- Abuse report of Rick C. Hodgin's spam - 15 Updates
- About placing member variable into a pure virtual base class - 6 Updates
- What is this constructor? - 4 Updates
| Juha Nieminen <nospam@thanks.invalid>: Oct 15 06:37AM > I am a witness for Jesus Christ in this world, No, what you are is an asshole, and a fucking spammer, plain and simple. |
| gazelle@shell.xmission.com (Kenny McCormack): Oct 15 11:48AM In article <pq1cjd$bv2$2@adenine.netfront.net>, >> I am a witness for Jesus Christ in this world, >No, what you are is an asshole, and a fucking spammer, >plain and simple. The two of you (Juha and Rick) have said the same thing - just using different terminology. Of the two, I prefer Juha's framing, but, again, the content is the same. -- "Only a genius could lose a billion dollars running a casino." "You know what they say: the house always loses." "When life gives you lemons, don't pay taxes." "Grab 'em by the p***y!" |
| "Rick C. Hodgin" <rick.c.hodgin@gmail.com>: Oct 15 05:34AM -0700 On Monday, October 15, 2018 at 2:37:42 AM UTC-4, Juha Nieminen wrote: > > I am a witness for Jesus Christ in this world, > No, what you are is an .. , and a .. spammer, > plain and simple. Juha, is there any project you are working on that I could lend my labor to, to help you achieve more in this world than you are able to do by yourself? I offer my services, skills, talents, knowledge, and abilities in computer programming, to help you in those things of interest to you. I want you to thrive, Juha. I'm willing to help you achieve that goal by giving you my labor. -- Rick C. Hodgin |
| "Rick C. Hodgin" <rick.c.hodgin@gmail.com>: Oct 15 05:36AM -0700 On Monday, October 15, 2018 at 7:48:18 AM UTC-4, Kenny McCormack wrote: > The two of you (Juha and Rick) have said the same thing - just using > different terminology. > Of the two, I prefer Juha's framing, but, again, the content is the same. Kenny, is there anything I can do to help you achieve more in this world than you are able to do on your own? Do you have any projects that you need computer programming assist- ance on? If so, please contact me and I will see if I can help you. I want to see you doing more and better than you're able to do by yourself. I offer to give you my service and labor to help you achieve more. -- Rick C. Hodgin |
| "Öö Tiib" <ootiib@hot.ee>: Oct 15 06:00AM -0700 On Monday, October 15, 2018 at 3:34:55 PM UTC+3, Rick C. Hodgin wrote: > I offer my services, skills, talents, knowledge, and abilities > in computer programming, to help you in those things of interest > to you. How can people measure your talents and abilities? Can you get your CAlive into state where someone can try to compile something with it? Some other fruit of yours? |
| Juha Nieminen <nospam@thanks.invalid>: Oct 15 01:10PM > Juha, is there any project you are working on that I could lend > my labor to, to help you achieve more in this world than you are > able to do by yourself? Yes: Stop spamming this newsgroup with off-topic posts. |
| "Rick C. Hodgin" <rick.c.hodgin@gmail.com>: Oct 15 09:12AM -0400 On 10/15/2018 9:00 AM, Öö Tiib wrote: > How can people measure your talents and abilities? Can you get > your CAlive into state where someone can try to compile something > with it? Some other fruit of yours? It's on its way. People will see the fruit from it in a year or so. As for my talents, I can be given tasks to do and see if I can do them. I won't blow smoke. I can tell people if I can do something, if it's outside of my abilities, etc. I just want people to realize my caring for them is sincere. -- Rick C. Hodgin |
| boltar@cylonhq.com: Oct 15 01:14PM On Mon, 15 Oct 2018 06:00:25 -0700 (PDT) >> in computer programming, to help you in those things of interest >> to you. >How can people measure your talents and abilities? Can you get They wouldn't need to. He'd probably turn up, tell them they'd all burn in hell unless they devoted themselves to jesus or hired him to help them see the true faith and the interview would be over. |
| "Rick C. Hodgin" <rick.c.hodgin@gmail.com>: Oct 15 09:18AM -0400 > They wouldn't need to. He'd probably turn up, tell them they'd all burn in > hell unless they devoted themselves to jesus or hired him to help them see > the true faith and the interview would be over. I'm not asking for a job. I'm offering my labor for free to help people. I'm willing to spend time helping Juha or Kenny improve their products / tasks / plans / goals, so they can accomplish more with the assistance than they could do on their own. It's not a for-fee service. It's a gift of free labor, because I truly care about people and want them to see that love in practice, and not just in words. -- Rick C. Hodgin |
| boltar@cylonhq.com: Oct 15 01:35PM On Mon, 15 Oct 2018 09:18:53 -0400 >It's not a for-fee service. It's a gift of free labor, because I >truly care about people and want them to see that love in practice, >and not just in words. Nothing is free, even from people like yourself. The price would be having to listen to you sermonising. You can't yourself. |
| "Öö Tiib" <ootiib@hot.ee>: Oct 15 06:39AM -0700 On Monday, 15 October 2018 16:12:43 UTC+3, Rick C. Hodgin wrote: > > your CAlive into state where someone can try to compile something > > with it? Some other fruit of yours? > It's on its way. People will see the fruit from it in a year or so. I'll ask again in 15. October 2019 then. |
| "Rick C. Hodgin" <rick.c.hodgin@gmail.com>: Oct 15 09:51AM -0400 >> and not just in words. > Nothing is free, even from people like yourself. The price would be having to > listen to you sermonising. You can't yourself. No sermons. Military people wear their various badges and insignia on their uniforms as they go. As I am known as a Christian, so I would be as I go as well, but it would not be the equivalent of me pointing to my badges and insignia and saying, "I was awarded this one for such-and-such activity and such-a-such place." I would simply be there, doing for the people. Who I am would naturally carry with me. Boltar, why do you think I reach out to people about the knowledge of Jesus Christ? Am I doing it to grow my posse? Am I doing it so people will send me money? Am I doing it for some self-serving reason? Or am I doing it because people have sin and I want them to be forgiven and to be saved? Have you ever truly thought about it? -- Rick C. Hodgin |
| "Rick C. Hodgin" <rick.c.hodgin@gmail.com>: Oct 15 09:53AM -0400 On 10/15/2018 9:39 AM, Öö Tiib wrote: >>> with it? Some other fruit of yours? >> It's on its way. People will see the fruit from it in a year or so. > I'll ask again in 15. October 2019 then. You're welcome to come on board and help me with my projects. I do not have just one. I'm trying to create a full hardware and software stack, one that is built not on profit-seeking, but one on honoring the Lord Jesus Christ first, and then being given to people for the cost of manufacturing and production second. In the software area, that is for free. In the hardware area, it will be the cost of what- ever it costs to have the technology manufactured. I'm honestly and truly looking to serve God in this world, by looking to the people He placed here and serving Him by giving to them in this world. I'm not duplicitous. I'm sincere in my love for Christ, and that love applied toward people. You are a great talent based on your posts, Öö Tiib. Your skills would be put to use if you're interested. -- Rick C. Hodgin |
| boltar@cylonhq.com: Oct 15 02:00PM On Mon, 15 Oct 2018 09:51:08 -0400 >reason? Or am I doing it because people have sin and I want them >to be forgiven and to be saved? >Have you ever truly thought about it? Your reasons are irrelevant. People don't want to be preached at whether it be from you or from the crazy in the street wearing an End of the World is Nigh sandwich board. |
| Dan Purgert <dan@djph.net>: Oct 15 02:40PM > Your reasons are irrelevant. People don't want to be preached at > whether it be from you or from the crazy in the street wearing an End > of the World is Nigh sandwich board. To be fair, that guy at least makes good sandwiches. And if it takes listening to a little bit of preaching while I wait for my turkey club, well... -- |_|O|_| Registered Linux user #585947 |_|_|O| Github: https://github.com/dpurgert |O|O|O| PGP: 05CA 9A50 3F2E 1335 4DC5 4AEE 8E11 DDF3 1279 A281 |
| Pavel <pauldontspamtolk@removeyourself.dontspam.yahoo>: Oct 14 11:47PM -0400 JiiPee wrote: > so can you give and example how to change that Animal class not to place > the data there As written, Animal::age_ is inaccessible from outside the class. When you decide who is to use it (what methods/classes), you will have better idea where to put it. If you want age to be accessible for every animal via a public or protected method, you can leave it in Animal. If all you are interested in for your interface is talking, you should probably rename your interface to Talker and move age_ to some LivingThing class -- that don't even need to have virtuals unless there are other requirements, thus having something like class Talker { public: virtual void talk() const = 0; ... }; class LivingThingData { int age_; public: int age() const { return age_; } }; then, class Cat: public Talker, private LivingThingData { /* assuming you only need age inside, say, to decide on the pitch for meowing.. */ }; Essentially, analyze business model and separate concerns. I don't think you can get once-and-for-all "correct" problem decomposition: good model for a set of requirements may become inadequate as requirements are added so you have to plan for some rework. The best you can do is to separate concerns that you know should be separate as of the *current* set of business requirements. HTH -Pavel |
| Juha Nieminen <nospam@thanks.invalid>: Oct 15 06:29AM > functions, it's no problem to inherit from several of them. If there > are members, you need virtual inheritance (which is complicated and less > efficient). No, you don't. You might need virtual inheritance in a case of diamond inheritance. However, interfaces aren't inherited from a common base class, so virtual inheritance isn't needed and would do nothing. |
| Juha Nieminen <nospam@thanks.invalid>: Oct 15 06:33AM > The term "interface" is terribly overused in various senses and in all > programming including in C++. It doesn't exactly help that "interface" means in other context simply the collection of public methods of a class (as in the "public interface" of the class). "Interface" in those language is really ill-named. Objective-C uses the term "protocol" instead, which is probably better. (Of course "protocol" itself is already in use in other contexts, but those contexts are different enough that it's justifiable.) |
| David Brown <david.brown@hesbynett.no>: Oct 15 08:46AM +0200 On 15/10/18 08:29, Juha Nieminen wrote: > You might need virtual inheritance in a case of diamond inheritance. > However, interfaces aren't inherited from a common base class, so > virtual inheritance isn't needed and would do nothing. You are right - you only need the virtual inheritance if you are getting the interface (and therefore the same data member) from more than one ancestor. For some kinds of interface classes that can be likely, so you might still end up using virtual inheritance just in case. |
| vambola@proekspert.ee: Oct 15 05:03AM -0700 On Monday, October 15, 2018 at 12:44:59 AM UTC+3, Thiago Adams wrote: > I am using this pattern (not the C++ version, but the C one, > with int as tag and switch cases) to implement an AST data > structure. Lets call your solution with some name? Something like "intrusive tags"? Then it is easier to reason when things have names. ;) Modern day often only profiler can tell for sure what has advantage or disadvantage in fastness in concrete usage situations. Thinking helps a bit but since compilers are quite clever now it is often wrong. If it is faster then it can be used as performance optimization. I don't see how it is simpler. > I cannot have fruit. If we had ICarPart then the glass or tire > is very coupled with the original problem and is difficult to reuse > and confuse as well. We can use some problem domain oriented name like "IPart" for possibly spare parts of different vehicles or "IConsumable" for stuff that can be possibly eaten by beings. But we will lose compile-time checking that "wing" is unacceptable "part" for "car" or "hay" is unacceptable "food" for "tiger" both with such interfaces and with your intrusive tagging. With visitor it is also likely run-time issue, with variant there is concrete set of types so it is compile-time issue. > In the AST, one sample is static_assert it can be in different > positions without any relationship with the other parts. Two things are important: if it is easy to unit-test and if less skilled programmers can take over its maintenance. I feel that intrusive tagging can be unit-tested OK but it takes yourself to be in control over it right now. With AST-like problems I would go with variant anytime, smaller types directly, bigger, polymorphic and/or recursive types as (smart)pointer in variant. We can concentrate on actual features and leave performance questions for later. Code of std::variant is huge but it is mostly compile-time meta-programming. The external tag of it is not such a speed or storage issue and the compile-time type safety defeats lot of issues with clumsy maintenance. > I will consider this solution in many other cases, > where I am the owner of the types. (I know all types) Yes, I think the weakness is that you have to own the types. With unknown, future types the options are always limited to usage of abstract interface classes, clever compile-time meta-programming, code generators, callbacks, lambdas and/or std::function. But your intrusive tags can be useful as performance optimization in implementation details. > The sample is avoiding coupling cause by interfaces. > I am not using this dynamic_select in any program, but I am > using the 'dynamic switch selection' in a C program. Yes, in C it makes lot of sense since every data in C is POD with certain binary layout. In C++ the intrusive tags are tricky for two reasons: 1) Same binary location of tags in objects of different types is hard to achieve. "Standard layout" is rather restrictive trait. Otherwise it is undefined behavior. Having special purpose interface to read the tag (like ITagged) would however defeat most efficiency advantages. Also dynamic (cross?)casting to what it apparently leads into can raise major violent controversy (AKA shitstorm) in some collectives. 2) Difference between different tags and equality of tags to same type in shared library and main program is hard to achieve automatically. There are no "link-time counters" or "link-time GUID" in C++. Linking is perhaps mentioned in less than handful of non-normative comments of standard. That issue leads either to language extensions or to code generators. |
| Thiago Adams <thiago.adams@gmail.com>: Oct 15 05:47AM -0700 > Lets call your solution with some name? Something like > "intrusive tags"? Then it is easier to reason when things have > names. ;) OK! > situations. Thinking helps a bit but since compilers are quite > clever now it is often wrong. If it is faster then it can be used > as performance optimization. I don't see how it is simpler. I could have used pointer to some function or global variable like vtable as intrusive tag. Then the linker would give me a value for the tag and I don't need to manage unique ids for all objects. But I believe that integers, preferentially sequencial, can give me the best performance because the compiler will see the switch case instead of if else if else .. Also something interesting for performance is when the virtual function is very small it can be inlined inside the switch. void Shape_Draw(struct Shape* pShape) /*default*/ { switch (pShape->id) { case 2: Circle_Draw((struct Circle*)pShape); //could be inlined break; case 1: Box_Draw((struct Box*)pShape); break; default: break; } } > With visitor it is also likely run-time issue, with variant > there is concrete set of types so it is compile-time > issue. Exactly. I realized the collection makes the problem worst because they create this problem of have a common base interface. Some hierarchies like iostreams doesn't need to worry about this because this problem would be difficult to happen in that scenario. In other software I did an interface to represent data source like IDataSource. Then I could add more data sources without to have to change the code that uses data source. In more than one situation I had to add new function in the IDataSource interface and implement this function in all types. But some data sources were not able to implement that function and I had to return an error. I think this scenario was a good use case for interfaces. But I am not sure I would try the 'intrusive tag' next time. The advantage is that we can implement some algorithm selectively. For instance, a need a new function F that only works for some types. Then I can implement this diferences (return an error for instance) in my switch cases without to change the classes that implement the interface the diferences will be clear at the algorithm itself. The disadvantage of having the switch is manually update. This is why I think we could have support from the language and do something like an template instantiation. I already did this in my transpiler. http://thradams.com/web/cprime.html See polimorphism sample. The switch is edited by the compiler. > Yes, in C it makes lot of sense since every data in C is POD > with certain binary layout. In C++ the intrusive tags are tricky > for two reasons: In C++ we can have a base class like TaggedObject. This doesn't create coupling because the object is very simple. |
| woodbrian77@gmail.com: Oct 14 07:29PM -0700 On Sunday, October 14, 2018 at 2:56:37 PM UTC-5, Öö Tiib wrote: > has not been posted. The default argument to A(int) uses comma > operator to achieve that temporary A with parameter 5 is constructed > first. I guess this is a "garbagio" constructor. Brian |
| Chris Vine <chris@cvine--nospam--.freeserve.co.uk>: Oct 15 10:07AM +0100 On Sun, 14 Oct 2018 19:29:32 -0700 (PDT) > > operator to achieve that temporary A with parameter 5 is constructed > > first. > I guess this is a "garbagio" constructor. It seems intended as an illustration of how sub-expressions work in relation to the noexcept keyword. So far as the comma operator is concerned, I don't find myself using it very often, but one use is in relation to copy constructors. Say you have a type T with a member 'x' protected by a mutex 'm'. The lifetime rules for temporaries mean that the following initialization of 'x' is atomic because the mutex will be remain locked while the initialization is taking place (the full expression in which the temporary is created does not expire until initialization of the new 'x' has completed): T(const T& t): x((std::lock_guard(t.m), t.x)) {} This avoids having to first default construct the new 'x' and then assign to it under the mutex in the constructor. I was somewhat surprised about this when I first came across it, but if you add some instrumentation it can be seen that it does actually work. |
| Fraser Ross <fraser.ross8ATbtinternet.com@com>: Oct 15 12:03PM +0100 > Brian > Ebenezer Enterprises - Enjoy programming again > http://webEbenezer.net Implicit conversions from int to A has not been mentioned by anyone. It is not specifically saying anything about that such as with this theoretical code: implicit A(int = (A(5),0))noexcept; Although there is the keyword explicit there is not implicit. Using implicit would at least suppress warnings if it existed. Fraser. |
| vambola@proekspert.ee: Oct 15 05:17AM -0700 On Monday, October 15, 2018 at 12:07:30 PM UTC+3, Chris Vine wrote: > T(const T& t): x((std::lock_guard(t.m), t.x)) {} > This avoids having to first default construct the new 'x' and then > assign to it under the mutex in the constructor. If it is really needed then it can lead to someone naively thinking that ... T(const T& t) : x((std::lock_guard(t.m), t.x)) , y((std::lock_guard(t.m), t.y)) {} ... is also thread-safe and typical surprises about how that T with broken invariant was constructed. ;-) |
| You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page. To unsubscribe from this group and stop receiving emails from it send an email to comp.lang.c+++unsubscribe@googlegroups.com. |
No comments:
Post a Comment