Analytics - Phase 4

It's that time of the year again. 

Just kidding. If there is one thing about this blog's schedule that is consistent, that is its inconsistency of schedule.

There has been this irregular tradition of looking at what kind of people (read "computers") chance upon my little corner.

In its latest iteration, we see a surprise surge in visits from the Cult of Mac! Also, Chrome has finally dethroned Firefox at the browser helm, over two years after it did so in the world's wider web.

Well, how have the interwebz shifted over the years then. We see quite a few interesting trends as time passed by. This is the 4th such post, so let's see how the balance has shifted over the years.

Since its early days, Chrome users had held their own in terms of traffic. In the last couple of years, however, it has grown from lingering in Firefox's shadow to claim its place under the sun, while IE, almost always on the verge of a comeback, will likely never quite catch on in these parts.

Also of note is the "other" section, fuelled almost entirely by esoteric KHTML and Dilo-like renderers in the early days, but replaced by more simplistic, yet consistently growing mobile-based alternatives.

Speaking of mobile alternatives, the practical absence of mobile visits can be better visualised by the OS share variance with time, with "other" accounting for a humongous 0 percentage points during the earlier years, while growing to a healthy 10% in recent times. 

Given it's FOSSy leanings, Linux-based user-strings have always been over-represented for this blog that for the web-at-large, but the sudden spurt in Mac visits is an unexplained surprise, which may or may not be explicable by the proliferation of "cheaper" Macs like the Air, the Mini and the Pro non-retina.

Note 1: All of these number are cumulative(since early 2009), so the trends would be more pronounced if we take each period in isolation. A year by year comparison might be in order, especially if compared to overall trends in market 

Note 2: Previous iterations lie here: Part 3 2 1

Petrol price misrepresentation on social networks

A misleading post regarding petrol prices in India and Abroad is doing the rounds (again!). 
Here's a copy of the post in question

Let's look at the accuracy of each of the above claims, just so we learn that the purpose of any media needs to go beyond the reconfirmation of the reader's preconceived notions.

Country Price in shared image(INR) Actual price in USD Actual Price in INR Difference in %age

Pakistan 26 0.6 40.2 55%
Cuba 19 1.41 94.47 397%
Nepal 34 0.94 62.98 85%
Afghanistan 36 0.77 51.59 43%
Bangladesh 22 1.29 86.43 293%
Italy 14 1.51 101.17 623%
Burma 30 0.55 36.85 23%
Sri Lanka 34 0.88 58.96 73%
India 82
60 -27%
Conversion rate used : 1 USD = 67 INR

As we can see clearly, not only is the data clearly misrepresented, but we can see how ridiculous the bias in "reporting" is. Italy and Cuba, purported to have the cheapest petrol on the list, actually have the most expensive. 

India, whose government the creator of the image so callously dismisses, is the only location with over-reported prices. Not to mention, prices vary way more based on crude prices, than on variation of taxation, given a fixed jurisdiction.

And let us not even forget the fact that petrol prices, governmental acumen and quality of life have little bearing on one another

A request to all that may chance across this post :
Please, at least try to fact check random posts by people about "facts" and "data", before you jump to share and retweet these posts, and further contribute to the spread and celebration of misinformation.

Note that these prices are mostly accurate as of early March 2016, and may vary considerably base on a variety of factors.

P.S. And what is the deal with the .00 after every price in the image. I though significant digits were, well, significant.


10.43252003274489856000 reasons why "X reasons Y" lists are boring

The barrage of sites with mind-numbingly stupid content and clickbait titles is neither new nor exhilarating.

However, ad-mongering, for the lack of a better word, and short-shorter-shortest attention spans of people, when paired with the ubiquity of technological access has made them garishly visible and overwhelming, sometimes to the point of frustration, and oftentimes, far beyond.

Here's a list of reasons why these word-bundles posing as "articles" are a force of evil. Also, the mandatory "You won't believe number 7!"
  1. Their titles offer little to no information about the actual content.
  2. Their content is usually unnecessary at best.
  3. Oversimplifications are not just made and brushed over, but celebrated.
  4. They are mostly opinion-pieces, with the opinion being "I am better than you" or "Yoohoo, You get a stereotype, you get a stereotype, everybody gets a stereotype! There's enough for everyone."
  5. The primary purpose of these articles is to entice with a title that tugs at heartstrings of potential readers, so that their clicks help earn ad revenue.
  6. The secondary purpose is to reinforce biases, stereotypes and self-indulgence.
  7. For some reason, a vast section of the populace believe other people share their biases and would love to read fluff pieces of no consequence.
  8. Most use cringeworthy images/gifs they don't hold the rights to publish.
  9. The authors are often deluded enough to think they have invested this new awesome format no one ever though of.
  10. The grammar is, well, horrendous on most occasions.
Notable trivia:
  • This post has the highest number of quotes I have ever had, scare or otherwise. If one chooses to be pedantic, yes, I mean quotation marks. There, there...
  • The list is 0.43252003274489856000 items shorter than the title. So Sue Me.
  • 43252003274489856000 is the total number of possible combinations on a standard Rubik's Cube.
P.S. I think "notable trivia" is an oxymoron, but I've been wrong before. 

Automatic headphone detection in Alienware M17xR4 for openSUSE Leap 42.1

All to often, people dismiss Linux if something tiny doesn't work out of the box. They claim that one does not face such problems in Windows, forgetting the extensive jumping through hoops known as installing a bunch of drivers that negates the whole "out of the box" premise.

I had just gotten a fresh install of openSUSE Edu Li-f-e, based on Leap 42.1, running on the behemoth that is Alienware M17xR4. All seemed fine until I realised that plugging a headphone into the relevant jack does not reroute the sound output through the headphones, while the speakers keep blasting away.

Pulseaudio, or rather pavucontrol, was helpful enough to switch between line-out and headphones on cue within the GUI, albeit with little effect, as the sound output remained unaffected.

The situation wasn't too different with the KDE's "Audio Volume" module, which largely seems to be a front end to pulseaudio.

As I was bungling through the settings, I realised that the issue wasn't detection of headphones, as it was the switching audio output from the speakers to  the headphone.

Enter, alsa!

alsamixer has a particularly unhelpful welcome display, which has no bearing on the level of customisability it offers.

AlsaMixer intro screen

We can, however, get much better controls, once we manually select the sound card (Press F6).
Note: If you fail to find the option for the given sound card, See Note 1 at the bottom of the post.

AlsaMixer - select sound card

We then navigate to the  HP/Speaker Auto Detect (in red in the image below) and toggle to ON (by pressing M).

AlsaMixer - change HP/Speaker Auto Detect

If you get stuck along the way,  help is just an F1 away.

Help menu for AlsaMixer

Note 1:

If you fail to find the option for the given sound card, add the line 
options snd-hda-intel model=alienware at the end of the file /etc/modprobe.d/50-alsa.conf .
If the file does not exist, look through the folder /etc/modprobe.d/ to check if any similar file exists with a different number preceding it, and edit the same.

Note 2:

Packages used:

Note 3:

There are multiple audio jacks in the M17xR4, so for the above process to work the headphone/ext. speaker needs to be plugged into the jack shown below

Audio jack in Alienware M17xR4 for auto headphone detection

openSUSE tweets