home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!fuug!funic!sauna.cs.hut.fi!cs.hut.fi!jtp
- From: jtp@cs.hut.fi (Jukka Partanen)
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: LockIBase
- Date: 30 Dec 1992 21:57:59 GMT
- Organization: Helsinki University of Technology, Finland
- Lines: 26
- Distribution: inet
- Message-ID: <JTP.92Dec30235759@cardhu.cs.hut.fi>
- References: <JTP.92Dec27193513@laphroaig.cs.hut.fi> <paulk.2z9x@terapin.com>
- NNTP-Posting-Host: cardhu.cs.hut.fi
- In-reply-to: paulk@terapin.com's message of 29 Dec 92 00:22:35 GMT
-
- In article <paulk.2z9x@terapin.com> paulk@terapin.com (Paul Kienitz) writes:
-
- > I ended up
- > working around this in my personal DMouse variant by grabbing the
- > MouseX and MouseY fields of the screen itself, so it activates the
- > window where the mouse was as of the PREVIOUS input event instead of
- > the current one.
-
- Well, that's about the same approach I use, and I've found it
- to be adequate. What I really would like to know is the
- correct locking protocol. First I LockIBase to browse through
- the screens and find the correct one, but then I must release
- to IBase lock to call WhichLayer. But the owner of the screen
- may close it after the IBase is released, so I must Forbid
- before unlocking the IBase and Permit after finding and
- activating the window. Yuck, I say. Isn't there a better way?
-
- > Does anyone know a way (preferably forward-compatible) to convert
- > arbitrary input event coordinates to screen coordinates?
-
- If you find out a cool way, tell me too!
-
- jtp
-
- Zippy quotation:
- YOW!! The land of the rising SONY!!
-