Jeri Ellisworth
potatohead
Posts: 10,261
I found this lecture at Stanford a great watch.
http://stanford-online.stanford.edu/courses/ee380/050518-ee380-100.asx
http://stanford-online.stanford.edu/courses/ee380/050518-ee380-100.asx
Comments
http://www.blip.tv/file/4055830
Thanks for that link.
I linked Jeri because I find her approach to new technology notable. There is lots of info online, and it's clear she consumes that where she can, or wants to.
That is combined with lots of play time to learn stuff.
During the Stanford talk, she mentioned mentors several times. Looking back on my own life, the mentors had a seriously big, positive impact. That remains true today --and in fact, some areas I have trouble in, I don't currently have a mentor.
The discussion here on "where are the docs?" is one I find very, very interesting. Some of my comments have been pointed, and I want to explain a bit of that.
Really, it's not a personal thing. Not that at all. Where I was headed was how people think, and why they do it the way they do, and what that means for the learning --or the "play" process all of us do to some degree when trying to grapple new technology.
The "just try it" idea has a simple power in that you will absolutely learn some stuff. Might cost something, but there is learning always. The Chuck Peddle interview fragment is also notable, in that the same kinds of dynamics are in play.
People, who have mastery of some particular discipline, have reached some internal model of it that's efficient and robust.
I am not sure that always works in a book. Sometimes it does, but often it doesn't, because we often internalize things very differently.
Perhaps it would be good if we asked more process kinds of questions of one another. The recent beginner puzzle threads are kind of fun, because it exposes a bit of that. I asked one about how people kind of get boot-strapped on some system.
There are some things that need to be established:
1. Tool chain. Gotta have a path from idea to executable.
2. Verification tools. Gotta have some way to look at what's going on. These vary widely! Some of us use the blinking light, others use scopes, still others use debuggers, and all sorts of things.
3. Process! For a given task, there is generally a repeatable way to arrive at a solid solution. This is a point of interest for me right now.
One of the most interesting properties of the Propeller is the ability to actually run code concurrently! There is considerable challenge there too.
In the brief reading I've done on process, including answers to the threads, and some looking back on older forum posts, and tech interviews, like the one linked here, I see having some broad understanding of what people expect to be true, then some narrow forming of tests and tools to verify truth, or simply behavior. There is also some thinking through the problems in layers.
I have trouble with that, often struggling with what to put at what layer and why, and how that connects the dots from idea to executable / active thing.
So, on this thread, please post your mentors, and talk a little bit about how watching them, interacting with them, impacted your own internal process. I think that's probably one of the biggest learning barriers --at least it is for me.
With programming, I know I need more process bits! And there is a ton to be had here --though indirectly. With circuits, the same thing is largely true as well.
There is this to say as well. For a lot of people, there are things that just "click", and things that really have to be hammered home a batch of times before they click. When we have some mastery, thought is action. When we don't there is the thought, then meta-thought, then action, and that gets hard.
In my own experience, I have trouble with chains of abstractions. If this is associated with that, through this other thing, checked by these things, then it's often hard to hold that model in my head. I think that's due to not having internalized these things to a point where they are just solid. A good contrast in my case is CAD modeling. I can hold very large numbers of constraints and parametrics in my head, seeing the whole web, building out behavior, no problem.
Probably that's a lot of manufacturing engineering in play there, plus observation of how things work and fit and change.
When programming, a pointer, to a pointer, through some logic to arrive at a result doesn't seem like something that's really all that different, but it's a lot harder, due to the lack of internalization.
Probably this post isn't making a lot of sense. Sorry about that. Again, if you feel inclined, post up some of your mentors, or stories about them, and your process experiences. How do you apply the scientific method, and what means do you use to internalize things to the point where they are as available to you as say, words are?
http://www.fatmanandcircuitgirl.com/live/
Live right now...
Currently invaded by Doug and I.. Need more Propellerheads..
OBC
http://www.ustream.tv/channel/jeri-ellsworth
http://www.youtube.com/user/jeriellsworth
Wish she would join this forum.
Parts 2 and 3 are on-line! Here are the links --
A Conversation with Chuck Peddle, Bil Herd, Jeri Ellsworth - part 2
http://blip.tv/file/4084084
A Conversation with Chuck Peddle, Bil Herd, Jeri Ellsworth - part 3
http://blip.tv/file/4084124