Silicon Valley Redneck

A very gray Saturday

Rainy this morning, mostly cloudy all afternoon, and supposed to rain again later.

I did some indoors stuff, and spent a chunk of the not-rainy afternoon destroying tiger habitat in the far backyard.  Didn't see any actual tigers, but a lot of the usual rodents and bunnies fled.  I'm kinda-sorta getting the hang of dealing with extremely tall grass.  Still need to develop a feel for what slopes are safe, though.

Corruption here, corruption there, corruption in Chrome user profiles....

A few weeks ago, I lost the ability to comment on some WordPress-based blogs using my WordPress account and Chrome - the login would just hang.

A few days ago, I lost the ability to comment on many Blogspot blogs; ReCAPTCHA kept telling me to change to a supported browser.  Like Chrome.  Which is what I was using.

A little rummaging around the Internet, and it began to look like my Chrome user profile had gotten corrupted.  Progressively corrupted, it must be, given that WordPress flaked out before ReCAPTCHA.

So!  Rummage around, find ~/.config/google-chrome, find various things with names like "Profile 1", try renaming a few of them while Chrome is not running, then restart Chrome, and all the profiles are still there.  So, I guess, I was looking in the wrong place?  Or those things are misleadingly named?

Exit Chrome again, make a copy of the config directory, restart Chrome, delete my normal profile, restart again, re-create the profile and wait for everything to drizzle back from The Cloud.  Now it seems to work properly.

Dashing through the grass

In a 26 HP open-station tractor...

... This isn't going to scan, is it?

Anyway, that was the early-afternoon activity: making a first (and perhaps sufficient) pass through the front not-lawn.  The grass there isn't as heavy as some areas out back, so I could actually go fairly fast without the mower clogging up, but there are still a lot of tall tough stems left behind amid the vast expanse of clippings.  I had to stop a couple of times to clear fluff off the grille, lest the engine overheat (the temperature gauge, while still well short of the red mark, was climbing alarmingly).

When I got back to the barn and applied the leaf blower, an amazing amount of the fluff emerged from the radiator.  Er.  Maybe I'd better check the air filter, too?

That leaves the front-side area, and most of the back.  Parts of it I'll probably use the lawnmower on.  Parts I'll likely ignore.  Or maybe I'll get someone out here with a proper brush hog.

Oh, and there was wildlife.  Plenty of assorted bugs, a brace of bunnies, and approximately 6.022×10²³ voles.  Perhaps, with the cover now diminished, the small birds o' prey will carry them away, and we'll never see our rodents any more.

Ah, word associations.

UPS is now letting me follow (approximately, anyway) the progress of the delivery truck on a map.

This time of day, the truck is always making morning deliveries around Catlettsburg.

Which... when I see that name, I just can't help thinking of the cliff of the Catlats from Galactic Patrol.

Not all nots are equal

And I'm not talking about the difference between a square not and a granny not, nor anything else to be found in the Boy Scout Book of Nots.

That pesky past-week-or-so slow-mo remote debugging session with regard to the ToasterCo project?  The one where something in either the C code or the Verilog code (unless it was somewhere in the hardware) had started mysteriously misbehaving after I implemented a special preheat mode for that one brand of bread?

Little difference between ~ and ! is what.  They're right next to each other on my keyboard, and they're both prefix operators in C, so there's no error message...

There's the place, see, where if we're in a particular operational state and the preheat timer has expired, we clear a bit in a control register, thusly:

    if (!special_preheat_timer) {
        TOASTER_CONTROL_REG &= !TCR_OROWEAT;
    }

Which... yeah.  Wrong not.  Instead of clearing a single bit (bitwise not of the mask = mostly ones), it was clearing the entire control register (logical not of the mask = zero).

Really hard to spot, that one.  Not so much as a warning from the compiler (I suppose lint might catch this sort of thing).  And it looks right... almost.

Disclaimer:  Oroweat has no connection to the project, and the appliance in question isn't actually a literal toaster.

It's heating season

That means the indoor humidity is kinda low.  Which isn't good for my sinuses.

It's also the time of year when Southmoon's superpower manifests itself: she shoots lightning from her ears!  Is that a Pokemon thing, or a Sith thing?  She can't shoot it very far, but it's powerful enough to get a crackle out of my desktop speakers all the way from the far side of the desk.

In other cat news, ongoing Top Hat / Sumatra strife is ongoing.  Sumatra wants to make everything his, and Top Hat doesn't appreciate it,  Yes, I know I haven't gotten around to introducing the new (as of February) cats here.  One of these days....

Tuesday report

A fine clear sunny fall day, aside from morning fog.

On the yard front, I did some more mowing around the unfinished earthworks; this is slow'n'careful stuff, being as how the ground is decidedly uneven.  Also, I managed to stall the tractor once, and a few subsequent times caught the signs early and prevented the stall.  (The 60" mower, in heavy grass, really is at the limit of what that little 26 HP engine can handle.)

Cleaning up the tractor and mower deck is very much easier if one has a leaf blower, which is one of the reasons I bought one (battery powered) a few weeks ago.  Not very effective on mud, mind you, but current operations aren't muddy.

Haven't done anything further with regard to setting up the mill, apart from ordering a miter-saw blade meant for cutting nonferrous metals (arrived today, and I installed it) and a ½" by 6" by 24" piece of aluminum (maybe coming tomorrow), which, cut to length and drilled for the mounting screws, should be a good underpinning for the mill while giving the required lift.

The forecast calls for rain much of next week, so this seems like a good time to be catching up on outdoors stuff.

On top of a tractor, all covered with hay...

Mowing the tall grass with the mower deck on the tractor is slow going, but possible (at least the parts that aren't too steep).  It's a lot better than trying to do the same thing with the zero-turn mower, not necessarily in terms of mowing capability but certainly in terms of my head being above most of the grass as I sit on the tractor.

And, yes, dry plant detritus was everywhere, despite the minimal wind.  Got all over me, and all over the tractor.  The grille seriously needed cleaning at the end of the (not very long) session.  No, the engine wasn't getting hot; I kept checking.

The taller the grass, the less of the mower width I can use per pass; with the really tall stuff, using more'n about the leftmost third of the nominal cutting width quickly results in big wads of cuttings barely making it out the discharge chute.

The tall grass also doesn't tend to cut cleanly, needing a second pass (at least) to take care of the leftover high clumps.

Ah, well.  Some progress today.  Sometime soon, I really ought to inspect the drainage area (part of what I cut today) and decide where to put a narrow drainage ditch with a bridge over it.

I remember hearing this story at a con, years ago.

Leslie Fish now recounts on her blog the origin story of the infamous Viper Militia.

(I'd leave a comment there... but Blogger says my browser isn't supported by reCAPTCHA.  It also says that Chrome, which is what I'm using, is supported.)

Anyway, 'tis a story to keep in mind when the headlines are filled with terrifying tales of domestic terrorist conspiracies foiled by the Feds.  Especially when a large percentage of the conspirators turn out to have been government infiltrators.  (Cue that Get Smart episode where Max infiltrates a KAOS cell.)

Sunday morning

Southmoon heeds the message of this YouTube preacher:

Out in the barn (now that it's light outdoors, and not so frosty), 'twas time to make preparations...

Oh, yes.  Yesterday's word is that the guy who normally harvests the hay (here and various other not-really-agricultural places) is under the weather, and also thinks he has enough hay for the winter, so he's not harvesting any more this year.  He did offer to come out (once he's feeling better) with a brush hog, but that leaves the cuttings lying around, just as a DIY mow operation would.  Could be the neighbors with the cows would be interested... but they have a small herd for this winter, and also have a large stockpile of hay already.  So, I'll put the mower deck on the tractor and see how it copes with the really tall stuff.  I won't be attempting any steep areas with that.

So, preparations.  Need to take the loader and box blade off the tractor.  Pull forward, remove loader, back up a few feet, remove box blade (on the concrete floor).  Removing the lower pins from the box blade, for what I hope will be the last time, is a royal pain despite the flat surface.  Pull forward again.  Set about installing the Pat's Easy Change dingus.  Determine that it seems to require both shims.  Decide to come indoors, have lunch, and look up a vid or two on the installation process.

Plan for after lunch: put the box blade back on with the PEC in place, then take it out and dump it in its storage spot.  Move the tractor back into the barn and install the mower deck.  Maybe even try doing some mowing in the jungle zone (max height, max speed, low feed (sorry: max RPMs, low forward speed), and keeping a previously-mown area to starboard so's there's a place for the detritus to go).  I'm thinking I don't really need to mow the whole back yard - it's not like an HOA inspector is likely to come looking - but I certainly want to do the front yard and the area around my unfinished dirt-rearranging project.

Happy Sunday!

Update:  After having to sort out a couple of mistakes (e.g.: it doesn't need both shims; it just needs to be installed with the hitch in the halfway-raised position), I think I now have the Easy-Change thingies installed and adjusted pretty near correctly.  And grabbing the lower pins of the box blade was so easy I thought it must have gone horribly wrong.  The original top link does extend far enough... I think.  At some point, probably soonish, I need to try hooking up the tiller and make sure the PTO shaft reaches.

Maybe I'll get the box blade stowed and the mower installed this afternoon.  Also got a bunch of other stuff to do, so no telling which tasks I'll end up getting around to.

Sherline NexGen mill: first impressions

The mill being here, in pieces, and a bench now being available for assembling said pieces onto, I'll be assembling it, and trying it out, over the next few days.  This post will, accordingly, be growing and otherwise being edited without notice as things progress.

This is an inch-pattern NexGen mill, model 5800, with Package A and the DRO option, so 5800A-DRO.  While I was at it, I ordered various accessories (rotary table with tailstock and right-angle adapter, tilting angle table, sensitive drilling attachment, and some fiddly stuff).

One thing I'd noticed before it arrived was the mounting holes: the documentation gives the pattern, but doesn't show the hole size.  (Turns out it's ¼".)  Now having the hardware in hand... Figure 18 shows the 5800 series having six mounting holes, but the hardware I see in front of me has only four.  Or... I guess they tried putting two sets of dimensions on a single drawing, instead of having one drawing for the 5800 and another for the 2000?

Anyway.  The base is well goobered with sticky grease, which is encrusted in turn with packing-material fragments.  Takes a bit of cleanup, and I'll probably want to replace the grease with something lighter.  Synthetic motor oil, maybe, or gen-u-wine way oil fit for a Bridgeport (whereof I have most of a jug).  Sez the manual:

Use a light oil such as sewing machine or 3-in-1
oil or grease on all points where there is sliding contact. This
should be done immediately after each cleanup. We grease
the slides at the factory to ensure the lubrication stays in place
during shipping, but light oil will work fine once you begin
using the machine. Do NOT use WD-40 for lubrication!

Methinks either motor oil or way oil falls somewhere on the "light oil or grease" spectrum.

For a mostly-aluminum machine, this critter is heavy.  Methinks I don't really want to get too far in assembling it until after the chip pan (Amazon calls it a "baking sheet") arrives on Sunday.  Assembly can wait until after the base is bolted down to the bench with the chip pan under it.

Speaking of chip pan: 'twould be a nice option for Sherline to offer, wouldn't it?  That and a dust cover sized for the 5800 series (it doesn't look like either of the sizes on offer is big enough).

Life would be easier if there were a list of what parts came in which bag; they're not grouped nicely by subassembly.  Also: the instructions for "2000- and 5800-Series Mills—Assembling and Mounting the Multi-Direction Column" are for the 2000 series; they don't match the 5800.  This is extra confusing.

I've already ordered some scaled-down tools for tasks like, e.g., tramming the head.  I have the proper dial test indicator and holder for tramming the head on a Bridgeport, but the spindle nose on the Sherline is kinda smaller, so a different holder is required.  Also on the way: a mini surface plate (I used to have a normal-sized one, back in the 90s, but that got left behind somewhere along the line and it wouldn't fit on the Sherline's table anyway).  Also ordered a 3/8" center/edge finder, as the ½" ones from back when are, again, a few sizes too big, and I figure a 3/8" should fit the endmill holder.

I did find those little lathe chucks and faceplates that I'd had since forever, and they do indeed fit the Sherline spindle thread.  They're a bit rusty, though.  I'll probably end up buying a brand shiny new Sherline 3-jaw to go with the rotary table.  Ah... the 3-jaw for my old Taig also fits, and is of similar (not identical) construction, and in decent shape.  Maybe I need to take a closer look at how the current Sherline chuck differs from these.

I don't think Sherline has gotten fully into the JIT mindset.  One of the accessories was packed with a piece of newspaper from October of 2011.

The DRO handwheel thingies came pre-installed, but the tach sensor/decal for the spindle is a DIY project.  This involves drilling a 1/16" hole in a location that's highly inconvenient after the headstock/motor unit is assembled, but would have been easy earlier.

It looks like M6 socket-head cap screws would be just about perfect for mounting the base to the bench, if I were confident about my ability to hold close tolerances in marking locations and installing the threaded inserts.  Being not so confident, I guess I'll go for M5 and washers.  Should be strong enough.  Bonus: it seems that the M5 inserts want just about a ¼" hole (I need to verify this on some scrap wood), which is handy if I use the base of the mill as a drilling jig.

Anyway: I now have it mostly assembled, but still in three pieces, each of manageable weight.  The motor speed control and spindle tachometer are verified working.  Further assembly awaits arrival of the chip pan, which should happen Sunday.  No point assembling the whole thing and then lifting it onto the pan....

Friday evening:  The 18" tooling plate came with documentation for a different 18" tooling plate.  Says it's a laser engraving tooling plate, black anodized (this'n is plain shiny aluminum), with a different hole pattern.  Oh, well: find and download a PDF of the corresponding document for the 7x18" mill tooling plate, which seems a better match.

Saturday morning:  No, I'm not working on it this morning, other than dragging the bench into something like its final location (after moving a clutter table and assembling a shelving unit to organize some of the clutter) and pondering the instructions.  There's a mysterious bent-wire-mesh thing, kinda Z-shaped, that doesn't seem to be on any drawing nor parts list.  I'd assumed it was meant to be a guard of some sort.  Maybe it's a stand for the DRO box?

Saturday evening:  The mail came very late today, but included all the things I'd been expecting from Amazon tomorrow, including the baking sheet chip pan, and one unreasonably heavy box that just barely fit in the mailbox (so I had to lug it all the way up the driveway; parcels too big for the mailbox get left by the garage, but the ones that do fit get left in the box, way down there at the curb).  What's so heavy?  Ah... a rock.  With a certificate of calibration.  96 cubic inches of shiny granite.  Um, call it 2.7 g/cc, and that works out to... 1573 cc, so about 9340 millipounds.  Sure felt heavier than that while I was carrying it.

The design of the chip pan is faulty; the rim stands too tall for easy use of the Y-axis handwheel.  It's almost like Nordic Ware didn't intend it to be used under this model of milling machine.  Guess I'll have to do some cutting and/or bending.

Sunday morning:  On closer examination, the right-hand side of the pan also gets in the way of fingers when cranking the X axis over.  So... I guess I need either a significantly shallower pan or some sort of standoff(s) for mounting the mill.  After it gets light, I'll rummage through some of those Bins in the Barn™ to see if I have aluminum stock suitable for the purpose.  Gotta be big sturdy standoffs, but I'm thinking that a couple of pieces of extruded aluminum a couple of inches wide by half an inch thick or so - by, of course, something more than the width of the mill's base - might be appropriate.  Or, heck, a couple of pieces of stainless steel of such dimensions.  (Oog.  But, if I use standoffs, I need correspondingly longer screws.  Or, I mount the standoffs, through the pan, to the bench, with offset holes for mounting the mill to the standoffs.)

Sunday afternoon:  Looks like lifting the mill by ½" will do the trick nicely.  Rummaging through barn bins, I find a piece of extruded aluminum of suitable size, assuming I can cut it neatly into appropriate lengths.  Also two cutoffs that would, as a set, be perfect were they but 1" longer each.

Well, no more time for this project today.  Maybe Monday.

Thursday morning:  I omitted to mention here ('twas in the Tuesday report) that I'd thought better of the idea of supporting the mill by the ends on piers, as it were, and instead ordered a chunk of aluminum of correct size.  Well, a bit oversize.  Said chunk of aluminum having arrived yesterday afternoon, I drilled some holes in it, put the nonferrous-metals blade on the Big Scary Saw, cut it to length, drilled holes through the chip pan and into and through the bench top (intentionally, that was), and, with some considerable fiddling, got it put together:

Yes, the garage is rather cluttered.  It's gradually improving, though.

The minimum spindle-nose-to-table distance (not shown) seems a bit much, but I haven't yet installed the tooling plate, the endmill holder, and an endmill; once I do that, I'll see how close it comes to carving up its own tooling plate.

I still need to sort out cable routing.  And set the backlash compensation on the DRO.  And, well, try actually carving some scrap piece of aluminum.

Since (according to the latest forecast) we only have a couple more dry days before a rainy week arrives, maybe I'd better put this on hold for a rainy day and go do some outdoors stuff instead.

Friday afternoon (23 October):  Got the tooling plate installed, and... that spindle-to-table distance is still looking very excessive.  Contact Sherline.  Apparently a standard-length endmill should end about ½" above the tooling plate, which says that I've somehow ended up with the column more'n an inch above where it's supposed to be.  It gets to be the end of business hours before this is sorted out.  Meanwhile, I set the backlash compensation on the DRO, and do some fiddling.

Saturday morning:  I want to try cutting something, dagnabbit!  Mount the vise to the tooling plate.  Look for the parallels... hmmm?  I ordered a parallel set... I thought I unpacked a parallel set... musta gotten misplaced...?  Oh, well: search later; for now, there's a small thin parallel set in the barn, recently extricated from one of the moving bins.  Fetch that.  Clamp a random pot-metal casting (broken) in the vise, and try making some cuts, with and without the DRO.

A Bridgeport it ain't, but anything even vaguely close to a Bridgeport wasn't happening this year.  It takes care and patience... but it does cut metal, smoothly and precisely.  Which is exactly what it's supposed to do.  And it looks like the DRO is going to be a big help.

Oh, and there's an uncommitted #10-32 hole in the left-hand side of the headstock.  I'm thinking of adding a mist coolant system and/or a light, so having an attachment point will come in handy.

The first couple of projects in the queue don't actually require resolving the spindle-height issue, just sorting out work-holding, so at this point I think I have a useful machine.

New conspiracy theory: I like my old version better.

There's this story in The Beastly Day (via Twitchy) about a weird conspiracy theory that Seal Team Six actually killed Osama bin Laden's body double, not Osama himself....

Which, so far, matches my own conspiracy theory, formulated shortly after the event.  But then it diverges; the new story holds that OBL is still alive somewhere.

Mine was better.  Mine held that the SEALs had killed (or, better, captured/rescued) the double... and captured many documents*, including proof that OBL had been dead for years.  Details of his demise would follow, surely involving a betrayal by his inner circle (i.e., those who were still visibly running Al-Qaeda).

* Along with OBL's private stash of camel pr0n, lovingly preserved by his successors.

Any software problem can be fixed in hardware.

That little saying of mine goes back to one of my first embedded-computing projects.  A valve was sometimes not changing state fast enough, but by the time we discovered this the development system was elsewhere so making firmware changes was problematic.  The solution?  A one-shot, triggered by the valve control line, pulling the READY line low for that extra millisecond!

Fast forward to this week, and my attempt to implement a new feature for ToasterCo.  One of the bakeries wants a special preheat cycle for their bread.  Like, really special.  This requires changes to the Verilog and to the C code for the AVR core.  And... not having a copy of the hardware here, I check in my changes to be tested there... and it misbehaves in two very interesting ways.  Whatever could be going on?

D'OH!  The hardware is derived from an older model that I designed, and has various safety interlock features designed in, one of which is there specifically to prevent exactly what the firmware is now trying to do.  Fixes that dangerous new firmware bug real good, though the results are somewhat mystifying if one doesn't take the hardware details into account.

So I guess the solution is to modify the hardware (removing one diode) to disable the interlock.

A note to linen-mongers

Sheets and towels get washed in hot water.  If the laundry tag says otherwise, the product is defective by design.

I think I mentioned this before, years ago, with regard to fitted sheets that (unlike the ones of my youth) cease to fit after a few washings on account of shrinkage.

Fully hot water is how sheets and towels get properly clean.  Give 'em the cold-water-no-bleach delicate treatment instead, and they soon become horribly raunchy.

This has got to be another case of nobody being properly in the loop.  I guess the people who'd be in a position to make decisions about such products don't use their own products and do their own laundry.

Google Finance: big makeover, but still weirdly wrong

So I check Google Finance, um, yesterday morning, I guess...

First thing in the morning, it's the old familiar.  Sometime later in the morning, it changes to an entirely new format.  And it looks like the positions of entries has gotten even more dynamic, so instead of things moving around every few days on the basis of unknown criteria, they move around constantly on the basis of market activity or something.

Totally new user interface, and it doesn't look like there's any "classic" option on offer.

But, y'know what hasn't changed?  Backend bugs is what.  Like, the dynamically-updating summary page may show a substantially different price for a given item than you'd see if you clicked through to the detail page for that item.  Maybe delayed by several minutes, or maybe just buggy.

It's like nobody involved in the development of such apps actually uses them.  Or little factors like usability and presenting correct and meaningful data take a back seat - nay, a trailer, and a shabby one at that - to the Make It Shiny Imperative.