Peter Donald wrote: > > On Mon, 12 Nov 2001 20:22, Ulrich Mayring wrote: > > Peter Donald wrote: > > > Hi, > > > > > > Whats everyone think about renaming server.xml to environment.xml ? > > > > Sounds reasonable to me. > > > > While we're on the subject of naming, what is *.xinfo supposed to > > achieve? Why not give it an XML suffix like all other files? Instead of > > myblock.xinfo we could name it myblock-info.xml or even myblock.xml. > > no idea. It was just what we decided at the time. I can't see anything wrong > with that though.
Well, we have three config files with an .xml extension. If we introduce a fourth, then for the sake of simplicity, the default should be to give it a name consistent with what's already there. If there is a compelling reason for a different naming scheme in this case, then the default can be overridden. > > I know that *.xinfo is > > for the block and assembly.xml for the .sar application. But it seems to > > me that in every .sar application I write, there is kind of an aggregate > > of the various *.xinfo files in my assembly.xml. It would be nicer, if > > this aggregation of information would be done automatically or not at > > all. > > again - need more context to figure out which part you are referring to. The *.xinfo file specifies which service this block offers and which services it depends on. The assembly.xml file specifies which blocks are included in the .sar application. For each block it is specified which class this block offers and which classes it depends on. Now, we have a difference here in that the services specify an interface and the classes an implementation. There could be several implementations of an interface - but there can be only one interface for an implementation. So once you specified an implementation it is IMHO redundant to specify its interface, no? Ulrich -- Ulrich Mayring DENIC eG, Systementwicklung -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
