Difference between revisions of "Talk:cpp/language/attributes/deprecated"
| Line 12: | Line 12: | ||
In addition, I 100% agree that "variable" generally includes static data members, but the standard does not always use it consistently that way. For instance, in [[https://eel.is/c++draft/dcl.init.list#note-1 list initialization]] it says that a variable can be list initialized but that clearly does not hold for static data members. | In addition, I 100% agree that "variable" generally includes static data members, but the standard does not always use it consistently that way. For instance, in [[https://eel.is/c++draft/dcl.init.list#note-1 list initialization]] it says that a variable can be list initialized but that clearly does not hold for static data members. | ||
| − | [[User:Hawkinsw|hawkinsw]] ([[User talk:Hawkinsw|talk]]) | + | Update: Sorry! I will continue to think out loud here. In the previous example I specified, I see that they used the word "definition" and that is in distinction to declaration which resolves the conflict that I was seeing in an example such as https://godbolt.org/z/EMscssY6T. |
| + | |||
| + | I guess my overall question is, "If the authors meant it to apply to declarations of both static and non-static data members, why did they specifically add non-static data member to the list? Variable seems like it would have sufficed." | ||
| + | |||
| + | [[User:Hawkinsw|hawkinsw]] ([[User talk:Hawkinsw|talk]]) 12:55, 13 January 2024 (PST)hawkinsw | ||
Revision as of 12:55, 13 January 2024
It looks like during the reformatting between [104386 and 104484], there was a typo introduced in where the [[deprecated]] attribute can appear. According to the standard, [[deprecated]] cannot appear in the declaration of a static data member.
Should this be corrected?
hawkinsw (talk) 23:44, 12 January 2024 (PST)hawkinsw
- A static data member is a variable, and thus can have this attribute.
- The standard says "a variable or a non-static data member" which is a bit misleading, but it actually does include static data members. --D41D8CD98F (talk) 00:19, 13 January 2024 (PST)
Thanks! The only reason I mention this is because there is a paper making its way through the committee that disagrees with that interpretation: [P2552]. I have emailed the author about the discrepancy.
In addition, I 100% agree that "variable" generally includes static data members, but the standard does not always use it consistently that way. For instance, in [list initialization] it says that a variable can be list initialized but that clearly does not hold for static data members.
Update: Sorry! I will continue to think out loud here. In the previous example I specified, I see that they used the word "definition" and that is in distinction to declaration which resolves the conflict that I was seeing in an example such as https://godbolt.org/z/EMscssY6T.
I guess my overall question is, "If the authors meant it to apply to declarations of both static and non-static data members, why did they specifically add non-static data member to the list? Variable seems like it would have sufficed."
