Registered Member
|
Currently, developers may to provide include list by themselves (it apply to C/C++ projects for example).
But these information could been obtained by pkg-config or others. Solution? Best way to obtain this information is run project build once user click on special button in project's include directories config page. This button will install special bash script (apply only to container, when not in container, we can use already installed one), which read pipe_output env variable, create pipe in pipe_output path to read information about needed include. Installed scripts will be compiler/linker wrappers, such like: gcc, g++, ld, etc. Once wrapper is called, it recognize switches responsible for setting includes, write it to created pipe and pass entire argument list to correct executable (gcc, g++, ld, etc.). Of course, we skip 0 argument. Why it is good solution? Because there is no problem with try to build project to gather project-related information in my opinion. Project should be able to build multiple times, no matter if it is already build. And of course - KDevelop would not do anything by themselves - user may to click special button to add additional include options.
Lachu, proud to be a member of KDE forums since 2008-Nov.
|
Registered Member
|
Does it really make sense to invest resources into kdevelop?
Don't get me wrong. I used kdevelop quite a lot. It was a great IDE especially because of its integration into the desktop. But that was long, long ago. kdevelop is unable to keep pace, it has fallen far behind and that is nothing that can be solved by implementing a few additional features. Today's IDEs, especially the Jetbrains variants are what is used for > 75% of all software projects. Reason for that? Not monopolistic structures but simply modern, fast adaption of changes in the programming landscape, a mighty ecosystem and most of all specialist IDEs for each purpose. That keeps IDEs usable, once can focus on the topic at hand instead of having to work through endless settings and features not at all useful for the current project. Which is why I would recommend to finally put kdevelop to a well deserved rest and concentrate on other, more pressing matters. |
Registered users: bartoloni, Bing [Bot], Google [Bot], Yahoo [Bot]