kossebau added a comment.

  In D25545#568090 <https://phabricator.kde.org/D25545#568090>, @dfaure wrote:
  
  > Looks good to me. Friedrich, what did we conclude about deprecating signals?
  
  
  From https://phabricator.kde.org/D24466#547023 & ff. I took with me that 
deprecating signals like methods should be fine, and will work WRT visibility 
and warnings at least with member-function-pointer-based connects.
  
  For the rest looks good also to me. Happy to see that 
ecm_generate_export_header usage can be applied as intended by somebody else 
than me :)
  
  Though one thing (which I also only fixed in later commits) is missing: 
instructing KApiDox & ecm_add_qch about the new deprecation macros (they need 
some help sadly). Cmp. e.g.
  R244:344565e7f6c6782b287d73373e7ce2319c80eab2 
<https://phabricator.kde.org/R244:344565e7f6c6782b287d73373e7ce2319c80eab2> for 
kcoreaddons, and see also 
https://community.kde.org/Policies/Library_Code_Policy#Deprecation_of_API

INLINE COMMENTS

> dialog.h:102
>       * The dialog won't be closed if you setBuffer() in slot connected to 
> this signal
> -     *
> +     * @deprecated Since 5.65, use spellCheckDone
>       * Also emitted after stop() signal

@deprecated Should be last in docs text,
Cmp. also 
https://community.kde.org/Frameworks/Frameworks_Documentation_Policy#Document_Public_and_Protected_Members

REPOSITORY
  R246 Sonnet

REVISION DETAIL
  https://phabricator.kde.org/D25545

To: mlaurent, dfaure, kossebau
Cc: kde-frameworks-devel, LeGast00n, GB_2, michaelh, ngraham, bruns

Reply via email to