- What's up with "Large Scale C++" - 3 Updates
- wstring_convert - 1 Update
- Adding a function to std::exception and company - 1 Update
- [Jesus Loves You] Time is approaching - 8 Updates
- Nasty gotcha in template code generation - 2 Updates
| Brian Wood <woodbrian77@gmail.com>: Dec 21 12:26PM -0800 On Monday, November 23, 2015 at 9:23:24 AM UTC-6, Puppet_Sock wrote: > on writing very large C++ programs would be gladly accepted. > If the project I am trying to sell goes ahead, it could easily > become very large. I don't remember seeing this poster recently, but the book arrived a year or so ago. I bought a copy of it and in a nutshell: I don't like how big and heavy it is, but have found a number of ideas that have been helpful to me and am not finished reading it. Also appreciate John for taking some stands that are contrary to the prevailing wisdom. He writes: "In our methodology, we have chosen to treat the valid identifier-character underscore (_) specially, reserving it's use for a higher purpose than mere word separation, for which we use camelCase or CamelCase in most logical names as appropriate." I'm heading in that direction also with my library for a few years now. Brian Ebenezer Enterprises https://webEbenezer.net |
| Brian Wood <woodbrian77@gmail.com>: Dec 21 12:30PM -0800 On Monday, December 21, 2020 at 2:27:04 PM UTC-6, Brian Wood wrote: > "In our methodology, we have chosen to treat the valid > identifier-character underscore (_) specially, reserving > it's use for a higher purpose than mere word separation, Oops, I added the apostrophe there by mistake. Brian Ebenezer Enterprises https://webEbenezer.net |
| Bonita Montero <Bonita.Montero@gmail.com>: Dec 21 09:38PM +0100 The major issue I have with writing large scale C++-applications is that exceptions that might be trown in C++ musst be carefully documented. I wish I'd have checked exceptions like in Java - that's the coronation of clean error-handling (although there are Java application developer hogs that derive exception that should be declared or packaged inside another exception (chaining) from RuntimeException). |
| Bonita Montero <Bonita.Montero@gmail.com>: Dec 21 08:14PM +0100 What are the means to convert UTF-8-strings to u16string-s with C++20. wstring_convert is deprecated. |
| Brian Wood <woodbrian77@gmail.com>: Dec 21 09:37AM -0800 On Saturday, December 19, 2020 at 3:22:42 AM UTC-6, Öö Tiib wrote: > has to have that terminating zero for what() and loggers do simple > linear processing so particular case does not benefit much from > length information. OK, but I don't think I claimed some big benefit. I'm not enthralled with 2020 C++. I mention this in the hopes it could be included in the next standard. > non-virtually (and even potentially inlined). Those occasions do > not allow compilers to remove families of functions from virtual > tables as rest of the calls remain virtual in meaningful program. Are you saying LTO is not relevant to meaningful programs? Brian Ebenezer Enterprises https://github.com/Ebenezer-group/onwards |
| "Chris M. Thomasson" <chris.m.thomasson.1@gmail.com>: Dec 20 03:24PM -0800 On 12/19/2020 6:13 PM, Mr Flibble wrote: >> rather easy to filter? > Oh the spammer Hodgin would probably see that as a victory .. you > putting "Gay Jesus Loves You Up" into one of your configuration files. LOL! One Jesus likes women, the other one likes men. Out of many one... Perhaps the BiJesus? ;^) |
| "Chris M. Thomasson" <chris.m.thomasson.1@gmail.com>: Dec 20 03:25PM -0800 On 12/20/2020 12:01 AM, Jorgen Grahn wrote: > There's already a header for that. It's called "Newsgroups:". > If you want to post nonsense, please remove comp.lang.c++ from > that header. Iirc, RH warned us that he would create a specific header to filter. [Jesus Loves You] ? |
| "Chris M. Thomasson" <chris.m.thomasson.1@gmail.com>: Dec 20 03:28PM -0800 On 12/19/2020 2:36 PM, jacobnavia wrote: > federation to restrain Trump! > I AM THE MASTER OF THIS WORLD! > :-) :^D |
| "Chris M. Thomasson" <chris.m.thomasson.1@gmail.com>: Dec 20 03:29PM -0800 On 12/19/2020 12:01 PM, Chris Vine wrote: > Verily, we are going through a time of great tribulation. We have both > you and Olcott spamming this newsgroup at the same time. > Woe, woe, woe. The Great Conjunction! |
| "Chris M. Thomasson" <chris.m.thomasson.1@gmail.com>: Dec 20 08:13PM -0800 On 12/20/2020 10:51 AM, Rick C. Hodgin wrote: > Not everybody visits "religious" forums. > Not everybody is exposed to knowledge of Jesus in their work or school > or home life. Have you every went to the local dump to tell people about the one true word? |
| gazelle@shell.xmission.com (Kenny McCormack): Dec 21 11:55AM In article <rrp7dj$17oo$1@gioia.aioe.org>, >> or home life. >Have you ever(y) went to the local dump to tell people about the one true >word? He probably has. That is what Xtians are supposed to do - go everywhere where there's even the slightest chance of picking up converts. And, for what it is worth, I actually think that newsgroups like this one are actually pretty good hunting grounds for religious hucksters like Rick. Think about it - geeky (and old) guys discussing some dorky old programming language. -- Joni Ernst (2014): Obama should be impeached because 2 people have died of Ebola. Joni Ernst (2020): Trump is doing great things, because only 65,000 times as many people have died of COVID-19. Josef Stalin (1947): When one person dies, it is a tragedy; when a million die, it is merely statistics. |
| "Öö Tiib" <ootiib@hot.ee>: Dec 21 08:03AM -0800 On Monday, 21 December 2020 at 13:55:33 UTC+2, Kenny McCormack wrote: > Think about it - geeky (and old) guys discussing some dorky old programming language. Also judging by questions posted on Stack Overflow massive amount of utter morons study the languages. |
| Jorgen Grahn <grahn+nntp@snipabacken.se>: Dec 21 05:04PM On Sun, 2020-12-20, Chris M. Thomasson wrote: >> that header. > Iirc, RH warned us that he would create a specific header to filter. > [Jesus Loves You] ? He did. But Rick doesn't decide policy here. When you're posting in those threads, you're just as rude and disruptive as he is. /Jorgen -- // Jorgen Grahn <grahn@ Oo o. . . \X/ snipabacken.se> O o . |
| Vir Campestris <vir.campestris@invalid.invalid>: Dec 21 10:05AM On 19/12/2020 09:47, Öö Tiib wrote: > Compiler vendors do very lot of things that are hard and not required, > but programmers are still expected to pay attention to what is their > fault and what compilers are not required to diagnose. Just FYI the reason this one bit me is that I was cleaning up some code being built for several platforms. What it used to do (and does again; I've given up at least for now) is have a base class with one derived class compiled in for each platform. We've got a company policy of moving to run-time dynamic behaviour where in cases like this all the classes exist, and the correct derived one is instantiated at runtime. The derived class files contain definitions of operator>> to read some config data in - but not for the class itself, but for one of the member data types. That's OK, except that under the hood the way this is called is to use lexical_cast (not Boost's, but something similar) to convert a config line into the data type. And internally that calls the operator>> that is in scope at the time the lexical_cast function was called. It wasn't obvious that this is what was going wrong until I put trace in place, and ran it on one of the platforms that was hitting the problem. The serialisation failed, and it then used default data rather than complaining. Andy |
| "Öö Tiib" <ootiib@hot.ee>: Dec 21 03:00AM -0800 On Monday, 21 December 2020 at 12:05:29 UTC+2, Vir Campestris wrote: > class compiled in for each platform. We've got a company policy of > moving to run-time dynamic behaviour where in cases like this all the > classes exist, and the correct derived one is instantiated at runtime. Odd desire. Platform to what we compile is known compile time so making it run-time dynamic feels absurd waste. > The derived class files contain definitions of operator>> to read some > config data in - but not for the class itself, but for one of the member > data types. Sometimes dynamic polymorphism is used only to improve testability. Through polymorphic interface it is easy to inject dummies, fakes, simulators and mocks for testing. Virtual calls are fine anywhere but in rare most performance-critical places. But that does not mean that all such alternatives may be really compiled into whole product. That is error if done. Also the interface of dynamic polymorphism being more elegant does not mean that static polymorphism should not be used. > place, and ran it on one of the platforms that was hitting the problem. > The serialisation failed, and it then used default data rather than > complaining. Yes added (unneeded) complexity and tricks can always result with added chances of bugs and other confusion. What is decided compile time can be also diagnosed compile time. Compile time diagnosed issues are cheapest to fix. Compiling for all supported platforms takes time but for that there is continuous integration. |
| 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