Hello!
> And whilst great you are fixing up this code, we are but a small fraction > of the world's code > that use java.io and I wonder if this is worth the compatibility risk ? > Why not introducing a separate writeBytes method like it was done for ByteArrayOutputStream (JDK-8199713)? With best regards, Tagir Valeev > -phil > > > > On 7/12/19 12:54 PM, Brian Burkhalter wrote: > > Here is new webrev incorporating the two changes below. > > > > http://cr.openjdk.java.net/~bpb/8187898/webrev.01/ > > > > Thanks, > > > > Brian > > > >> On Jul 12, 2019, at 12:01 PM, Brian Burkhalter < > brian.burkhal...@oracle.com> wrote: > >> > >>> On Jul 12, 2019, at 11:17 AM, Roger Riggs <roger.ri...@oracle.com > <mailto:roger.ri...@oracle.com>> wrote: > >>> > >>> Would it be appropriate to add @Override to the new method (and > perhaps existing overridden methods). > >> Yes, I think so. > >> > >>> Previously, calling FilterOutputStream.write(byte[]) would delegate to > write(byte[], 0, length). > >>> The proposed change duplicates the code and changes the ways that > overridden classes might see the call. > >>> What's the benefit of duplicating the code and calling out.write(buf)? > >> Probably nothing. Probably better to call write(b,0,b.length) directly. > >> > >>> Not my call, but in PSPrinterJob.java (970-977) it might be cleaner to > just delete the unused code and comment. > >> I agree but I was waiting for a suggestion from a 2D reviewer. > >
