Well, I'd already bought the CANbus pod from Total Phase, and in the wee hours of the last business day of the year I just ordered the EAGLE upgrade - 1 seat, professional, from v5 to v6 - turns out it's $549 for the upgrade; with the compulsory donation to the Glorious People's Republic of California, it ends up being a few cents under $600, which is rather steeper than I'd expected, right around twice what the v4 to v5 upgrade cost three years ago.
Ah, well. There's a reason I have the professional edition: I put it to revenue use on a fairly regular basis. Doesn't take much of a project to pay for the software upgrade, though, thinking more objectively, it'll probably take quite a while for the feature enhancements to save enough of my time to pay for it.
It does look like I'll be needing to run a split environment for a while, as there's no "save as v5" capability, and I need to maintain the toaster-controller designs in legacy format so's the client can deal with them (though actually the "run as freeware" option seems to allow full read access, just not modifications outside the freeware limits). Meanwhile, I should be converting my other designs, and my library, to the new format and making use of the new capabilities - hey, look, arbitrary pad shapes!
Well, what with the heating-season air's effects on my respiratory passages, holiday stress, an overdose of sugar, and what seemed to be an 18-hour virus, I ended up spending X-mas indoors, all safe from Robot Santa.
Apart from not yet updating Iggle, I finally did get around to updating Flash Player (version 11 seems to handle Youtube fine; haven't yet tried it on any of the sites that weren't working with 10 and claimed that they needed version 10) and Firefox.
I'm not sure I like the rapidly-changing major version numbers, but Firefox 9.01 eliminates at least some of the annoyances - and especially the performance issues - of version 3.mumble, so I may in fact stick with FF instead of switching to Opera when I move to the new workstation (which might have masked the performance issues anyway). Now I need to figure out how to migrate my cookies, account info, and suchlike to other copies of Firefox, which I guess is easier than migrating them to Opera. Apparently that's what this newfangled "Sync" thing is for, but that opens a new can of worms with regard to things like online banking, so I'll have to take a close look at the security issues.
Oh, and I started mucking about with virtualization. So far, I haven't managed to virtualize the Windows XP installation from my old office Winbox (which also has my copy of Office installed, and some other stuff like the company copy of OrCAD), and it looks like I may need to stick the drive back in the old machine and make some changes before trying again: when I try to boot it under Virtualbox or AQEMU/KVM, even in safe mode, it blue-screens and then reboots before I can see what it's complaining about*, and repair mode wants the Administrator password, so I guess I need to (re)set that password. If I can get that sucker virtualized, the big old Athlon box can go away. Then I'll be wanting to do the same trick with the Windows 7 installation on the newish laptop....
*Hm. Maybe I could use a video camera as a debugging tool.
Got lotsa shiny new goodies, some of which I (or one client or another) might find highly useful. Diff pairs, trace length matching, BGA escape routing, ... ooh, and arbitrary pad shapes!
And going to XML format for design files promises to make life easier for sundry future activities, even if the files do get rather bulkier.
But: no "save as version 5 format" capability. So, I still need to use 5.11 on any files that get sent to the client that bought a version 5 license just last year.
Also, it's still a 32-bit binary, so running it on the new workstation, with 64-bit Debian Wheezy, requires, apart from finding libpng14.so.14 somewhere, installing a few other libraries in /usr/lib32 (this was confusing, as, e.g., libssl.so.1.0.0 was clearly present... just not in the 32-bit flavor).
And I still need to find out what it's gonna cost to do the upgrade, when I get around to it. There's an online system for doing that... which is kinda confusing at first glance... and it won't tell what the "upgrade v5 to v6" price is until the end of the process, which calls for filling out more forms that I feel like completing just now.
So, I guess I'll play around with the freeware version a bit, and get a feel for what the new features look like, and see how well versions 5 and 6 coexist. In My Copious Free Time, of course.
Once I'm comfortable with it, I assume I'll get the upgrade... and probably agitate for JCM to get a several-seat license; we can get the whole shebang for everybody for the price of one seat of schematic capture from another leading brand, making it ever so much easier to pass designs around among ourselves (if not with clients and vendors).
On morning walkies Monday, wildlife activity seemed normal: various deer grazing near the trail and watching the passersby, and a flock or two of turkeys.
Tuesday, no deer and no turkeys. Well, it was a frosty morning, and I'd broken out the winter wardrobe (Long sleeves! Long pants! Socks!), including something to keep my ears warm, viz, a deerstalker cap, and maybe the deer didn't appreciate being stalked.
Wednesday, warmer; I was back to my usual attire, but still no sign of deer nor turkeys. Did they all fly south for the winter?
Thursday, rain. Also, I got a slow start on the morning, and had Stuff To Do. No walkies, so wildlife not observed.
Friday, normal attire, normal route, normal-ish time; several turkeys right by the parking lot, and none further in. Two deer in one of the usual spots, acting skittish and hastening from view.
So... I guess something's got the wildlife spooked, and I don't really think it's me. Maybe next week I'll carry the big camera, in case of big kitties. Also maybe the pocket cutlery, in case of uppity big kitties.
Golly, Richard Dean Anderson's car broke down, and he couldn't fix it with a paperclip and whatever bits were lying by the roadside. Surprise!
I watched the pilot episode of that show, and a couple of others at random, and, well...
Not only was the tech completely wrong from beginning to end, often dangerously so (as in: not only will this trick not work, but if you try it in real life you'll blow your hands off), but such improvisations as didn't violate any laws of physics were still Infinitely Improbable.
Think: MacGoofus needs to improvise a gizmo. He grabs a few random bits that happen to be lying around, and they all fit together just so. You, on the other hand, order a replacement water pump for your truck, specifying the make, model, year, and engine type, wait three days for express delivery, go to install it, and discover that halfway through the model year they changed the bolt pattern and nothing lines up.
And what's with never using guns? Right near the beginning of the pilot episode, as I recall, he used a cheap glass hand-mirror to make a high-powered laser self-destruct. This Doesn't Work; the laser will destroy your cheap mirror long before its own high-quality optics are damaged. What does work for sure is breaking the laser's output mirror, e.g., with a bullet. (Your superior intellects are no match for our puny weapons!) A proper go-everywhere fixer ought to have a pistol and a few demolition charges with him at all times.
(No, I'm not in the habit of lugging around a pistol and a pocket full of explosives, but my style of fixing doesn't often call for that sort of thing. I'm more likely to need a soldering iron, an oscilloscope, a set of jeweler's screwdrivers, a variety of small pointy objects, and of course ductelectrical tape. If the job calls for a milling machine, I'll just have to take it home to work on it, as the inflatable Bridgeport remains elusive.)
In the wee hours of the morning, I was having a series of dreams, all vaguely connected, after a fashion... here are some bits I can recall.
As the series began, some friend and/or business associate was showing off something he'd learned from the Internet: how to make a late-model Datsun run on acetylene.
(Yes, Datsun. Not Nissan. And somehow I think this was inspired by Robert Zubrin and this guy. What's the octane number for acetylene, anyway? 43? -27?)
Anyway: this somehow seemed like a good idea, though I suggested using MAPP instead of acetylene (being somewhat less insanely dangerous). I commenced calling around to welding suppliers to get a handle on what filling up with MAPP would cost. Somehow, it didn't occur to me to inquire about the general availability of MAPP filling stations.
Meanwhile, with none of the issues resolved, and solely on the basis of a demonstration that a minimally-modified car could run on acetylene, we commenced equipping a factory to build our own, designed-from-scratch, welding-gas-fueled cars.
Then, after a discontinuity and some driving around in a half-built vehicle, a posse appeared, bent on rounding up one of the partners in the venture, a big guy with delusions of being Jack Harkness. Fortunately, they'd learned all about being the good guys from old cowboy movies, and he wasn't really a bad guy, so what ensued was comedy rather than tragedy as he taught them what was wrong with their methods (which mainly seemed to involve ropes, which they didn't really know how to handle).
And, somewhere along the line, I woke up to a world that doesn't really make all that much more sense. (Gee, I wonder if I could get a government subsidy to work on acetylene-fueled "alternative fuel" cars...?)
If I were actually getting into the car business, I'd probably go for a small-battery diesel-electric hybrid with a motor at each corner. Could probably get great results in both fuel efficiency and traction. Makes a lot more sense than re-inventing the entire vehicle just to change the gas tank.
Now, if the test was in the same form as the sample questions provided by WaPo, being four-option multiple-choice, the average turnip ought to have gotten 15 out of 60 right. I don't see how anyone with a halfway-functioning brain and the slightest knowledge of arithmetic and geometry could score less than 90% on such a test.
OK, so maybe it wasn't that test, but the dreaded FCAT (sample material is helpfully linked by a guest commenter at TSM). Looking through that big PDF, flipping pages at random, I find a mixture of questions, not all straight multiple-choice, some of which are no-brainers, some I'd want a pencil for, and the occasional one that actually calls for a calculator, or at least a slide rule (such as the ones that involve taking square roots and actually filling in a number rather than picking one of four very different answers). Still: basic arithmetic, geometry, and logic will get you through most of 'em.
Yeah, yeah: I'm just a (a) iggnerant wingnut with only a BA degree from a stupid state university, or (b) sooper-jenius egghead with no grasp of the real world; pick one. Thing is, basic arithmetic and geometry are routinely useful in the real world, and I've even been known to use trigonometry from time to time (typically in sewing projects).
And it's truly frightening to contemplate the existence of a highly-credentialed ruling class, making decisions about, e.g., economics (simple principles*; big, complicated math!) while being entirely innumerate.
* I don't trust the ruling class to have any grasp of the simple principles, either. Look how the Parable of the Broken Window translates into policy proposals.
So here I am trying to come up with a workaround for the disappearance of Gnome Canvas from the latest Debian....
It looks like I can draw text with Pango. OK. I've even figured out how to draw it on a Layout of specified width. So:
Now what? It needs to end up superimposed on a base image, which is in a Gdk::Pixbuf. It seems like a Gdk::Drawable ought to be involved somehow. But... the relationships among Pixbuf, Drawable, and Layout are somewhere between unclear and nonexistent.
I'm remembering why I used Canvas in the first place. And why I decided to switch to Qt for future work.
But, I've got these existing little utilities written using GTK+, and they need to put text on little background images, and it has to work with both current Linux and old Windows installations. So I need to make this work, somehow, without Canvas.
Time to dig out a Book, printed on real paper, and see if that offers any insight.
Update: 'tis ugly, on account of sundry things don't Just Work and/or are exceedingly nonobvious. Here's the new, canvas-free code snippet, and using an ad-hoc fudge factor that happens to give decent vertical centering in my application:
# Without canvas if (!@font) @font = Pango::FontDescription.new("Sans bold 10") end pingo = Pango::Layout.new(Gdk::Pango.context) pingo.width = width * Pango::SCALE pingo.alignment = Pango::ALIGN_CENTER pingo.font_description = @font pingo.text = text
On the fields of Trenzelor, at the fall of the Eleventh, when no living creature can speak falsely or fail to answer, a question will be asked—one that must never be answered. And Silence must fall when the question is asked.
Now, what can this mean?
One might surmise that the question and the answer are mutually exclusive, and that if both exist in the same Universe they will cancel each other out and everything will go silent.
On the other hand: "Silence must fall" could also mean "The Silence must fall", as in: the religious order dedicated to killing the Doctor to prevent the question being answered shall surely collapse if the question is answered.
And "the fall of the Eleventh" could mean "when Matt Smith's contract isn't renewed", what with him being the Eleventh and all. So, end of a season, maybe the next one.
And it's also been suggested what the question is, though the question offered doesn't seem to match the clues. Still, if Dorium has it right, and revealing the Doctor's true name would bring about dire consequences, there's the potential for a knock-knock joke to go horribly wrong.
[When no living creature can speak falsely or fail to answer]
Except that the spaceship itself need not be huge, nor cylindrical; that could just as well be the shadow of the protective field of a much smaller craft. Say, something the size of a police box.
Yeah, it's just the TARDIS, right where the Doctor parked it on April 22. Another fortnight or so, and he'll get bored and go find another adventure, just wait and see.
Or, wait: does Mercury have a magnetic field? That hole in the CME doesn't look quite right for a planetary magnetic field (though I'm not sure of the perspective), but a big black magnetic anomaly (like the one We Weren't Told About a decade ago) lying on the surface might just do the trick.
What the heck are we doing sending sooper-sekret drones to Iran without destruct packages?
(Yeah, a dramatic destruct package would make ground handling that little bit more exciting, but something with just enough energy to slag the sensitive components needn't be too much of a hazard.)
And... the Iranians overrode the control signal and landed the drone? For reals? The command channel isn't, like, encrypted or anything? Not even off-the-shelf TLS?
Or is that thing the Iranians are showing off really a captured drone? Is it the right model? Is the paint job right? (I would have expected countershading, just offhand.) Did they run down to the hobby shop for a life-size plastic model? (I see that some experts are suggesting that it's a parade float.)
Just looked at the fine print on the datasheet for a dinky MOSFET.
Point 1: it's intended for use in consumer products.
Point 2: it's on some sort of Japanese export-control list relating to weapons of mass destruction.
Gosh, I hope we don't end up violating any Japanese export regulations, though I guess transistors aren't covered by the Lacey Act. (But wait: they come on tape. Which is made of paper. Which is a plant product. Oopsie!)
Well, maybe the product needs a warning label:
WARNING: this product is not to be exported to Iran, Syria, or North Korea without a destruct package attached.
Destruct package, warhead, whatever.
(No, the product isn't a missile guidance system... but, should I have occasion to design a guidance system, for a drone or whatever, I'll have to remember to include such a label.)
A couple of hours ago, I realized that my Android mail was not updating.
It talks to the same IMAP server as my webmail and the Thunderbird on my workstation.
And yet: it resolutely refused to display anything newer than 0900. And it was getting to be rather past 1700, and Thunderbird was showing more mail, some of which I had in fact already read and replied to, using one of the more normal clients (i.e., one with a keyboard attached).
I enabled wi-fi, and tried again. The IMAP log showed the client connecting from the inside network. The recent messages were still not there.
I installed K9, and it could see the newer messages. So, my IMAP server wasn't refusing to send newer material to the phone.
When I moved five messages, which had arrived one after the other at 0904, out of INBOX, suddenly droid-o-mail could see the newer messages.
And: Thunderbird had been displaying a timestamp of 0905 for those messages, but when I moved them to the other folder, they started showing a timestamp of 1824, which is about when I moved them.
AND!!! Those five messages were bounces of the five messages I'd sent to various people at a client company on 9 Aug 2010. The bounce messages were from postini.com, which as far as I can tell has nothing to do with the client. The bounce message was:
550 ... Relaying Denied. Please check mail before sending.
I don't, on quick scan, find any responses to those particular messages, so they may in fact have gotten lost. In August. Of last year.