This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author serhiy.storchaka
Recipients serhiy.storchaka
Date 2017-10-29.09:21:27
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1509268888.25.0.213398074469.issue31891@psf.upfronthosting.co.za>
In-reply-to
Content
NetBSD perhaps is the last open OS that don't use ncurses. For now the _curses module is not compiled on NetBSD. Its curses implementation doesn't contain several ncurses functions. <term.h> included for non-ncurses implementations defines a couple of macros that conflict with local variable names. The hardcoded signature of setupterm differs from the signature in <term.h>. The fallback implementation of has_key() depends on KEY_* constants that are absent in NetBSD curses too. It seems to me that curses is broken on NetBSD for very long time.

The proposed PR fixes building of the _curses module and makes curses tests passing. This of course doesn't guaranties that curses works correctly on NetBSD, our curses tests are rudimentary.

There were other NetBSD specific guards in the code. They are no longer needed because the guarded functions already are supported in NetBSD for long time. There are issue9667 and issue21457 for removing unneeded conditional compilation.
History
Date User Action Args
2017-10-29 09:21:28serhiy.storchakasetrecipients: + serhiy.storchaka
2017-10-29 09:21:28serhiy.storchakasetmessageid: <1509268888.25.0.213398074469.issue31891@psf.upfronthosting.co.za>
2017-10-29 09:21:28serhiy.storchakalinkissue31891 messages
2017-10-29 09:21:27serhiy.storchakacreate