Timeline for Simple HTTP Server
Current License: CC BY-SA 4.0
7 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Apr 21, 2019 at 23:46 | comment | added | Edward |
@RolandIllig Project (e.g. boost ) and system (e.g. sys ) are different because they are well known, widely available and stable. I am referring to files that one writes for one's own particular project. I tend to aggregate logical portions of my projects into either shared or static libraries. I find that this makes them easier to thoroughly unit test. For those reason, If I had a file with 50 dependencies, or a project include file named common.h , I would consider it an design failure or an implementation flaw and would rewrite into more granular, better named pieces.
|
|
Apr 21, 2019 at 22:32 | comment | added | Roland Illig |
@Edward Do you also use -I/usr/include/sys so you don't have to write #include <sys/types.h> anymore? In another example, #include <boost/...> clearly says from which project the include file comes. How do you known which of the 50 dependencies provides the <common.h> header?
|
|
Apr 21, 2019 at 2:45 | comment | added | Edward |
Having paths in include files isn’t wrong necessarily but it makes the code less flexible and harder to maintain because it imposes a directory structure that is harder to change. I find it’s better to put that directory structure in one place (e.g. the Makefile ) than spread across every file.
|
|
Apr 21, 2019 at 2:17 | comment | added | Radagast | Thank you for the review! I'm using g++ -I to include paths, but I let the folder's name to be more clarifying, am I wrong? I will make some repairs and I will certainly take your advices, thank you so much! | |
Apr 20, 2019 at 23:49 | history | edited | Edward | CC BY-SA 4.0 |
added 17 characters in body
|
Apr 20, 2019 at 23:16 | comment | added | Konrad Rudolph | I disagree with your first item. And your third item is plain incorrect. | |
Apr 20, 2019 at 22:50 | history | answered | Edward | CC BY-SA 4.0 |