What makes a conference talk a good talk?

What makes a conference talk a good talk?

Now; I've spent two days at #oscon in Europe. A commercial event, but mainly about stuff that concerns open source as you would expect. It has been a while since I have been to a conference, so i kind of feel a little bit rusty. I am in no sense of the word a seasoned conference attendee, even if I have had a few talks myself at #JavaZone and some other minor events, participated in quite a few conferences both in Norway, US, and Europe. Short version; I always get excited by people that seems to know what they are talking about and as an added bonus excel in giving presentations.

The funny thing is the delicate balance between being able to deliver a message that is persuasive, concise, composed by recognition points aligned with some fragments of prior knowledge. Being enthusiastic, charismatic, scientific and visible might trump the latter, but it is rarely the case that one or the other is exposed in singularity. I think, being able to convey the essence of something that is important to you is also about wanting you to learn more about the subject or the scientific field. This is the subtle relationships between immediate versus longer term perceived properties of a talk. A talk that might be entertaining at first might not be the one that you appreciated the most in retrospect. For that matter; to evaluate sessions as you go might not be a just way to give feedback when you evaluate sessions at a conference. Note to self, wait a day or two before evaluating talks. Humbleness, being able to refer to real-world cases and observations, good and bad will increase credibility. I think some people, including myself tend to like talks where I may say something like "He/She has experienced the same thing as myself!". Well, it does not bring much new stuff to the table does it?. That someone else feels like you does not make it a scientific fact. However, did you need recognition from someone else to state the same thing that you already knew? The bottom line should be, what is the origin of the problem that you have encountered? Is this solved? If not, what is the alternative?. It is equally important to know what is known, and what is not known. And if the ratio between known and unknown gets to high, it is caught in the sales-rep filter.
I will take some of these note to self points into account when attending my next conference.