to proportion and arrangement Asymmetric • Unbalanced or Exceptional Implies edge-cases or surprising behavior “at-scale” Symmetric • Harmonious or Balanced Implies high predictability (when pattern asymmetry in C Language: These are probably “obvious” to experienced developers (but sometimes surprising to new developers) • struct (but not array) may be returned from a function • Default is pass-by-value of) “Some” asymmetries in Your System is probably fine Asymmetry tends to cause edge cases and surprising behavior std::string_view vs. std::span<> • Similar motivations and use cases: • Non-owning

accumulated tiny arithmetic errors visible. Given our previous remarks, this should no longer be surprising; the general solution is to round results to fewer than 16 decimal digits before printing. Various to when you switch on a torch on a sunny day; it will not really become brighter. Perhaps more surprising is that calling echo() with some simple float literals will print a different value, such as 66.04000000000001 for Nim v2.0, while with Python3 we get 66.04 exactly. However, this is only surprising for people who do not fully understand what a statement like echo 66.04 does: We already know

accumulated tiny arithmetic errors visible. Given our previous remarks, this should no longer be surprising; the general solution is to round results to fewer than 16 decimal digits before printing. Various to when you switch on a torch on a sunny day; it will not really become brighter. Perhaps more surprising is that calling echo() with some simple float literals will print a different value, such as 66.04000000000001 for Nim v2.0, while with Python3 we get 66.04 exactly. However, this is only surprising for people who do not fully understand what a statement like echo 66.04 does: We already know

"UNDEFINED" VALUES 74 NO SURPRISING NULL VALUES FROM DICTS 75 76 NO SURPRISING NULL VALUES FROM DICTS 77 NO SURPRISING NULL VALUES WHEN UNPACKING 78 NO SURPRISING NULL VALUES WHEN UNPACKING UNPACKING 79 80 NO SURPRISING RESULTS FOR "2" + 8 81 THE LAST PARAGRAPH OF FLUENT PYTHON 82 COMMUNITY 83 84 PYTHON BRASIL 2018 BY JULIO MELANDA (INSTAGRAM @JULIOMELANDA) 85 86 87 88

accumulated tiny arithmetic errors visible. Given our previous remarks, this should no longer be surprising; the gen­ eral solution is to round results to fewer than 16 decimal digits before printing. Various to when you switch on a torch on a sunny day; it will not really become brighter. Perhaps more surprising is that calling echo() with some simple float lit­ erals will print a different value, such as 66.04000000000001 for Nim v2.0, while with Python3 we get 66.04 exactly. However, this is only surprising for people who do not fully understand what a statement like echo 66.04 does: We already know

accumulated tiny arithmetic errors visible. Given our previous remarks, this should no longer be surprising; the gen­ eral solution is to round results to fewer than 16 decimal digits before printing. Various to when you switch on a torch on a sunny day; it will not really become brighter. Perhaps more surprising is that calling echo() with some simple float lit­ erals will print a different value, such as 66.04000000000001 for Nim v2.0, while with Python3 we get 66.04 exactly. However, this is only surprising for people who do not fully understand what a statement like echo 66.04 does: We already know

accumulated tiny arithmetic errors visible. Given our previous remarks, this should no longer be surprising; the gen­ eral solution is to round results to fewer than 16 decimal digits before printing. Various to when you switch on a torch on a sunny day; it will not really become brighter. Perhaps more surprising is that calling echo() with some simple float lit­ erals will print a different value, such as 66.04000000000001 for Nim v2.0, while with Python3 we get 66.04 exactly. However, this is only surprising for people who do not fully understand what a statement like echo 66.04 does: We already know

accumulated tiny arithmetic errors visible. Given our previous remarks, this should no longer be surprising; the gen­ eral solution is to round results to fewer than 16 decimal digits before printing. Various to when you switch on a torch on a sunny day; it will not really become brighter. Perhaps more surprising is that calling echo() with some simple float lit­ erals will print a different value, such as 66.04000000000001 for Nim v2.0, while with Python3 we get 66.04 exactly. However, this is only surprising for people who do not fully understand what a statement like echo 66.04 does: We already know

accumulated tiny arithmetic errors visible. Given our previous remarks, this should no longer be surprising; the gen­ eral solution is to round results to fewer than 16 decimal digits before printing. Various to when you switch on a torch on a sunny day; it will not really become brighter. Perhaps more surprising is that calling echo() with some simple float lit­ erals will print a different value, such as 66.04000000000001 for Nim v2.0, while with Python3 we get 66.04 exactly. However, this is only surprising for people who do not fully understand what a statement like echo 66.04 does: We already know