Profile

unixronin: Galen the technomage, from Babylon 5: Crusade (Default)
Unixronin

December 2012

S M T W T F S
      1
2345678
9101112131415
16171819202122
23242526272829
3031     

Most Popular Tags

Expand Cut Tags

No cut tags

August 28th, 2009

unixronin: Pissed-off avatar (Pissed off)
Friday, August 28th, 2009 10:49 am

Yesterday was ... an eventful day.  Not in a good way.

Let's start with the Silly Goose's loft bed, which we got off Craigslist (or maybe Freecycle) last ... fall, I think.  Nice full-size mahogany loft bed frame; she'd been asking for a loft bed for some time.  But it turns out to have some transverse stiffness issues (as in the end frames were coming apart because they were held together only by dowels).  I researched the problem some, and concluded that the best fix was to put some 5-16/18 machine screws through the corner posts into barrel nuts in the end beams to tie the end frames tightly together.

The barrel nuts, I could find easily.  But the bolts were another story.  Have you tried to buy 6" 5/16-18 machine screws in a US hardware store lately?  You can't.  They're a slow-selling item, so they don't carry them at all.  If you manage to find them at all, they'll be slotted round head.  (Why anyone still uses slotted-head screws for any job bigger than glasses frames is beyond me.)  I managed to find 5" ones online ... in stainless steel at $5.80 each.  The hell you say.

Finally, last week, my father managed to find a box of 6", 5/16-18, countersink Philips head machine screws at Fasteners Inc. in Pasco, Washington.  I rejoiced, and went out to buy the 24 barrel nuts the job would need.  Lowes Hardware only had 13.  I asked them to order more.  They said they would.  Yesterday, with the bolts sitting on my workbench, I went back to get the rest of the barrel nuts I needed.  They hadn't ordered them.  I called Home Depot in Tilton.  The hardware department guy there wasn't entirely sure what I meant, but was pretty sure he'd never seen any such thing in the store.

If hardware stores continue phasing out their slowest-selling items, pretty soon the only things you're going to be able to buy at hardware stores are light bulbs, duct tape and drywall screws.  And then people will fix everything with duct tape and drywall screws because there'll be nothing else they can buy.

(OK, so I'm exaggerating a little there.  But the principle stands.)

But that wasn't the biggest fun of the day.  Oh, not by a long way.  You see, we discovered earlier this month that our front porch needs work.  [livejournal.com profile] freetrav commented that it felt spongy, and lifting up a couple of the no-longer-attached tiles, we found that the plywood substrate under them was rotten.  So the front porch went on the fix-it list.  Yesterday, [livejournal.com profile] cymrullewes and I having discussed various options for rebuilding, we picked up the tiles and tore up the rotten plywood to see what lay below.  We were hoping to find a little bit of 2x4 framing over the top of the concrete slab, in which case the fix would have been dead easy:  tear out the framing and lay 2"-thick concrete cobble pavers directly on the slab.  (Well, OK, over a little mortar to hold them in place.)

What we actually found under the three layers of rotten plywood was floor joists and a yawning hole directly down into the crawlspace.  Worse, one of the floor joists is also rotten, because no attempt whatsoever had been made to seal the porch against rainwater aside from the layer of tile on the surface.  There wasn't even a plastic or Tyvek moisture-barrier layer bbetween the plywood and the joists.  Worse yet, the rot has spread in past the front door and is into the entry hall floor, and possibly into the floor of Pirate's bedroom on the other side of the wall.  And worst of all, it was obvious that within the last two or three years at most — which is to say, shortly before we bought the house — someone had discovered the rot, and instead of repairing it, had just roughly shored the ends of the joists up from below as a temporary fix, and then tacked styrofoam sheet insulation in place to conceal the problem.  So a $75-$100 "tear out and repave" job has just turned into a $4000-$5000 "call in a contractor to rebuild the entire front entryway and porch" job.

Fortunately, our neighbor across the street is a building contractor.  We already have the wheels in motion to get the job done through his brother's construction company.  For now, we have a couple of pieces of plywood covering the hole and some heavy plastic sheet across the top of that to keep any rain out.

We could have done without this problem.  But at least we discovered it before somebody went through the floor of the porch.

unixronin: Galen the technomage, from Babylon 5: Crusade (Default)
Friday, August 28th, 2009 10:53 pm

On babylon5, I run procmeter3 for general system activity monitoring.  But in my new-to-me laptop, whitestar, the version I have has a compatibility issue with something that's causing it to crash as soon as any NFS volume is mounted by any method.  So until I resolve the problem, I'm temporarily running gkrellm on whitestar.

It seemed to me like whitestar's CPU seemed ... kinda busy for an idle machine.  So I looked at the CPU usage stats.  Then I looked at cumulative CPU usage against uptime. Procmeter3 on babylon5 had consumed 34 seconds of CPU time in 11 days and 2 hours. Gkrellm/d on whitestar had consumed 56 minutes of CPU time in not quite 33 hours. Per unit time, gkrellm and gkrellmd together on whitestar are consuming almost a thousand times as many CPU cycles as procmeter3 on babylon5, on the same processor architecture running at the same speed, to display the same information.

Any time you have a tool that does exactly the same job as someone else's tool, but takes a thousand times as much CPU time to do it, it should be a sign that you wrote some really crappy code.

unixronin: Galen the technomage, from Babylon 5: Crusade (Default)
Friday, August 28th, 2009 10:53 pm

On babylon5, I run procmeter3 for general system activity monitoring.  But in my new-to-me laptop, whitestar, the version I have has a compatibility issue with something that's causing it to crash as soon as any NFS volume is mounted by any method.  So until I resolve the problem, I'm temporarily running gkrellm on whitestar.

It seemed to me like whitestar's CPU seemed ... kinda busy for an idle machine.  So I looked at the CPU usage stats.  Then I looked at cumulative CPU usage against uptime. Procmeter3 on babylon5 had consumed 34 seconds of CPU time in 11 days and 2 hours. Gkrellm/d on whitestar had consumed 56 minutes of CPU time in not quite 33 hours. Per unit time, gkrellm and gkrellmd together on whitestar are consuming almost a thousand times as many CPU cycles as procmeter3 on babylon5, on the same processor architecture running at the same speed, to display the same information.

Any time you have a tool that does exactly the same job as someone else's tool, but takes a thousand times as much CPU time to do it, it should be a sign that you wrote some really crappy code.

Update, 0110:  Turns out the procmeter3 issue I was running into was an ACPI bug that is fixed in v3.5b. Problem solved, procmeter3 all happily configured on whitestar and working.