home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!gatech!darwin.sura.net!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!ucbvax!VNET.IBM.COM!lybrand
- From: lybrand@VNET.IBM.COM (David Lybrand)
- Newsgroups: comp.sys.pen
- Subject: Re: Slow wake-up with 3125
- Message-ID: <9209011147.AA12933@ucbvax.Berkeley.EDU>
- Date: 1 Sep 92 11:48:10 GMT
- References: <1992Aug31.231535.5596@morrow.stanford.edu>
- Sender: usenet@ucbvax.BERKELEY.EDU
- Organization: IBM Corp, Boca Raton, FL
- Lines: 35
-
- In <1992Aug31.231535.5596@morrow.stanford.edu> Richard Acuff writes:
- >- "PP might be checking for configuration changes (eg. new floppy)."
- >I would suggest that these checks should be done in the background if
- >they take more than .5 seconds (like Iain Cooke).
-
- Currently the PenPoint "resume" is a single process, so it waits for
- each device to reinitialize on the resume call. I've noticed that
- even if the device tells penpoint to stage on interrupt or time,
- the resume process doesn't go on to the next device until that
- device returns "done" on the resume (after the staging). Some rework
- in PenPoint here has potential to speed up the resume process.
-
- >- "PP might be refreshing it's buffers from disk." Why? All the
- >data's still in RAM, right?
-
- On ThinkPad and on GO's Hyde, the "disk" is removable Solid State File
- media. The card could have been removed from the machine and accessed
- in a different machine during standby (just as a diskette could be).
-
- >- "PP is repainting the screen." If the hardware is keeping the screen
- >memory across standby (which seems to be the case), why? If the
- >hardware doesn't save screen memory, wake-up time could be saved by
- >copying out screen memory as part of stand-by and blting it back upon
- >wake-up.
-
- PenPoint does do this (it actually saves the VGA memory to a RAM buffer).
- But the blt does as a second or so to startup time.
-
- >Since we mostly seem to agree that fast wake-up is important, it would
- >be nice to get to the bottom of this so at least we'll know where
- >improvements are needed.
-
- I fully agree. I hope that GO finds the time for performance improve-
- ments in this and other areas.
- DPL
-