ROsman%ASS%SwRI05@D15VS178A.SPACE.SWRI.EDU (08/31/90)
Is there any such thing? I know that there's a number of Public Domain Drivers. Is the NFS spec public? Thanks with putting up with my stupidity... Oz
srodawa@vela.acs.oakland.edu (Dr. Srodawa) (09/01/90)
In article <9008311436.aa19497@louie.udel.edu> ROsman%ASS%SwRI05@D15VS178A.SPACE.SWRI.EDU writes: >Is there any such thing? I know that there's a number of Public Domain >Drivers. Is the NFS spec public? > >Thanks with putting up with my stupidity... > >Oz There didn't USED to be. But, the 4.3BSD-reno distribution claims to have one. They state (paraphrasing now) in the release notes that it was written from scratch using only publicly available protocol documentation. -- | Ronald J. Srodawa | Internet: srodawa@unix.secs.oakland.edu | | School of Engineering and CS | UUCP: srodawa@egrunix.UUCP | | Oakland University | Voice: (313) 370-2247 | | Rochester, Michigan 48309-4401 | |
epsilon@wet.UUCP (Eric P. Scott) (09/03/90)
In article <9008311436.aa19497@louie.udel.edu> ROsman%ASS%SwRI05@D15VS178A.SPACE.SWRI.EDU writes: >Is there any such thing? I know that there's a number of Public Domain >Drivers. Is the NFS spec public? There are "free" NFS servers, but I've yet to see a PD NFS client. The spec is public--RFC 1094+RFC 1057+RFC 1014. -=EPS=-
srodawa@vela.acs.oakland.edu (Dr. Srodawa) (09/05/90)
In article <1491@wet.UUCP> epsilon@wet.UUCP (Eric P. Scott) writes: >In article <9008311436.aa19497@louie.udel.edu> > ROsman%ASS%SwRI05@D15VS178A.SPACE.SWRI.EDU writes: >>Is there any such thing? I know that there's a number of Public Domain >>Drivers. Is the NFS spec public? > >There are "free" NFS servers, but I've yet to see a PD NFS >client. The spec is public--RFC 1094+RFC 1057+RFC 1014. > > -=EPS=- This was true until a few weeks ago. In the public announcement for BSD4.3-reno it is stated that a PD NFS is included--both client and server. It was written independently using the RFCs as the only source of information. -- | Ronald J. Srodawa | Internet: srodawa@unix.secs.oakland.edu | | School of Engineering and CS | UUCP: srodawa@egrunix.UUCP | | Oakland University | Voice: (313) 370-2247 | | Rochester, Michigan 48309-4401 | |
epsilon@wet.UUCP (Eric P. Scott) (09/07/90)
In article <2884@vela.acs.oakland.edu> srodawa@vela.acs.oakland.edu (Dr. Srodawa) writes: >This was true until a few weeks ago. In the public announcement for >BSD4.3-reno it is stated that a PD NFS is included--both client and server. That's all fine and well, but it probably doesn't help THIS audience. -=EPS=-
ROsman%ASS%SwRI05@D15VS178A.SPACE.SWRI.EDU (09/29/90)
Did I miss it, did someone actually answer the original question? The original question(s) were: (1) Does a PD version of Sun's PC-NFS Client exist? (2) Where can I FTP it from. I saw a message that indicated that it was included with 4.3BSD-reno, but rooting around in UCBARPA and UCBVAX hasn't revealed anything. Please copy me on any response, we're having some mail troubles. Oz (Rich Osman, WB0HUQ) INTERNET: ROsman%ASS%SwRI05@D15VS178A.SPACE.SwRI.EDU (512) 522-5050 (w); (512) 699-1302 (h, merciless machine) (512) 522-2572 (just the fax)
srodawa@vela.acs.oakland.edu (Dr. Srodawa) (10/02/90)
In article <9009282024.aa21819@louie.udel.edu> ROsman%ASS%SwRI05@D15VS178A.SPACE.SWRI.EDU writes: >Did I miss it, did someone actually answer the original question? The original >question(s) were: > > (1) Does a PD version of Sun's PC-NFS Client exist? > > (2) Where can I FTP it from. > >I saw a message that indicated that it was included with 4.3BSD-reno, but >rooting around in UCBARPA and UCBVAX hasn't revealed anything. Please copy me >on any response, we're having some mail troubles. YES! The Berkeley announcement of 4.3BSD-reno stated that it includes a PD client and server for NFS. But, they aren't releasing portions of 4.3BSD-reno separately..you have to buy the package. The cost is low (to universities) and requires an existing 4.3BSD license. One would have to implement the BSD interface used by the NFS code to complete a PC-NFS client. Of course, this might be able to be borrowed from other PD packages. So.. 1. A PD NFS client exists, but it would have to be converted to run in PC-land. 2. It can't be ftp'ed right now..but probably could be made available by sites which have purchased 4.3BSD-reno. -- | Ronald J. Srodawa | Internet: srodawa@unix.secs.oakland.edu | | School of Engineering and CS | UUCP: srodawa@egrunix.UUCP | | Oakland University | Voice: (313) 370-2247 | | Rochester, Michigan 48309-4401 | |
ljm@MERCURY.TWG.COM (10/05/90)
>>Did I miss it, did someone actually answer the original question? The original >>question(s) were: >> >> (1) Does a PD version of Sun's PC-NFS Client exist? >> >> (2) Where can I FTP it from. >> > >YES! The Berkeley announcement of 4.3BSD-reno stated that it includes a PD >client and server for NFS.... Well, yes but no. As we, and I assume FTP Software, Beame and Whiteside, and SUN all found out, the difficult part of writing an NFS client for DOS is not the NFS, it's writing the redirector. Actually, it is fairly straightforward to write an NFS client *application* from the RFCs (NFS - 1094, RPC - 1057, XDR - 1014 for those who care). I'm sure the UNIX client code will be of some use in clarifying the RFCs, but its existance doesn't mean that a PD NFS client for DOS is right around the corner. enjoy, leo j mclaughlin iii ljm@twg.com
mra@srchtec.uucp (Michael Almond) (10/05/90)
In article <9010031749.aa22040@Mercury.TWG.COM> ljm@MERCURY.TWG.COM writes: > >>> (1) Does a PD version of Sun's PC-NFS Client exist? >>> >...is not the NFS, it's writing the redirector. Actually, it is fairly >straightforward to write an NFS client *application* from the RFCs >(NFS - 1094, RPC - 1057, XDR - 1014 for those who care). I'm sure the >UNIX client code will be of some use in clarifying the RFCs, but its >existance doesn't mean that a PD NFS client for DOS is right around the >corner. I've heard the term redirector before and assume you mean routines which intercept OS requests to a drive and hand them over to the NFS routines. If this is what you mean by redirector, couldn't you use some ramdisk software as a starting point? (BTW, I'm very interested in writing this software. IMHO PC networking software is over priced.) --- Michael R. Almond mra@srchtec.uucp (registered) search technology, inc. emory.edu!stiatl!srchtec!mra Atlanta, Georgia (404) 441-1457 (office) .'.'.'.'.'.'.'.'.'.'.'.'.'. A Georgia Tech Alumnus .'.'.'.'.'.'.'.'.'.'.'.'.'
romkey@ASYLUM.SF.CA.US (John Romkey) (10/14/90)
Date: 5 Oct 90 14:17:30 GMT From: Michael Almond <decwrl!ucbvax.berkeley.edu!att!emory!emcard!stiatl!srchtec!srchtec.uucp!mra> Organization: search technology, inc. Sender: pcip-request@udel.edu I've heard the term redirector before and assume you mean routines which intercept OS requests to a drive and hand them over to the NFS routines. If this is what you mean by redirector, couldn't you use some ramdisk software as a starting point? No. Ramdisk software for PC's tends to operate by supplying PC-DOS with a new disk driver that reads and writes memory instead of disk. A "redirector" operates differently. It doesn't see disk block requests; it sees file operation requests. Redirectors either grab int 21 and check if the DOS calls are for their network drives or for DOS (passing the request through to DOS in that case), or sit under DOS using the calls through DOS that Microsoft provides so that their redirector can do something similar. I don't believe this interface is documented, but at least one person has successfully reverse engineered it. The problem with writing a redirector is that it's a pain in the ass to mimic the semantics of the DOS filesystem. You have to return all the correct error codes in all the correct situations, deal with the undocumented systrem calls, and in general act like DOS acts. Ick. (BTW, I'm very interested in writing this software. IMHO PC networking software is over priced.) Believe me, once you add in the development costs, the costs associated with tracking the changes in protocol specifications and supporting that new version of DOS or new display card or new network interface, and building up a support organization that can handle calls from users who have problems with other vendors' software, you're spending lots of money... - john romkey USENET/UUCP: romkey@asylum.sf.ca.us Internet: romkey@ftp.com
zeeff@b-tech.ann-arbor.mi.us (Jon Zeeff) (10/14/90)
> I've heard the term redirector before and assume you mean routines which > intercept OS requests to a drive and hand them over to the NFS routines. > If this is what you mean by redirector, couldn't you use some ramdisk software > as a starting point? > >No. Ramdisk software for PC's tends to operate by supplying PC-DOS Actually, if you aren't going to have multiple machines accessing the files at the same time, yes - it would be much easier to just read/write blocks from another machine. A single file on your server could be the entire dos file system. You might have utilities to extract dos files from it on the unix machine. The whole thing would be less useful than NFS, but might have some applications (like making backups of dos systems easier or saving hard disk costs). -- Jon Zeeff (NIC handle JZ) zeeff@b-tech.ann-arbor.mi.us
geoff@hinode.East.Sun.COM (Geoff Arnold @ Sun BOS - R.H. coast near the top) (10/15/90)
Quoth zeeff@b-tech.ann-arbor.mi.us (Jon Zeeff) (in <5Q7*L9F@b-tech.uucp>): #Actually, if you aren't going to have multiple machines accessing the #files at the same time, yes - it would be much easier to just #read/write blocks from another machine. A single file on your server #could be the entire dos file system. You might have utilities to extract #dos files from it on the unix machine. The whole thing would be less useful #than NFS, but might have some applications (like making backups of dos #systems easier or saving hard disk costs). It's been done - RVD (remote virtual disk) was part of the early PC TCP work at MIT (John - can you add more details/lore/dirt?) I don't know of any currently-available package that includes this. Another equivalent mechanism was ND (network disk) which was used for booting diskless Suns long ago, before we all figured out how to boot, dump and swap to NFS files. -- Geoff Arnold, PC-NFS architect, Sun Microsystems. (geoff@East.Sun.COM) -- *** "Now is no time to speculate or hypothecate, but rather a time *** *** for action, or at least not a time to rule it out, though not *** *** necessarily a time to rule it in, either." - George Bush ***
engwall@mtha.usc.edu (Ralph Engwall) (10/16/90)
In article <2920@jaytee.East.Sun.COM> geoff@east.sun.com (Geoff Arnold @ Sun BOS - R.H. coast near the top) writes: >It's been done - RVD (remote virtual disk) was part of the >early PC TCP work at MIT (John - can you add more details/lore/dirt?) >I don't know of any currently-available package that includes this. I thought RVD was part of PC TCP that never got implemented. I haven't seen a working version of RVD. If someone knows where to FTP such an animal form please do tell. -- +-------------------------+-----------------------------------------------+ | Ralph Engwall | Q: Why am I banging my head against the wall? | | engwall@mtha.usc.edu | A: Because it feels *so* good when I stop. | +-------------------------+-----------------------------------------------+