Comments on: Finding a Web Browser for constant page reloading

By: laur.ie » MonitorControls – Utilties for monitor management on Windows

Thu, 25 Nov 2010 15:49:05 +0000

[...] Utilties for monitor management on Windows 25th November 2010, 03:49 pm TweetWhen I ended up using Windows to power the overhead information screens at Last.fm, I lost the ability to have a one line crontab entry [...]

By: Sebastian Oliva

Sebastian Oliva — Fri, 19 Mar 2010 18:58:57 +0000

You have the option in Opera to reload pages in a stipulated interval, Check right click > reload every

That way you only need to change the content of the page to reload, and get rid of the JavaScript

By: John Kav

John Kav — Fri, 06 Nov 2009 04:28:56 +0000

Hi Laurie,
Thanks so much for sharing your experience. I have been trying to build a Ubuntu box (to be an appliance later) that runs a browser listening to an internet radio station. As the box will be installed in the comms cupboard running 24/7 it needs to be no maintenance. So I have been working down the constant page display road and have found also the wonderful mem leaks. Strangely though my original config of Ubuntu 9.04 Seamonkey 1.1.17 and adobe flashplayer 10 was very solid. We ran it many times over a 24-36 hour period and witnessed a 30-40 MB mem reduction over a 15 min period, however it also released that mem on a 15 min cycle as well. There was a period of weeks where we have not had access to the target hardware and were having problems replicating our results on different hardware, we assumed it was the different h/ware causing the issues. We have been down a number of other roads that have been much less stable and recently with access again to the orig h/ware went back to the config that ran well. NOW this config bleeds mem at the rate of knots and eventually stops. The web page source I have been told has had little change however we cannot now make the orig config operate as before…… Obviously there has been a subtle change in the web page, however the end game being so sensitive makes it pretty much useless as a production device. I hate to say it however I am going to have to try your solution, for now.
Thanks again JK

By: Laurie

Laurie — Tue, 02 Jun 2009 07:15:29 +0000

Aaron: Thanks for the reply. I have no reason to doubt that you’re absolutely correct, and that’s definitely worth looking at, but it’s still weird that Chrome is so much better at cleaning up after itself even in that situation.

By: Aaron C

Aaron C — Tue, 02 Jun 2009 00:43:37 +0000

I set up a similar kiosk and also had leaking memory issues. After wrestling with several different browsers, I rewrote my code a fixed the problem. It turns out that while some browsers definitely have a smaller memory footprint (Google Chrome wins!), browsers only leak memory when the code they’re running makes them do it. In the case of Firefox, a plugin could be the culprit, but with no plugins installed Firefox, just like all the others browsers, will only leak memory if the page they’re on is causing them to do it. It’s true. Don’t believe it? Try loading a blank page and watch memory consumption.

Moral of the story… Your monitoring system is the one leaking memory, not the browser it runs in. This “JavaScript Frame” you speak of is probably adding a new head and doctype to the frame with every page load, causing the browser to pull double DOM duty (then triple, quadruple, etc.) or something.

By: Laurie

Laurie — Mon, 01 Jun 2009 22:05:47 +0000

Max: It’ll be nice one day to have a crack at Chromium on Linux. When it’s ready, anyway

By: Max Howell

Max Howell — Mon, 01 Jun 2009 21:46:16 +0000

In my cross platform experience, when it comes to memory stuff, Windows wins. I’m not sure why, but it’s something to consider.

By: BB

BB — Sun, 31 May 2009 21:54:50 +0000

Very interesting read. Thanks for the post. This is the type of thing I would get carried away with!

By: martind

martind — Fri, 29 May 2009 08:30:20 +0000

just use javascript! e.g. with jquery http://ajaxian.com/archives/blink-at-the-marquee-one-more-time