonline.com/airports-amp-route ... y-shutdown

Disabling NoS does not fix the problem. I tried both links in a Fx profile that had only four extensions, including NoS; it was not until I removed NoS and restarted Fx that I could get the articles to work properly. I discovered the problems when trying to use the Read Now bookmarklet, which requires javascript for some sites. (The atwonline article does not really need Read Now, but it was the only RSS feed that I had to use as a test before posting this.)

I have notified both sites. ATWOnline has responded, and sent my emails to their technical department. I just sent a report to ABC News, so no reply as of now.

[therube]

You need to put your links in [url] tags to avoid breakage.

> Disabling NoS does not fix the problem

Then NoScript isn't an issue here, is it?

> Read Now bookmarklet

What version of NoScript are you running?
Lots of changes of late affecting bookmarklets.

[Fatal Sites]

Sorry, frustrated and spending a great deal of time chasing down the causes, did not even think about the url brackets.

If I remove NoS, and the problem goes away, that tells me that, yes, NoS is the problem. I'm open to other solutions, but as that is the only one I have found, that's why I consider NoS to be very much an issue.

NoS version 2.6.8.17

Problem(s) have only arisen over the past few weeks, which would jibe with your comment about recent changes affecting bookmarklets. I forgot about the Readability extension, which I stopped using when the bookmarklet was suggested as being better; I will give that a try.

[Fatal Sites]

Brief test: I changed to the Fx profile from which I had deleted NoS, reinstalled it, along with the Readability extension.

I brought up the ABC link that I had posted, told NoS to allow all temporarily, clicked on the Read Now button provided by the extension, that worked. I then found a longer ATW online feed, repeated allowing/clicking, and that, too, worked.

Based on this sample of two, I'm going to install the Readability extension in my regular Fx profile, delete, the bookmarklet, and see what happens.

[Fatal Sites]

I simply don't understand: I installed the Readability extension, removed the bookmarklet, restarted Fx.

I was looking at RSS feeds, selected a long one from ATWonline,and wanted to use Read Now. As ATW requires javascript, I instructed NoScript to temporarily allow all on that page, which sent Fx into paroxysms of not-responding, etc. Killed Fx, started the test profile, entered the ATW URL, instructed NoS to temporarily allow all on page, worked correctly.

Something or somethings is/are amiss, but what it might be, how I can fix it to allow news items to work properly--as they almost invariably did until a few weeks ago--is quite beyond me. The regular Fx profile in use was created 28Feb14, in an attempt to fix the problems with a somewhat older profile.

[Thrawn]

Instead of Temporarily Allow All This Page, can you try allowing domains one at a time, and see whether there is a particular domain that causes the problem?

[Fatal Sites]

A good deal of time spent on further testing has apparently shown that, and not for the first time, evidence that points rather strongly to a specific extension, is not entirely, if at all, correct.

Yesterday, I created a new Firefox profile, testing it each time as I added files (such as bookmarks) and extensions, using as my test a news feed from abcnews.go.com, that I knew had caused problems with the Firefox profile I was using on a regular basis. I intentionally did not install NoScript. The test passed every time, until I had finished creating the new profile, to my dismay, my having no idea why it had worked and then stopped.

Earlier today, I installed YesScript, and entered the ABC news url and the AWTonline URL, two sites that I knew caused the problems; when I read news feeds from those two sites, there were no problems, other than that Read Now would not work on them, a result that I expected.

A few minutes ago, a feed from Politico fouled up Firefox, so I added it to YesScript. I then had to use Task Manager to stop Firefox, as the misbegotten javascript does more than just causing the not-responding and other oddities I had been observing, it also prevents Fx from closing normally. I should mention that, as I have no knowledge of javascript programming, I have no way of knowing if the problem(s) are with the news sources or with Fx's decoder, or both. I have submitted a few Firefox feedback forms, and now that I have some specific and positive information, shall pass that along in another feedback form.

I have concluded that this is one of those instances where quite a lot of testing and time have pointed to, in this case, NoS, where I now know that NoS is not what was causing my on-going frustration. At least now I can read news feeds that, until recently (actual time span not known) caused no problems, and, as I have no actual knowledge of javascript nor had I encountered the new problems, I did not know how to easily fix.

I am still waiting to hear back from AWTonline, which did tell me that they had sent my email to their tech support; if I do hear, I will now be able to give them my precise findings, and shall do the same thing if ABC News answers my email. I need to see if I can find a way to contact Politico as well.

My thanks to those who have assisted on this.

[Thrawn]

Thanks for letting us know.

If you can provide exact URLs where problems occurred, then it might be possible for us to create surrogate scripts that will allow you to avoid the problem.

[Fatal Sites]

As I have blocked the sites, I don't have any specific URLs, and I really don't want to unblock them, given the serious problems they gave me.

What I do have is the site names that I have put into YesScript to avoid them:

abcnews.com
abcnews.go.com
atwonline.com
www.politico.com

These are sites that are accessed regularly as RSS feeds. Not every single item causes faults, but you might be able to just go to the sites and find articles that are, for whatever reason, what I would call faulty.

Please let me know if I may assist--so long as I don't have to unblock the sites, that is; one problem is even figuring out how to capture the URLs, as at least some of the articles cause Fx to go into irremediable not-responding cycles, for example.

[Thrawn]

You can capture URLs with ABE:
In Options-Advanced-ABE, add the following to the USER ruleset,

Code: Select all

Site .politico.com
Deny

then allow scripts for politico.com. ABE will block the requests and log them to the Browser Console.

[Fatal Sites]

I first had to figure out what ABE is; having done that, I now have to ask for an explanation of your comments about using it. I can easily capture URLs for blocked sites, so just using ABE for that purpose seems to be of no avail.

You say that I can unblock Politico, as an example, and then, using ABE, try to access it, but I don't understand what you mean by "ABE will block the requests and log them to the Browser Console." If blocking the request and capturing the URL is all that is all that it does, I'm not going to bother using ABE. Am I missing something about what it is that using ABE would do?

[Thrawn]

You said that you don't have specific URLs because you blocked the sites (and, understandably, you don't want to unblock them). ABE would give you specific URLs while keeping them blocked.

[Fatal Sites]

I'll just do it the simple way the next time that I get some RSS feeds, by merely copying some URLs, and then posting them here, complete with [url] tags; no need for my adding the confusion of using ABE.

It is obvious that I won't have any idea if those URLs will cause problems for anyone willing to test them, given that they will come from sites blocked by YesScript.

[Fatal Sites]

Question before I proceed: I put a link inside the [url] brackets, but when I previewed the message, the link did not show as a link, just as the code as I had entered it. Do such links only appear as clickable after the message is posted, or am I doing something incorrectly?

[Fatal Sites]

Please do not have read my previous message: The link I tried had lost one of the /s after http: , so it was not a URL.

Here are a few to try:


http://abcnews.go.com/International/wir ... a-22893293

http://abcnews.go.com/International/us- ... d=22894802

http://atwonline.com/finance-amp-data/t ... rofit-2013

http://atwonline.com/safety/malaysia-ai ... inaccurate