In substandard transliteration of Sanskrit, the difference between short (hrasva, e.g: a, i, u) and long (diirgha, e.g: aa, ii, uu) is almost always, (actually, by definition?) ignored, so to speak.
There are several ways to indicate the long vowels, for instance A, I, U; aa, ii, uu; (favored by linguists?:) a: , i: , u: ; a_ , i_ , u_, etc. Lets take as an example, YS I 11: anubhUtaviSayAsampramoSaH smRtiH .. 11.. (anubhuuta-viSaya-asampramoSaH smRtiH) .. 11.. A typical substandard transliteration might be: anubhutavishayasampramoshaH smritih .. 11.. How would that affect the translation? Word-for-word of the correct transliteration might be: Experienced (anubhuuta) object (viSaya) not-letting-drop ([from memory] asaMpramoSaH) [is] memory (smRtiH). The substandard: anubhuta: no such word in Sanskrit as per CDSL; object (vishaya: ignoring the difference between a retroflex [S] and a palatal [in H-K: z] sibilant) loss (sampramosha) [is] memory (smritih). So, ignoring the long 'u' (U, uu) makes that word meaningless in Sanskrit, and ignoring the long 'a'(A, aa: viSaya+asaMpramoSaH, read without hiatus between those two words: viSayaasaMpramoSaH) turns the meaning upside down, if you will!
