For archive (Updated on February 17, 2019)
In the early days, it was expensive to store and/or transmit data (e.g. text, sound, picture) in raw form.
Compression techniques were developed and used to reduce the size of the data.
- “Lossless Compression” requires that the original data can be recovered without any loss.
- “Lossy Compression” techniques are used to reduce the size of the data as much as possible (e.g. by stripping off minor details). The original data cannot be recovered fully. The recovered data would have some loss (e.g. in quality).
- Telegraph messages (and subsequently SMS messages) have limits on the number of words (or characters). So, it makes sense to compress a message by (a) leaving out some letters in a word (b) combining words into a phrase (c) using abbreviations and acronyms.
- For example, the following were first used by the military to report situations: SNAFU (Situation Normal. All Fouled Up.) and FUBAR (Fouled Up Beyond All Recognition).
- Run Length Encoding (RLE) is a simple and straight forward way to encode characters. A string of repeated characters can be represented by a pair (Character, Number of consecutive occurrences).
- JPEG and MPEG (e.g. MP3, MP4) are commonly used to compress video and audio files.
Compression of Burmese words — for smart phones and messaging — had been taken to such an extreme that some no longer know or care about the correct spelling and usage. So sad.
U Khin Maung Zaw (KMZ) added :
Our good friend Htay Lwin Nyo (EP74, UCC – GBNF), told me when I met him after he moved to SJSU, San Jose State University, that he made some killings either selling or licensing the data compression algorithm. Neither did he elaborate nor did I ask him more details at the time.