51

Re: Re: the lame logic

Hello, Dair, you wrote: D> Setters and  are necessary for normal redefinition in descendants. D> for example, in a setter it is possible to pull still the notification message someone about parameter change. This wrong sharing on model-view-controler. It is the controler should pull the notification message, instead of model in setter D> Though in case of any simple things like Point/Rect I, of course, would do by their structures with all public fields, certainly. But it I as already spoke above, carp already. The full normally operating code made on another and anything - there live people

52

Re: Re: the lame logic

Hello, ononim, you wrote: O>. The person with two underlinings in  and why their two told? More precisely even so - __ kot2 it is the normal identifier in With ++?

53

Re: Re: the lame logic

Hello, Maniacal, you wrote: M> the class Title is mandatory begins with ' a C ', structures with ' S',  with "m _", in the beginning of a title of a variable a letter designating its type: n - for int, b for bool, sz for an ASCIIZ-line, ' a ' for an array, "g _" for global variables, "gc _" for global constants M> m_bTerminating, m_nCount, m_szName, etc. with the advent of templates at people broke  how to name such variables because they  to be absolutely any type and then already the inspiration came that the type is not important,  about it more even hinders. Oils to fire were added by that the size ceased to be int and became size_t - which not clearly with what prefix to write and somewhere from the middle 90 all began to refuse the Hungarian notation. That it is live till now in 2016 to year in any company - , it is exact not plus to it in a karma

54

Re: Re: the lame logic

__> Hello, ononim, you wrote: O>>. The person with two underlinings in  __> and why their two told? Probably, that was longer __> more precisely even so - __ kot2 it is the normal identifier in With ++? Than the child would not play, if only not goto

55

Re: Re: the lame logic

Hello, flt, you wrote: M>> the Suffix "_m" is used at Microsoft for members of classes with any visibility. The same WTL, and MFC like too F> Only not the suffix, and a prefix, and not "_m", and "m _" Well, not "Volga", and hundred roubles, did not benefit, and I already lost couple of years in WTL did not look, forgot partially that there as

56

Re: Re: the lame logic

Hello, Kodt, you wrote: But  the policy of purity of the code - since purity of style - does the light work. Basically, from public code style  and truth the most made. Certainly, confuses that they did not master an exception and on trifles, but in general it is better remaining And here it - nafig-nafig! You cannot make function, at which or-zero-or-two parameter. It turns out, you will involuntarily resolve record Point (x) which designates Point (x, 0). Points with zero ordinate are specially necessary to Someone? Moreover and without explicit, it generally the unguarded sex. Yes, it agree. It is argument to do two designers. I personally in absolutely small project will prefer one (suddenly I will add a new field and  only in one designer, about the second I will forget - such happened), but by operation to us the big project the big command two I am better to have, by the way, saw somewhere a code sample in style struct S {int x = 10;} ; It is a part of the new standard? Which? Or it for simplicity of arguing was? Uniformity with other geometrical classes, especially - with a rectangle. Otherwise would be rect.width and rect.right () or rect.width () and rect.right. Well, it too as though, is possible, for example rect to present an occasion to arguing in a pair type pos, dimensions - to sootv th pos will be Point. That is will be r.pos.x - it is not necessary to fence uniformity, it arises in the natural image. But, probably, in certain cases it is necessary to have any agreements about x () and y () though it is more logical to have more likely the agreement about pos () Under specifications. Snakes for , pascals for the bulky. The strange specifications. Most likely are written C# by companions. On a course came from attempts   to With ++. It was necessary to them is better  to find for a spelling style guide the Hogwash just clear, - finding of the left upper point, for example, at join and intersection of rectangles. And I here did not guess Why not in the form of function LeftTopMost (const Point& a, const Point& b) is  a narcotism. So there is no need to encourage a narcotism, even if it ! It is necessary to be at once to cane to the one who added it. It was explicit not from the beginning in a class

57

Re: Re: the lame logic

Hello, __ kot2, you wrote: __> is more exact even so - __ kot2 it is the normal identifier in With ++? On hands to beat Each name that contains a double underscore __ or begins with an underscore followed by an uppercase letter (2.12) is reserved to the implementation for any use.

58

Re: Re: the lame logic

Hello, Ops, you wrote: Ops> On hands to beat Ops> Ops> Each name that contains a double underscore __ or begins with an underscore followed by an uppercase Ops> letter (2.12) is reserved to the implementation for any use. Here the same system -  kot at all forums it is occupied, and __ kot - it is free and reserved specially for me

59

Re: Re: the lame logic

Hello, __ kot2, you wrote: __> I, by the way, saw somewhere a code sample in style __> __> struct S __> {__> int x = 10; __>}; __> __> it is a part of the new standard? Which? Or it for simplicity of arguing was? It non-static data member initializers, have been added in a C ++ 11: http://en.cppreference.com/w/cpp/langua … ialization And class Point could look so: struct Point {int x = 0; int y = 0; constexpr Point () = default; constexpr Point (int x, int y): x (x), y (y) {}...}; Also there is a variant with the delegating designer: Point:: Point (): Point (0, 0) {} it is Still strange that  selected a way of duplication of the code and steam Point/PointF, Size/SizeF, etc. instead of templates did.

60

Re: Re: the lame logic

Hello, Kodt, you wrote: PM>> it is Still strange that  selected a way of duplication of the code and steam Point/PointF, Size/SizeF, etc. instead of templates did. It is not strange.  persistently like to push functions in separate compilation units to reduce swelling  and to avoid a hell of static variables at the component assembly. They, for example, had recently a paranoia attack - explicitly  all designers generated by the compiler.//.h class Foo {> public: Foo (const Foo&);//= default.... private: int x _, y _, z _;};>//.cc Foo:: Foo (const Foo&) = default; With templates such focus, matter of course, you will not turn. Also pay attention what Point associates has attribute GFX_EXPORT is old kind __ declspec (dllexport) from the unit gfx.dll And unless explicit instantiation in.cc a file does not work? Type such://.hpp hpp template <typename T> struct PointT {T x = {}; T y = {}; PointT (); PointT (T x, T y);}; using Point = PointT <int>; using PointF = PointT <float>;//declare explicit instantiation extern template PointT <int>; extern template PointT <float>;//.cpp template <typename T> struct PointT <T>:: PointT () = default; template <typename T> struct PointT <T>:: PointT (T x, T y): x (x), y (y) {}//define explicit instantiation template PointT <int>; template PointT <float>; It will be necessary with dllexport to try.

61

Re: Re: the lame logic

:>> Also pay attention what Point associates has attribute GFX_EXPORT is old kind __ declspec (dllexport) from the unit gfx.dll PM>> And unless explicit instantiation in.cc a file does not work? Type such: Works, likely (?). It will Work, but here to me interesting another: what saving of the size  can turn out for the account of speeding up of real function invocation which is simple, how two copecks?

62

Re: Re: the lame logic

Hello, Kodt, you wrote: From source codes . These dregs from CSS are pulled... background-size: 300px 100px;/* width-> 300, height-> 100 *//* but! */padding: 1.6em 20px;/* top and bottom (vertical) 1.6em, left and right (horizontal) 20px margin */padding: 1em 3px 30px 5px;/* top 1em padding *//* right 3px padding *//* bottom 30px padding *//* left 5px padding */

63

Re: Re: the lame logic

Hello, c-smile, you wrote:> From source codes . CS> These dregs from CSS are pulled... Anything, anything good from web did not come. Kill it with fire.