Hi, i would like to ask your opinion on the duration of the greek diphthongs. long or short and why?
I don’t think it’s a matter of opinion. Diphthongs are long. I’ll let the experts tell you why. We’re talking Classical Greek, right?
In Epic poetry and elegiacs, some diphthongs scan as short to suit the meter. The shortening of a long vowel or diphthong in the arsis of a foot before an initial vowel of the following word is called epic correption and is very common.
Miltiades, what motivates your question?
That’s a good question annis. I think it is a matter of opinion and here’s my thought.
if diphthong is defined as the co-occurence of two vowels in a single syllable and is traditionally concerned as long, could we say it’s the co-occurence of 2 short vowels making a long syllable meaning: /oi/ lasts x*2.
But if we say so, the same thing is true for the duration of two short syllables in speech, eg. /pedon/ (valley) lasts x2 (x+x)the difference being made by the existence of the consonants that make the distinction between syllables. This, shows, i think, that diphthongs may not be concerned as lasting not x2 but x and that could make the difference between other co-occurences of vowels that are not traditionally recognized as “diphthongs”, such as ‘ea’ in /dasea/ (forests)(lasting x+x) and this giving - through internal reconstruction (contr- long e /dase:/ (lasting x*2).
Thus, could we say that diphthongs could be seen as short-lasting? At least in Modern greek, when it comes to diphthongs, we can recognize some that in speech last no more than a single syllable, knowing that modern greek has no length distinction (all syllables are short). Such claims, if proved plausible, may cause trouble in the metric patterns of classical greek poetry.
Except your definition of a diphthong is not correct. A diphthong is defined as a sequence of vowel + glide interpreted as a single syllable. In the case of Greek, the glides are ι, υ.
Also, from the standpoint of linguistic typology - the study of linguistic features common across languages - in those languages which distinguish long and short syllables, rarely is the ratio exactly 1:2. More common is something like 1:0.8 or so.
Such claims, if proved plausible, may cause trouble in the metric patterns of classical greek poetry.
Not just poetry, the entire morphological package, and accenting. There are a bunch of grammatical processes in which diphthongs and long vowels cause one behavior, and short vowels a different behavior. So I don’t think such claims are plausible. 
I don’t know why the old-timers considered their diphthongs to be long; my guess is that they played it by ear and heard themselves taking longer to say them than short vowels. We know that they considered them to be long because they left us metrical prose and poetry.
I’ll bet you anything that when you speak Modern Greek you make the vowels in stressed syllables longer than in unstressed syllables. You just don’t need to teach kids that some vowels are longer than others because they just come out at the right length naturally. I’ll also bet anything that you can tell whether someone is from Athens or Crete in part by how long it takes them to say the “ou” in ouzo. Someone from Athens (a supposition, I’ve never been in Greece) will make the “ou” 1.3 times longer than the “o” in the second syllable, and someone from Crete will make it 1.9 times longer.
Meter is simply applied phonetics.
Καλημέ?α, Μιλτιαδη! Τι κάνεις;
It doesn’t seem to pertain to your question, but from my primitive knowledge of Greek and the workings of accentuation, final diphthongs such as οι and αι are considered short, as with ἄνθ?ωποι versus ἀνθ?ώπους, πα?αδείγματος χά?ιν.
In any case, δεν ξέ?ω. Συγνόμι.