Mike the Mad Biologist

Which for those of you who don’t know what the Mad Biologist’s Pentultimate Political Philosophy is, it’s very simple: people have to like this crap. Recently, I upgraded to Firefox 4 and I’ve been having ‘stability issues’, although they seem to have decreased in frequency somewhat. Which brings me to this excellent post about the increasing unreliability of personal computers:

Here’s one we all know well: you visit a page you visit everyday, probably a page you visited just minutes ago. Nothing has changed on your end, but suddenly the page locks up. The little egg timer tells you the page is loading, but it never finishes loading. At some arbitrary point — there’s no indication of when to give up, you just have to guess — you decide to give up and try to close the page. But the page won’t close. It’s THAT stuck. Is it really that difficult to close a page that is stuck? If so, fix that shit guys. It’s been going on for years and it’s gotten worse, not better.

I use a Mac, but this is also familiar (italics mine):

Of course one can always go into Task Manager, and manually close the page. And nine times out of ten when we do that we get the message “The program is not responding.” Reeeeally? You mean, it’s acting like it’s locked up or stuck maybe? You know, the whole reason I had to resort to task manager in the first place? If the PC can detect the program is not responding, and you have asked that program to stop because it is not responding, I don’t need to be told it’s not responding, I just need it to close. That’s the whole reason I opened up the goddamn task manager and tried to close it. The solution is obvious, the program should be stopped. Break the goddamn connection, and can you do that without shutting down every open tab? Fix that shit guys, it’s ridiculous and it makes people not like you.

Amen Selah to that. There is one good thing to come from all of this though:

On the lighter side, I guess we needn’t worry about Artificial Intelligence taking over the world when computers can’t close a stuck webpage on their own.

Comments

#1 Mark
June 13, 2011

The webpage isn’t stuck, the browser itself is. Chrome uses individual threads for the tabs so that on tab locking up (usually) does not cause problems to any others. The usually has to do with things that are shared like flash.
#2 Who Cares
June 13, 2011

There is a good reason for the pop-up telling the user (again) that a program isn’t responding are you sure you want to kill the program. The reason is data integrity. If a program is writing data to permanent storage you don’t want that file to be corrupted because the program is closed in the middle of that write action.

But that is not the case here. Darksyde is talking about a browser trying to make contact/download data from a site and locking up a page.
To answer his question: Depending on what that page was trying to show it can be that difficult to close the page.
But that has to do with the components that the browser has to use so it can render the page. Just one has to have some form of shutdown protection and it can block the page (and in some cases like Flash the entire browser) from responding if for some reason it never exits this protective mode. This is the reason that Firefox runs all add-ons , plugins and what else in a separate process. So that at least the browser can (or to be more precise should be able to) respond even if there are components misbehaving.
#3 Eric Lund
June 13, 2011

It’s not just Firefox. I use Safari, and every once in a while some page I visit routinely will cause Safari to throw out the spinning beach ball. In extreme cases (it’s even happened on the ScienceBlogs page) Safari crashes completely. And I can’t do anything else in Safari as long as the beach ball is spinning.

While we are on the subject: For more than a decade now (dating back to when Netscape ruled the browser world), web browsers have had something called a cache, the ostensible purpose of which is to cause pages to load more quickly because items are stored locally and don’t have to be retrieved from some distant corner of cyberspace. So why is it that, when I push the “Back” button, the browser seemingly has to go through the ritual of completely reloading the page? More specifically, if the page in question is loading very slowly or even timing out (yes, I’ve observed that symptom too, and I have to tell the browser to stop loading a page that has gotten stuck in a timeout situation), why can’t the browser use the cache to display the page as I saw it just a few minutes or even seconds ago? Typically, I want to go back and follow some other link from the page I’m backing up to–in this case I don’t care if I get a slightly old version (that’s what the reload button is for), I just want to see the previous page so I can follow the next link.
#4 Hercules Grytpype-Thynne
June 13, 2011

Not sure what you mean by “pentultimate”. If you mean “penultimate”, I may have to get all Inigo Montoya on your ass.
#5 Who Cares
June 14, 2011

@Eric Lund:
The browser can only restore a page from cache if the page in questions has been completely loaded.

Even then you can get a non responsive page when doing the back action. If the page has a component on it (preferably at the top so that it stops the page from displaying from the start) that when activated notices that it is out of date and needs to get data from a website that just happens to be not very responsive. Eventually your browser calls it a day and times out. Seeing that it never got the entire new page description (even though it is functionally the same as the old) and that the page you went back to isn’t the the same (due to that one component claiming it isn’t the same) it can’t rebuild it from cache.
Even if the page is merely slow instead of unresponsive the browser has to load in the entire page again due to that one component saying it’s different making it so that the page it has in cache isn’t the same at the one it is trying to display.

It is just that the only way to find out that just 1 part of a page has changed is by reloading the entire page and then comparing new and old.

The only good thing in all these situations is that the cache has certain components (mostly images) already on your computer so that when the website says I need the image from location http://farm1.static.flickr.com/27/64797071_cadf3e2637.jpg your browser instead of following that link to download the image gets it from the cache.
#6 Art
June 14, 2011

Firefox 4.0.1 is definitely a step back from Firefox 3.xxx, or whatever directly preceded it. It loads pages more slowly. Is less stable. Often ends with the page in a white on white blank with a progress bar stuck at 90% or more. It clearly locks up more and is generally less responsive. Anyone thinking about upgrading to it … don’t.

To add insult to injury they did away with the status bar. They don’t warn you. And they don’t make not having a status bar optional. They just take it away, so you add-ons disappear. After poking around on the ridiculously slow to load Foxfire site for an hour you might hit the exact combination of search terms that reveals that yes, you may indeed have your status bar back … if you download another add-on and restart Firefox once again.

This is getting to be like Windows 98 where every little change requires a restart. I guess they figure it is painless to restart the program … I mean like everyone who counts has a 30MBps connection so reloading those seven windows and ninety-seven tabs are reloaded in a thrice.

Yeh … reloading ninety-seven tabs takes several forevers over a slow connection. Particularly when the damn browser won’t render shit on most of the pages until the entire, every last little stinking bit, of the page is tucked into bed and read a bedtime story. So you can’t be reading ahead or stay entertained waiting for a page to load. Thanks Firefox.

I got Firefox when it was still quick, feisty, nimble and responsive. Now it is scarcely more svelte than your typical MS bloatware and is rapidly becoming just as arrogant and controlling.

I downloaded Chrome and found out why Firefox got rid of the status bar. Bottom line is that Chrome has a semi-transparent status bar that jumps around so the folks at Firefox just had to have one also. Evidently they are all the rage while simple and functional status bars are just sooooo 2008. Never mind that they do useful work. Baaaa.

Bottom line here is that no amount of eye candy and features can make up for the basic product not working. No feature or improvement can make up for the loss of basic functionality. It has to work.

This last release has been a huge disappointment.

Good luck getting a straight and cogent answer on their support site.
#7 tenine
June 14, 2011

I use Safari and find that the “spinning beach ball of death” can be neutralized by opening Activity Monitor and force quitting flash. PCs may have fewer such problems as flash is better optimized for windows.
#8 alan
June 18, 2011

“At some arbitrary point — there’s no indication of when to give up, you just have to guess — you decide to give up and try to close the page.”

Yep, the coder has the same problem and the same solution.
#9 Reinste ventures
June 20, 2011

Actually the problem is not in the webpage but browser has the problem. So stop blaming the webpage .