Next book edition almost done

Eleven days ago, I said

Now all I have to do for the next release of the book is do the standard final checks (page breaks, spell checks, and URL checks).  This will probably take me another week.

I have now gotten the page breaks fixed and checked all the URLs (only 8 of the 215 distinct URLs needed fixing).  I last checked them about a year ago, so that is a link-rot rate of only 4%/year (a half-life of about 18 years).  In the process of fixing the page breaks, I noticed and fixed a few minor typos, as well as tightening the text in a couple of places (to improve the page breaks).  I found one instance of “the the” with my tandem-word checks (probably introduced since the last released edition).

I still have to do the spell checks.

I did release one new video last night: https://youtu.be/vLece-VKfkQ, which talks about providing a constant current for electroplating (see the post Controlling current if you don’t want to waste time watching an 11-minute video).

Forty-seventh weight progress report and 2020 year-end report

This post is yet another weight progress report, continuing the previous one, part of a long series since I started in January 2015.

I’ve been putting on weight since my big diet of 2015, with occasional attempts to correct course.

This year has been a particularly bad one, with my weight reaching the highest value ever—touching the “overweight” range. I was doing ok in the first quarter—gradually dropping towards my desired weight, but once the pandemic started and I was staying home snacking instead of bike commuting up the hill to my office, I packed on the pounds at an alarming rate.

It probably did not help that I spent a good chunk of the summer and fall experimenting with bread recipes. A few of the experiments are recorded on this blog, but a lot of the more recent baking has not involved any recipes to post to the blog (making whole-wheat sourdough using minor variants of the bread-machine recipe), and I’ve been too lazy to photograph.

My exercise has been very limited—I averaged only 0.85 miles a day of bicycling in November and December and probably not much more in walking (about 80–87k steps per month—way down from 230k in January). Because bike commuting was my main source of exercise, I remain concerned about how I’m going to get fit again—exercising at home does not seem to happen, even when I promise myself that I’ll do some today.

I did get in some recreational bicycling today—bicycling up Empire Grade to see the dozer line at the top of campus (not really very visible any more, as the grass has sprouted) and even higher up to see the edges of the CZU burn.  I’m out of shape enough that I turned around after only 7 miles and 1150 feet of climbing.  There was not much to see—if I want to see the burned areas, I’ll either have to go further up Empire Grade or choose a different part of the burn to look at—perhaps taking a flatter, but longer, route up Highway 1.

Other milestones for the year: I got all the videos done and the closed captions edited for BME 51B, and I’m almost finished with the videos for BME 51A for Winter (only 10–12 more to go, or about 3 hours worth of videos), though the caption editing has barely started.  The videos are available on YouTube as two playlists: Part A (for BME 51A, 108 videos totalling 24 hours so far) and Part B (for BME 51B, 49 videos totalling almost 12 hours). I did get a new release of the textbook out on December 28th, in time for the new class that starts on Jan 4.  The book is available (as always) from https://leanpub.com/applied_analog_electronics, and anyone who has bought it in the past (even with a free coupon) can get the latest edition free by logging in with their LeanPub account.

I’ve also managed to keep my backyard mowed this summer (it used to always be a jungle of head-high weeds). I’ve almost finished clearing the ivy and blackberries from the area behind the garage—I’ve only got about 25 square feet left to clear—about one more week’s greenwaste can, though the rain may bring back a lot of the blackberries, as I can’t remove the roots from under the concrete. In March, I didn’t think I’d be able to get this far. 

My son and I acted together for the first time, doing a short promo video for Santa Cruz Shakespeare.  I’ll post a link to it when they finish adding the title and donation info at either end.

When compared to my to-do list from September 2019, my accomplishments for 2020 don’t seem so great—a lot of the stuff on that list is still not done.  Oh, well—something to do after I retire in June.

PteroDAQ installation video

As part of my lecture videos for the first half of the electronics course, I’ve created a video on installing PteroDAQ:

and put in the playlist Applied Analog Electronics Part A.  The instructions are only for macos, but similar steps are needed on Linux and Windows machines.

Videos for Fall done!

I think that I finished the videos for BME 51B last night—53 days before classes start.  There are 49 videos totalling 11 hours and 49 minutes (averaging 14 minutes, 28 seconds).

So far, only one of them has had the captions edited (https://youtu.be/s36BWhteeho), but the student doing the caption editing is beginning to get the hang of it and I expect the caption editing for subsequent videos to be done more rapidly.  The process is somewhat simpler than I expected, as we don’t need to do anything with time stamps.

1.  Download the auto-generated transcript of the video using https://downsub.com/, which can provide the transcript in a simple text format without time stamps.  For some reason, YouTube does not provide an easy way to get this with their native interface—only providing a file in VTT format, which has lots of duplication of words in order to simulate scrolling.  The VTT format is much richer than the simple text format, but very hard to edit to correct bad line breaks and incorrect punctuation.  (I saved the VTT files, in case I need to restore the autocaptioning, but I don’t expect to use them.)There is a way to get the transcript on the native YouTube interface, but it involves screen-scraping and doing a cut-and-paste to a new file.  The downsub.com download just needs the YouTube URL and keeps the title of the video as part of the file name, making a much more usable interface.
2. Edit the text file, keeping all lines to 45 characters or fewer.  The point of the editing is to correct mistranscribed words (there were two or three in a 20-minute video, and for one of them even the student couldn’t figure out from my mumbling), to eliminate verbal glitches, to correct punctuation, and to put the line breaks at semantically meaningful places (where possible) to make reading the captions easier. Music at the beginning and end should be labeled as “[Music]”.
3. Upload the edited transcript, as described in https://support.google.com/youtube/answer/2734799?hl=en&ref_topic=7296214 and https://support.google.com/youtube/answer/2734796#upload.  One step seems to be missing there—the upload silently failed the first time I did it, with YouTube still showing the auto-generated captions the next day.  I tried again, deleting the autocaptions when uploading the edited transcript file, and that seems to have worked.

The student hired by the university does steps 1 and 2.  I do one more editing pass over the transcript, correcting any places where the student was unsure of the transcription and looking for punctuation errors (mainly comma splices and hyphenation errors).  The comma splices are a natural result of spoken speech not being obviously broken into sentences—run-on sentences are a common speech pattern.  Judicious punctuation makes the captions easier to read correctly without changing the words of the utterances.  On the first video, I fixed half a dozen punctuation and capitalization errors, corrected one word that the student was unsure of, and corrected another phrase that YouTube had mistranscribed and the student had not caught.  It took me much less time and pain to do that final editing pass than it would have to do all the editing, so it was definitely worthwhile having the student do the initial editing.

I have not yet estimated how many videos I’ll need for BME 51A for the Winter, but I think that there will be closer to 20 hours of video than the 12 hours for BME 51B, so I don’t think I’m halfway yet through the whole course. I’ll probably not quite have finished them by October 1, when Fall classes start, but I might be able to get them done before the grading load for BME 51B crushes me.

I’ll also need to do some videos that won’t go up publicly on YouTube: 10 quiz solutions each for BME 51A and BME 51B.  I’ll probably also do an intro video for each course, calling student attention to important parts of the syllabus—the intro videos may go up on YouTube, but not be part of the playlist.

I’m probably going to rename the Applied Analog Electronics playlist to Applied Analog Electronics Part B, and create a new playlist Applied Analog Electronics Part A.  That will make it easier for students to find the videos they need, and it will make it easier for me keep the videos in order—it is much easier to add a new video at the end of a playlist than to add it somewhere near the beginning.

Fall 2020 plans and tools

I swapped my Fall sabbatical for a Spring one, so that I could have time to figure out how to move my lab course into an at-home lab. I also took a university-sponsored course this summer on remote instruction to help.  That course on remote-instruction has now ended, but I still have some work to do on converting my course.  USCS will still be providing some help—they’ll be paying for a student (one of my graders from last Winter) to edit the closed captions on my videos.

I’ve made a list of some of the tools I’ll be using for the Fall:

• I’ll be using Canvas for collecting assignments (as I have for a couple of years) and SpeedGrader or Gradescope for grading them (the badly-named SpeedGrader for written assignments, as I’ve been doing, and Gradescope for quizzes, which will no longer be on paper).
• I have my syllabus on my own website on the University server, so that it is public (I hate secret syllabi buried inside a learning-management system—they make it nearly impossible for students to know what a course is about before they have committed to taking it). All assignments, reading schedules, … are there. Only the assignment due dates are duplicated in Canvas.
• I use Piazza for having students ask questions, as I’ve been doing for a couple of years. It is a much better interface than the Canvas discussion forums, and students are willing to use it.
• I’ll be using Zoom for the synchronous parts of the course—lab times and office hours. That is new for me, but I’ve been in enough Zoom meetings, classes, and webinars now to have some idea what is reasonable. I’ll undoubtedly find all sorts of new problems in the first 2 weeks of class (particularly with breakout rooms, which have not worked very well any time I’ve been in them).
• The small stipend I got for converting my course to online required that the lectures be fully asynchronous (and that any synchronous activities be attendable even by people in a time zone 15 hours different from mine). So I’m recording mini-lectures (6 minutes to half an hour) using OBS. I’ll probably have about 35–40 of them (considerably less lecture time than the usual 32.5 hours of lecture for the course, but some of that time would normally be taken up with quizzes). I’ve set up my desktop computer at home (shared with my wife) as a recording studio, with a green screen, a document camera, and (starting today, if it arrives as scheduled) a cardioid desktop microphone.
• I’m also planning to have video answers to the quizzes that are unlocked by submitting the quiz—those I’ll have to do at the last minute, as I usually write each quiz after seeing what students got wrong on the previous quiz. The quizzes are fairly low stakes (all 10 quizzes add up to about 13% of the grade), so I’m not going to worry that a few of the students are going to cheat like hell on them—I’ll be saddened by the cheating, but I’m not going to proctor. If I get obvious cheating (like identical very wrong answers that can’t be easily explained except by copying), I’ll still do the academic-integrity reporting and fail the students.

I’ve recorded 31 mini-lectures so far, and I’m gradually getting better at using OBS and lecturing into the void, but it is very, very different from my usual style, which involves 50′ chalkboards and is an improvisational performance in response to student questions. I’m glad that I only have to do this online stuff for one year (I’m retiring in 2021).

UCSC is  almost fully online this Fall (21 in-person courses out of 1300, mostly small lab or grad courses, so less than 0.5% of total seats).  I expect that Winter quarter will be much the same, though some optimists expect more in-person course (up to maybe 5% of total seats).

I still have to work out (with the Baskin Engineering Lab Support staff) the logistics of shipping stuff to the students. The parts that were ordered for the cancelled Spring course are available in storage at UCSC, but we’ll have to add to the list to provide some duplication for mistakes (as students can’t get same-day replacements from the BELS supply room, nor can they borrow easily from classmates), plus providing resources that were previously communal (inductors, wire, solder, soldering irons, solder suckers, safety goggles, stainless-steel electrodes, electrode holders for Ag/AgCl electrodes, …).  I’ll be making up a list soon of the changed needs for the parts kits.  We’ll also have to work out which things the students need to ship back over winter break, for distribution to BME 51A students in Winter.

We’ll probably freeze the enrollment in the class in mid-September (no late adds this year!), so that there is time for shipping.  I currently have only 32 students in the class, down considerably from the 50 who passed the first half in Winter 2020.  Some students graduated in Spring 2020 (we granted some emergency course substitutions for students whose required courses were cancelled in Spring), but I was still expecting about 40 students this Fall.