Roman Kennke <[EMAIL PROTECTED]> writes: > > The attached patch adds support for CSS borders. This implementation is > actually much better than Sun's because it not only recognizes the > border-width attribute (which Sun doesn't, except for 0 and 1 borders), > it also recognizes the border-top-width, etc for installing different > borders on different edges. I also added support for border-XYZ-color > and border-XYZ-style (both of which Sun doesn't support yet). I am > missing most border styles though (Sun too btw).
Here's a can of worms: Do we actually *want* to be better than Sun's implementation, featurewise? For the same reason that we as Free runtime users and developers hate it when people write code which (inadvertently or not) relies on features that are only in Sun's implementation (eg the undocumented sun.* bits that Free runtimes have no intention to duplicate), IMO it's not such a good idea to end up in a situation where people might inadvertently rely on features that only Classpath provides, making their code not portable to other implementations. It's one thing if we put extra stuff into Classpath-specific packages so that people will need to make a conscious choice to use it. But adding CSS support that will silently fail to work on Sun's implementation seems dangerous. What do others think? Stuart. PS None of this of course is intended in any way to disparage the great work that Roman is doing! :)
