28 January 2020

The Wisdom of the Marginally Intelligent

Stupid is as stupid does.
-- Forrest Gump/1984

Yet another report of the Lemming Reagan Democrats abandoning the Democratic Party for the Trump Party in today's 'Times', proving that Gump was right. Not that he's a real person, of course.

As has been mentioned here before (although not all that recently), the uneducated, sick, poor, white folk in the hollers are in that condition because they insist on voting for rich, white, right wingnuts whose agenda is to keep them uneducated, sick, and poor. Just a tad less so than their black folk neighbors, of course, just enough to instill that measure of superiority.

Once again, here's the balance of power (number of years) in DC from Reagan to now:

House: R 20 D 20
Senate: R 22 D 18
Pres: R 24 D 16
SC: R 40

Why they would think Reagan was on their side, much less The Manchurian President who gave away $1+ trillion to his fellow fat cats, can only be explained one way.
As Gump said: Stupid is as stupid does.

21 January 2020

I Told You So - 21 January 2020

Another bit of reporting on 5G
T-Mobile (TMUS) said last month it achieved a nationwide 5G network because, rather than using high-band spectrum, T-Mobile used mostly lower frequency airwaves to build its network. Those signals cover much wider areas and are better at traveling through walls and trees, but "low-band spectrum" doesn't provide the dramatic benefits we think of when we think of 5G.

For now, T-Mobile's 5G network provides, on average, a 20% increase in download speeds compared to 4G LTE, according to a company spokesperson. That's a stark difference from the 100 times-faster-than-4G speeds on high frequency 5G networks.

Don't hold your breath. And consider this: is the difference between downloading a movie in a few minutes rather than less than a minute on the same scale as having some kind of telephone versus none at all? Progress, such as it is, really is just incremental improvement on amazing tectonic inventions of the late 19th and early 20th century. Moreover, this may well be the first time that 'invention' tries to flummox the laws of Nature.

19 January 2020

Drug Addicted

A brand new bit of news that explains, yet again, where that $2 billion figure comes from. Drug companies that keep throwing good money after bad really, really have to be punished in such a way that doesn't punish the rest of us.
Spectrum Pharmaceuticals Inc. shares fell 47% in premarket trade Thursday after the company said a mid-stage trial of a treatment for non-small cell lung cancer, or NSCLC, missed its main goal. The Henderson, Nevada-based company said the phase 2 trial dubbed Zenith20 evaluating poziotinib in previously treated NSCLC patients was not met in the first group of 115 patients receiving 15 mg a day. "While the response rate of Cohort 1 in this trial was lower than we expected, the positive signals observed for this cohort provide support for the continued clinical evaluation of poziotinib in this patient population with significant unmet medical need," Chief Executive Joe Turgeon said in a statement.
[my emphasis]

Note that history shows that Phase 3 trials always do poorer than preceding trials.

18 January 2020

One Classy Dame - part the second

So, the SCM piece got me musing, again, on the notion of RDBMS with SCM. It would appear to be a perfect fit. But, rather than musing in the dark, I let my fingers do the walking through the Yellow Googles. Turns out that there are real products, within an inch or so past vapourware at least. Here's some bling on Oracle:
Exadata X8M uses Xeon SP CPUs, Optane DIMMs and RoCE (Remote Direct Memory Access across Converged Ethernet) over 100GbitE. RoCE enables Oracle's database to directly access persistent memory, thus bypassing the OS, network, and IO software stack.

RDBMS which do file management under the OS are not new, so Oracle's not breaking ground with that part. It also means they don't have to wait for linux to directly support such; in the sense that SCM isn't just another file.

There remains conflict on how to use SCM: either as a direct data (row in RDBMS terms) store or as an 'intermediate' filesystem surrogate such as this paper.
... replacing hard drives with SCMs often forces either major changes in file systems or suboptimal performance, because the current block-based interface does not deliver enough information to the device to allow it to optimize data management for specific device characteristics such as the out-of-place update.

As you might expect, I'll vote for something called 'object' store, or 'persistent buffer store', etc. The thrust being to eliminate all that 'impedance mismatch' that some coding folks like to throw out at RM advocates. All industrial strength RDBMS know how to do transactions within their buffers; some do the subsequent I/O to disk management themselves while others use the OS's filesystem to handle that. But both are doing the translation from 'row object' to 'file'. Why? Of course, because most (AS/400 et seq. possibly excepted) OS store data as files. It's also worth noting that the original 360, and successors, were not file oriented in the sense of *nix and successors. They're based on CKD protocol, which, if you twist your neck just right, can be viewed as a row store.
It is a self-defining format with each data record represented by a Count Area that identifies the record and provides the number of bytes in an optional Key Area and an optional Data Area. This is in contrast to devices using fixed sector size or a separate format track.

There is at least one book addressing the question. Here's a snip from the Amazon page:
Existing DBMSs are unable to take full advantage of this technology because their internal architectures are predicated on the assumption that memory is volatile. With NVM, many of the components of legacy DBMSs are unnecessary and will degrade the performance of data-intensive applications.

From the authors' earlier paper they hit the nail on the head (and confirm the notion that led my fingers to go walking through the Yellow Googles in the first place):
Consider a transaction that inserts a tuple into a table. A DBMS first records the tuple's contents in the log, and it later propagates the change to the database. With NVM, a DBMS can employ a logging protocol that avoids this unnecessary data duplication. The reason why NVM enables a better logging protocol than WAL is two-fold. The write throughput of NVM is more than an order of magnitude higher than that of an SSD or HDD. Further, the gap between sequential and random write throughput of NVM is smaller than that in SSD and HDD. Hence, a DBMS can flush changes directly to the database in NVM during regular transaction processing [15, 14, 12, 64, 40, 62, 80].
[links active in the cite]

The crux of the matter: the community has based the notion of transaction of 'slow' disk drives and 'fast' memory, with the transaction happening in memory, but only durable when flushed to disk. This boundary layer impacts to such an extent that many/most/all industrial strength RDBMS have offered the choice to do all the I/O under control of the engine, ignoring the OS facility. In the *nix world this is referred to as 'raw I/O'.

Here's another take on the meaning/purpose of SCM
This is good, of course, but it got me wondering whether it followed from requirement A — no rewrite of applications — that the solution B automatically follows — that the likes of Optane persistent memory must reside in I/O space. After all, such persistent memory was created to be directly attached to the processor chips, and be byte addressable just like RAM. Think paradigm busting, outrageously fast commits of data to persistent storage. Said differently, can a processor complex be created with directly attached persistent memory and where the typical use of that system does not require changes to the applications?

And, surprise surprise, this author remembers AS/400!
Another key — and here very applicable — concept basic to the IBM i operating system is that single-level storage. Even decades ago with the System/38, SLS meant that when your application used a secure token as an address to access data, it did not matter whether that data then was first found on disk or in RAM. Even after a system restart — say occurring due to a power failure — you restarted using exactly the same address token the address the same data.

Finally, this author, likely not by intention, stabs MVCC in the gut (fine by me)
See the difference, along with the impact on throughput as a result? The locks, required in any case since time is passing, are held for a minimum of time. The probability of any subsequent transactions seeing these locks decreases significantly. Subsequent transactions don't as often need to wait, and when they do their wait time is far less. In our train metaphor used earlier, a train doesn't even get built anywhere nearly as often. Life is good.

Die MVCC, DIE!!!

15 January 2020


Have you wondered how it is that Australians would continue to elect governments that considered their continued viability on the land just an annoyance? I certainly have. Scott Morrison is nearly as demented as The Manchurian President. There has to be a reason.

Turns out, it's pretty obvious from a single bit of demographic data:
Australia is one of the most urbanised nations in the world, with an oft-quoted figure of 85% of the population living within 50km of the coast of our very large continent, with vast empty areas in the middle.

IOW, so long as it's the bush (literally, the mostly empty quarters) that's burning, the city folk don't give a shit. Since most of the voters do live in the urban metropoli, it's no skin off their noses. This is a case where the tyranny of the majority will, in due time. destroy that majority.

You Go Girl

By now any of dear readers who are even a little interested in hoops knows about the new CBA twixt the WNBA and the players. The top-line reporting is that average contract value will go up by a bunch. Some reporting said double. Some reporting said that the double was for the stars. What surprised the hell out of me is that a sports pundit actually gets it. I don't keep a pad and pen around when I watch the TeeVee, generally, so I can't guarantee that I get this exactly right.

Yesterday, Pablo Torre on 'High Noon' did his punditry about the CBA, and took some exception to the reporting. He stated, correctly, that average (mean) salary boost doesn't reflect the real impact. He actually said, "median"!! And he didn't even bother to explain the difference. Now, here's what the wiki says about his background (before I go, I'm going to guess some post-graduate...):
He graduated from Harvard College magna cum laude with a degree in sociology in 2007, and was inducted into the Phi Beta Kappa Society.

OK, not a Ph.D. or J.D.. But I'm going to guess at least one class in social science stats. My first wife's mother was a Phi Bet at Middlebury. I was at dinner at their house one evening when it was revealed; the kids were all adults by that point and never knew growing up. I never did find out why it was such a secret.

10 January 2020

Boeing Boeing - part the twelfth

Well, well. May haps this new release points to at least some at Boeing possibly asserting that those engines need to be under the wing. Ya think?
Some assessments of the 737 Max's design were particularly crass. Messages that were emailed in April 2017, a month after the first version of the plane was certified, show one employee described the airplane as "designed by clowns, who in turn are supervised by monkeys." Another comment added: "piss poor design."

These messages refer to the design, not the MCAS software. I would expect that aeroengineers are concerned with the physical characteristics of the aircraft. And, may haps, we'll find someone at Boeing stating what is now obvious: put those damned engines where they belong. Finding such confirmation must lead to the Worst Case Scenario.

06 January 2020


Having grown up in western Massachusetts, sports affiliation was split between Boston teams and New York teams. Not least because the Boston crowd treated the population west of 128 as tax revenue, otherwise left to their own devices. And the 'big city', Springfield, was always jealous of the more prosperous Hartford some miles to the south. In the football world, before the Boston Patriots, we had the New York Football Giants. The Patriots were inept for the first couple of decades of their existence, and if Belichick hadn't lucked out finding a 6'5" quarterback available in the 6th round (199 overall), they would still be. Well, since yesterday, they are again. Cheetahs always cheat. They deserve to fail.

Just an aside. During one of the NFL studio shows over the weekend, one of the commentators said that Pete Carroll is the oldest head coach. I was stunned. Compared to Cheatin' Bill, he looks like a spring chicken. Bill looks like what comes out of the back end of a hippo. I guess clean living has its reward; well, if one ignores the unpleasantness at USC, of course.

Now what? Brady's skill level no longer exists, he's said to have declined to give the Patriots a 'home town discount', and who wants a decrepit 43 year old QB? Some of the pundits on-air and in-print have attributed the team's so-so play to a lack of stellar wide outs, but who's fault is that? Who's fault is it that there's no Brady-in-waiting? Jimmy G's still playing, as I recall.

If a team does have a couple of young, stellar wide outs, why would they opt for a decrepit 43 year old QB? What's the point? Yes, some of the pundits cite Peyton Manning/Broncos as precedent, but that's materially different. Manning was coming off injury, wasn't near 43, and didn't actually propel the Broncos over the season; he was that damning-with-faint-praise 'game manager'.

What passes for 'analytics' in sports, isn't stats or data science; it's merely arithmetic, calculating population parameters. All the observations are recorded, and the parameters figured out. There's no inference to be had. If a QB can't make passes over 10 yards for a full season, then that's all you know. May be, if he gets 'faster' (a la Tom Terrific) receivers, he can then make deep balls. May be not; it may be that his arm is just shot.

02 January 2020

Release the Hounds!!

Last night, during one of the myriad bowl games, I saw the T-Mobile advert touting that it's 5G covered miles and not the blocks of rival telecoms. Woo hoo!!! Of course, they don't say they do this amazing engineering simply by running a 'sub-6' 5G. Verizon, et al, will have counter-adverts real soon now. Of course, the only place that mmWave 5G will be effective is in arenas and office buildings with Godzilla wifi fed by a building top antenna. Whoever gets incrementally better than 4G/LTE will win. It's always been so. Ya cannough fool Mother Nature.

01 January 2020

One Classy Dame

OK, the title was, likely, used by a couple of hundred bad guys in 20s, 30s, and 40s gansta films. Today's topic is storage class memory, yet again, wafted on the raft of an AnandTech piece, in which ex-Toshiba seeks to denigrate SCM, in favor of its fashion of NAND. Bah. From the point of view of 5NF RDBMS, nothing is better than an SCM datastore. As Codd decreed, all data changes at once; there is no row-by-agonizing-row nonsense out to disk.
The ability to offer both high-density DRAM at order-of-magnitude performance levels or extremely low latency storage in a single product indicates its utility, rather than a lack of optimization for one or the other.

On the whole, the comments make mincemeat of the argument.