Desk Descriptions

It’s been one week since I switched to a sit/stand desk setup. It’s time to report back.

magic buttons

Here are some of the points to note if you switch to a similar setup.

Comfortable shoes are a must!
If you can splurge a bit, you can get one of those fancy standing mats. I’ve tried standing barefoot, with thin socks, with thick socks, and in peshawari chappals. All of them gave me foot pain. I switched to running shoes and all was good.

Alternate between sitting and standing!
Standing too long leads to fatigue (just like sitting too long leads to fat-igue). “Too long” may mean anywhere between 1-3 hours. I’ve found switching after 1-1.5 hours works for me. Today, I was zoned-in and realised after two hours that I was working while standing. Luckily, my desk has buttons to adjust its height so switching doesn’t take long.

Rest your feet!
Standing too long on both legs, one leg, or heels leads to leg pain. From what I’ve read, foot stools help — I just use my chair or a foot rest (that you can see in the picture above).

Don’t forget you are on a sit/stand setup!

Last Monday morning, my desk was down and over the weekend, I had forgotten that I had to switch to standing mode. I noticed it after lunch and by that time, half of the day had already passed. The trick I try to use now is to leave the desk up when leaving for home. It also helps me the next day when I can start my day right away, standing, after my 25-30 mins of commute.

I’ve noticed that my sluggishness level has lowered a bit during the last week, specially post lunch.  Any effects on weight will be noticed later (and will not be reported). Having said that, one week may not be long enough to notice anything so the experiment continues!

Sit mode
Desk slightly raised. (P.S. I need recommendations for tea-bags closest in taste to Tapal Danedar)

A survey on Nastaleeq Font preferences for Urdu

tl;dr – A survey I carried out shows that 55% of the respondents considered Jameel Noori Nastaleeq to be the most readable out of six commonly used Urdu fonts on the web.

Update: These stats were updated on Aug 4, 2015


The data below is from a small survey I carried out in order to learn about font preferences of Urdu readers on the Internet. Here’s how it was conducted.

  • 5 prevalent Urdu Nastaleeq fonts were selected.
  • Arial (Unicode) was added as the 6th font.
  • Top 500 ligatures (by frequency) from a corpus of 19.3 million words were selected and concatenated to form a paragraph of text.
  • For normalization, the text was formatted to fit in a box covering approximately half of an A4 page (about 12 lines, single spaced) and font point size was adjusted accordingly.
  • 6 such boxes were created, one for each font.
  • Each box was assigned a letter.
  • An online survey was created in a ‘blind’ setting and the voting link was shared over social networks.
  • Votes were collected over a period of 8 months.

Results: Out of the 281 responses recorded, 55% preferred Jameel Noori Nastaleeq while 24% thought Alvi Nastaleeq was more readable.


Font Responses
Jameel Noori Nastaleeq 156
Alvi Nastaleeq 67
Arial 10
Noto Nastaleeq 30
Urdu Typesetting 15
Nafees Nastaleeq 3
Grand Total 281


Possible Errors:

I couldn’t find any published research on readability of Urdu fonts on the web. This survey can be taken as a starting point but please use the results of this survey only as a crude quantifiable measure of font preferences. (In other words, don’t cite this survey in your research, the experiment settings have oversights).

  • No formal definition of ‘readability’ was provided. No example of readable or unreadable text was given.
  • The participants may not be ‘experts’ in Urdu. There may be some non-native respondents in the survey.
  • There may exist a bias towards Jameel Noori Nastaleeq due to the mere-exposure effect. This font is used in most Urdu based print publications, including daily newspapers and magazines.
  • While the height was same for all fonts, each used a different point size (11-15).
  • It was noticed later that due to an error in conversion, the image for Nafees Nastaleeq was slightly smaller than the rest of the fonts, which might have caused some votes to change in favour of other fonts.
  • Some respondents have commented that comparison between fonts would have been better/easier if the text was in the form of readable phrases or sentences (rather than ligatures). There were reasons of not doing that. Firstly, I was lazy and didn’t want to find a set of phrases which covers all high frequency ligatures in a minimum number of sentences. Secondly, IMHO, form should be distinct from meaning and how things look should not be effected by what they mean. I can be convinced otherwise with respect to the latter reason but for the former one, I will not budge 🙂

Comments/suggestions/criticism welcome.

LDA-Based Topic Modelling in Javascript: An Update

I’ve just pushed a Javascript version of LDA on my github account. It’s based on my no-longer-functioning earlier work. For testing, I use a subset of the SMS Spam Corpus available here (and thus take no responsibility of the inappropriateness of the text within 🙂 ). Each topic is represented as a word cloud; the larger a word, the more weight it has in the topic. The source sentences are displayed again with a bar which shows the percentage distribution of topics for that sentence. Hovering on each area in the bar would show you the words in the topic. You can of course replace it with any other text, change the number of topics using the slider, and press the ‘Analyse’ button to see it work.

Click here (or on the image below) to startlda.js