Hello Dr Bal,

Thanks for the question. It is indeed important to get this clarified.

When we are doing the analysis, we cannot and should not limit ourselves to any specific font. At the same time this is also true that we cannot take into account all the possible fonts.

We cannot and should not limit because across the Operating Systems there are various fonts present. Even within the same operating system, across its versions, different fonts can be used. Even if we talk about a particular OS and a particular version, among different browsers different fonts could be used to show the address bar content. If we further fix a particular browser, browsers are not the only place where URLs are used. They can be a part of an email where all the possible fonts can be used. In essence, font specific analysis cannot yield great results.

Having said the above, analyzing all the fonts is impossible too.

It actually means that there is no fixed way to take this up. In my opinion, we should go by the common glyph shaping behavior within the respective scripts and chart our road ahead. Even across various fonts, styles there is always a common denominator which we can consider. That is what we are doing.

Thanks for asking this question. It brings this aspect of our approach, out.

Regards,

Akshat Joshi

-- 
Regards,
Akshat Joshi
C-DAC GIST