On Thu, Feb 15, 2007 at 04:11:01PM -0800, Mike Kupfer wrote: > >>>>> "sch" == Stephen Hahn <[EMAIL PROTECTED]> writes: > > sch> I agree with basing the operation on the "putback -n" result, > sch> unless Mike was after some specific state from wx? > > I would rather not re-invent, or even borrow, the code from wx to parse > "putback -n" output. I'm fine with doing a "wx update" at the start of > the script in order to get the most recent information. > > While it's true that wx is not required for general ON development, I > believe it is widely enough used, and sufficiently robust, that it's > reasonable to require its use for this conversion script. If necessary, > the script can offer to run "wx init" for the user if there's no wx > state. > > Of course, this is all assuming I write the script. If someone else > wants to write it, I certainly have no objections to their parsing > the output from "putback -n" in the conversion script.
While I'm not volunteering at this point to write it either I will say that wx is not very efficient (a result of hackery and paranoia). It would be smart to look at the wx_update() function to see if a more efficient way of determining workspace change can be created from wx. For example there is a call to deal_ws_renames() which doesn't really need to be done if the putback -n is going to happen. BTW, if I had know the insane about of mods I would end up making to wx I would have recoded the whole thing in perl. Something to think about. -- Will Fiveash _______________________________________________ tools-discuss mailing list tools-discuss@opensolaris.org
