The GTK3 port has been in the making for a very long time. Long before anyone even mentioned GTK4. Porting an application to a different GUI toolkit is a lot of work.
And it shouldn’t be. Sure, there are some new features you may want to take advantage of, but it’s lamentable that GTK doesn’t try harder to maintain backwards compatability.
You know who does major version changes well? Go. Excellent backwards compatible over a decade of very active development, and when there are recommended or required changes, the compiler provides tooling to update source code to the new API.
GTK2->GTK3 was a major leap. For something like a GUI toolkit, changes and advancements are inevitable. A GTK4 port would be much less difficult, as the developer-facing changes are an order of magnitude smaller.
According to the GTK team, trying to maintain backwards compatibility dragged the whole project down. I agree that a basics’ automatic porting tool would’ve been nice.
Yes, it shouldn’t. Unfortunately, the developers of GTK thrived on changes to the API during the GTK3 era. I don’t know why Go devs don’t (and I am indeed very glad that they don’t). Perhaps it’s because of the different structures of the development teams or perhaps because GTK has more hazy goals. 🤷♂️
It makes a lot of sense to use GTK4, but I guess they wanted to respect the chronological order for the GUI library?
The GTK3 port has been in the making for a very long time. Long before anyone even mentioned GTK4. Porting an application to a different GUI toolkit is a lot of work.
And it shouldn’t be. Sure, there are some new features you may want to take advantage of, but it’s lamentable that GTK doesn’t try harder to maintain backwards compatability.
You know who does major version changes well? Go. Excellent backwards compatible over a decade of very active development, and when there are recommended or required changes, the compiler provides tooling to update source code to the new API.
GTK2->GTK3 was a major leap. For something like a GUI toolkit, changes and advancements are inevitable. A GTK4 port would be much less difficult, as the developer-facing changes are an order of magnitude smaller.
According to the GTK team, trying to maintain backwards compatibility dragged the whole project down. I agree that a basics’ automatic porting tool would’ve been nice.
Yes, it shouldn’t. Unfortunately, the developers of GTK thrived on changes to the API during the GTK3 era. I don’t know why Go devs don’t (and I am indeed very glad that they don’t). Perhaps it’s because of the different structures of the development teams or perhaps because GTK has more hazy goals. 🤷♂️