A log equation is an exponent statement in disguise. log2x=5 says "the power of 2 that gives x is 5", which means x=25=32. Rewriting in exponential form states the same fact the other way round, and once the log is gone the equation is ordinary algebra: log2(x+3)=5 becomes x+3=32, so x=29.
With two logs, combine first. log2x+log2(x−2)=3 becomes log2(x(x−2))=3, then x2−2x=8, which factors as (x−4)(x+2)=0. The algebra offers x=4 and x=−2, but log2(−2) does not exist, since no power of 2 is negative. Only x=4 survives, and it checks: log24+log22=2+1=3. The extra root slipped in when the two logs merged, because the product x(x−2) is positive at x=−2 even though both factors are negative.
Compare 2x=10. Both kinds of equation mix logs and powers, but there the unknown is the exponent, so you take a log of both sides to bring it down. Here the unknown is inside the log, so you do the opposite and exponentiate. The deciding detail is where x lives: up in an exponent calls for a log; inside a log's parentheses calls for exponential form.
Think of it like this. A log is a locked box holding the answer to "which power of b?". The key is the base itself: raising b to both sides opens the box, since blogb(stuff)=stuff. Whatever comes out must be positive, though, because the box could only ever have held a positive number.
Core idea. Rewrite logb(stuff)=c as bc=stuff to free the variable from the logarithm.