On Thu, Jun 26, 2008 at 12:05 PM, Garrett D'Amore <gdamore at sun.com> wrote:
> Mark Martin wrote: > > >> >> Is there any harm in leaving those interface definitions in? I can >> certainly remove that whole "Imported Interfaces" table, but I'm (privately) >> debating the value of and feasibility of a project that might be able to >> scrape some of that information out of the entire public caselog. In this >> case, that's very detailed import information that may get lost. Or perhaps >> scrapping the published caselog documentation has no value and automated >> tools to check the actual code/binaries would be of more value anyway. >> > > There's no harm, that I can see, but little value either. Since the API is > public, we'd never be able to remove or change it anyway, because we > couldn't find all consumers. > > In fact, we can't find such consumers anyway via case logs, since the > normal practice is not to declare public Committed APIs that are imported. Fair enough. Thanks again. Updated revision here -- http://cr.opensolaris.org/~devnull/PSARC/2008/097/psarc_2008_097_draft_opinion.txt -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/opensolaris-arc/attachments/20080626/5869d844/attachment.html>
