Exercise restart: day 4 overview

The exercise log said it was time to call an outright restart on April 26 2015: the prior daily workout (core exercises and plank-type static holds) was March 17 2015.  I'm seeing better compliance this time on the daily workout (i.e., now well enough to actually schedule two ten minute blocks/day; would need a third 20-minute chunk to bring up the foam roller reliably.)

It looks like that the 45-degree and 60-degree incline bench presses are "close enough" to the shoulder press that they destructively interfere, so something needs adjusting in the split schedule for the dumbbell exercises.

The core exercise block is currently:

• crunches: 30 reps (i.e., no strength building, just maintain endurance)
• oblique crunches: 30 reps
• bicycles: 30 reps (had to recondition this, first day was short).  This is an augmented oblique crunch.  The limiting part of this exercise is a static hold of neutral lower spine.  This exercise was adapted from Pilates, where the recommended reps is six.
• Russian twists: This is in strength building mode; bodyweight is currently about 75% of my estimated 1RM.  This is another augmented oblique crunch, and involves a dynamic hold of neutral lower spine.  It's even more taxing on the lower back than bicycles.

Once Russian twists are at 30 reps I'll consider adding another exercise in.

The plank exercise block is non-optimal and I haven't quite worked out correct footwear for it yet; the compression socks don't give enough traction for either mountain climbers, or commando rows.

3 month report -- exercise program

Overview

• Finally on split schedule.  The weight lifting routine is ordered so that just about anywhere is a reasonable split point.  In practice, the split point is close to when the bench press series ends (now have: flat, decline, and 3 inclined bench press/fly supersets).
• I ended up taking off three weeks outright; came down with something on the 18th (2 days after the split workout Jan 15th/16th).  Ten days under six hours/night of sleep, then started catching up on the 29th.  Restart date is Feb 5th.
• I'm not sure whether the 3 week hiatus led to some deconditioning, or whether estimating 1 rep maxima from an AMRAP of 30 repetitions is just unreliable.  Both the second and third incline bench press/fly supersets (45 degree and 60 degree incline, respectively) were failures -- I ended up taking a pound off the third set and did not attempt the intended 4th set, as the left shoulder simply could not keep up; both the 2nd and 3rd sets of these inclined bench press/flys were grinders  The split session should resume with the shoulder press.
• As I don't think I really can afford to gain a pound of fat for each pound of muscle, I'm having to add in explicit aerobics.  Since I'm in a rehabilitation strength range, I don't expect to measurably slow down strength gains by doing this.
•  FitnessBlender has a decent (high volume) Youtube channel; the login on the their domain mainly provides a search index to their Youtube videos.  They mention that they're having new features built out for their site, so there is some intention to make their site provide reasonably compelling features.  Right now the search feature is the only thing making their site more competitive than collating exercise videos in a playlist.

2 month report -- exercise program

Overview:

• Frequency of workouts declined from 3/week to 1/week, roughly.  Time cost per workout is up to 90 minutes.  Typical time costs 70 minutes strength training, 4 minutes crunches, 16-ish minutes flexibility.
• The retests of estimated 1RM are still going up, typically 1 or 2 lbs/month.  In percentage terms this isn't bad (going from 14 to 15 lbs estimated 1RM is a ~7% gain).
• The folklore indicates that the postural problems I'm having at the shoulder level are from unbalanced shoulder muscles.  Cutting through the jargon is a problem here -- there are actually two layers (deltoids are the outer layer, rotator cuff is the inner layer).  And the exercises for the rotator cuff are exotic (cf. Omar Isuf ; ~5 minutes)
• New exercises to start on the rotator cuff: Cuban press, shoulder horn.  For the Cuban press, I'm using the variant I saw on Buff Dudes.  A bodyweight check indicates that my shoulders and/or upper back are too far out of alignment to even consider the WWE variant.  Two more exercises are indicated for complete coverage.
• The base exercise which gets me into the Cuban press is the Dublin press.  As a shoulder isolation exercise, the Cuban press requires a much lower weight than the Dublin press. The Dublin press should use a slightly lighter weight than the shoulder press, as it has more torque (the upper arm actually ends up horizontal, which is when the Cuban press does the rotation forward).  Currently, Dublin press is using 10lbs while the shoulder press is using 11lbs.
• The deltoid exercises are at a slightly higher weight than the rotator cuff exercises (5lb vs 4lb, usual 4-4-4-4 scheme)
• One of my early Christmas gifts was a foam roller; my interest in this exercise gadget is flexibility, not Crossfit like  balance training.  Results have been good so far.

1 month report -- exercise program

It has been one month since my first strength-training workout in the new attempt to regularly exercise (Nov 2, 2014).  The final inspiration was realizing that the core exercises of a powerlifting program (examples: Westside , Wendler 5-3-1 ) all had dumbbell variations -- and I already had a dumbbell set from my last attempt at a regular exercise program in 2010, which was Pilates-based.

Basic overview of Wendler 5-3-1: T-Nation Wendler interview , Muscle&Strength Wendler 5-3-1 overview .  Corresponding videos on Youtube : Brandon Campbell, ~9 minutes , FlexForAll2, ~17 minutes .

The four key exercises are:

1. squat (arguably long-term medically contraindicated for everyone due to the loading on the knees; I'm including this as a form check rather than trying to build strength)
2. deadlift
3. bench press
4. overhead press

These all have dumbbell version per Muscle&Strength's exercise catalog .  Also note that Wendler provides a standard way to estimate 1RM (1 rep maximum) from an AMRAP (as many repetitions as possible) set: [weight]*[repeititions]/30 + [weight] , where if repetitions exceeds 30 use 30.

So:

• Start with these four exercises
• Use 5-5-5 until I get around to measuring 1RM
• use 4-4-4-4 if I have the dumbbells for 80%-90% of 1RM or  6-6-6-6 if I only have dumbbells for less than 80% of 1RM .  These are inferred from Prelepin's table .  I'm oversimplifying slightly to keep the rule of thumb easy to memorize.
• add other exercises to cover holes
• keep the total time budget under an hour per workout

Eventually the time cost will force me into a split routine (too many exercises to do in one hour), but that hasn't been needed yet.

Right now, I'm not considering additional cardiovascular exercise; as long as I don't rest long enough to stop breathing hard, I should be doing low-intensity cardiovascular work during the dumbbell workout section.  I do want to deal with some inflexibility (e.g., I can't get into the starting position for a stiff-legged deadlift).  I'm moderately deconditioned so my estimated 1RMs are weak.

I don't know why I haven't read about this anywhere, but there is a natural relation of limiting weights between various dumbbell exercises, based on which exercises get you to the starting position for other exercises.  (The following doesn't work for barbells; normally one doesn't need to personally get the weight to its starting position with barbell exercises.  The Olympic lifts would be critical for getting the barbell into the starting position for a front squat or shoulder press, as they allow higher weight than a curl.)

• Deadlift
• Squat -- starting position is standing.  (This is not true for a barbell front squat or back squat; the barbell is at shoulder height for these.)
• Shrug -- starting position is standing
• Bent-over row -- pause at 45-degree inclination of the back
• Pendlay row -- pause at 90-degree inclination of the back
• bentover kickback -- let forearms go vertical while holding upper arm parallel to chest
• Curl -- starting position is standing
• Shoulder press -- start from top of curl
• Wrist curl -- pause at midpoint of curl
• Bench press -- have to hold onto the dumbbells while getting on the bench)
• Fly -- starting position is the top of the bench press

All of the above were in my Dec. 2 workout, but not in the above order.  Everything potentially involving the biceps or triceps was put after the shoulder press

Overly precise arithmetic

Bit-twiddling in the name of numerical accuracy, and speed, don't mix.  This matters mainly for quantum chemistry and similar numerical theoretical physics (where numerical error actually limits result accuracy), and n-body simulations.  I happen to need the latter to know what Iskandra's year is, rather than handwave it.  (It's also a credibility check for a proposed climax for the biographical novel about Zaitharak, the Iskandran Badger Copernicus/Galileo/Kepler.)

As Python's memory management model requires extreme contortions to minimize thrashing the garbage collector, it's technically disqualified even though it has arbitrary-precision floating point.  That leaves Ruby (speed factor of two slower than C for the Maya n-body simulator), and C++ .

The distraction for today was a first draft of a C++ class to wrap the algebraic manipulations needed to minimize subtractive cancellation errors, evade overflow, etc.  I suspect it will build on some prior work from 2006.  The medium-term objective is to guarantee calculating the Lorentz metric of a (four-dimensional) vector (x,y,z,t) in Minkowski space with maximal accuracy.

Science fantasy: editing progress

Two weeks of editing finally paid off last night: one chapter of working-title "Ill-Starred Romance" restructured properly, with a thoroughly inevitable complication documented.

But not properly expanded yet.  I wasn't medically literate enough to push the new draft to a stable version.

Z.C++: augmenting automatic memory management.

The recreational task for yesterday was replacing two non-owned raw pointers in zaimoni::MetaToken with a new type loosely modeled after boost::flyweight, zaimoni::flyweight.  I got a noticeable speedup in the test suites.

Specification
1) reference-counted; when reference count drops to 0, automatic cleanup desired.
2) This use case was C strings (automatic cleanup with free ok), immutable between initial setup and destruction.

This was a natural for boost::flyweight.  Unfortunately, I didn't have the background to immediately understand the boost::flyweight documentation; I ended up prototyping a partial reimplementation to make sure I understood the design decisions, etc. behind boost::flyweight. (Specifically omitted: multi-thread safety.)