Ping Tim Wescott: YouTube [long]

Hi Tim,

I prepare lots of "tutorials" -- text & multimedia encapsulated in a PDF container -- for various subjects/technologies. Typically as a way to explain an algorithm/technology outside the confines of "comments in source listings" or "in the margins of a schematic". These are typically

25-50 pages and usually exploit the container's capability to include "attachments" (so, I can include long source code listings, PCB artwork, etc. without adding to the document's page count).

These can be viewed as LECTURES -- information flows ONE-WAY. They are intended to speak IN MY ABSENCE. I.e., I have *capped* the time that I will spend on that "descriptive task" (I don't have to be available for Q&A from clients and others, "indefinitely")

Some colleagues who've seen many of these (I have a couple thousand pages) have pestered me to "present" some of the material covered, "live" -- to cover more ground in a more interactive form (i.e., this gives them a chance to pose what-if's and probe stuff that I've not covered -- in reality, I suspect it will turn into a brainstorming session for each of their projects as they individually hypothesize how the material could be applied to their individual problems).

These can also be viewed as lectures -- but with the potential for a two-way exchange of information (Q&A, at the very least).

I have *tentatively* agreed to do this. Now we're arguing over how, when and how much! (amazing how eager folks are to spend *your* time! I wonder if they'll be as eager to help me install the picture windows in the back of the house?? :> )

I presented an outline to an informal gathering ~a week ago to see if it addressed the material they thought appropriate. I settled on a "lab" sort of presentation whereby I can present content and let them stumble on the problems with each idea -- giving me a natural path to introduce the *next* idea.

[E.g., imagine teaching a course on control theory where folks can quickly hack together each control algorithm you've presented -- but *you* can manipulate the plant to draw attention to shortcomings in each of those (and lay the foundation for introducing the next concept/algorithm -- bang-bang, proportional, add integral term, derivative, SPC, etc.)]

I think this is the easiest way to cover a lot of material in a "session". Sitting, quietly, while someone lectures gets tiring, fast. And, they can consult the PDF's for more detail, later.

Anyway, someone seized on the "brilliant" idea of filming this and posting as YouTube videos. (Why does EVERYTHING have to be on-line? I'm waiting for folks to start having "virtual weddings" -- friends sit at home in the comfort of their OWN living room and watch them exchange vows. Then, tap away on the enter key to applaud when the bride gets kissed! Hit the up-arrow to throw virtual rice. Later, everyone gets drunk -- on their *own* booze -- and skypes video of themselves tripping over each other...)

I think this is a BAD idea:

- the "sessions" are too long to just be a passive observer (time passes quicker when you are actually *in* the "lab" vs. watching it)

- and they are too many in number (the topic is pretty broad; we can always adjourn a session for a group dinner and then resume a week later)

- the approach is too interactive; you'd end up having to watch over someone else's shoulder as they tried to solve the problem

- there's a different "feel" when you're on the outside looking in

- it interrupts the flow of the "class" (has to be done in a single shoot; no Mulligans as that's not fair to the "live participants")

- it's more work overall (producing the video... lighting, audio, capturing live screen images off workstations, etc.)

At the very least, I'd have to completely rethink the courseware to drive it back to more of a "lecture"/demonstration format. I've watched a LOT of videos in the past week and can't see how they can be anything other than "lectures" (i.e., why not just let folks read the PDF's and omit all this overhead? and, save me the additional time and effort!)

So, what comments do you have re: what *you* have found as the limitations/drawbacks (and advantages!) of YT presentations? E.g., imagine how you'd present one of *yours* "in person" (to a live audience) vs. as a YT video. How would it alter your presentation (for the better? or worse??) and how do you think it would alter the "reception" of the material presented?

Would you just rattle off "this is how it is done"? And these are all the problems you will encounter? And this is how you address each of them? (this is what my PDF's do -- knowing the reader can move freely through the document to revisit issues of particular interest and gloss over others with which he may already be familiar/understand).

Or, would you get folks involved (assuming you had the physical resources) so they can each *discover* problems -- that you can then explain and steer towards a "better" solution?

I.e., how is the medium constraining your approach to problems? Granted, you can edit out any blemishes in a presentation; but, you wouldn't want to hide them in the *implementation* -- as that's how folks learn... what NOT to do! Why would you *want* to opt for a YT/lecture approach, if other alternatives were available? Is there any way to make a video NOT be a "lecture"?

Thx!

Reply to
Don Y
Loading thread data ...

Hey Don -- now you've gone and made me think!

The short story is that when I do this in person, it's in chunks ranging from 45 minutes to 1-1/2 hours and I encourage questions from the audience. When I do this on YouTube, my intention is to keep it to shorter segments -- and based on you making me think, I think I'm going to start each video by encouraging questions in the comments.

Live is always going to be a higher quality experience for the people who can show up at the right moment without trammel or care. YouTube is always going to be a higher quality experience for people who just need to know a little bit, or who's schedules and/or pocketbooks forbid them from traveling someplace to take a class. Papers are always going to be a higher quality experience for people who learn by reading (I'm a "reading learner" -- if I want to really get the juice from a lecture I need to take notes).

--
Tim Wescott 
Control systems, embedded software and circuit design 
 Click to see the full signature
Reply to
Tim Wescott

Hi Don, (in answer to "not a lecture") I've got a video or two "in mind" that I need to do for customers. These are mostly just demonstrations. How do you "drive" this particular piece of equipment.. and make it do this "thing".

I that regard I really like the video format:

1.) customers may not have good English language skills.. but a picture works. 2.) It's hard to put all the things you do into a written document, and even if it's all written down, there can still be lots of confusion.... "Which knob are you talking about turning?"

George H.

T
Reply to
George Herold

(sigh) A mind is a terrible thing to waste! ;)

Yes, people get antsy when just sitting passively, "soaking it up". The same is true of a written document. I suspect most people "skim" them, first -- then go back and read with more focus. Finally, return to it when they feel "fresh", again.

If you're attentive to your posting, then there can be greater immediacy to the feedback; a viewer stuck on a point halfway through your presentation can get an answer and return to re-view it, again.

OTOH, if you aren't attentive, he may lose interest before he sees your reply; it's a gamble with a YT presentation as the viewer doesn't know how quickly to expect that response; in a "classroom", they can be pretty sure that their question will be acknowledged (even if not ANSWERED) when it is asked.

I can do the same with "print" documents -- revise them as folks bring up other issues after having read them. In my case, that relies on the contact between me and the reader (I'm not just "posting" PDF's on a web site but, rather, exchanging them, specifically, with named individuals).

Without the feedback mechanism, you have to anticipate everything that the reader/viewer might ask and present that, at some logical place, in your video/PDF. Or, risk them having those issues unaddressed.

Yes. In my case, these are people that I see reasonably often (though I suspect some of *their* peers/workmates will end up in attendance).

I think another advantage to a live (or video) presentation is that it poses less of a hurdle to viewers. I.e., you have to REALLY want to know something to sit down and digest a written document. There's no one spoon-feeding it to you AND you know there's no one you can ask for clarification; *you* have to sort out what is being said.

So, folks who might not be inclined to wade through pages of descriptive text might think nothing of sitting through a few hours of a "class" (esp if it's "free") as hedging their bets against possibly needing that information in the future.

I don't rely only on "print" but, instead, leverage multimedia, as appropriate. For example, it's hard to explain a particular sound with words; much easier to have a sound clip that folks can listen to at their leisure (esp if there are subtleties in the sound that might not be readily apparent). And, your own (i.e., *my* own) ignorance may lead to misunderstandings in that presentation.

For example, if I try to describe the 'a' sound in "mash", folks in different parts of the country will think of different sounds (because "mash" has regionalized pronunciations). OTOH, "*this* is what it sounds like" avoids any ambiguity.

But, returning to my original comment...

I was trying to suggest that you have different opportunities when presenting "live". It's not just that Q&A can be more interactive. But, rather, how you actually present the material can differ, significantly.

E.g., suggestion about a "lab" that presents control theory instead of as a lecture.

Imagine letting participants code different control algorithms and apply them to a simulated plant to observe how their algorithms perform, "live". Introduce a bang-bang controller as a household thermostat (two lines of code?) and they can see the control working to turn the heat on each time the house cools, then off once it reaches the desired temperature.

Having seen this, *you* can tweek the plant so the house cools quickly (i.e., an open window). With this, show how the simple control is overly active; calling for heat, then shutting off as soon as the temperature climbs above the setpoint, then AGAIN calling for heat as it falls just *below*.

"Ah! We need a deadband!" and introduce the concept of hysteresis, etc.

Let them tweek their trivial bang-bang controller implementation and then they can see how it behaves differently in the same "open window" scenario. What has been gained? What has been lost?

But, now the heat might overshoot as the furnace has lots of latent heat (because it had to run longer to move the temperature from one end of the deadband to the other). Maybe we need to introduce an anticipator!

Etc.

You can present all of this in a video/print "lecture". But, it starts to get nit-picky as you lay out all the caveats. OTOH, someone stumbling over them "as exercises" sees them (I think) as small, incremental additions to their knowledge store. Logical steps in a progression to "realizing" the issues involved (instead of lots of details that just clutter up a presentation).

And, because they are actually *doing* something, I think the time passes quicker. They're actively solving (small) problems and continually moving forward in their understanding of the source material (I'm planning on 3 hour sessions with a break in the middle; if we cut it into smaller pieces, it results in too many "sessions".)

If, later, they revisit the "notes" (my PDFs), they can be reminded of these smaller issues -- as well as having still finer points introduced that weren't worth the time to address "in class".

Think about how you would "teach" a group BUILDING your pendulum; what problems would they likely encounter (that you *could* alert them to but might choose NOT to -- so they get a chance to encounter them first-hand)? They'll learn *why* you made design choices instead of just *accepting* those choices. And, perhaps, pose other alternatives (that you can expound on or dismiss -- for a *reason*).

Reply to
Don Y

[Note that I consider a video to be a lecture in much the same way as that of a "print" publication. The "lack of LIVE interaction" being the distinguishing issue -- I can make an interactive PDF but the interactions are all canned at the time I create it]

Yes. Not an issue for me but if you are addressing a larger, unknown population, that could be a factor.

But, if the instrument (?) *you* are using doesn't NEED that adjustment made (e.g., "focus" on a 'scope), then it's possible you'll overlook it, as well. OTOH, if each "participant" has a 'scope in front of them, you're more likely to encounter one that is out of focus, needs its probes compensated, etc. You end up covering more issues without having to spell each one of them out, individually, in your presentation (video or print). They seem like natural "asides" when they are encountered "live".

Any presentation without feedback is difficult because you have to imagine the issues that will stump the viewer/reader and find a way of addressing them -- without burdening the savvier viewer/reader. It's like the challenge of creating an index for a publication: which topics will the FUTURE reader be interested in? How will they GO LOOKING for that information? (will they search for "potentiometer"? Or, "volume control"? Or, "knob"?)

Now, imagine that *every* reader had to wade through the index in order to "read" the publication!

Good luck with your demos! Hint: try a dry run (even before filming) on someone who knows the material with an eye towards "what will the viewer *not* know -- that you and I take for granted?"

E.g., many decades ago, I came across a woman I worked with patiently waiting by her computer -- several minutes. When I asked if she was having a problem, she replied, "No, I'm just waiting for it to boot up..." (red flag!) "Um, how long have you been waiting??"

Turns out, she had inserted the (8 inch!) floppy rotated 90 degrees. She'd still be sitting there all these years later as it hadn't occurred to anyone to make it clear that there is a right way and a wrong way to insert the floppy! (they're square!)

Reply to
Don Y

If I were teaching in a college environment, where I had a quarter or a semester to go through the material, I could do the lab thing. I'm simply not going to stand over someone in a room and give them half an hour to code something up -- that works when things go well, but if I had such an assignment I'd want to do the initial work where I could curse and throw things.

In a short-course, I think you need a much more managed environment, and possibly one that uses some interpreted language (so that they can type in the "lines of code" into a terminal and have them execute immediately, without waiting for a build).

I don't know if it could be done successfully.

--
Tim Wescott 
Control systems, embedded software and circuit design 
 Click to see the full signature
Reply to
Tim Wescott

Using the sorts of languages/IDE's that we're used to *building* with, yes. But, you can abstract away alot of the silly syntax that is required in most current languages because *they* have to address a wide variety of application domains.

OTOH, if you were creating an environment specifically to code up "control algorithms", you could have operators that were better suited to these sorts of tasks -- so folks aren't busy looking for unbalanced parens or missing puncuation, etc.

(When you pseudo-code, how strict is the syntax that you impose on that code? Yet, you *know* what it is intended to convey)

Exactly. You wouldn't build a PID loop using simple arithmetic operators -- too much trivial coding time would be wasted. Instead, you'd have an "integrator" available with a reset pin, etc. You'd be able to examine its current value and use that to prevent windup, etc.

You'd have a means whereby the user/coder could *watch* the value of variables (graphically) over time so he could see where things weren't behaving properly -- instead of requiring him to select "watch variables" that he monitored *numerically*.

I'm gambling that it can. I've a few months to hack something together. Note that I don't have to worry about performance, efficiency, boundary conditions, etc. -- things that often make real-world coding "tedious".

Thanks!

Reply to
Don Y

This is already a thing, but only in Sweden (so far):

formatting link

The actual participants not having to be in the same "brick 'n' mortar" location also has international legal precedent:

formatting link

For the attendees it's a lot like a lecture environment...

Mark L. Fergerson

Reply to
Alien8752

I'm sure the honeymoon is quite "unsatisfying"!

Reply to
Don Y

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.