Jump to content
Forum upgrade is live! Read more... ×
Sign in to follow this  
Master_Scythe

Browser caching plugins?

Recommended Posts

Just curious;

 

I recall from years gone by, the 'speed' of the internet, or more specifically, of browsers.

Remember when you'd have at least 25 "Back" steps that just INSTANTLY loaded? and if you went back too far you could FORWARD and be at the page INSTANTLY?!

 

Well, I'm finding these days that with the way I operate (with a billion tabs, and re-checking numbers and such) that this is annoying.

Every time I hit forward or back, thanks to the internets more active elements these days, the whole page basically reloads.

 

Anyway I'm rambling, you get the idea.

 

Is there an about:something browser setting I'm missing in firefox?

Or perhaps a decent plugin that will correct the caching and ignore more live elements.

 

I'm already running Disconnect, AdBlockPro and uBlock Origin; as well as have the MVPS hosts file added to my windows machine.

 

anything we know of?

Share this post


Link to post
Share on other sites

The move has been towards a synchronous type of 'net experience. Some is good but some utterly shits me.

Not much a fan of predictive typing that tries to out-guess you as you type into Youtube or Google Search. Definitely not a fan of the Stop button going away (I have an FF extension that brings it back).

 

Even though a page might have that bit of delay and appear to be doing nw activity it might not be necessarily. I just tried loading another thread (last page) then selected the previous one so there'd be 2 buffered. Then I unplugged my cable and was still able to freely use Back/Forward in the open tab to go between the two. But yeah, it does seem that these days the rule seems to be network activity even if there's no new data to be had.

 

Whether anything can be done about it, I sort of doubt it. Maybe an extra layer of caching for what that's worth, probably more bother than benefit.

Share this post


Link to post
Share on other sites

Not sure how such a plugin would work with most pages being pretty much dynamic especially crypto currency sites.

Share this post


Link to post
Share on other sites

Doesn't look like he's done a Firefox version.

 

I'm still on Waterfox - I can get Quantum almost to the functionality I need, and would probably need to piss around with CSS mods to get some of the rest.

Such a PITA that so many extensions haven't been ported over yet. I keep a laptop updated and periodically check.

Extended support I think only goes until middle of the year so can't stick to the legacy versions much longer anyway.

Share this post


Link to post
Share on other sites

Not sure how such a plugin would work with most pages being pretty much dynamic especially crypto currency sites.

 

I would hope it would just show me old data, unless i forced a manual refresh?

 

One such example would be just before, in some down time, when I visited HobbyKing.

The WHOLE PAGE reloads, to check closest warehouse, update pricing, and so on.

 

However, I LITERALLY hit back, and went "actually, I want to check that whole category" and hit forward, within 3 seconds.

If my data is 3 seconds old, so be it, HobbyKing is a great example of a 'live' site that takes up to 20 seconds to load a page during peak times.

The re-load took a full 10 seconds.

Share this post


Link to post
Share on other sites

Long waits probably thanks to their refreshing policy... trivial transactions requesting an entire reload so it's using exponetially more resources than it needs to.

Share this post


Link to post
Share on other sites

I noticed mobile browsers just excessively refresh compared to desktop ones.

Share this post


Link to post
Share on other sites

The big difference there though is that all workspace is RAM based and it's a scarce resource to begin with.

Share this post


Link to post
Share on other sites

And then we have... Google Calendar.

 

Bad enough that just to create an event from an ICS file I have to goto settings, click Import/Export then click again and select a file.

Then you press the back arrow to return to your calendar and the event isn't there. You have to hit Refresh. Fark.

Share this post


Link to post
Share on other sites

Long waits probably thanks to their refreshing policy...

That's a good point.

 

Scythe, look in the offending pages' HTTP headers for the Cache-Control directive.

 

The "default" is 86400 seconds, or a day, so going back to a tab days or weeks later will result in the page reloading when you hit forward/back.

 

Also some content providers might not honour the If-Modified-Since directive. If the content predates the date specified, the server is meant to return HTTP 302, and the browser needn't re-download the page.

 

I can only assume that the reasons already given explain why it seems to be worse in more recent times.

Share this post


Link to post
Share on other sites

 

Long waits probably thanks to their refreshing policy...

That's a good point.

 

Scythe, look in the offending pages' HTTP headers for the Cache-Control directive.

 

The "default" is 86400 seconds, or a day, so going back to a tab days or weeks later will result in the page reloading when you hit forward/back.

 

Also some content providers might not honour the If-Modified-Since directive. If the content predates the date specified, the server is meant to return HTTP 302, and the browser needn't re-download the page.

 

I can only assume that the reasons already given explain why it seems to be worse in more recent times.

 

 

I'd assume you're exactly correct.

But surely there's something that will let my local PC use more cache space, and NOT ask the content provider for an update?

Why should my local machine care what a provider has asked for, that's what I'd love to override.

 

If you have a look at any product on hobbyking.com does anything jump out as being a key offender?

it's just one page I know does it excessively, and there's just too much JS for me to try and figure out.

Since reading became a chore, code is my downfall :( used to be my passion :(

Share this post


Link to post
Share on other sites

It isn't so much what the remote server "asks for", rather how they have configured their servers to instruct your browser, and how standards-compliant browsers honour those instructions.

 

As for Hobbyking...

 

dhvs5MH.png

 

Said the web server to your browser: "Don't cache shit!" Actually, it just says don't cache the actual page. Every resource requested by HTTP can be controlled.

 

Some of the sub-requests, like for images and stuff, have different header requests and responses, like so:

 

PYFPLyF.png

 

I don't know if Firefox's dev tools are any good, but in Chrome, open the dev tools, go to the Network tab, reload the page, and inspect the headers for each request. Check the directives that I've highlighted above. You can also tell if the request was served from memory or disk cache.

Edited by SquallStrife
  • Like 1

Share this post


Link to post
Share on other sites

wow thats simultaneously deceptively simple and needlessly complicated.

 

Thanks! I'll be inspecting some random pages and seeing if I can scriptmonkey them to behave better.

Share this post


Link to post
Share on other sites

I wouldn't say "needlessly complicated", it looks hard, but really it's just a "Cache this? Yes/No" instruction for every item the browser needs to retrieve from the server.

 

The issue, I think, is that HobbyKing has a lazy CloudFlare set up.

 

It's reasonable that they'd like the HTML re-downloaded on any navigation, so that, for instance, your shopping cart shows the right number of items, or in-stock status on a hot item is current. But there's no reason they couldn't use "Cache-control: private max_age=20" or similar on their catalogue listing pages to make quick successive navigations less painful.

 

No reason other than laziness.

 

I'm pretty sure the paid-for CloudFlare tools let you configure headers per-page or per-URL. If not they can exclude those pages from CloudFlare and set those headers on their own web server with header() in PHP or equivalent.

Edited by SquallStrife

Share this post


Link to post
Share on other sites

I wouldn't say "needlessly complicated", it looks hard, but really it's just a "Cache this? Yes/No" instruction for every item the browser needs to retrieve from the server.

 

I guess it's the evolution of users and online thinking.

The reason I was thinking 'gee that's complicated' is because I grew up (as I know you did) in a time when if you want the latest, you re-query.

If you cart has changed, you refresh the page, don't wait for it to update via an active element.

 

I still live in that mindset of "Refresh the page for latest version" or more specifically FORCE (no cache) refresh to be sure.

But I understand thats not the norm.

 

The more exposed to 'modern' tech I am, the more I think we've given up too much control for lazyness.

We could certainly have given some, some things were FAR too manual, but there's a level of control:convenience I think we're passing....

Share this post


Link to post
Share on other sites

i would never want to go back, but this is definitely a downside of all this real time shit.

 

its why i sometimes open a bunch of tabs and then turn off javascript (via FF ext button), or even resort to pathetically taking screenshots

 

what i dont understand is why some tabs still lose their shit even after js is off...

Share this post


Link to post
Share on other sites

I haven't even bothered with JS off extensions.

 

Plenty of sites use it and many just become unusable when you disable it.

What shits me though is in the modern day there's still plenty with NFI how to program. Like using CPU hogging delay loops instead of the proper way that uses timer based wait facilities.

And the fact that practically all browsers run all the tabs in a single thread makes it worse.

Share this post


Link to post
Share on other sites

I haven't even bothered with JS off extensions.

 

Plenty of sites use it and many just become unusable when you disable it.

What shits me though is in the modern day there's still plenty with NFI how to program. Like using CPU hogging delay loops instead of the proper way that uses timer based wait facilities.

And the fact that practically all browsers run all the tabs in a single thread makes it worse.

 

And the fact that people think Chrome == Chromium, and "I dont care" about the extra DNS hop EVERY PAGE AND LINK goes through, when using Google search or Chrome.

"It might be milliseconds, but its literally DOUBLING your DNS query time!" *nerd rage* lol

Share this post


Link to post
Share on other sites

Yeah, annoying with Google. They used to link direct but now in order to capture their stats (and provide a referrer?) they have that middleman which in fact probably adds 2 seconds to the process.

Share this post


Link to post
Share on other sites

That's not Chrome doing that though, it's Google itself, symptom is there in Edge and IE too (no Firefox on work PC to test).

 

Try it in Opera or whatever. Search for something, right click a result, copy link, paste to Notepad.

 

I have an addin in Chrome that makes the search result links go directly to the actual page,

Share this post


Link to post
Share on other sites

I know that it's the page and not a browser specific thing. Nice that you can get an extension to overcome it. Don't have Chrome installed on this machine.

 

Strangely enough... it looks like Vivaldi is doing it direct by default (it's a Chrome based browser).

Share this post


Link to post
Share on other sites

I was more responding to this:

 

And the fact that people think Chrome == Chromium, and "I dont care" about the extra DNS hop EVERY PAGE AND LINK goes through, when using Google search or Chrome.

"It might be milliseconds, but its literally DOUBLING your DNS query time!" *nerd rage* lol

It's only Google search that does this, not the Chrome browser:

 

gKNMN0t.png

 

Couple of spurious hits to *.google.com after navigating the forums for a bit. Probably for ads, I don't have an ad blocker on this PC right now. Certainly not "EVERY PAGE AND LINK".

Edited by SquallStrife

Share this post


Link to post
Share on other sites

I was more responding to this:

 

And the fact that people think Chrome == Chromium, and "I dont care" about the extra DNS hop EVERY PAGE AND LINK goes through, when using Google search or Chrome.

"It might be milliseconds, but its literally DOUBLING your DNS query time!" *nerd rage* lol

It's only Google search that does this, not the Chrome browser:

 

Couple of spurious hits to *.google.com after navigating the forums for a bit. Probably for ads, I don't have an ad blocker on this PC right now. Certainly not "EVERY PAGE AND LINK".

 

 

Perhaps they calmed it down; but there was a time when every DNS query was passed through google when using chrome.

I use DuckDuckGo and StartPage, and it was still "google hopping' before taking me to the page I asked for.

 

If they calmed that down, perhaps chrome browser isn't so evil... but I'll still stick with any of the Foxes (water, fire, etc)

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×