[seqfan] Longest substring is n

[Éric Angelini]

Hello SeqFans,
a little idea with some tricky traps:
"Lexicographically earliest seq of distinct 
terms such that the longest substring 
shared by a(n) and a(n+1) is n".

S = 1,12,23,34,45,56,66,78,89,109,110,112,
1213,1314,1415,1516,1617,1718,1819,1920,
2021,221,223,2324,2425,2526,2627,2728,
2829,2930,3031,3132,332,334,3435,...

The "tricky traps" (well...) are the above
terms like 110, 112, or 221, 223, or 332, 334, etc.
I fell myself into the trap when it came to
extend the seq after 9798,998,10099,
10100 and... now _not_ 10102 because the
"longest substring" shared by the last two 
terms is now _not_ "n" but 1010 (n =100 here). 
The correct extension should  be with 
101102, if I am not wrong. Is this seq 
worth computing and submitting? I guess 
it will become less and less obvious 
"im Laufe der Zeit", "au fil du temps", 
"nel tempo", over time.
Best,
É.




à+
É.
Catapulté de mon aPhone