Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!sun-barr!decwrl!shelby!portia!bir7 From: bir7@portia.Stanford.EDU (Ross Biro) Newsgroups: comp.unix.i386 Subject: Re: ESIX Networks and X Summary: Good Product But It has a couple of Problems Keywords: ESIX X Message-ID: <7294@portia.Stanford.EDU> Date: 5 Dec 89 01:37:59 GMT References: <10855@megatest.UUCP> Sender: Ross Biro Reply-To: bir7@portia.Stanford.EDU (Ross Biro) Organization: Stanford University Lines: 40 I'm running ESIX 3.2 rev c on a 386/20 with 4 meg of page memory, a WD1007-WA2 ESDI controller connected to a 340 meg drive. Overall ESIX has performed well, but I have found a few problems: With 4 meg of memory, X is all but unuseable. I can run one copy of xterm without serious problems but after that memory becomes full and everytime I move the mouse I have to wait for something to be swapped. The networking software does not support domain resolution (i.e. my machine is connected to internet, but I can only ftp etc. to .stanford.edu machines.) The dos support (mdir etc.) only supports 12 bit FATs so it is all but impossible to access a dos partition on a hard drive. The floppy driver doesn't sense the media type on 3.5" disks, it just assumes that they are the maximum density that the drive will support, thus it is impossible to use 720k disks in a 1.44 meg drive without changing your setup parms, and the 1.44 meg disks are treated as 1.2 meg disks so they will not be compatible with dos. There may be a problem with interactions between the select function and an Alarm handler. This doesn't cause problems with any of the software supplied with ESIX but it may cause problems when porting software from bsd boxes. ---- The problem with the 12 bit FATs and the floppy driver were told to me by ESIX technical support so they are aware of them, but I am under the impression that they have no plans for fixes in the near future. The technical support people were unaware of any problem with select, so it could just be my imagination (but I did send them some code which I think demonstrates the problem.) The only other problems I have encountered when running ESIX are do to lack of memory. When I compiled the GNU c compiler the make took three days while most of the CPU time was spent swapping pages to and from the hard disk. I think with another 4 meg it would have taken about 2-3 hours. Ross Biro bir7@portia.stanford.edu (uunet!stanford!portia!bir7)