The Big Programming Thread - Page 956

Prev 1 954 955 956 957 958 1032 Next

Thread Rules
1. This is not a "do my homework for me" thread. If you have specific questions, ask, but don't post an assignment or homework problem and expect an exact solution.
2. No recruiting for your cockamamie projects (you won't replace facebook with 3 dudes you found on the internet and $20)
3. If you can't articulate why a language is bad, don't start slinging shit about it. Just remember that nothing is worse than making CSS IE6 compatible.
4. Use [code] tags to format code blocks.

Nars_

31 Posts

April 21 2018 00:33 GMT

#19101

delete this post

Silvanel

Poland4756 Posts

April 21 2018 07:32 GMT

#19102

On April 20 2018 23:35 Excludos wrote:

Show nested quote +

Or the quote my boss gave me when I told him we desperately needed to implement some unit testing in our consistently breaking system. "You should have just made it better to begin with" and "We don't have time to do testing right now, the system needs to work first". The irony of the last statement was lost on him.

On a completely different topic: I'm starting in a new job next week xD

GZ on the new job.

In regards to "testing" this approach is only possible in small companies/projects where there might be no place for dedicated tester. In the end that means developers end up being their own testers. Which isnt very effective.
The larger the project, the more people involved, the larger and more numerous dependencies the greater is need for good testing strategy. And that usually involves people dedicated only to testing. Not to mention that if Your customer is another company with at least a little bit of knowledge about SW development they might just ask "What is Your testing strategy", and they wont take "we just write good code" for an answer.

bo1b

Australia12814 Posts

April 21 2018 08:29 GMT

#19103

Off topic question, but where do you guys think the future of computer science education will go. Currently it is possible to get a job of varying levels without a university education, some things are locked behind doors for the most part, such as machine learning jobs, or computer vision, or other things with a high level of research behind it.

Yet I go online and I find a variety of moocs with what looks like fairly excellent layouts for a cs education, such as:

https://github.com/P1xt/p1xt-guides/blob/master/cs-java-focus.md
https://github.com/ossu/computer-science
various online university courses, such as cs 50x

Do you think that completing some of those pathways will ever lead to a more difficult to get to job? Completing the ossu from start to finish is no joke, that really closes in on a university level education, plus a bit.

Silvanel

Poland4756 Posts

April 21 2018 08:52 GMT

#19104

Well from my experience the knowledge is held by specific companies not universisties (unless we talk about pure research stuff), there are some areas which only select few companies develop and only way inside such field is to join such company. And once inside the bar for changing position and learning is significantly lower.

sc-darkness

856 Posts

April 21 2018 10:09 GMT

#19105

I expect to re-learn quite a lot of basic stuff once quantum computer is mainstream. Will it invalidate a lot of our existing knowledge taught at computer science courses? Will our CS degrees be almost worthless?

Excludos

Norway8257 Posts

April 21 2018 10:18 GMT

#19106

On April 21 2018 16:32 Silvanel wrote:

Show nested quote +

I disagree wholeheartedly. The most effective tester is the programmer himself. He knows exactly what every piece of his code/program is suppose to do, and can quickly dive in and fix it. That doesn't mean you don't need external testers as well, but those as more to test the program from a user's point of view rather than a tech savvy programmer's point of view (As well for the fact that a dedicated tester will have a lot more time for exactly that, while a programmer also has to..program). And all programmers do test, even if it's a bare minimum: They compile the program after writing a piece of code. But you could do so much more without eating into your feature list schedule, and automated testing is excellent for that. Unit testing and front end testing tools like Cypress are golden for making sure that all your user stories works like they are suppose to. So whenever your/my boss tells you that "There's no time for testing because nothing is working", he is showing a fundamental failure to understand what testing does: Making sure that stuff works. At that point the only solution is to find another job. Because I sure as hell aren't going to push untested features which could literally end up killing people or animals (We have already done the latter and have had some damn close near misses on humans too. I won't stick around for it)

Silvanel

Poland4756 Posts

April 21 2018 14:31 GMT

#19107

Well You can go to jail for something like that (like someone who makes shitty bridge and ends up killing people) so yeah better to avoid something like that. Not to mention it is unethical.

ShoCkeyy

7815 Posts

April 21 2018 17:15 GMT

#19108

I'm a programmer who works on strategy/architect for improved UX/UI and conversion now, with no CS degree (Thanks for that resource b1ob, taking a look), but have worked for multiple fortune 500 companies. The next thing is testing, I test all the time, I test other peoples code as well. Testing is important, just because you built it, doesn't mean you know everything. I've built many things that I figured always worked, but when I started to user test my applications, I quickly realized how users tend to break things anyways. So in the end, testing is extremely important. And don't always test yourself, you want others to test for you too.

Adding an extra edit: I do however believe getting a degree helps in a sense of experience building from nothing when you know nothing. It took me longer learning to build from scratch thats for sure.

sc-darkness

856 Posts

April 22 2018 20:51 GMT

#19109

On April 19 2018 02:48 sc-darkness wrote:
Ok, what the fuck is the problem with std::string and std::wstring on Linux and Windows?

I see that std::string is somehow magically "UTF-8 ready" on Linux. On the contrary, Windows prefers std::wstring for unicode. Yet, std::wstring is 4 bytes on Linux and only 2 bytes on Windows. How does that make std::wstring less desirable on Linux? Why? Clearly wstring is larger so it could hold more characters? And why is this fucked up on an OS level? I just don't understand all this mess.

Link: https://stackoverflow.com/questions/402283/stdwstring-vs-stdstring?utm_medium=organic&utm_source=google_rich_qa&utm_campaign=google_rich_qa

Edit: Are we meant to use these now?

Show nested quote +

Fuck the hacky 8-bit ASCII (a.k.a Windows code pages) backwards compatibility support on Windows and the shit it caused to Unicode support.

Code:
+ Show Spoiler +


#include <string>
#include <iostream>

int main()
{
   std::string chinese_str = "新年快乐 / 新年快樂";
   std::cout << chinese_str << std::endl;

   std::cout << "Size of char: " << sizeof(char) << std::endl;
   std::cout << "Size of wchar_t:  " << sizeof(wchar_t) << std::endl;

   return 0;
}

Linux:
+ Show Spoiler +

新年快乐 / 新年快樂
Size of char: 1
Size of wchar_t: 4

Linux uses UTF-8 by default.

Windows:
+ Show Spoiler +

???? / ????
Size of char: 1
Size of wchar_t: 2

Warnings:
+ Show Spoiler +

warning C4566: character represented by universal-character-name '\u65B0' cannot be represented in the current code page (1251)
warning C4566: character represented by universal-character-name '\u5E74' cannot be represented in the current code page (1251)
warning C4566: character represented by universal-character-name '\u5FEB' cannot be represented in the current code page (1251)
warning C4566: character represented by universal-character-name '\u4E50' cannot be represented in the current code page (1251)
warning C4566: character represented by universal-character-name '\u6A02' cannot be represented in the current code page (1251)

Windows uses UTF-16 by default but only for wstring not for string.

If we replace string with wstring, it still doesn't work with wcout. More work is necessary to make Windows Command Prompt to display it for you. Thank you Microsoft for screwing Unicode support. :D

Excludos

Norway8257 Posts

April 22 2018 21:16 GMT

#19110

On April 23 2018 05:51 sc-darkness wrote:

Show nested quote +

Fuck the hacky 8-bit ASCII (a.k.a Windows code pages) backwards compatibility support on Windows and the shit it caused to Unicode support.

Code:
+ Show Spoiler +


#include <string>
#include <iostream>

int main()
{
   std::string chinese_str = "新年快乐 / 新年快樂";
   std::cout << chinese_str << std::endl;

   std::cout << "Size of char: " << sizeof(char) << std::endl;
   std::cout << "Size of wchar_t:  " << sizeof(wchar_t) << std::endl;

   return 0;
}

Linux:
+ Show Spoiler +

新年快乐 / 新年快樂
Size of char: 1
Size of wchar_t: 4

Linux uses UTF-8 by default.

Windows:
+ Show Spoiler +

???? / ????
Size of char: 1
Size of wchar_t: 2

Warnings:
+ Show Spoiler +

Did I mention that I lo-o-o-ove Qt? All of these shitstained compatibility issues worked out behind my back without me having to worry about it.

sc-darkness

856 Posts

April 22 2018 21:29 GMT

#19111

Well, I have to maintain MFC at a second job in a row. Oh well, we plan to get rid of it at some point.

Khalum

Austria831 Posts

April 22 2018 22:49 GMT

#19112

On April 23 2018 05:51 sc-darkness wrote:

Show nested quote +

Fuck the hacky 8-bit ASCII (a.k.a Windows code pages) backwards compatibility support on Windows and the shit it caused to Unicode support.

Code:
+ Show Spoiler +


#include <string>
#include <iostream>

int main()
{
   std::string chinese_str = "新年快乐 / 新年快樂";
   std::cout << chinese_str << std::endl;

   std::cout << "Size of char: " << sizeof(char) << std::endl;
   std::cout << "Size of wchar_t:  " << sizeof(wchar_t) << std::endl;

   return 0;
}

Linux:
+ Show Spoiler +

新年快乐 / 新年快樂
Size of char: 1
Size of wchar_t: 4

Linux uses UTF-8 by default.

Windows:
+ Show Spoiler +

???? / ????
Size of char: 1
Size of wchar_t: 2

Warnings:
+ Show Spoiler +

I ported a ton of code to use boost::filesystem::path for paths instead of a seemingly random combination of char*, std::string, QString, ... last year. I don't know how I retained my sanity.

Excludos

Norway8257 Posts

April 23 2018 07:34 GMT

#19113

On April 23 2018 07:49 Khalum wrote:

Show nested quote +

I ported a ton of code to use boost::filesystem::path for paths instead of a seemingly random combination of char*, std::string, QString, ... last year. I don't know how I retained my sanity.

Why not use QPath or QFile when you already have Qt installed? Not that boost is bad, but I don't see a reason to use both of them as, to my knowledge, Qt covers everything boost does.

ddengster

Singapore129 Posts

April 23 2018 07:43 GMT

#19114

On April 23 2018 05:51 sc-darkness wrote:

Show nested quote +

Fuck the hacky 8-bit ASCII (a.k.a Windows code pages) backwards compatibility support on Windows and the shit it caused to Unicode support.

Code:
+ Show Spoiler +


#include <string>
#include <iostream>

int main()
{
   std::string chinese_str = "新年快乐 / 新年快樂";
   std::cout << chinese_str << std::endl;

   std::cout << "Size of char: " << sizeof(char) << std::endl;
   std::cout << "Size of wchar_t:  " << sizeof(wchar_t) << std::endl;

   return 0;
}

Linux:
+ Show Spoiler +

新年快乐 / 新年快樂
Size of char: 1
Size of wchar_t: 4

Linux uses UTF-8 by default.

Windows:
+ Show Spoiler +

???? / ????
Size of char: 1
Size of wchar_t: 2

Warnings:
+ Show Spoiler +

Read up utf8everywhere

As a rule of thumb, you use should using utf8 (std::string or char, preferably char arrays

) instead of wchar types in your code; only convert them when the windows API demands it. UTF8 is an encoding that contains all the ascii character plus the extended ones, see #7 of the article in the link. UTF16 also has a number of disadvantages, and the article lists that down as well.

There are open source code/libraries that do the conversion from utf8 to utf16 and vice versa, go find them. If you introduce another wstring type in your code, you'll have to spend a whole bunch of time writing versions of code for wstring types.

Edit: Also, do note that you need to adjust accordingly if you do string manipulation or want to use string lengths with utf8 characters.

nunez

Norway4003 Posts

April 23 2018 10:08 GMT

#19115

On April 23 2018 16:34 Excludos wrote:

Show nested quote +

On April 23 2018 07:49 Khalum wrote:

On April 23 2018 05:51 sc-darkness wrote:

3) UTF-8 encoded string literal. The type of a u8"..." string literal is const char[].
4) UTF-16 encoded string literal. The type of a u"..." string literal is const char16_t[].
5) UTF-32 encoded string literal. The type of a U"..." string literal is const char32_t[].

Fuck the hacky 8-bit ASCII (a.k.a Windows code pages) backwards compatibility support on Windows and the shit it caused to Unicode support.

Code:
+ Show Spoiler +


#include <string>
#include <iostream>

int main()
{
   std::string chinese_str = "新年快乐 / 新年快樂";
   std::cout << chinese_str << std::endl;

   std::cout << "Size of char: " << sizeof(char) << std::endl;
   std::cout << "Size of wchar_t:  " << sizeof(wchar_t) << std::endl;

   return 0;
}

Linux:
+ Show Spoiler +

新年快乐 / 新年快樂
Size of char: 1
Size of wchar_t: 4

Linux uses UTF-8 by default.

Windows:
+ Show Spoiler +

???? / ????
Size of char: 1
Size of wchar_t: 2

Warnings:
+ Show Spoiler +

I ported a ton of code to use boost::filesystem::path for paths instead of a seemingly random combination of char*, std::string, QString, ... last year. I don't know how I retained my sanity.

Why not use QPath or QFile when you already have Qt installed? Not that boost is bad, but I don't see a reason to use both of them as, to my knowledge, Qt covers everything boost does.

boost::filesystem was merged into standard c++ since c++17.

Khalum

Austria831 Posts

April 23 2018 11:22 GMT

#19116

On April 23 2018 19:08 nunez wrote:

Show nested quote +

On April 23 2018 16:34 Excludos wrote:

On April 23 2018 07:49 Khalum wrote:

On April 23 2018 05:51 sc-darkness wrote:

Fuck the hacky 8-bit ASCII (a.k.a Windows code pages) backwards compatibility support on Windows and the shit it caused to Unicode support.

Code:
+ Show Spoiler +


#include <string>
#include <iostream>

int main()
{
   std::string chinese_str = "新年快乐 / 新年快樂";
   std::cout << chinese_str << std::endl;

   std::cout << "Size of char: " << sizeof(char) << std::endl;
   std::cout << "Size of wchar_t:  " << sizeof(wchar_t) << std::endl;

   return 0;
}

Linux:
+ Show Spoiler +

新年快乐 / 新年快樂
Size of char: 1
Size of wchar_t: 4

Linux uses UTF-8 by default.

Windows:
+ Show Spoiler +

???? / ????
Size of char: 1
Size of wchar_t: 2

Warnings:
+ Show Spoiler +

I ported a ton of code to use boost::filesystem::path for paths instead of a seemingly random combination of char*, std::string, QString, ... last year. I don't know how I retained my sanity.

Why not use QPath or QFile when you already have Qt installed? Not that boost is bad, but I don't see a reason to use both of them as, to my knowledge, Qt covers everything boost does.

boost::filesystem was merged into standard c++ since c++17.

Aye. We're on c++11 but once we decide to switch we can use std::filesystem functionality -> less 3rd party dependencies -> good.

nunez

Norway4003 Posts

April 23 2018 11:37 GMT

#19117

yes, exactly.

sc-darkness

856 Posts

April 23 2018 17:36 GMT

#19118

Wasn't there a plan to support networking too? Shame it didn't make it.

WarSame

Canada1950 Posts

April 30 2018 03:39 GMT

#19119

I need a lot of help understanding Dagger 2. I've read a lot of guides and I don't understand DI in general.

You have an object you want to inject. You Provide it with @Provides inside of an @Module. You then use it somewhere else with @Inject. Is this correct so far?

If it is I'm failing to apply it to my cause and getting errors that aren't google-able.

In my case I am trying to inject an instance of a javax.crypto.KeyGenerator. I created a Module AppComponent with an @Provides KeyGenerator provideKeyGenerator(String walletName). I then want to @Inject this into another Class with certain parameters but I can't get that to work. I've tried a number of things, but keep running into problems injecting it, because it runs into problems injecting into a static field.

Does anyone have a good, basic guide for doing something like this? I feel like I'm running into a wall with this. I'll try to post some sample code tomorrow night to give this post an actual purpose.

Danehardin

0 Posts

April 30 2018 21:51 GMT

#19120

--- Nuked ---

Prev 1 954 955 956 957 958 1032 Next

Please or register to reply.

The Big Programming Thread - Page 956

Completed

Ongoing

Upcoming