Skipping Extension Installation Delay

Firefox has a delay when you install extensions as a security mechanism. This is done because it would otherwise be pretty easy for a website to trick someone into installing an extension before they even realize what they are doing (which is obviously a bad thing). See Bug 162020 for details and even an example.

I’ve seen a few sites publicize how to disable this security feature, though I’d point out this is really not a good idea. It’s 5 seconds people. Even if you have 20 extensions installed, your talking about 100 seconds, less than 2 minutes of your life. Seems like a reasonable compromise for the extra security.

SSL Bug In Firefox 3b5

I’ve encountered this bug I just can’t quite figure out, so I figured I’d put it here. Hopefully with a broader audience someone else had encountered it and perhaps this will lead to the root cause being identified.

For some reason Firefox 3 can’t access Webmin on port 10000, which is how it’s setup on a box I have. It worked in Firefox 2.0, but not 3.0. I’m not sure if it’s something to do with Perl’s Net::SSLeay, which Webmin uses for SSL support, or the port number being 10000. I’ve tinkered a little bit with SSL settings, but so far haven’t been able to figure out exactly what’s going on. It seems to be a regression in NSS.

Anyone notice a regression like this using nightly builds somewhere else? This is the only case I’ve personally experienced it. If you have, then visit bug 423499 and let us know.

Edit [5/4/08 @ 11:30 PM EST]: No idea what’s going on here, but apparently nobody else can reproduce, so calling it quits for now.

Summer Of Code 2008

Google announced the project lists for Summer Of Code 2008. Some of the more interesting projects from my perspective:

Adium

Dojo Foundation

FFmpeg

Gallery

Inkscape

Joomla!

The Mozilla Project

MySQL

PHP

Pidgin

WebKit

WordPress

Over Logging

Linksys On Southpark

Southpark last week featured an internet outage as a plot. Pretty clever though I was disappointed to not see 1 reference to the series of tubes. I’m not sure if the reference to Linksys (Cisco) being responsible for the Internet being down was a complement or an insult. Though those Linksys boxes are infamous with just dying like that until you power cycle. Any other brand seems to have figured out how to not have that issue. Linux firmware on a Linksys also seems to remedy it. References to “Independence Day” and “Close Encounters of the Third Kind” were clever.

You can watch it online by clicking on the screen grab.

Shell Stats

Since it seems like everyone else on Planet Mozilla is doing it… My twist: multiple systems. Actually I found it interesting too see the variation based on what I use them for.

Home

$ history | awk '{a[$2]++}END{for(i in a){print a[i] " " i}}' | sort -rn | head
114 ls
91 cd
63 sh
41 ssh
37 sudo
14 pico
13 exit
12 ping
10 ./gl_tail
8 top

Home Server

$ history | awk '{a[$2]++}END{for(i in a){print a[i] " " i}}' | sort -rn | head
68 ls
64 dig
55 cd
45 whois
27 ps
24 clear
23 sudo
14 pico
12 top
12 exit

Work

$ history | awk '{a[$2]++}END{for(i in a){print a[i] " " i}}' | sort -rn | head
67 top
65 ssh
31 ls
29 sudo
26 dig
23 cd
20 ps
19 svn
17 php
8 ping

Don’t Let It XPire

Seems everyone who tries Windows Vista comes to at least one of two conclusions (if not both):

  1. Please don’t let Windows XP Expire – There’s even a petition for those in this camp. And it’s getting press.
  2. Mac Time – Enough said. Mac OS X 10.5 isn’t perfect, but is anything? It’s about as close as anyone has gotten.

It will be interesting to see the fate of XP.

EnergyGuide For Computers And Servers

Energy Star Energy GuideMost appliances sold in the US are now required to ship with the infamous EnergyGuide Label. That yellow label is most associated with major appliances, but it could be utilized beyond that. It’s a pretty simple idea in principle. The label makes it very clear how much power the device consumes, and based on average usage how much it will cost the consumer to operate on a yearly basis.

Perhaps it’s about time to adopt a similar convention for computers, servers, displays, printers, and networking equipment. Doing so would have several benefits:

  • Make consumers aware of the variety in efficiency – Different products consume different amounts of power. Being able to easily compare isn’t a bad thing.
  • Encourage manufacturers to be more competitive – While things such as processors have become more efficient (Pentium D consumption:performance ratio was worse than say the Core 2 Duo), computer manufacturers aren’t really pressed to adopt the most power efficient technologies since they may increase component costs.
  • Weed out poor power supplies and power bricks – A poorly kept secret is that power supplies and those power bricks we all have under our desk aren’t the most efficient devices out there. There are however some newer ones that have improved greatly like 80 PLUS certified power supplies. By making power consumption more visible, the really cheap power supplies become a negative in computing rather than a silent component.
  • Lower costs for businesses – It’s no secret businesses (like google) are trying to lower costs in their data centers. By making power consumption more obvious, it would be easier for them to make better choices to reduce their power costs and get the most out of their budget.

The real problem is that it’s tough to go to a manufacturer’s website or a store and tell what this device will cost you to operate, and compare to another device. That doesn’t really seem necessary. It should be easy to tell.

I think such a label should include the following:

  • Estimated Yearly Electric Use – in kWh for 24×7 power consumption pattern, additionally 8×5 for desktop computers, non-networked printers, displays. For 8×5 should also show the estimated yearly electric use in standby for the remainder of the time. That’s 1 number for servers, network equipment (which operate 24×7), maximum of 3 for everything else.
  • Estimated Yearly Operating Cost – in $ based on the above kWh data.
  • Standby Power Consumption Use – How many watts the device uses when in standby. This is really to see how close it comes to meeting the one watt initiative.
  • Standby Activation Modes – Several badges indicating how standby is initiated. Including: idle timeout, user initiated, timer, auto-wake (comes out of standby when device is needed). These let the consumer know how the device can minimize power consumption.
  • Upgradable Power Supply – Is the power supply connected to the device using an industry standard so that it could be replaced/upgraded during the device’s lifetime?

The data should be made available by easily removable signage attached to the device on shipping, made available on request in stores, and available on the manufacturers website linked from the product page.

By making power consumption more obvious, it would be easier to make a decision based on TCO (total cost of ownership) that accounted for power consumption. The bigger advantage is that it would encourage manufacturers to consider power consumption as a potential way to sell a product, and that means improving power supplies and standby consumption.

It seems like an easy way to spur some innovation and allow for better decisions by consumers. Considering the cost of power is expected to rise, and the need for IT is rising, it seems like a decent proposal. Making the most of the watts available requires knowing consumption. That way keeping servers and desktops powered on is as efficient as possible.

Image from FTC