Youtube comments of Lawrence D’Oliveiro (@lawrencedoliveiro9104).
-
1700
-
1700
-
1300
-
1300
-
1000
-
456
-
407
-
357
-
352
-
254
-
253
-
238
-
201
-
198
-
192
-
Stanislaw Lem once wrote a story about an inventor who created a machine that could make anything beginning with the letter N. He told it to make nimbuses, noodles, nuclei, neutrons, naphtha, noses, nymphs and naiads, and it did them all. Then he asked it to make natrium.
“Never heard of it,” said the machine.
“What? But it's only sodium. You know, the metal, the element...”
“Sodium starts with an S, and I work only in N.”
“But in Latin it's natrium.”
“Look, old boy,” said the machine, “if I could do everything starting with N in every possible
language, I'd be a Machine That Could Do Everything in the Whole Alphabet, since any item
you care to mention undoubtedly starts with N in one foreign language or another. It's not
that easy. I can't go beyond what you programmed. So no sodium.”
169
-
167
-
167
-
158
-
156
-
147
-
145
-
140
-
139
-
139
-
130
-
130
-
128
-
128
-
121
-
119
-
115
-
113
-
111
-
106
-
106
-
105
-
101
-
99
-
95
-
93
-
91
-
90
-
90
-
90
-
84
-
81
-
81
-
81
-
81
-
80
-
79
-
79
-
78
-
78
-
76
-
75
-
74
-
74
-
74
-
73
-
72
-
71
-
70
-
69
-
68
-
65
-
65
-
64
-
63
-
62
-
62
-
62
-
61
-
61
-
61
-
59
-
59
-
59
-
56
-
56
-
55
-
54
-
53
-
52
-
51
-
49
-
49
-
48
-
48
-
48
-
47
-
46
-
46
-
46
-
46
-
46
-
46
-
46
-
46
-
45
-
45
-
45
-
45
-
44
-
44
-
44
-
44
-
44
-
43
-
43
-
43
-
42
-
42
-
42
-
41
-
41
-
40
-
40
-
40
-
40
-
39
-
39
-
39
-
38
-
38
-
38
-
38
-
38
-
38
-
38
-
38
-
37
-
37
-
37
-
37
-
37
-
36
-
36
-
36
-
36
-
36
-
35
-
35
-
35
-
35
-
35
-
34
-
34
-
34
-
34
-
34
-
34
-
33
-
33
-
33
-
33
-
33
-
33
-
33
-
32
-
32
-
32
-
32
-
32
-
32
-
32
-
31
-
31
-
31
-
31
-
30
-
30
-
30
-
29
-
29
-
29
-
29
-
29
-
29
-
29
-
29
-
28
-
28
-
28
-
28
-
28
-
28
-
28
-
28
-
28
-
28
-
28
-
28
-
27
-
27
-
27
-
27
-
27
-
27
-
27
-
27
-
A couple of significant notational differences between C and Pascal are, when declaring variables, C inherits the Algol syntax of type-spec before variable name: “«type-spec» «var-name»” whereas Pascal invented its own syntax: “«var-name» : «type-spec»”. And also pointer dereferencing: C uses a prefix operator “*«pointer-val»” whereas Pascal has a postfix one: “«pointer-val»^”.
This may seem very tomayto-tomahto, but consider how C++, which inherits C syntax lock stock and barrel, is now trying to deal with newer features involving dependencies between types, for example between the arguments to a function and its result type. It now finds that, in addition to the C syntax of putting the return type before the function name, it also has to add an alternative syntax where the return type comes after the argument list, à la Pascal.
As for the pointer issue, this may not occur very often, but back in my Mac programming days, more than once I had to write an expression like “thePort^^.theGDevice^^.gdPMap^^.pmTable”,
which in C would have had to become “(**(**(**thePort).theGDevice).gdPMap).pmTable”. (These were structures defined for interfacing to the Mac system, not created by me.) Which would you prefer?
27
-
27
-
26
-
26
-
26
-
26
-
26
-
26
-
26
-
26
-
26
-
3:49 FFmpeg is actually a suite of tools. Besides the “ffmpeg” executable, there is also “ffprobe” for analyzing the structure of media files, and “ffplay”, which is a basic command-line-based media player. Plus there are a bunch of underlying libraries (libavfilter, libavcodec, libavformat, libswscale etc), that apps can link directly against.
Myself, I have written a few programs that made use of FFmpeg. In every case so far, it was easier to spawn the standard FFmpeg executables and feed them suitable command lines, than to try to make use of the libraries myself.
26
-
26
-
26
-
26
-
25
-
25
-
25
-
25
-
25
-
25
-
25
-
25
-
25
-
24
-
24
-
24
-
24
-
24
-
24
-
24
-
24
-
24
-
23
-
23
-
23
-
23
-
23
-
23
-
23
-
23
-
23
-
23
-
The three basic objects in Git are blobs, trees and commits. A blob represents the contents of a file, and is identified by the SHA-256 hash of its contents. A tree associates names with blobs, and represents a snapshot of source tree contents at a point in time. A commit holds a pointer to a tree, and to zero or more parent commits, as well as information about who did the commit and when, an explanatory comment etc.
Trees and commits are also represented by SHA-256 hashes of their contents. This is why a commit holds pointers to its parents, not to its children; because once the commit has been created, it cannot be changed (e.g. to add child pointers) without losing its original SHA-256 ID.
Branches and tags are just pointers to particular commits. They have no associated SHA-256 hashes, so they can be added, removed and changed at any time.
23
-
23
-
23
-
23
-
23
-
23
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
22
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
21
-
20
-
20
-
20
-
20
-
20
-
20
-
20
-
20
-
20
-
20
-
20
-
20
-
20
-
20
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
19
-
18
-
18
-
18
-
18
-
18
-
18
-
18
-
18
-
18
-
18
-
18
-
18
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
17
-
16
-
16
-
16
-
16
-
16
-
16
-
16
-
16
-
16
-
16
-
16
-
16
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
15
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
14
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
Back in the 1990s, there were these things called “Unix workstations”. What made them “workstations” rather than “desktops”, besides the higher-end hardware, was that they included both client and server functionality together in one box.
Somehow, Microsoft managed to persuade users that they didn’t need such a wide range of capability. Windows NT “Workstation” had its server functionality deliberately crippled (and being a bit cheaper also helped), and yet it still managed to wipe out the Unix workstations. So now what users had on their desktops was just a “desktop” OS. If they wanted server-style functionality, they would have to pay extra for a separate product: Windows Server.
And not just that, but they would have to pay even more on top of that, for “CALs”, one for each user of a server function. So really the “desktop” segment could be seen as a creation of Microsoft’s marketing department, as an excuse for nickel-and-diming its customers to death. And somehow that has come to be accepted as the way things should be.
Linux, on the other hand, is essentially the last standing “workstation” OS. It doesn’t play by the “desktop” rules. It offers full server-side functionality in the same box, with no “CALs” limits. Think of it as Usain Bolt, and the “desktop” market as a three-legged race; would you really complain that an Olympic-class runner would do so badly in such a pointless and trivial competition?
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
13
-
One of the key things I learned from my CS courses is abstraction and layering. You start with the bare hardware, which is the most general machine possible because it hasn’t been specialized in any way. You put down the first layer of programming, which creates a new “abstract machine” that is better suited to some purpose you have in mind, but is likely less suited to something that someone else may have in mind. Then another layer on top of that, and so on.
And every stage, each layer is replaceable and can be substituted with something else. For example, you get to the layer of the command-line shell, which can do a range of things. You can customize it by writing a shell script, which performs some function that is useful to you.
But once you get to the GUI layer, that’s the end of all your abstractions. That’s because the GUI layer cannot be automated. So you cannot build any more “abstract machines” on top of it. It is designed to be directly controlled by humans, and that’s it.
13
-
13
-
13
-
13
-
13
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
12
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
11
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
The “intuitive” way to think about rotations in 3D might be to express them in terms of latitude and longitude on the surface of a sphere. Starting from long 0° and lat 0°, you might move x° east or west, and y° north or south. This works OK starting from (0°, 0°), but it doesn’t work so well when you start from, say, the North Pole. Because every direction from there is south, how do you define how far east or west you want to go, and from what point?
This phenomenon is called “gimbal lock”. It is a phenomenon well known in computer graphics if you try to specify rotations in terms of “Euler angles” (as the longitude/latitude angles are known). It was also a known hazard during the Apollo moon landings, as it imposed limits on how their spacecraft could maneouvre without confusing the orientation gyroscopes.
Quaternions are really nice because they avoid gimbal lock. Related to this, they also interpolate nicely between any two orientations, which is handy for animation purposes in CG.
Trouble is, many people working in CG are afraid of quaternions, because “math is hard”. This includes some developers of CG software.
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
10
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
9
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
8
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
“On top of the operating system is the rest of the system software. Here we find the command interpreter (shell), window systems, compilers, editors, and similar application-independent programs. It is important to realize that these programs are definitely not part of the operating system, even though they are typically supplied preinstalled by the computer manufacturer, or in a package with the operating system if it is installed after purchase. This is a crucial, but subtle, point. The operating system is (usually) that portion of the software that runs in kernel mode or supervisor mode. It is protected from user tampering by the hardware ...”
-- Tanenbaum & Woodhull, “Operating Systems Design And Implementation”, third edition, page 3
That is the well-known “MINIX Book”, by the way. So you see, by that definition, Linux is very definitely an ”operating system”.
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
7
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
There was a saying in the old days, “nobody ever got fired for buying IBM”. The obvious update for today is “nobody got fired for buying Microsoft”. But that doesn”t really sound convincing enough. You need something that sounds more reassuring, don’t you? Something that makes it clear how much worse the alternatives can be.
How about, “nobody ever got tied to an anthill and smeared with honey for buying Microsoft”.
Or, “nobody had their skin flayed off, one little cut at a time, with a blunt razor, for buying Microsoft”.
Or, “nobody ever got tied in a tank with a tap dripping on their head, driving them insane before it drowned them, for buying Microsoft”.
Do you think that makes Microsoft buyers feel better?
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
6
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
But ... but ... but ... the Second Amendment is sacred, that’s why you have to trample on the First Amendment instead, and restrict people’s rights to play video games.
As I’ve said before, the Second Amendment is a wonderful decoy that the Government can use anytime it wants to take away any of your important rights: just make a feint towards that, and while everyone is worked up and distracted and shouting “from my cold, dead hands!”, pass a law to restrict some other freedom instead. All this nonsense of being able to use guns to protect yourself against Government oppression is a complete lie: when your police has access to military-grade weaponry, and all the citizenry has is peashooters by comparison.
Yes, you can still do grievous harm to each other with those peashooters, but point them at a representative of the Government, and you will discover what real harm can be like.
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5:17 Actually, Stallman was not keen on Linux at all. He was working on his own “Hurd” kernel, which was going to be have a “microkernel” architecture, which would supposedly make it much easier to develop, test and maintain. Better than an old-fashioned “monolithic” kernel, like Linux.
Well, here we are, about 30 years later, and we’re still waiting for Hurd to reach production quality. Meanwhile, Linux started becoming popular just a couple of years after that initial announcement, and spread to a whole bunch of non-x86 architectures at the same time.
5
-
5
-
5
-
5
-
5
-
5
-
10:35 One important difference, which still remains true to this day, is that on Unix and Linux systems, process creation is carefully designed to be a very low-overhead operation. Thus, a shell script can spawn any number of processes to implement the stages of a pipeline, run concurrent background tasks etc, and do so quickly and easily. Whereas on other common systems, including Windows, process creation remains expensive and complicated. Thus, when software originally developed on Linux is ported to Windows, it tends to work less well and less reliably for this among other reasons.
Even the Mac, supposedly a Unix system, has had performance issues with multitasking.
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
5
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
13:01 Probably a good time to mention another useful feature of tab completion, namely if you type a wildcard spec and then press tab once or twice, Bash will show you all the expansions of that wildcard spec, so you can confirm that your command will be acting on what you thought it would be acting on.
Another useful trick is to first use the wildcard with something harmless like an “ls” command, to confirm that it only shows the files you want to affect. Then press uparrow to bring the command back, change the “ls” to “rm”, and turn it loose.
Because, you know, having suffered the occasional embarrassment of inadvertent deletion or overwriting of files I didn’t mean to over the years, you learn a few tricks for, shall we say, “defensive command-lining” ...
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
8:57 That so-called “Unix philosophy of do one thing and do it well” is a half-truth. Look at any Unix/Linux system, and you will a mix of large pieces and small pieces. And the small pieces would not work without the help of the large pieces. Consider:
* The Linux kernel, currently at over 20 million source lines and counting. Which “one thing” does it do well? It is successful because it does so many things well.
* The X server, which was notorious in earlier times for practically being an OS unto itself, with its own drivers reaching deep into the hardware, running with root privileges. Which “one thing” does it do well? (To be fair, it should eventually be replaced with Wayland.)
* Good old Bash. I counted over ¾ million lines in its source code. Its man page is over 5000 lines long. I use it every day, and I know lots of other people do, too. What “one thing” does it do well? It’s basically a large monolithic mish-mash of miscellaneous functionality, accreted over the years. But at least it is useful miscellaneous functionality.
I could go on. Big pieces make little pieces possible. systemd makes possible the writing of very short config files to manage custom-developed services—unlike the dozens or hundreds of lines of boilerplate commonly needed with sysvinit. So your code can concentrate on “doing one thing, and doing it well”, thanks to systemd!
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
There are 3 well-known ways of representing signed integers using binary notation: sign-magnitude, ones-complement, and twos-complement.
Sign-magnitude is probably the first one that anybody would think of: reserve one bit for the sign, and use the rest of the bits to represent the magnitude.
Let’s say we have 5-bit integers, for simplicity. After reserving one bit for the sign, we have 4 bits left for the magnitude. If we put the sign bit at the left and use 0 for positive and 1 for negative, then +3 could be represented as
00011
while -3 would be
10011
Addition and subtraction requires complicated logic to extract the sign and magnitude, and do different things with the magnitudes depending on whether the signs are same or different, then deciding what sign to attach to the result.
In ones-complement, +3 is represented the same as before, but instead of negating by flipping only the sign bit, we flip all the bits. So -3 becomes
11100
Note that complementing applies to zero as well. Thus, both
00000
and
11111
represent zero. The latter can be considered to be “negative zero”, since its sign bit is 1. When we add +3 and -3:
00011
+11100
=11111
which is indeed one of the representations of zero.
Twos-complement gets rid of the redundant representation of zero, by redefining negation to be done by flipping all the bits, then adding 1. Thus, -3 is now represented as
11101
So adding +3 and -3:
00011
+11101
=100000
And throwing away the 1 that has gone into the 6th bit (because we can only represent 5 bits) gives zero, as expected.
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
4
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
6:16 But then, if you are always filling rectangular areas rather than plotting individual points, then you will always be specifying the bounds of the areas to be filled. Thus, filling the whole playing field would mean filling a rectangle with corners (0, 0), and (width, height), not (0, 0) and (width - 1, height - 1).
In other words, stop thinking of pixels as little squares. You want squares, think in terms of squares, not pixels.
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
There are several problems with the living-in-a-simulation idea. Here is one:
* Mathematicians seem to have become afraid of self-reference. Because whenever a mathematical theory is capable of self-reference, it tends to introduce paradoxes (e.g. Russell’s Paradox). So they prefer it when each theory is kept in its own watertight compartment, limited to its own particular set of axioms, because connections between theories can be a back-door for reintroducing self-reference.
* However, when scientists study reality, they find it it does not divide itself up into compartments. Everything that is real is part of an interconnected reality. And connections between what started out as separate scientific theories happen all the time.
* Therefore, it follows from the first two propositions that no single mathematical theory is capable of modelling all of reality.
* But if our “reality” is indeed such a simulation, then it must be built on some such mathematical theory (as embodied in a computer program, if you want to be specific about it). And that theory must be able to model our entire “reality/simulation”, without introducing any paradoxes.
* Given that no such mathematical theory can exist in our “reality/simulation”, it follows that what we live in must be reality itself, not a simulation.
How’s that?
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
The way I see it, most of those predictions from, say, the 1950s/1960s or so, were predicated on having large mounts of power available--personal jet packs, holidays on the Moon and so on. Where was this power going to come from? Obviously, nuclear energy. But we took a wrong turn on that, concentrating on nuclear reactors that could produce material for nuclear bombs and required complex safety systems, instead of investing in designs (e.g. thorium-based) that were primarily peaceful and inherently safer. And so, the promised power that was “too cheap to meter” never eventuated.
What we got instead was improvements in technologies for handling information--which was something that hardly anybody predicted (except perhaps Vannevar Bush). Hence the whole range of computer-centric technologies, with lots of processing power, large amounts of storage, and also fast transmission of information around the world.
These did not take much power to run. But as the computers get more and more powerful, their power requirements are also increasing. Already, the single largest expense in running a data centre is not the high-tech hardware, but the electricity bill. And so we are approaching the power limit that we previously dodged, and we will have to deal with it at some point.
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
Here’s the kind of stuff I have written, from my Python wrapper qahirah_xcffib, for keeping track of initialized X11 extensions:
self._ext_inited = dict \
(
(key, {"query" : def_query_version(key, args), "done" : False})
for key, args in
(
(xfixes.key, (xfixes.MAJOR_VERSION, xfixes.MINOR_VERSION)),
(xrender.key, (xrender.MAJOR_VERSION, xrender.MINOR_VERSION)),
(xshape.key, ()),
)
)
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
6:26 Not just an odd number, but a prime number: 7. One that cannot be evenly divided up into anything else!
Before byte addressability became de rigueur , computers had word lengths like 24 bits, 36 bits, 60 bits -- all numbers with lots of integer divisors including powers of 2 and 3, even 5, so they could be divided up into equal-sized portions in many different ways. But with byte addressability, the basic unit is 8 bits, and the natural machine word length (so far) is 2, 4 or 8 times this. So the only factors you have are powers of 2, which limits the ways you can divide them up.
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
3
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
Microsoft used to be very different. I have a book called The Microsoft Way , written in 1996 by Randall E Stross, who was a trained historian. Somehow Microsoft gave him free reign to wander about the company with a “No Escort Required” ID badge, and talk to whomever he wanted, and then publish the results without getting their OK on anything he wrote.
Back then, the company was big, but it still had the hunger of a small company. Stross describes how the entire Windows code base would be rebuilt every night. And if the build aborted due to some bug, the developer who caused that bug would fix it pronto, before the next night’s build.
The company is not like that today. Nowadays, it can take a year or more for a new feature request to bubble up from the developer who proposes it, to the higher levels of management to approve it, and then back down to the build team to include the code.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
17:55 MultiFinder was very useful, but it was such a hack. Windows belonging to different applications were kept in separate “layers”, and only those belonging to the foreground application could be moved/resized/interacted with. The ones in the background were effectively just painted onto the desktop, not real windows at all.
Interestingly, this “layer” concept seemed to carry over to some extent to OS X, which you’d think, being a true multitasking OS, wouldn’t be bound by such obsolete restrictions. While OS X let you interleave windows belonging to different applications, it had this weird quirk where, when you closed the frontmost window, the next window to come to the front would come from the same application (if there was such a window), in preference to one from another application, even if the latter was next in stacking order.
I think this limitation still exists in OS X/macOS today.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
Fred Brooks, in his classic The Mythical Man Month from a few decades ago, made the point that the only way forward to significantly improve programmer productivity was to move to higher- and higher-level languages, which abstracted away more and more of the lower-level details -- what he called “metaprogramming”. His example, AppleScript, unfortunately has not stood the test of time, but others, like Perl, PHP and Python, certainly prove the point. Look at the huge accumulation of toolkits around these languages, and how much work they save you.
Python is easily the best example, simply because its core language is so small, yet so versatile, that the kinds of things you can do with library code is just amazing.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
Science is about confronting the tough questions, not shying away from them.
If the Earth is flat, are the other planets flat, too? If so, why do they look round?* If not, why is the Earth special?
Remember, Galileo was the first human to see other planets (apart from the Moon) as round disks through his telescope, as opposed to dots* with the naked eye.
** Though I think Venus, when it gets close enough to us, gives a hint of being something more than a dot, in its crescent phases.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
Evolution is a real phenomenon: otherwise, why are human bones and dinosaur bones never found together?
Quantum physics is real: otherwise, why does the double-split experiment behave the way it does? If those aren’t “particles” leaving their traces in the detectors, then what are they?
Our theories about these real physical phenomena may vary over time. That is because we keep uncovering new physical phenomena. For example, if we ever discover rabbit bones in Precambrian strata, that would (quite literally) turn evolution on its head. But our current predictions say that will never happen.
Scientists are never satisfied with existing theories, though. They keep trying to find new data, just to see if that agrees with the theories, or not. That’s the difference between science, which is the study of reality, and ideology, which insists that reality conform to some preconceived ideas.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
I notice you only talked about function decorators, not class decorators. They have their uses, too.
I think the most extensive use I have made of custom decorators so far has been in my DBussy package, a pure-Python wrapper for libdbus. The information defining the D-Bus interface to a Python class is specified via a function decorator on each Python method implementing a D-Bus interface method, with a class decorator to tie it all together. Then when you register your class to receive messages, the dispatch mechanism knows how to decode incoming D-Bus method calls, convert the arguments and dispatch to the appropriate Python method.
As a bonus, this same information is used to automatically generate the D-Bus XML introspection format, when another D-Bus peer sends you the introspection message. And conversely, when you introspect another D-Bus peer, this info is used to dynamically generate a “proxy interface class” -- a Python class whose methods simply send the appropriate D-Bus messages and return the responses.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
3:00 Adobe Illustrator was breathtaking when I first saw it in 1988. And the demo disk had Bach’s Sixth Brandenburg on it, which was also pretty cool.
However, techno-geek that I am, I soon discovered that it would not let you easily change the type of a control point between a smooth and corner join. When Aldus FreeHand came out soon after, it did have that feature, so I preferred it.
Then, sometime in the 1990s, Adobe acquired Aldus, and Macromedia along the way as well. It kept PageMaker and Flash, and possibly also Fontographer, but it got rid of FreeHand.
These days, I use Inkscape. Which has all these features, and more. And interoperates nicely with other tools, like GIMP and Blender.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
Back in the day when Windows Vista was about to come out, I bought a little Asus Eee 701 PC. You may remember that brief era of the “Linux netbook” -- this was one of the better-known ones. Small (7” screen), cheap, low-powered (900MHz single-core Celeron), little (4GiB) SSD instead of a hard drive. KDE 4 was still in its early days at that time, but I booted it up off an SD card, and enjoyed the full 3D effects, transparency, rotating desktops, wobbly windows and all the rest of it, running nicely in real time.
If you remember, the “Aero Glass” effects in Vista would not run on anything but the highest end machines at the time. In fact, this limitation led to the infamous lawsuits over the meaning of “Vista capable” versus “Vista ready”, and whether consumers were misled into thinking they were getting capabilities that they were not.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
Actually printf is an interesting case, because it shows up the drawbacks of trying to do everything as a library routine within the limited type system offered by C. The trouble is that printf needs to be polymorphic -- accept arguments of any mixture of different types -- which C really cannot handle. So it simply gives up and doesn’t bother checking the argument types.
There is another approach, which is illustrated by the memory-allocation API in Modula-2. The language defines built-in “NEW” and “DISPOSE” routines, which are fully type-safe, because they are constructs known to the compiler. But the language itself doesn’t define any memory-allocation routines. Instead, these constructs are translated to lower-level calls to routines named “ALLOCATE” and “FREE”, which are type-unsafe, and not defined anywhere in the language. Instead, they have to be provided somewhere in the program, or imported from some library. This way you get the best of both worlds -- type-checking at the high level, but full versatility of implementation at the low level.
Imagine if something similar could be done on a more general scale, for I/O and possibly for other things. C++ uses a similar idea for its memory-allocation mechanism, but unfortunately its “<<” and “>>” I/O operators are a bit rubbish. And its printf support is just as unsafe as in C.
GCC tries to do some type-checking of printf format strings. But this cannot work if the string is dynamic, since that requires the argument type information to be available at run-time.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
7:28 The wobble is not about speed, it’s about tracking. In order to ensure the laser is correctly focused on the groove, you can do a comparison of focus error on opposite sides of the sensor; if one is further out, you know which way you have to move the focusing mechanism to bring things back into alignment again. The problem is, the comparison system could develop an offset error over time, from mechanical misalignment, contamination or whatever other cause. To work around this, instead of a steady tracking, you do an oscillating system instead. So now instead of measuring an error signal, you measure the wobble in the readout signal itself, which is more resistant to static offset error.
Sony invented a clever system in their MD disc where, instead of the mechanism implementing the wobble, it was built into the shape of the grooves instead. This is the same wobbly-groove idea you are seeing here.
(Source: The Art Of Digital Audio , 2nd ed, by John Watkinson, page 577 or thereabouts.)
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
It’s no coincidence that the sound is so similar to that made by telephone modems, since they are both trying to solve nearly identical problems, namely how to use a medium designed for audio to transfer digital data.
The obvious differences being:
* Modems can ask for a retransmit when they hit an error, whereas the cassette player cannot be stopped and rewound to replay a section under computer control (unlike the tape drives on the bigger computers of the time, for example).
* But on the other hand, even cheap audiocassette machines would be capable of better frequency response than the 4kHz standard bandwidth for telephone lines.
Another thing is, modem coding technology continued to progress a lot after the days of using audiocassettes for data, reaching a peak of a 33.6kb/s transfer rate in either direction before being eclipsed by broadband technologies like ADSL. If similar trellis-coding techniques were retrofitted onto cassette encoding, it should be possible to load a program of, say, 128kB size in a matter of seconds rather than minutes. Or would that take more processing power than an authentic 8-bit machine could manage?
And another thing to look at might be the incorporation of robust error-correction codes, like that on CDs and DVDs. But again, maybe this would require too much additional hardware or processing power to fit within the authentic 8-bit experience ...
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
If you want to go back to the First Gulf War to justify invading Iraq, remember that did not arise in a vacuum. Before then, Saddam was seen to be on the side of the West, as a bulwark against Iran. He attacked Iran in a proxy war on behalf of the West, which Iran survived, but which destroyed the economy of Iraq. So then he figured he would take over Kuwait to use its resources to shore himself up.
Before that First Gulf War, Saddam massacred thousands of his own Kurdish population with poison gas in Halabja, yet nobody in positions of power in the US or anywhere else in the West described him as a “brutal dictator” for that. It was, you know, a minor peccadillo of an important ally, or something like that.
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
What about the block-level interface?
Conventional hard drives present an interface where the OS can reliably read and write numbered blocks in an arbitrary order, and build a conventional disk-based filesystem on top of that.
Flash storage, on the other hand needs wear-levelling to try to reduce the chance of blocks going bad, and also prefers to do writes in larger chunks. SSD controllers go to a lot of trouble to hide these peculiarities, and present an interface to the OS that looks just like a disk.
However, some OSes (like Linux) have purpose-built filesystems that are designed to deal with the characteristics of flash storage, including explicitly taking wear-levelling into account. That means they don’t need the SSD controller providing that disk-emulation layer, which just adds overhead and gets in the way.
Are there “bare bones” flash drives, if you like, that do away with the disk emulation, and offer a simpler, more direct controller interface? Would they offer better performance when used with such filesystems?
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
2
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
I had a look at the source code. Counting up the number of lines in files in subdirectories called “lib” and “src” for both projects, curl comes in at about 190,000 lines, while wget is only slightly smaller at about 180,000 lines.
So while wget handles fewer protocols, it has more functions for working with them. Like in order to do recursive mirroring of a website, it has to do some HTML parsing.
I would say, all the other protocols that curl implements, that wget does not, are unnecessary. For example, media streaming: what is the point, when ffmpeg is purpose-built not only for handling the connections, but also the necessary media packaging? Why bother with SCP/SFTP, when SSH clients already include that? Or SMB, when you have smbclient? And Telnet, of all things--what is the point with that?
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
Windows Phone was very much emphasizing form over function. For example, the idea of making those on-screen “tiles” display actual content, rather than just being static icons, seems like a good one, until you see how they implemented it. I remember during a demo by a Nokia exec, you would see things like the email icon dancing around in a distracting way, even when there were no new messages to display. This kind of useless “noise” in the UI is just going to put people off.
And there were too many limitations in the actual OS. For example, Windows Phone 7 didn’t allow apps to access the actual file system, whereas you could do this on Android. Windows Phone 8 I think relaxed this restriction, but then the app build process was done in a strange way which involved accessing a Microsoft server, so you couldn’t actually do builds just on your own facilities.
There were hardware limitations, too. Windows Phone 7 could only run on a single core, while Windows Phone 8 required dual-core hardware. This when the newest versions of Android were available on devices with anything from one to four cores.
Also I remember HTC (the king of the Android heap at the time) brought out it’s “One” series of models, in both Android and Windows Phone versions. The Android versions had hardware features beyond those available in Windows Phone. It’s clear the only reason HTC was bothering with Windows Phone was because Microsoft was paying it to do so.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
3:20 Near as I can tell, Kildall’s PL/M compiler was always a cross-compiler: it never ran native on any 8-bit microprocessor. To use it, you had to have access to a big (and expensive) DEC PDP-10 machine.
I think this ran the TOPS-10 OS, which was a pretty typical DEC OS. A filespec consisted of a device name, a directory part, a name part, an extension and a version number. Clearly this was the inspiration for CP/M filespec syntax: the device name (which could be multiple characters) was simplified down to a single character, the directory part and version number were omitted, the name limited to 8 characters, and the extension to 3.
And aspects of that syntax were carried over to MS-DOS, and still persist in Windows to this day. Particularly the single-character drive names, which seem pretty ridiculous on modern hardware.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
Remember why long shots, particularly in spaceship propulsion, are worth pursuing: it’s because space travel with rockets (as at present) is complicated, expensive, and dangerous. That was true half a century ago, it’s still true today. So if some idea, no matter how wacky, promises some short cut to the process, it’s worth at least looking at, because the cost of dismissing it out of hand, should it turn to be right, is so high.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
5:47 It’s worth clarifying what the whole point of this “parent/child” relationship, and consequent possibility of “orphan” processes, is about: when a process terminates, it becomes a “zombie”, which means it still takes up a process slot, even though no process is running any more. At this point, it is up to its parent to pick up a notification of the termination of the child, via a call like wait(2) or similar; once it gets this notification, then the zombie process disappears, and that process slot is freed up to hold a new created process.
So, if you write a program on Linux or other POSIX-type system, and you spawn lots of processes, then you have to remember this rule to avoid your system running out of process slots: always gobble your zombie children.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
With Python, it’s a very powerful combination. Here’s the sort of dynamic query criterion I have constructed:
list
(
"%(name)s like %(value)s"
%
{
"name" ∶ field[0],
"value" ∶
sqlstring("%" + escape_sql_wild(params․getvalue(field[1])) + "%"),
}
for field in
(
("make", "search_make"),
("model", "search_model"),
("details", "search_details"),
("serial_nr", "search_serial"),
("inventory_nr", "search_invent"),
("requisition_nr", "search_requis"),
("machine_name", "search_machine_name"),
("os_updates", "search_os"),
)
if params․getvalue(field[1]) != ""
)
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
“On top of the operating system is the rest of the system software. Here we find the command interpreter (shell), window systems, compilers, editors, and similar application-independent programs. It is important to realize that these programs are definitely not part of the operating system, even though they are typically supplied preinstalled by the computer manufacturer, or in a package with the operating system if it is installed after purchase. This is a crucial, but subtle, point. The operating system is (usually) that portion of the software that runs in kernel mode or supervisor mode. It is protected from user tampering by the hardware ...”
-- Tanenbaum & Woodhull, “Operating Systems Design And Implementation”, third edition, page 3
That is the well-known “MINIX Book”, by the way. So you see, by that definition, Linux is very definitely an ”operating system”.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
@zacharythatcher7328 Threads are not an alternative to callbacks. Threaded event loops (in fact, threaded everything) were all the rage in the 1990s, until it was realized that they made the code very hard to debug. That’s why we now stick to single-threaded event loops, select(2)/poll(2)/epoll(2) etc. But then, time-consuming tasks required the use of callbacks to maximize responsiveness, until the old technique of coroutines was revived more recently as a way to make such code more readable. This has required language changes, but a lot of languages (Python included) have adopted those changes.
What Arjan is using threads for here, and what I have done with them myself, is to turn blocking calls into non-blocking ones. That’s all. Luckily, in POSIX, “blocking” normally means “thread-blocking”, not “process-blocking”. That’s why we can get away with it. Again, it’s all in the cause of maximizing responsiveness, without getting back into the hell of full-on multithreading.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
Yup, that’s the right way to think of it. Say you have a function which evaluates a cubic Bézier curve through given control points p₀, p₁, p₂, p₃ at a point given by a parameter t. Instead of expressing it all as a single function
p ← B(p₀, p₁, p₂, p₃, t)
you split the function call into two stages
p ← B(p₀, p₁, p₂, p₃)(t)
so you specialize it once for particular control points
B’ ← B(p₀, p₁, p₂, p₃)
and then the specialized function can be evaluated repeatedly for the specified parameter values
p ← B’(t)
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
“Lorem ipsum” is the traditional text. But how about “Corporate ipsum”, e.g.
Professionally whiteboard market positioning sprints. Enthusiastically embrace long-term high-impact vortals, then proactively harness ubiquitous technologies, to credibly extend disaggregated manufactured products, to uniquely exploit content-free ideas.
Conveniently scale plug-and-play services, then energistically evolve fully tested epicentres of excellence, then enthusiastically evolve premium innovation, then enthusiastically initiate 24/365 clouds, and progressively e-enable high-quality e-commerce.
Appropriately benchmark long-term high-impact infrastructures, to rapidiously impact resource-sucking leadership skills, to continually drive open-source human capital, then fungibly scale innovative strategic theme areas. Holisticly promote backward-compatible portals.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
“On top of the operating system is the rest of the system software. Here we find the command interpreter (shell), window systems, compilers, editors, and similar application-independent programs. It is important to realize that these programs are definitely not part of the operating system, even though they are typically supplied preinstalled by the computer manufacturer, or in a package with the operating system if it is installed after purchase. This is a crucial, but subtle, point. The operating system is (usually) that portion of the software that runs in kernel mode or supervisor mode. It is protected from user tampering by the hardware ...”
-- Tanenbaum & Woodhull, “Operating Systems Design And Implementation”, third edition, page 3
That is the well-known “MINIX Book”, by the way. So you see, by that definition, Linux is very definitely an ”operating system”.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
How about a recursive generator? Here’s one that yields all permutations of a given sequence:
def permute(seq) :
"generator which yields successive permutations of the elements of seq."
if len(seq) == 0 :
yield ()
else :
for i in range(0, len(seq)) :
for rest in permute(seq[:i] + seq[i + 1:]) :
yield (seq[i],) + rest
♯end for
♯end for
♯end if
♯end permute
Example use:
>>> list(permute(("pears", "oranges", "feijoas")))
[('pears', 'oranges', 'feijoas'), ('pears', 'feijoas', 'oranges'), ('oranges', 'pears', 'feijoas'), ('oranges', 'feijoas', 'pears'), ('feijoas', 'pears', 'oranges'), ('feijoas', 'oranges', 'pears')]
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
Remember, Microsoft is not extending Linux with any new capabilities, it is (trying to) extend Windows with more Linux-like capabilities. Also, it tried very hard to “extinguish” Linux many years ago, and failed badly. (For some amusement, go look up Microsoft’s old “Get The Facts” campaign, and what happened with the London Stock Exchange switching to Windows NT, with much great fanfare, and then having to abandon that.)
So fine, what is the next step after “embracing and extending” Windows? My guess is, Microsoft will realize that continuing to maintain the Windows kernel is just getting too difficult and expensive for too little reward, so that will be abandoned at some point in favour of a Windows compatibility layer on top of the Linux kernel.
That would work better than WSL1, which tried to put a Linux compatibility layer on top of the Windows kernel.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
“On top of the operating system is the rest of the system software. Here we find the command interpreter (shell), window systems, compilers, editors, and similar application-independent programs. It is important to realize that these programs are definitely not part of the operating system, even though they are typically supplied preinstalled by the computer manufacturer, or in a package with the operating system if it is installed after purchase. This is a crucial, but subtle, point. The operating system is (usually) that portion of the software that runs in kernel mode or supervisor mode. It is protected from user tampering by the hardware ...”
-- Tanenbaum & Woodhull, “Operating Systems Design And Implementation”, third edition, page 3
That is the well-known “MINIX Book”, by the way. So you see, by that definition, Linux is very definitely an ”operating system”.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
A lot of people are saying, believe me, that’s just what I heard, let’s face it, everybody’s talking about it, in all fairness, the most tremendous thing ever, they are saying it’s tremendous, it’s phenomenal, everybody’s talking about it, likes of which nobody has seen before, to be perfectly honest, people think it’s going to happen. #TalkLikeTrump
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
“On top of the operating system is the rest of the system software. Here we find the command interpreter (shell), window systems, compilers, editors, and similar application-independent programs. It is important to realize that these programs are definitely not part of the operating system, even though they are typically supplied preinstalled by the computer manufacturer, or in a package with the operating system if it is installed after purchase. This is a crucial, but subtle, point. The operating system is (usually) that portion of the software that runs in kernel mode or supervisor mode. It is protected from user tampering by the hardware ...”
-- Tanenbaum & Woodhull, “Operating Systems Design And Implementation”, third edition, page 3
That is the well-known “MINIX Book”, by the way. So you see, by that definition, Linux is very definitely an ”operating system”.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
If you want to see a much-hyped solution in search of a problem, look at quantum encryption. This uses entangled particles to communicate information across arbitrary distances. Which is theoretically entirely possible. What makes it special is the idea that it is proof against eavesdroppers, because the slightest attempt to interfere with the state of one particle will instantly show up in the (mis)behaviour of the other particle. And this will set off alarm bells etc.
Now, if you understand anything about computer security, you will know that the security of the entire current Internet is built on protocols that simply don’t care about eavesdroppers. Spies can collect as much as they like about the encrypted information going back and forth, it will be of essentially no help to them to decipher any of that content. (Yes, there are still potentially useful things like traffic analysis. I’m ignoring that for now.)
So with quantum encryption, you have a system that is so sensitive to interference, you could potentially trigger an eavesdropper alarm without actually bothering to eavesdrop at all. In computer security, we call this a “denial-of-service attack”. Do this often enough, and the supposedly “secure” quantum communication channel becomes unusable.
You see the problem ...
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
Umm ... why do you need hard disk images on Linux, exactly? The OS doesn’t care about the precise layout of its files on disk. Just do a file-level copy, and you can restore from that, too. Let me recommend rsync as a robust file copy tool that can cope reliably with hundreds of gigabytes of data, working across (possibly unreliable) network connections etc. It has clever tricks like deduping, being able to make an incremental backup that looks, for restore purposes, like a full backup.
For example, to move a Linux install to a new hard drive, just copy the files across, re-setup the boot loader, edit /etc/fstab for the changed device names, and that’s it. I’ve upgraded hard drives countless times that way, and never had to reinstall.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
3:38 The reason why Edge is so tightly coupled, not just to Windows 10, but your current version of Windows 10, is a really stupid one: it’s because it’s written as a “Universal Windows Platform” app. UWP, if you remember, was Microsoft’s latest in a long string of attempts to "modernize” its antiquated Windows APIs, from MFC, through Dotnet and Silverlight, through WinRT, through ... how many other dead and gone Microsoft technologies can you name? But if anything is proof of the uselessness of UWP, this whole Edge fiasco has to be it.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
Worth mentioning Jacob Bronowski’s classic 1973 documentary series on the history of science, The Ascent Of Man . May seem a bit old now, but he’s still fascinating as a presenter. Episode 6 dealt in part with Galileo. Bronowski visits the actual room where Galileo was tried--it had been turned into a post office by that time. He was also granted access to the secret Vatican archive where a remarkable series of historical documents are kept, including the (in)famous divorce application by Henry VIII of England, the refusal of which led to the foundation of the Church of England.
While the trial of Galileo took place in 1633, the dossier collected on him by the Inquisition starts as far back as 1611. There was also claimed to be a document (never produced at the trial) which was supposed to prohibit him from teaching geocentrism. Bronowski found that document in the archive, and it was at best an unfinished draft which was rejected, or at worst, a complete forgery.
As part of Galileo’s sentence, besides the recantation and the house arrest, he was also to be shown the instruments of torture. You can imagine the effect on a scholar, a man unused to physical abuses, such a display would have.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
15:36 OK, here’s an interesting thought: one situation where, it seems, this issue of prior probabilities is explicitly ignored, is in the legal system. Here in 🇳🇿, we had a high-profile murder case where all details of the accused had been kept suppressed until just the last few days. It turns out the reason was that he was implicated in the deaths of some other victims, and the court wanted to be sure that these other cases would not influence the verdict in this one. Whereas it is now clear he is a serial offender, who has done this sort of crime several times, and will likely do it again.
In short, it seems the legal idea that each criminal trial should be run entirely independently, without regard to what might be happening in other trials of the same perp, is directly counter to Bayes’ Theorem.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
Try solving this without recursion: computing all the permutations of a sequence.
def permute(seq) :
"generator which yields successive permutations of the elements of seq."
if len(seq) == 0 :
yield ()
else :
for i in range(0, len(seq)) :
for rest in permute(seq[:i] + seq[i + 1:]) :
yield (seq[i],) + rest
#end for
#end for
#end if
#end permute
>>> list(permute((1, 2, 3, 4)))
[(1, 2, 3, 4), (1, 2, 4, 3), (1, 3, 2, 4), (1, 3, 4, 2), (1, 4, 2, 3), (1, 4, 3, 2), (2, 1, 3, 4), (2, 1, 4, 3), (2, 3, 1, 4), (2, 3, 4, 1), (2, 4, 1, 3), (2, 4, 3, 1), (3, 1, 2, 4), (3, 1, 4, 2), (3, 2, 1, 4), (3, 2, 4, 1), (3, 4, 1, 2), (3, 4, 2, 1), (4, 1, 2, 3), (4, 1, 3, 2), (4, 2, 1, 3), (4, 2, 3, 1), (4, 3, 1, 2), (4, 3, 2, 1)]
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
It’s a tricky thing to explain precisely because it’s such a tricky concept.
It used to be simpler: until about the late 1980s, faster CPUs could be matched with faster RAM, so the higher the clock speed, the faster everything went. But then improvements in CPU speed outstripped increases in RAM speed. So for a while you had “wait states”, then finally they added caches: first one level, then two, nowadays even three separate levels to bridge the ever-widening gap between CPU speeds and RAM speeds.
So now the sizes of those caches, and the size of the pipes for transferring data between them, the CPU and main RAM, have become big factors in the speed of the system as well.
And then within the CPU, you have design decisions like how many concurrent instructions can be in flight at once, with separate arithmetic units assigned to each one, and how efficiently data can be moved around the chip, that further complicate the relationship between clock speed and CPU performance.
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1