1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
|
% \enablemode[print]
\usemodule[pre-stepwise,present-tiles,abr-02]
\definecolor[maincolor] [darkgray]
\definecolor[othercolor][r=.3,g=.3]
% \setupinteractionscreen
% [option=max]
\setupbodyfont[15pt]
\startdocument
[title={Speed:\\\\can we make\\it any faster},
subtitle={Hans Hagen\\EuroBacho\TeX\\May 2013}]
\StartSteps
\starttopic[title={Speed}]
\startitemize
\startitem speed matters in a edit-run-preview cycle although this is mostly perception \FlushStep \stopitem
\startitem the nicer the interface, the slower it gets, but you seldom set something up \FlushStep \stopitem
\startitem everything you provide gets used at some point, also in inefficient ways \FlushStep \stopitem
\startitem lots of local (grouped) tweaks leads to many mechanisms kicking in unseen \FlushStep \stopitem
\startitem wrong use of functionality can have drastic and unexpected speed penalties \FlushStep \stopitem
\stopitemize
\stoptopic
\StopSteps
\StartSteps
\starttopic[title={Pages per minute}]
\startitemize
\startitem we try to speed up baseline performance (in pages per second) \FlushStep \stopitem
\startitem identify and optimize critical routines, both at the \TEX\ and \LUA\ end \FlushStep \stopitem
\startitem of course the machine (Dell M90, SSD, 4GB, 2.33 Ghz T7600, Windows 8) and versions if \LUATEX\ (0.72+) and \CONTEXT\ matter \FlushStep \stopitem
\stopitemize
\blank
\starttyping
\dorecurse {1000} {test \page}
\stoptyping
\FlushStep
\blank
\starttabulate[|r|r|r|r|]
\HL
\NC \bf \# pages \NC \bf Januari \NC \bf April \NC \bf May\rlap{\quad(2013)} \NR
\HL
\NC 1 \NC 2 \NC 2 \NC 2 \NC \NR
\NC 10 \NC 15 \NC 17 \NC 17 \NC \NR
\NC 100 \NC 90 \NC 109 \NC 110 \NC \NR
\NC 1000 \NC 185 \NC 234 \NC 259 \NC \NR
\NC 10000 \NC 215 \NC 258 \NC 289 \NC \NR
\HL
\stoptabulate
\FlushStep
\stoptopic
\StopSteps
\StartSteps
\starttopic[title={What happens}]
\startitemize
\startitem load macros and \LUA\ code is loaded from the format \FlushStep \stopitem
\startitem the system gets initialized, think of fonts and languages \FlushStep \stopitem
\startitem additional (runtime) files are loaded \FlushStep \stopitem
\startitem text is typeset and eventually gets passed to the page builder \FlushStep \stopitem
\startitem pages are packaged, this includes reverting to global document states \FlushStep \stopitem
\startitem the \PDF\ representation is created \FlushStep \stopitem
\startitem each of these steps has its bottlenecks \FlushStep \stopitem
\startitem the more we don, the more \LUA\ gets involved \FlushStep \stopitem
\stopitemize
\stoptopic
\StopSteps
\StartSteps
\starttopic[title={What we can do}]
\startitemize
\startitem avoid copying boxes where possible \FlushStep \stopitem
\startitem only enable initializers and finalizers when functionality is used \FlushStep \stopitem
\startitem be clever with fonts, in usage as well as in supporting features \FlushStep \stopitem
\startitem use trial runs in multi||pass mechanisms \FlushStep \stopitem
\startitem avoid too much macro expansion (only matters for tracing) \FlushStep \stopitem
\startitem accept that more functionality has a price \FlushStep \stopitem
\stopitemize
but
\startitemize
\startitem don't compromise functionality \FlushStep \stopitem
\startitem avoid too obscure code \FlushStep \stopitem
\startitem forget about optimization by means of combining functionality \FlushStep \stopitem
\stopitemize
\stoptopic
\StopSteps
\stopdocument
|
