Here's my recommendation Phil: Rename the .as files that have Class in them to .lzs files (use `svn move`) and rewrite them to use class declarations as you have elsewhere.
Note that you will have to change the library.lzs files that include the renamed files. You could put a comment there, and in the file, noting that although it is written in lzs dialect it is a (swf) platform-specific file. On 2006-09-06, at 20:58 EDT, Max Carlson wrote: > I'm with Henry. Let's get rid of the shims, and worry about > rewriting the swf kernel later. > > -Max > > Henry Minsky wrote: >> On 9/6/06, *P T Withington* <[EMAIL PROTECTED] >> <mailto:[EMAIL PROTECTED]>> wrote: >> [Adding max, henry and laszlo-dev] >> On 2006-09-06, at 15:34 EDT, Philip Romanik wrote: >> > Hi Tucker, >> > >> > There are some references to Class from the swf kernel, as >> well as >> > some *.as files that aren't in the kernel ( LzConnection.as >> for >> > example). Does this mean that we need the shim for a while? >> Interesting. >> Our policy is that the kernel API cannot depend on using our >> class >> model. But I guess that does not preclude using it internally >> to a >> particular kernel. >> We really do want to remove the shim, because it is unnecessary >> overhead. >> The question is, do we not want to use the class declaration >> in .as >> files and/or the kernel? If not, then these need to be >> rewritten in >> js1/as2 classic constructor/prototype pattern. >> I think that it is ok to use the class system in the swf kernel, I >> think of it as "grandfathered in". >> Someone writing a new kernel won't use it, probably, but why kill >> ourselves to massively rip up working code. >> But we ought to rewrite the sources to use the new class syntax, >> so we can get rid of the shim. >> Can you give us a list of the files, whether they are in the >> kernel >> or not, and what classes they involve? Then I'll ask Max and >> Henry >> to weigh in on how these should be converted. I'm guessing >> some will >> want to become .lzs files and rewritten to use class >> declarations, >> and some will have to be rewritten in classic style. >> -- >> Henry Minsky >> Software Architect >> [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> _______________________________________________ Laszlo-dev mailing list [email protected] http://www.openlaszlo.org/mailman/listinfo/laszlo-dev
