Checking the API status for the variable “Private d” of the QVariant class
-
Are we complaining about loss of performance due to an additional integer increment and decrement?
This can occasionally happen, can't it?
…, what is causing your aversion to how QVariant works?
I got special software development opinions. The QVariant class is useful. I am just looking for a few extensions there.
… but you mention you are using QStandardItemModel …
This is Qt standard software which provides known functionality. I would like to apply it as it would mostly be usual.
-
@elfring said in Checking the API status for the variable “Private d” of the QVariant class:
This can occasionally happen, can't it?
Can, should and will.
I got special software development opinions.
Forcing your opinion on the framework is possible but don't expect it to be easy.
Forcing the framework to adopt those opinions is a titanic endeavour.This is Qt standard software which provides known functionality.
True and I always suggest to use it as it's super convenient but it must be clear that it's very far from being an implementation of a model optimised for performance
-
-
@elfring said in Checking the API status for the variable “Private d” of the QVariant class:
… I always suggest to use it as it's super convenient but it must be clear that it's very far from being an implementation of a model optimised for performance
Will such information trigger any further software evolution?
That's unlikely.
If you want performance, you have more abstract classes at your disposal, where more coding is necessary (less convenient) but more fine control over details is possible.
-
If you want performance, you have more abstract classes at your disposal,
How can they really help here?
where more coding is necessary (less convenient) but more fine control over details is possible.
Can it eventually help more to improve (or extend) also the implementation of the class “QVariant”?
-
@elfring said in Checking the API status for the variable “Private d” of the QVariant class:
If you want performance, you have more abstract classes at your disposal,
How can they really help here?
where more coding is necessary (less convenient) but more fine control over details is possible.
Can it eventually help more to improve (or extend) also the implementation of the class “QVariant”?
Feel free to improve QVariant, if you have the means. Qt is an open project, if your contribution is judged beneficial, it will be included upstream.
-
-
@elfring said in Checking the API status for the variable “Private d” of the QVariant class:
Feel free to improve QVariant, if you have the means.
How do you think about to support the addition of a member function like “(std::)get”?
I'm not a Qt developer. I'm not sure if STD is allowed in public Qt API. And lastly, I don't care about std::get ;-) But I may be alone in this.
-
And lastly, I don't care about std::get
Will you care more for recent software extensions from the standard “ISO/IEC 14882” in the near future?
-
@elfring said in Checking the API status for the variable “Private d” of the QVariant class:
And lastly, I don't care about std::get
Will you care more for recent software extensions from the standard “ISO/IEC 14882” in the near future?
What does that have to do with this topic? Are you interviewing me for a job now or something? :-)
Sorry mate but I see no point in continuing this conversation.
-
Sorry mate but I see no point in continuing this conversation.
Some software developers agreed that they need the class “variant” for working with type-safe unions. They chose another useful programming interface. These software design decisions might influence other class libraries, don't they?
The Qt software library is using also a variant class for the support of the meta-object protocol, isn't it?
