1 INFO-VAX	Sat, 25 Oct 2003	Volume 2003 : Issue 591       Contents: Re: Coming to Bootcamp? / Re: DHCP serving more than one subnet (longish) / Re: DHCP serving more than one subnet (longish) / Re: DHCP serving more than one subnet (longish) / Re: DHCP serving more than one subnet (longish) / Re: DHCP serving more than one subnet (longish) / Re: DHCP serving more than one subnet (longish) / Re: DHCP serving more than one subnet (longish) / Re: DHCP serving more than one subnet (longish) / Re: DHCP serving more than one subnet (longish) @ Re: Did someone ask HP for VAX to Itanium cluster possibilities? Digital Memorabilia  Re: Do we need a DCL debugger? RE: Do we need a DCL debugger?( Re: Dynamically expand a Files-11 volume Re: EV79 CANCELED !!!!!!!!!  Re: EV79 CANCELED !!!!!!!!!  Guess that's now then?/ NFS mount of ODS5 disk fails on extendend names " Re: OpenVMS Freeware V6.0 Contents2 Re: OpenVMS I64 ISV application count now over 5002 Re: OpenVMS I64 ISV application count now over 5002 Re: OpenVMS I64 ISV application count now over 5002 Re: OpenVMS I64 ISV application count now over 5002 Re: OpenVMS I64 ISV application count now over 500? Re: OpenVMS Licensing on VAX emulators (was: Re: VAX emulators) ? RE: OpenVMS Licensing on VAX emulators (was: Re: VAX emulators)  Re: OpenVMS Roadmap  Re: OpenVMS Roadmap  Re: Product problem  Re: Spam from Hewlett-Packard  Re: Spam from Hewlett-Packard  Re: Spam from Hewlett-Packard  Re: Spam from Hewlett-Packard + Swan Song For The Alpha Processor - TechWeb / Re: Swan Song For The Alpha Processor - TechWeb  Re: Time server protocole? Re: Time server protocole? Re: Time server protocole? Re: Time server protocole? Re: Time server protocole?D Re: Tru64 unix Layeredproducts won't install on my OpenVMS machines!D Re: Tru64 unix Layeredproducts won't install on my OpenVMS machines!
 VAX emulators 2 Re: VAXscan now available on OpenVMS Alpha systems. Re: [survey] why do you still use VAX systems?  F ----------------------------------------------------------------------  # Date: Fri, 24 Oct 2003 18:20:57 GMT A From: "Colin Butcher" <colin_DOT.butcher_AT@xdelta_DOT.co_DOT.uk>   Subject: Re: Coming to Bootcamp?; Message-ID: <d4emb.394$1V7.6535310@news-text.cableinet.net>   J Yup. As Alan says - one of the best events I've EVER been to. Great stuff,* great people, great fun. Darned good week.   See you there.   --     Hope this helps, Colin. ) colin DOT butcher AT xdelta DOT co DOT uk L Systems Archaeologist - Investigation & troubleshooting of older systems and	 networks.    ------------------------------  # Date: Fri, 24 Oct 2003 17:37:56 GMT - From: "John E. Malmberg" <wb8tyw@qsl.network> 8 Subject: Re: DHCP serving more than one subnet (longish)> Message-ID: <Urdmb.125649$qj6.8039080@news1.news.adelphia.net>   Michael T. Davis wrote: B > 	I probably should have including our routing information when I, > described our DHCP server's configuration: > 5 > Type  Destination                           Gateway  > 9 > AN    0.0.0.0                               192.168.0.1 7 > AH    127.0.0.1                             127.0.0.1 9 > AN    192.168.0.0/22                        192.168.0.5 9 > AH    192.168.0.5                           192.168.0.5 9 > AN    192.168.4.0/24                        192.168.4.5 9 > AH    192.168.4.5                           192.168.4.5  > I > The 0.0.0.0 is our "default" route to the Internet.  Recall that WE0 is E > our server's physical intrface, bound to 192.168.0.5, and WEA1 is a ) > pseudo-interface, bound to 192.168.4.5.  > F > 	The response from John E. Malmberg brought up an interesting point.K > The above routes don't seem to provide for a direct route for the pseudo- K > interface (WEA1, bound to 192.168.4.5) to get to the Internet.  Might the M > functionality we're looking for be obtained by running two DHCP servers (on < > separate machines), each configured for a specific subnet?  A All DHCP servers that can be reached by a client must be able to  3 properly service the request with the correct data.   E So as long as you have two subnets on the same network segment, that   alone will not help.  F As I have not done DHCP stuff for over 3 years, and even then nothing F that complex, I do not know what specific steps are needed to do what 	 you want.   F When the DHCP request comes in for the first time, it is coming in as C 0.0.0.0. as a broadcast request.  No subnet information is present.   E So I would configure the DHCP server to have a list of MAC addresses  F that are fixed to specific I.P. addresses/subnet/default gateways for E the 192.168.0.0/22 subnet, and let the rest of the addresses get the   dyanmic information.  H As long as the clients have a route to the DHCP server before and after 5 they have been assigned an address, this should work.   C All it depends on is if your DHCP server can give out the specific  C subnet and default gateway information for the fixed MAC addresses.   H Again this is not something that I have ever done, so I do not know how = hard it would be to implement, if your DHCP server allows it.   E Multiple DHCP servers that give out equivalent data are possible and  I very useful, but they either have to know what addresses that the others  @ have given out, or they must not have any overlapping addresses.   -John  wb8tyw@qsl.network Personal Opinion Only    ------------------------------  % Date: Fri, 24 Oct 2003 14:18:26 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>8 Subject: Re: DHCP serving more than one subnet (longish)) Message-ID: <3F996CED.9717ED2D@istop.com>    "Michael T. Davis" wrote: I > The 0.0.0.0 is our "default" route to the Internet.  Recall that WE0 is E > our server's physical intrface, bound to 192.168.0.5, and WEA1 is a ) > pseudo-interface, bound to 192.168.4.5.   N DHCP packets are really ethernet packets whose contents are formatted as TCPIPR packets to allow  a host's ethernet layer to pass the packet onto the TCPIP stack.  G When a client sends an initial DHCP request, it doesn't know its own IP E address nor that of a DHCP server. The request is sent as an ethernet N broadcast. The response from the server is sent as an ethernet packet adressedL only to the client. The ethernet packet contents may be formatted as a TCPIPM packet, but IP routing information is bogus since the destination doesn't yet  have an IP address.   H If both subnets are on the same physical ethernet segment, then the DHCPK server will see requests coming from machines on both subnets. The solution K would then consist of putting the list of ethernet hardware adresses in the ' "nodes" section (one item per machine).   L You can then just associate each hardware address with a group that containsM ip ranges and all the other parameters, or specify all of the DHCP parameters # for each specific ethernet address.   N However, if there is a router between the old subnet and the DHCP server, thenN it gets dicey. Initial packets have the IP adresses set to 0.0.0.0 so a routerM won't know where/how to route those packets. And even if it routes, them, the N problem starts when a server responds. If the server sends the ethernet packetN to the client's ethernet address, the router won't receive the packet and thusI won't transfer it to the other side. And if the packet is adressed to the H router, then what will the router do with it since the destination IP is
 unknown ?   J You need a Relay Agent in your router. This relay agent  intercepts a DHCPL request and adds relay agent information in the options field. The server isL expected to send responses that contain the information that the relay agentJ had inserted in the options. The relay agent, upon reception of a responseN from a server, will examine the relai-agent option which will tell it to which6 ethernet address to send the packet on the other side.    = DHCP protocol : ftp://ftp.rfc-editor.org/in-notes/rfc2131.txt L Relay Agent information is at: ftp://ftp.rfc-editor.org/in-notes/rfc3046.txt  H I am not sure if the VMS DHCP server (from the now defunct Join company)M supports the mirroring of the relay agent information obtained from a request  and copied to the response.    ------------------------------  % Date: Fri, 24 Oct 2003 12:31:14 -0700 0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>8 Subject: Re: DHCP serving more than one subnet (longish)' Message-ID: <3f991b92$1@cpns1.saic.com>    JF Mezei wrote:  > "Michael T. Davis" wrote:  > I >>The 0.0.0.0 is our "default" route to the Internet.  Recall that WE0 is E >>our server's physical intrface, bound to 192.168.0.5, and WEA1 is a ) >>pseudo-interface, bound to 192.168.4.5.  >  > P > DHCP packets are really ethernet packets whose contents are formatted as TCPIPT > packets to allow  a host's ethernet layer to pass the packet onto the TCPIP stack.  F DHCP packets are formatted the same as any other IP packet.  The only @ real difference is that the source address is all zeros and the G destination address is all ones (which means it is a broadcast packet).   I > When a client sends an initial DHCP request, it doesn't know its own IP G > address nor that of a DHCP server. The request is sent as an ethernet  > broadcast.  G IP Broadcast.  When encapsulated into an ethernet packet, the ethernet  1 destination is also set to the broadcast address.   E > The response from the server is sent as an ethernet packet adressed  > only to the client.   F Optional.  The response can be either broadcast or addressed directly.  : > The ethernet packet contents may be formatted as a TCPIPO > packet, but IP routing information is bogus since the destination doesn't yet  > have an IP address.   B I think you have a semantic issue here.  The response packet is a F properly formatted IP packet.  If it weren't, no router would pass it H and DHCP requests can certainly be relayed via a router.  Nor would the & host IP stack be willing to accept it.  J > If both subnets are on the same physical ethernet segment, then the DHCPM > server will see requests coming from machines on both subnets. The solution M > would then consist of putting the list of ethernet hardware adresses in the ) > "nodes" section (one item per machine).  > N > You can then just associate each hardware address with a group that containsO > ip ranges and all the other parameters, or specify all of the DHCP parameters % > for each specific ethernet address.  > P > However, if there is a router between the old subnet and the DHCP server, thenP > it gets dicey. Initial packets have the IP adresses set to 0.0.0.0 so a routerO > won't know where/how to route those packets. And even if it routes, them, the P > problem starts when a server responds. If the server sends the ethernet packetP > to the client's ethernet address, the router won't receive the packet and thusK > won't transfer it to the other side. And if the packet is adressed to the J > router, then what will the router do with it since the destination IP is > unknown ?  > L > You need a Relay Agent in your router. This relay agent  intercepts a DHCPN > request and adds relay agent information in the options field. The server isN > expected to send responses that contain the information that the relay agentL > had inserted in the options. The relay agent, upon reception of a responseP > from a server, will examine the relai-agent option which will tell it to which8 > ethernet address to send the packet on the other side.  G Not quite.  The relay agent simply enters info into the gateway field.  F This tells the DHCP server what subnet the requesting host is part of C and also how to return the packet.  It has nothing to do with DHCP  ! options which are left untouched.   B Note that the relay agent is only going to use the address of the E primary network on which the host resides, it will not reference any  E secondary network.  Thus, the DHCP server has no way of knowing that  ' there are multiple subnets on that LAN.   ? > DHCP protocol : ftp://ftp.rfc-editor.org/in-notes/rfc2131.txt N > Relay Agent information is at: ftp://ftp.rfc-editor.org/in-notes/rfc3046.txt > J > I am not sure if the VMS DHCP server (from the now defunct Join company)O > supports the mirroring of the relay agent information obtained from a request  > and copied to the response.   G All supported IP stacks for VMS come with a DHCP server.  All of these  ) DHCP servers can handle relayed requests.   I The bottom line: dynamic addressing and multiple subnets on the same LAN  G do not mix.  Only one of the subnets can be dynamic, the other(s) must  G be static.  There are some configurations where the static part can be  D in a DHCP server rather than on the host itself but, one way or the I other, you must manually associate an address with each specific host on   all the other subnets.  
 Mark Berryman    ------------------------------  % Date: Fri, 24 Oct 2003 15:24:10 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>8 Subject: Re: DHCP serving more than one subnet (longish)) Message-ID: <3F997C52.DB06F6EC@istop.com>    "John E. Malmberg" wrote: D > All it depends on is if your DHCP server can give out the specificE > subnet and default gateway information for the fixed MAC addresses.   K The VMS DHCP server can provide DHCP responses that are fairly complete and I cover far more than IP, subnet mask, default gateway, DNS which most ISPs E provide. You can also provide server names for various servers, MAIL,  X-windows, printing, time etc)  M The VMS DHCP server has 4 different tabs. The Basic DHCP paremeters, the DHCP K parameters (more complete), the Join Parameters (even more options) and the  SUn parameters (even more)   ------------------------------  % Date: Fri, 24 Oct 2003 17:45:35 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>8 Subject: Re: DHCP serving more than one subnet (longish)) Message-ID: <3F999D6C.B6065F1B@istop.com>    Mark Berryman wrote:G > DHCP packets are formatted the same as any other IP packet.  The only A > real difference is that the source address is all zeros and the I > destination address is all ones (which means it is a broadcast packet).   I Responses from the server however rely on the ethernet address only since B there is no IP address to send the response to. (responses are notK broadcasts). So while it may have the same format as an IP packet, it isn't 7 one per say when it comes to issues related to routers.   H > Optional.  The response can be either broadcast or addressed directly.  M Nop. Responses are sent to the client's ethernet address or to the ip address I of a relay server if the giaddr field is not null. In the later case, the M relay server will issue an ethernet packet on the other ethernet segment that # is adressed to the client directly.   C > I think you have a semantic issue here.  The response packet is a G > properly formatted IP packet.  If it weren't, no router would pass it < > and DHCP requests can certainly be relayed via a router.    H Nop. DHCP requests cannot be passed by a ROUTER. You need a relay serverI (which I assume can be integrated into a router, but it isn't part of the  basic role of a router).  H > Not quite.  The relay agent simply enters info into the gateway field.  M I stand corrected. giaddr is filled with the relay server's IP address on the G other side of the client's lan. Interstsingly, the relay server RFC did B mention the addition of options in the client originated messages.  G > This tells the DHCP server what subnet the requesting host is part of & > and also how to return the packet.    H It doesn't directly tell it what subnet it is coming from. For instance,L consider a router that have 4 interfaces. The IP address in the giaddr wouldJ be the IP address of the interface talking to the DHCP server and the DHCPJ server would then not know which of the other 3 subnets the client requestP originated from. (I think that this is where those relay server options come in)    J > The bottom line: dynamic addressing and multiple subnets on the same LANH > do not mix.  Only one of the subnets can be dynamic, the other(s) must > be static.  M Not quite. You should be able to define a whole bunch of ethernet adresses as N belonging to a group, and that group has its own range of dynamic IP adresses.M However, on the VMS DHCP server, I haven't quite figured out how to associate  a group with an IP range.    ------------------------------  + Date: Sat, 25 Oct 2003 00:00:18 +0000 (UTC) ? From: Graham Burley <burley.not-this@encompasserve-or-this.org> 8 Subject: Re: DHCP serving more than one subnet (longish)9 Message-ID: <3F99C9CF.757061F2@encompasserve-or-this.org>    JF Mezei wrote:  >    > J > Nop. DHCP requests cannot be passed by a ROUTER. You need a relay serverK > (which I assume can be integrated into a router, but it isn't part of the  > basic role of a router). >   9 BOOTP and DHCP relaying is normally performed by routers.    > J > It doesn't directly tell it what subnet it is coming from. For instance,N > consider a router that have 4 interfaces. The IP address in the giaddr wouldL > be the IP address of the interface talking to the DHCP server and the DHCPL > server would then not know which of the other 3 subnets the client requestR > originated from. (I think that this is where those relay server options come in) >   G The giaddr is the address of the first relayer. In the case of a router E with multiple interfaces it will be the address of the interface that B received the request. The giaddr is preserved as the packet passesC across routers so that a BOOTP or DHCP server can configure clients  that are multiple hops away.     Graham   ------------------------------  % Date: Fri, 24 Oct 2003 18:12:48 -0700 0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>8 Subject: Re: DHCP serving more than one subnet (longish)' Message-ID: <3f996ba1$1@cpns1.saic.com>    JF Mezei wrote:  > Mark Berryman wrote: > G >>DHCP packets are formatted the same as any other IP packet.  The only A >>real difference is that the source address is all zeros and the I >>destination address is all ones (which means it is a broadcast packet).  >  > K > Responses from the server however rely on the ethernet address only since D > there is no IP address to send the response to. (responses are notM > broadcasts). So while it may have the same format as an IP packet, it isn't 9 > one per say when it comes to issues related to routers.   C Not correct.  DHCP works over media other than ethernet.  The DHCP  H server knows nothing about any ethernet packet, all it ever sees is the B UDP packet.  The ethernet (or other media) encapsulation has been D stripped off by the time the DHCP server sees the packet.  The DHCP H server gets what you are calling the "ethernet address" from the chaddr D field of the DHCP packet.  If the client identifier option has been I included in the packet by the client then the DHCP server may ignore the  G "ethernet address" entirely.  The DHCP packet itself is a fully built,  I fully functional UDP packet.  It most definitely is NOT an ethernet-only   packet.    > H >>Optional.  The response can be either broadcast or addressed directly. >  > O > Nop. Responses are sent to the client's ethernet address or to the ip address K > of a relay server if the giaddr field is not null. In the later case, the O > relay server will issue an ethernet packet on the other ethernet segment that % > is adressed to the client directly.   + Sorry, wrong again.  To quote from the RFC:   F     If the 'giaddr' field in a DHCP message from a client is non-zero,I     the server sends any return messages to the 'DHCP server' port on the F     BOOTP relay agent whose address appears in 'giaddr'. [This will be9     the case if the request is relayed through a router].        If the 'giaddr' D     field is zero and the 'ciaddr' field is nonzero, then the serverG     unicasts DHCPOFFER and DHCPACK messages to the address in 'ciaddr'. =     [This will be the case when an address is being renewed.]   F     If 'giaddr' is zero and 'ciaddr' is zero, and the broadcast bit isE     set, then the server broadcasts DHCPOFFER and DHCPACK messages to H     0xffffffff. If the broadcast bit is not set and 'giaddr' is zero andD     'ciaddr' is zero, then the server unicasts DHCPOFFER and DHCPACKC     messages to the client's hardware address and 'yiaddr' address.   G     Note the option; the broadcast bit determines whether a response is F     unicast or broadcast and the setting of that bit can be controlled)     within the DHCP server configuration.   C >>I think you have a semantic issue here.  The response packet is a G >>properly formatted IP packet.  If it weren't, no router would pass it < >>and DHCP requests can certainly be relayed via a router.   >  > J > Nop. DHCP requests cannot be passed by a ROUTER. You need a relay serverK > (which I assume can be integrated into a router, but it isn't part of the  > basic role of a router).  H Sorry, you are 0 for 3 here.  A DHCP request is a standard UDP datagram F and can be passed by any router.  You may be confusing the need for a G relay agent on the router to fill in the giaddr field but lack of that  E agent does not prevent the router from passing the packet (and I can  F even configure a network so that DHCP will work if the router doesn't  have such an agent).  H >>Not quite.  The relay agent simply enters info into the gateway field. >  > O > I stand corrected. giaddr is filled with the relay server's IP address on the I > other side of the client's lan. Interstsingly, the relay server RFC did D > mention the addition of options in the client originated messages. >  > G >>This tells the DHCP server what subnet the requesting host is part of & >>and also how to return the packet.   >  > J > It doesn't directly tell it what subnet it is coming from. For instance,N > consider a router that have 4 interfaces. The IP address in the giaddr wouldL > be the IP address of the interface talking to the DHCP server and the DHCPL > server would then not know which of the other 3 subnets the client requestR > originated from. (I think that this is where those relay server options come in)  E The giaddr field will be filled in with the address of the interface  D that is on the same LAN as the requesting client, not the interface F closest to the DHCP server.  The DHCP server will then use this field I to, among other things, determine what network (or subnet if you prefer)  6 the client is on (you really need to re-read the RFC).  J >>The bottom line: dynamic addressing and multiple subnets on the same LANH >>do not mix.  Only one of the subnets can be dynamic, the other(s) must >>be static. >  > O > Not quite. You should be able to define a whole bunch of ethernet adresses as P > belonging to a group, and that group has its own range of dynamic IP adresses.  G Sorry, I wasn't clear here.  Having to manually maintain a list of MAC  D addresses is the same thing as having to maintain a bunch of static I configurations in my book.  However, your statement is correct.  You can  F create a group of hosts and dynamically allocate within that group if ) you willing to do the manual maintenance.   O > However, on the VMS DHCP server, I haven't quite figured out how to associate  > a group with an IP range.   D What are you calling the "VMS DHCP" server?  There are several DHCP  servers that run on VMS.  
 Mark Berryman    ------------------------------   Date: 25 Oct 2003 01:03:13 GMT6 From: DAVISM@er6.eng.ohio-state.edu (Michael T. Davis)8 Subject: Re: DHCP serving more than one subnet (longish): Message-ID: <bnci4h$4fu$1@charm.magnus.acs.ohio-state.edu>  5 In article <3f991b92$1@cpns1.saic.com>, Mark Berryman $ <Mark.Berryman@Mvb.Saic.Com> writes:   >[...] > I >The bottom line: dynamic addressing and multiple subnets on the same LAN G >do not mix.  Only one of the subnets can be dynamic, the other(s) must G >be static.  There are some configurations where the static part can be D >in a DHCP server rather than on the host itself but, one way or theI >other, you must manually associate an address with each specific host on  >all the other subnets.   K         If I understand you correctly, this is precisely how I'd like to be E able to implement things.  Our old subnet would provide for static IP I addresses, either by manual configuration or by DHCP through the use of a K permanent lease; the new subnet will be used strictly for dynamic addresses $ from a pool.  So my questions are...  >         Is this possible with the TCP/IP Services DHCP server?  : 	If this isn't possible under TCP/IP Services, what stacks8 	and/or DHCP servers that run under VMS will support it?  > 	If this is possible under TCP/IP Services, what's the missing) 	piece that's preventing it from working?   ( 	Where is this "DHCP relay" to be found?  I FWIW, we only provide service for clients for which we have MAC addresses F registered (registered_clients_only).  Host name assignment is via DNS (assign_name_by_ipaddr).   >  >Mark Berryman >    Regards, Mike --K              Michael T. Davis              |    Systems Specialist: ChE,MSE N   E-mail: davism@er6.eng.ohio-state.edu    | Departmental Networking/ComputingJ            -or- DAVISM+@osu.edu            |     The Ohio State UniversityJ http://www.er6.eng.ohio-state.edu/~davism/ |     197 Watts, (614) 292-6928   ------------------------------  % Date: Fri, 24 Oct 2003 21:53:14 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>8 Subject: Re: DHCP serving more than one subnet (longish)) Message-ID: <3F99D768.16D161C4@istop.com>    Mark Berryman wrote:
 > The DHCPI > server knows nothing about any ethernet packet, all it ever sees is the  > UDP packet.   N But the UDP packet is "faked" since the client doesn't yet have an IP address.N And the DHCP RFC clearly states that they've had to imginatively interpret the( IP protocol definitions to make it work.  M The deal is that an ethernet broadcast is made whose contents have the IP/UDP L protocol bytes set as well as port 67 (server) or 68 (client) set. This way,M all ethernet interfaces on the lan receiving that packet would pass it to the S TCPIP stack which would pass it to the DHCP server if such is defined in that node.     H > "ethernet address" entirely.  The DHCP packet itself is a fully built,J > fully functional UDP packet.  It most definitely is NOT an ethernet-only	 > packet.   J It isn't fully functional because the source IP address is set to 0.0.0.0.I This means that A router cannot route a response with normal IP handling.   H >     If 'giaddr' is zero and 'ciaddr' is zero, and the broadcast bit isG >     set, then the server broadcasts DHCPOFFER and DHCPACK messages to J >     0xffffffff. If the broadcast bit is not set and 'giaddr' is zero andF >     'ciaddr' is zero, then the server unicasts DHCPOFFER and DHCPACKE >     messages to the client's hardware address and 'yiaddr' address.   L The keyword here is "if the broadcast bit is set". If not, responses go backM to a specific ethernet address if there are no IP adresses (client or router)   I > Sorry, you are 0 for 3 here.  A DHCP request is a standard UDP datagram " > and can be passed by any router.  B Request yes because it is , by definition, a broadcast. (actually, DHCPDISCOVER and REQUEST).M But responses by DHCP server are not necessarily a broadcast and a UDP packet / with IP destination of 0.0.0.0 is not routable.   M Now, if the router has DHCP relaying, then it is different, But the router is F no longer a "router", it becomes specialized machine that examines theJ contents of a packet to find out where the packet should be sent, at whichH point, it will generate an ethernet packet (with UDP formatted contents) destined to the actual client.      H > Sorry, I wasn't clear here.  Having to manually maintain a list of MACE > addresses is the same thing as having to maintain a bunch of static  > configurations in my book.  L Nop. If you point each individual MAC address to a group in the DHCP server,L then if you change your DNS server for instance, you only need to change theB group definition once and this will be used by all subsequent DHCPW requests/renews from those known MAC addresses. It makes for easier network management.    ------------------------------  # Date: Fri, 24 Oct 2003 18:44:21 GMT & From: jlsue <jefflsxxxz@sbcglobal.net>I Subject: Re: Did someone ask HP for VAX to Itanium cluster possibilities? 8 Message-ID: <bisipv435nlnv7m8lbjod2kgmk0105t2ip@4ax.com>  I On Thu, 23 Oct 2003 11:04:14 +0200, Michael Unger <unger@decus.de> wrote:   + >On 2003-10-23 08:12, "Keith Parris" wrote:  >  >> [...] >>  H >> So I'd guess the most risky area is probably code being ripped out onI >> Itanium as being no longer necessary to talk with Alpha, but needed to F >> talk to VAX.   Because HP isn't testing VAXes with Itanium, such asE >> case could slip through testing without being noticed.  Of course, D >> having common code between Alpha and Itanium makes this much moreG >> unlikely, as it would presumably have to be conditionalized code for ; >> Itanium only.  But you can see how it might be possible.  > H >Thanks for the extensive explanation. Now it is pretty clear to me whatH >*will* work and what just *might* work and what will be *supported* and >*why*.  >   ? Note, too, that most of what he discussed was in the VMScluster  communications area.  F There may be other inconsistencies as well, including things like ODS5F support, XFC, cluster-wide logical names, etc.  All these could impact4 compatability for cluster-wide sharing of resources.   ------------------------------  # Date: Fri, 24 Oct 2003 23:41:24 GMT # From: "John Smith" <a@nonymous.com>  Subject: Digital MemorabiliaJ Message-ID: <EMimb.281496$ko%.102394@news04.bloor.is.net.cable.rogers.com>  J I was cleaning out a drawer today and found a CD that has about 1200 SPD'sK from 1995 on it (total of about 45Mb of *text*) and a large number of other I goodies like Powerpoint slideshows for various products, some interactive K demos (DECwrite, DEC Athena. Tuxedo, and more), and tabular info similar to L the following (which will be totally unrecognizable if you aren't using someI sort of Microsoft news client). Is there any interest in all this ancient  history from a defunct company?     )       OpenVMS VAX System Comparison Chart       1                   System(      MicroVAX 3100 Models 40, 85, and 95      VAX 4000 Model 105A'      VAX 4000 Model 505A and Model 705A       VAXft System Model 810 
      VAX 6000          Model 510 and 610 
      VAX 7000          Models 710 through 760         Performance       Model 40: 32 TPS          Model 85: 110 TPS          Model 95: 165 TPS       181 TPS      Model 505A: 185 TPS         model 705A: 280 TPS       Model 810: 90 TPS      Model 510: 50 TPS         Model 610: 102 TPS      Model 710: 314 TPS   #       Relative Processor Perfomance          xVAX-11/780       Model 40: 5         Model 85: 16         Model 95: 32      32       Model 505A: 32          Model 705A:45       Model 810: 32 MHz      Model 510: 13         Model 610: 32       Model 710: 50         CPU Clock Speed       Model 40: 25 MHz          Model 85: 62 MHz         Model 95: 72 MHz
      83.3 Mhz       Model 505A: 83 MH         Model 705A:111 MHz      Model 810: 32 Mhz      Model 510: 63 MHz         Model 610: 83 MHz       137.5 Mhz         In-Cabinet CPU UpgradeL      Model 40 upgrades to Model 85 or 95;Model 80 upgrades to Model 85 or 95.      VAX 4000 Model 100 and 100A to Model 105AH      Each VAX 4000 system upgrades to any higher VAX 4000A series system9      Model 110 and 410 can be box upgraded to a Model 810 @      Each VAX 6000 system upgrades to any higher VAX 6000 system@      Each VAX 7000 system upgrades to any higher VAX 7000 system         Alpha-ReadySystem Upgrade /      System upgrade to AlphaServer 1000 or 2000 '      System upgrade to AlphaServer 2000 :      System upgrade to DEC 4000 Model 700 AlphaServer 2100      N/A6      System upgrade to DEC 7000 Model 700 Alpha system>      In-cabinet CPU upgrade to DEC 7000 Model 700 Alpha system         Maximum Memory Capacity       Model 40: 32 MB         Model 80: 128 MB         Model 95: 128 MB      128 MB       512 MB       Model 810: 256 MB      Model 510: 512 MB         Model 610: 1 GB       3.5 GB          Maximum Disk Capacity       17.5/29.4 GB 
      75 GB      500 GB       Model 810: 64 GB       16 GB/over 8 TB      42 GB/over 10 TB          Maximum I/O Throughput
      4 MB/secV      8.5 MB/sec0      12.5 MB/sec      Model 810: 16 MB/sece      80 MB/sec      400 MB/secs         I/O Supports      1 SCSIh         1 Ethernet      4 DSSIe  8       (2 embedded, 2 Q-bus), 1 Q-bus, 1 SCSI, 3 Ethernet      6 DSSI   0       (4 embedded, 2 Q-bus), 2 Q-bus, 3 Ethernet&      4 DSSI, 8 (4 redundant), Ethernet      1 XMI,s         4 CI,n         12 DSSI,  
       2 FDDI,r         6 Ethernet,b         5 VAXBI,         2 VMEm      4 XMI,s         10 CI,         24 DSSI,  
       8 FDDI,b         16 Ethernet,         6 VAXBI,         8 VMEu         VAXcluster System Supporte
      Ethernet       Ethernet,         DSSI,   
       FDDI      Ethernet,         DSSI,x  
       FDDI      Ethernet,DSSI      Ethernet,         DSSI,N  	       CI,!  
       FDDI      Ethernet,         DSSI,   	       CI,   
       FDDI         System SoftwareV      OpenVMS VAX      OpenVMS VAX      OpenVMS VAX      OpenVMS VAX#      Model 510: OpenVMS VAX, ULTRIXv         Model 610: OpenVMS VAX      OpenVMS VAX  *       Network Application Support Software      NAS 200      NAS 200      NAS 200      NAS 200      NAS 200, 300, 400      NAS 200, 300, 400   ------------------------------  % Date: Fri, 24 Oct 2003 21:02:09 +0200R" From: Didier Morandi <no@spam.com>' Subject: Re: Do we need a DCL debugger?r2 Message-ID: <3f997732$0$228$636a55ce@news.free.fr>  ( Man, I missed that thread... How come???  E AS there are more than 50 replies, I will not go through all of them  % now. So, I will just post my request.   D I would be pleased to have a way to know the line nr of a given DCL F instruction. This would allow for example an error handler to display @ "gracefully" the line nr of the source where the error occurred.  B I wrote a pre-processor to have this because I *love* fancy error H management. The preproc adds a line with a counter before every line of F source. When an error occurs, the counter gives the line nr. I do not A tell you (as we say over here) the mess with continuation lines,  7 DECK/EOD and other IF THEN ELSE nestings, but it works.r  E When you're done with your 10,000 lines source, you preproc the .COM oC which will rename the original file to .COM_SRC and produce a .COM  
 "executable".s   Very useful.   D.   Guy Peleg wrote: > D >The DCL engineering team is currently evaluating new features to be. >added in future version of VMS (post V7.3-2). >OH >We were wondering if folks think that developing a DCL debugger will be% >useful. It won't be as sophisticated H >as the VMS debugger, I was thinking about implementing step,examine and >set break.r > 8 >Please post a reply with your comments or send me mail. >e >Thanks in advance,u >t
 >Guy Peleg >OpenVMS Engineering >$I >** Please note that this note is no commitment for actually developing aa2 >DCL debugger, this is only a survey that will not* >necessarily result in actual product..... >  >    ------------------------------  % Date: Fri, 24 Oct 2003 12:46:54 -0700.# From: "Tom Linden" <tom@kednos.com>.' Subject: RE: Do we need a DCL debugger? 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIIEDDIFAA.tom@kednos.com>   D In that case, what you really want are ON-conditions which allow you3 recover through error handlers  like REXX  see e.g. ? http://publibfi.boulder.ibm.com/epubs/pdf/hcse2a30.pdf page 163 D regina has been ported to VMS I read somewhere, but I don't know anyC more about it.  Noter that REXX provides scoped handlers, depending2# upon whether you use CALL or SIGNALt     >-----Original Message----- * >From: Didier Morandi [mailto:no@spam.com]( >Sent: Friday, October 24, 2003 12:02 PM >To: Info-VAX@Mvb.Saic.Com( >Subject: Re: Do we need a DCL debugger? >f > ) >Man, I missed that thread... How come???I >rF >AS there are more than 50 replies, I will not go through all of them & >now. So, I will just post my request. >eE >I would be pleased to have a way to know the line nr of a given DCL  G >instruction. This would allow for example an error handler to display tA >"gracefully" the line nr of the source where the error occurred.a >nC >I wrote a pre-processor to have this because I *love* fancy error CI >management. The preproc adds a line with a counter before every line of oG >source. When an error occurs, the counter gives the line nr. I do not  B >tell you (as we say over here) the mess with continuation lines, 8 >DECK/EOD and other IF THEN ELSE nestings, but it works. > F >When you're done with your 10,000 lines source, you preproc the .COM D >which will rename the original file to .COM_SRC and produce a .COM  >"executable". >2
 >Very useful.t >e >D.  >t >Guy Peleg wrote:e >> yE >>The DCL engineering team is currently evaluating new features to bee/ >>added in future version of VMS (post V7.3-2).n >>I >>We were wondering if folks think that developing a DCL debugger will bev& >>useful. It won't be as sophisticatedI >>as the VMS debugger, I was thinking about implementing step,examine ande >>set break. >>9 >>Please post a reply with your comments or send me mail.m >> >>Thanks in advance, >> >>Guy Pelegt >>OpenVMS Engineeringg >>J >>** Please note that this note is no commitment for actually developing a3 >>DCL debugger, this is only a survey that will nott+ >>necessarily result in actual product.....s >>   >> h >  >---' >Incoming mail is certified Virus Free.-; >Checked by AVG anti-virus system (http://www.grisoft.com).2B >Version: 6.0.530 / Virus Database: 325 - Release Date: 10/22/2003 >  ---n& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.530 / Virus Database: 325 - Release Date: 10/22/2003l   ------------------------------    Date: 24 Oct 2003 17:37:04 -0700+ From: seanobanion@attbi.com (Sean O'Banion)o1 Subject: Re: Dynamically expand a Files-11 volumep< Message-ID: <f883d5a4.0310241637.9a3ccec@posting.google.com>  B Pay close attention to the fact that setting the maximum size of a< volume can only be done when the volume is mounted privately (non-system and non-cluster).   D The current size of the volume can be changed when mounted system orE cluster wide.  Note that a shadow membership capacity change does not F automatically cause the shadow volume to change size: the current size8 must be changed explicitly with some form of Set Volume.  D This means you need think about the maximum size you want the volumeC to become without dismounting the volume from the system or clusterm wide mount to private.  + Also, note the new fields on Show Dev/Full.   B Once these are understood, and I'm probably making this sound more. complicated than it is, this works quite well.     Sean    r Karl Rohwedder <extern.karl.rohwedder@volkswagen.de> wrote in message news:<bnarmu$fg422@doiweb4.volkswagen.de>... > Rudolf Wingert schrieb: 
 > > Hello, > >  > > Keith Parris wrotes: > >  > > Q > > or having the DCL commands and other support to dynamically expand a Files-11h > > volume on-the-fly. > >  > >  > > Where can I find this? > > " > > TIA and regards Rudolf Wingert > > T > > P.S. There a RAID sysstem, with can dynamically expand the volume size. In cas e > > of this the dynamicallya' > > expanding of FILES-11 is importent.e > > b > See the VMS V7.3-2 announcements (OpenVMS homepage) and check for 'dynamic volume expansion' and  > 'dissiimlar device shadowing'. >  > -- i > - > mit freundlichen Gren | with best regardso > " > Karl Rohwedder		iT-Ingenieurteam& > Ellernbruch 11		D-38112 Braunschweig > G > mailto:it-ingteam(at)t-online.de | mailto:rohwedder(at)decus.decus.de / > mailto:extern.karl.rohwedder(at)volkswagen.de    ------------------------------  # Date: Sat, 25 Oct 2003 00:12:29 GMTr5 From: rdeininger@mindspringdot.com (Robert Deininger)w$ Subject: Re: EV79 CANCELED !!!!!!!!!L Message-ID: <rdeininger-2410032022330001@user-105n8f9.dialup.mindspring.com>  2 In article <3F98AA90.E184BCC8@istop.com>, JF Mezei" <jfmezei.spamnot@istop.com> wrote:    K >Lets not forget the rumours that EV7 was ready well before it was actuallyiO >released, with much speculation that HP didn't want to release EV7 before IA64 M >so that IA64 could have its 15 seconds of fame before it was outperformed bye >Alpha.   I You ought to forget those rumors (didn't you start them?), since they areu pure, undiluted bullshit.e  I The Marvel systems went out the door with the best EV7 CPUs that could bezG produced at the time.  The chip and systems engineers spent most of thelE previous year finding and fixing serious problems. A couple of months B before ship, it was far from clear that there would not be furtherE delays.  Another revision of the CPU chip might have been needed, andeF would have meant months of delay.  Fortunately, the bugs were fixed or" avoided the systems were released.    K >And then, when EV7 was allowed out, its performance was less than had been 
 >anticipated.l  F How much longer would customers have waited for systems?  They got theJ best available at the time.  Holding back until now would have allowed theH first systems to be somewhat faster.  Waiting until next year would have; made them even faster.  Would that be good business for HP?     r6 >> The OpenVMS port to Itanium has continued and is on >> schedule. >  >Very few doubt this.  >aB >> as EV79.  As the project progressed, it became clear in workingG >> closely with our CPU chip supplier that the EV79 chip would not meetn6 >> our expectations for performance or time to market. >cM >If their expectations were that IA64 would outperform Alpha, then of course,u+ >the EV79 wouldn't meet their expectations.e  D After several passes of EV79 prototype chips, they did not _work_ atI anything approaching the design speed.  They were absolutely not ready to G ship.  Additional prototype passes would have been required, and it was G virtually certain that nothing could have been given to customers untileI post-2004.  The problems were largely out of HP's control.  Throwing tons E of money at it wouldn't have fixed the schedule problem, at least nott enough to matter.     E >> However, the manufacturing process maturity for EV7 has progressed G >> very well, producing good yield and making it possible to produce anoE >> EV7 part (EV7z at 1.33GHz) with significantly improved performance  >> over our current EV7. >sN >So, what this means is that EV7 won't even need any changes, they'll just use >a faster clock.  = You think that's easy?  Why don't you start up a CPU company?   > >And it may also mean that EV7 was purposefully given a slowerO >clock, and now, instead of doing EV79, they are just going to remove the speeds! >restriction they had put on EV7.   J There was never a "speed restriction" on EV7, except the requirements thatD it could be fabbed in useful quantities and that it work reliably in systems.    L >Sorry of I appear sceptical, but it was up to Comapq and HP to work hard to! >regain trust, which they didn't.e >sI >> Therefore, we have made the decision to provide our customers with theTF >> performance improvements that we had committed based on EV7z rather
 >> than EV79.p    K >In the end, HP breaks the "roadmap", "commitments" etc. They promised EV79tB >which was a process shink. And they will not honour that promise.  F EV79 was NOT a "process shrink", whatever that vague term might mean. I Bleeding-edge CPU fabbing is more art than science, and changing ANYTHINGpE in the design invites unexpected problems in the real chips when theynI arrive.  It's a very long distance from a CPU that works perfectly in theqJ simulator to one that works in silcon.  EV79 had real, algorithmic changes in the functional parts.    I Other verndors are not immune to this.  Why do you think Intel keeps 3 ortG 4 CPU design programs running in parallel?  If (when) one has problems,t< the others still have a chance to keep new chips coming out.   ------------------------------  % Date: Fri, 24 Oct 2003 21:32:12 -0400H* From: JF Mezei <jfmezei.spamnot@istop.com>$ Subject: Re: EV79 CANCELED !!!!!!!!!) Message-ID: <3F99D27B.D88302A1@istop.com>t   Robert Deininger wrote:nK > The Marvel systems went out the door with the best EV7 CPUs that could bes > produced at the time.a    M Yet, a few months later, HP admits that the current EV7 can be boosted by 15% H because existing productions of EV7 has shown its speed can be increasedL without changes. Either the engineers were totally incompetent in estimatingT the highest possible speed for their new EV7, or HP decided to lower the clock rate.  F > After several passes of EV79 prototype chips, they did not _work_ atK > anything approaching the design speed.  They were absolutely not ready tou > ship.   K And EV79 wasn't meant to ship until next year. Heck,. EV7 systems have only S now beund to ship, you'd think they'd wait a while before unleashing its successor.a  I Unless, of course, their goals are not to get the most profits from theiro
 alpha assets.:  < > post-2004.  The problems were largely out of HP's control.  N Sorry, I don't buy that. Compaq/HP should have known exaxctly what is involvedN in a process shrink expecially with their supposedly so cosy relationship withL Intel. So when they commotted to EV79, they would have known what to expect.  M If IBM can do it with Power, Sun with Sparc and Intel with 8086, HP should betH able to do it with Alpha. Note that EV79 was to have been just a process! schrink, not a logic/core change.d   >  Throwing tonsG > of money at it wouldn't have fixed the schedule problem, at least notr > enough to matter.e  J Why are they in such a hurry ? They just unleashed EV7. Customers would beY more than happy to see Alpha's lifetime extended, especially since IA64 is so lackluster.0  G > EV79 was NOT a "process shrink", whatever that vague term might mean.o  N Well, how come the roadmap, "plan of record" and other propaganda attemps from. Compaq/HP spoke of a process shrink for EV79 ?   ------------------------------    Date: 24 Oct 2003 18:02:42 -0700% From: a.greig@virgin.net (Alan Greig)  Subject: Guess that's now then? < Message-ID: <af3b9b31.0310241702.f8b7031@posting.google.com>   Shirley that's it? === Forwarded message === 2 From: Terry C. Shannon (terryshannon@mediaone.net)4 Subject: Re: Why Compaq won't succeed on IA64 either# View: Complete Thread (78 articles)i Original Format  Newsgroups: comp.os.vmse Date: 2001-06-29 13:35:52 PSTi  2 "Alan Greig" <a.greig@virgin.net> wrote in message$ news:3B3CDCAA.DBDEF417@virgin.net... >o >e > "Terry C. Shannon" wrote:  >tM > > But there is some encouraging behind-the-scenes marketing stuff going on. L > > CPQ actually set up a "War Room" to deal with the recent turn of events. > >t >aO > Hmm, can I get them extradited to the Hague for war crimes? ;) Chipicide muste > be a crime Shirley?t  K Shirley is my aunt, but whatever... the chipicide is not yet committed, nor J will it be until EV79 is past prime. But it might be time to start drawing% up an indictment for use in Den Haag.Y   ------------------------------  % Date: Fri, 24 Oct 2003 23:05:55 +0200t From: ML <lemkemch@t-online.de>e8 Subject: NFS mount of ODS5 disk fails on extendend names/ Message-ID: <oprxkfj5w23durha@news.t-online.de>c  J We've just started to experiment with an ODS5 volume.  We simply convertedI and old one with set volume/struc=5.  The volume is exported via NFS to adF Linux box.  If I access a file with an extended new file name I get an error on the Linux box:t  
 stein> ls -lst$ ls: test.test.ps: File name too long total 35K    18 drwxr-xr-x    2 ai26     remeis        512 Oct 24 22:58 awstats-5.dircL    18 drwxr-x---    2 ai26     remeis        512 Oct 24 22:58 awstats-59.dir     On VMS this looks fine:i   A400# dir tmpdisk:[ai26]   Directory TMPDISK:[AI26]  8 AWSTATS-5.DIR;1     AwStats-59.DIR;1    test^.test.ps;42   Total of 3 files.f   In the NFS server log I find    D %%%%%%%%%%%%  NFS$SERVER00952  24-OCT-2003 22:58:56.03  %%%%%%%%%%%%C %TCPIP-E-NFS_BFSCAL, operation LOOKUP   failed on file test.test.psf6 -TCPIP-S-NFS_CLIENT, uid=235 gid=200 host_name = stein- -SYSTEM-W-BADFILEVER, bad file version number     % Did we forget to configure something?o   A400# tcpip show ver  B    DIGITAL TCP/IP Services for OpenVMS Alpha Version V5.0A - ECO 3@    on a COMPAQ Professional Workstation P running OpenVMS V7.2-1   A400# sh sys/noproI OpenVMS V7.2-1  on node A400  24-OCT-2003 23:01:19.01  Uptime  8 22:52:29      Thanks,  Michaelr   ------------------------------  % Date: Fri, 24 Oct 2003 12:01:05 -0700 3 From: Alan Frisbie <Usenet01REMOVE@Flying-Disk.com>.+ Subject: Re: OpenVMS Freeware V6.0 Contentsz, Message-ID: <3F9976F1.10903@Flying-Disk.com>   Hoff Hoffman wrote:EK >   Attached below my sig file is the abstract for all submissions that are,I >   likely present on the OpenVMS Freeware V6.0 distribution, the releaseP7 >   that is expected to ship with OpenVMS Alpha V7.3-2.d  G Thank you very much for all your work on this.   I see several packagess@ that will be very useful, and I eagerly await its arrival.   The4 V7.3-2 kit will be an added bonus.   :-)   :-)   :-)  ? It is obviously a labor of love (love/hate?) for you, so I wantoA you to know how much we, the users, appreciate it.   Every singlen@ Freeware CD has had something (usually multiple somethings) that has made my life easier.   Thanks,r- Alan                 Remove "REMOVE" to reply    ------------------------------  # Date: Fri, 24 Oct 2003 17:53:54 GMTq& From: jlsue <jefflsxxxz@sbcglobal.net>; Subject: Re: OpenVMS I64 ISV application count now over 500 8 Message-ID: <gppipv0th774upp2ta2r745kpfo5dpcr0m@4ax.com>  F On Fri, 24 Oct 2003 02:47:55 GMT, "John Smith" <a@nonymous.com> wrote:    K >don't you think that a little advertsisng and promotion of VMS could boost K >those numbers. I'm pretty sure you could use a raise/bonus, and if the VMSFK >division doubled its profit by selling more VMS systems.......think of thes >possibilities....  G I believe that most of us in here have already stated that we'd love tooJ have more VMS advertising.  But continuing to expect that somehow we folksC in the trenches can do something about this will only increase yourt
 frustrations.o   ------------------------------  # Date: Fri, 24 Oct 2003 17:56:33 GMTa& From: jlsue <jefflsxxxz@sbcglobal.net>; Subject: Re: OpenVMS I64 ISV application count now over 500 8 Message-ID: <1tpipv02700ngrak0rdcntimcq97ujd1o7@4ax.com>  E On Fri, 24 Oct 2003 10:39:30 +0100, Andrew Harrison SUNUK Consultancyn. <Andrew_No.Harrison_No@nospamn.sun.com> wrote:  ; >Because with the exception of some of the more rabid Linuxa> >newsgroups comp.os.vms contains one of the biggest collection> >of BS merchants and as you also know you figure pretty highly >in that collection.  G What a crock.  Absent your constant bs discussions (and a couple of theeA other constant naysayers in here), 90% of this newsgroup would beg! VMS-related technical discussion.r   > G >No but thats mainly because people on the Solaris/AIX newsgroups don't H >spend most of their time as you Rob, Keith, Fred etc do fudding anotherH >vendor in an apparent attempt to hide the obvious deficiencies of their >platform/vendor.n  J Most of which is in direct response to your own fud and BS; without which,@ again, this group stay more on topic of VMS-related discussions.   ------------------------------  # Date: Fri, 24 Oct 2003 18:47:45 GMT # From: "John Smith" <a@nonymous.com> ; Subject: Re: OpenVMS I64 ISV application count now over 500eG Message-ID: <ltemb.38878$h61.9214@news01.bloor.is.net.cable.rogers.com>v   jlsue wrote:H > On Fri, 24 Oct 2003 02:47:55 GMT, "John Smith" <a@nonymous.com> wrote: >d >sG >> don't you think that a little advertsisng and promotion of VMS couldyD >> boost those numbers. I'm pretty sure you could use a raise/bonus,A >> and if the VMS division doubled its profit by selling more VMSF/ >> systems.......think of the possibilities....' > F > I believe that most of us in here have already stated that we'd loveF > to have more VMS advertising.  But continuing to expect that somehow@ > we folks in the trenches can do something about this will only > increase your frustrations.t    D Ah...so what you're saying (to paraphrase loosely and in the currentJ vernacular) is that carly (tm) needs to be 'bitch slapped upside the head'I for her action/inaction with respect to VMS? As a shareholder, that seemstK like a very appropriate thing to do to the HP CEO in this case. [aside: I'mhG sure that carly is a nice individual and probably throws a great dinner> party]  I It seems to me that if my company had a product line that was bringing intI between $400-800 million in profits annually, I'd want to be selling lots K more of it to more and more customers, new and exisiting, and I'd be taking K positive steps to ensure that, including using advertising in creative waysg4 to stimulate demand. Why doesn't HP want to do that?  J And as to the flip-side of that, if my company had a product line that wasI deemed crucial to its public image with major customers and wasn't makingtG money, I'd find ways of ensuring that it did become profitable and that E would include making every attempt to increase sales of that product.e    ! So to ask a few simple questions,k  K a) Just what does your management say to the troops when the question, "Whyu' aren't we advertising VMS?" gets asked?i  C b) Or reading between the lines of your response above, does nobodyi> internally ask that question for fear of being fired for gross insubordination?  I c) Is there a HP culture of "don't rock the boat" when it comes to askingcH about VMS advertising, marketing, and increasing sales to new customers?  G d) Or is it that Marcello, Gorham, Fortier, Blatz, Blackmore, Stallard, I carly (tm), et al. are just too busy counting the days left to retirements# and the big pension to do anything?   H e) Or is it that the people mentioned in d) don't believe in VMS and its future in the market or at HP?    K I think that most, if not all, of HP's OpenVMS customers would like to know  the answers to these questions..  E Feel free to forward this post any and all internally at HP for theiriE comments and responses. Invite them to participate here in c.o.v. andiH interact with customers directly. If they have persuasive rationales andL arguments then I'm sure that their points will be thoughtfully considered by	 all here.e   ------------------------------  % Date: Fri, 24 Oct 2003 15:29:55 -0400y* From: JF Mezei <jfmezei.spamnot@istop.com>; Subject: Re: OpenVMS I64 ISV application count now over 500 ( Message-ID: <3F997DAA.C4C883A@istop.com>   jlsue wrote: jlsue wrote:L > Most of which is in direct response to your own fud and BS; without which,B > again, this group stay more on topic of VMS-related discussions.    L Humm, who decided to cancel the promised EV79 ? Is is the naysayers here, or HP ?  L Are you saying that whenever HP makes a big mistake, that we are not allowedM to complain about that mistake and point the fingers on how HP is mismanaginge VMS ?u  I A few months ago, when HP specifically excluded mentioning VMS in all its J presentations, do you deny that HP admitted that it had resulted in enoughB complaints that HP would start to make an effort to mention VMS ?   H You may dislike the naysayers, but without those multiple complaints, HPL wouln't mention VMS anywhere, anytime and hHP probably would have hidden theA VMS web site in a very deep drak and humid place on its web site.o  H Like it or not, the constant pressure on HP is having a large success inM maintaining at least the minimal presence it has. Without it, it would have 0nG presence, or worse, continued statements from Scott Stallard stating HPSL expects VMS customers to migrate to HP-UX over time. Or did you conveniently forget that one already ?    ------------------------------    Date: 24 Oct 2003 16:43:23 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)a; Subject: Re: OpenVMS I64 ISV application count now over 500t3 Message-ID: <UR0jRBks1MT2@eisner.encompasserve.org>s   In article <bnb8gj$e3p$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes:  A > Check my past posts I have also never suggested that OpenVMS isr? > crap, rubbish, terrible, slow, insecure etc all words used byrA > members of the choir describe Solaris, Linux, UNIX. Mostly usedl@ > by people who have absolutely no experience of the system that > they are describing.  B    I think they find those of us who do have the experience fairly    reliable.  B > I am carefull not describe OpenVMS in these terms because I have< > no current experience of OpenVMS, I used VMS when I was at= > univercity and later when I used to install DECNET on SunOSi? > at Digital/Sun sites. Its a great pity that the vast majority > > of Solaris slaggers off who fall directly into this category) > as well don't follow the same strategy.a  D    VMS has changed a lot of the past 20 years, so your experience as    way back is hardly relavent.   L    An end user can hardly be considered an expert on the topics of security,D    reliablilty, API design, maintainability, TCO, or appropriatenessH    for a particular application.  And those are what we slam eunichs andE    Windoze on.  And those are what many of us have experience with on9    the systems we slam.-   ------------------------------  # Date: Fri, 24 Oct 2003 23:24:46 GMT<# From: hoff@hp.nospam (Hoff Hoffman)RH Subject: Re: OpenVMS Licensing on VAX emulators (was: Re: VAX emulators)2 Message-ID: <2ximb.7791$Mg7.3708@news.cpqcorp.net>  _ In article <CIEJLCMNHNNDLLOOGNJIKEDHIFAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> writes: ; :HP's web site contains part numbers for transfering OS andn< :layered product license to CHARON-VAX.  There has been some; :discussion in the past about other emulators like simh andk< :eVAX.  I don't think these are in a commercial form, but if= :they were, and you wanted to use one would the same transfero! :license as for CHARON-VAX apply?T  K   This is centrally a question for the HP corporate software license folks,tJ   and probably not one that can ne answered within the newsgroups.  AFAIK,J   you can purchase an OpenVMS license for an unsupported platform, so thisJ   doesn't look to be much of a stretch from there, but this is not my callK   and not my decision to make.  Again, this is centrally a question for therI   corporate license folks.  (The folks that you will want/need to discusse$   this with do not read newsgroups.)     Licensing information:  3     http://licensing.hp.com/swl/view.slm?page=indext  0   There's an email address at the above webpage.  N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.como   ------------------------------  % Date: Fri, 24 Oct 2003 17:11:59 -0700 # From: "Tom Linden" <tom@kednos.com> H Subject: RE: OpenVMS Licensing on VAX emulators (was: Re: VAX emulators)9 Message-ID: <CIEJLCMNHNNDLLOOGNJIMEDKIFAA.tom@kednos.com>t   >-----Original Message-----R+ >From: Hoff Hoffman [mailto:hoff@hp.nospam]t' >Sent: Friday, October 24, 2003 4:25 PM  >To: Info-VAX@Mvb.Saic.ComI >Subject: Re: OpenVMS Licensing on VAX emulators (was: Re: VAX emulators)d >e >t? >In article <CIEJLCMNHNNDLLOOGNJIKEDHIFAA.tom@kednos.com>, "Tomd! >Linden" <tom@kednos.com> writes:s< >:HP's web site contains part numbers for transfering OS and= >:layered product license to CHARON-VAX.  There has been somea< >:discussion in the past about other emulators like simh and= >:eVAX.  I don't think these are in a commercial form, but ifC> >:they were, and you wanted to use one would the same transfer" >:license as for CHARON-VAX apply? > L >  This is centrally a question for the HP corporate software license folks,K >  and probably not one that can ne answered within the newsgroups.  AFAIK,eK >  you can purchase an OpenVMS license for an unsupported platform, so thissK >  doesn't look to be much of a stretch from there, but this is not my calltL >  and not my decision to make.  Again, this is centrally a question for theJ >  corporate license folks.  (The folks that you will want/need to discuss% >  this with do not read newsgroups.)  >s >  Licensing information:j >k4 >    http://licensing.hp.com/swl/view.slm?page=index >01 >  There's an email address at the above webpage.3  K Yes, thanks Hoff, I had already sent off an email, but I thought this mighto have surfaced here before. >b1 > ---------------------------- #include <rtfaq.h>e >-----------------------------L >    For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq3 > --------------------------- pure personal opiniong >---------------------------F >        Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.com >, >---' >Incoming mail is certified Virus Free.z; >Checked by AVG anti-virus system (http://www.grisoft.com).bB >Version: 6.0.530 / Virus Database: 325 - Release Date: 10/22/2003 >t ---a& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.530 / Virus Database: 325 - Release Date: 10/22/2003    ------------------------------  # Date: Fri, 24 Oct 2003 18:44:48 GMTl, From: "warren sander" <warren.sander@hp.com> Subject: Re: OpenVMS Roadmap2 Message-ID: <Aqemb.7779$A37.4910@news.cpqcorp.net>   I know it's an html page.-  G but if you open a powerpoint presentation and do a 'save as a web page'hL this is the best option that I've found for setting up the various paramters allowed.  4 If you can come up with a better method let me know.2 Right now I give you the ppt and the html version.I On some presentations I also give PDF and PS (print via powerpoint of the 2 slide show via the adobe generic postscript driver level 1 postscript),  H I didn't say that the page is for an 800x600 display but that powerpoint) calls it that in it's save as html dialogw      7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in message # news:3F9822A5.893CB8EB@istop.com...i > warren sander wrote:J > > not want to dump on MS but when converting a ppt to html the best size to( > > pick is the 800x600 because it makes > > a fixed size.M >l >0J > That particular page isn't a PPT issue. It is an HTML issue. With proper HTML, I > the frames would be adjusted proportional to the screen size, and then,  one7F > would use scroll bars to view the 800-600 content inside of a frame. >3I > The way the page was designed, the unnecessary fluff was in frames thatS tookJ > up all of the screen real estate, and the frame where the PPT would have been! > displayed was totally occluded.: >2G > Use percentages to size frames, not fixed point sizes. HTML  must not  assume a > specific screen size.5 > K > Buy yourself a mac where you can easily change screen size and you'll see  howoF > your pages behave at different screen sizes. )called "resolution" in windows 5 > world even though screen resolution never changes).l   ------------------------------  % Date: Fri, 24 Oct 2003 15:50:33 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com> Subject: Re: OpenVMS Roadmap) Message-ID: <3F99827E.60624FD8@istop.com>e   warren sander wrote: >  > I know it's an html page.g > I > but if you open a powerpoint presentation and do a 'save as a web page' N > this is the best option that I've found for setting up the various paramters
 > allowed.  . One line of interest in the generated HTML is:; <frameset rows="210,*,35,120" framespacing=1 frameborder=1>0  C Inside this frameset, you expect at least one frame to be 275 high.e  G The top frame is very large and serves no purpose for the presentation.   K Also, removing the "noresize" from all <frame tags would greatly help usersaH reduce unnecessary frames, leaving more room for frames that contain theL information you seek. When you specify the "noresize" as is the case in that> page, users are not allowed to make a frame smaller or larger.   ------------------------------  # Date: Fri, 24 Oct 2003 21:20:00 GMT 1 From: Bob Blunt <robert.blunt@hp.donotspamme.com>  Subject: Re: Product problem2 Message-ID: <4Igmb.7786$xd7.2570@news.cpqcorp.net>   Hank Vander Waal wrote:p  > Yes the files are really there > N > I booted off a 7.1 cd and tried to remove Decnet V and it came back and saidM > that dra0 is NOT a system disk...  Tried to show the installed products and. > it errors outd > N > Can the PCSI files become corrupt?  If so is there any way to fix them?  Can) > one remove them and have them rebuilt??  >  >  > -----Original Message-----D > From: Bob Koehler [mailto:koehler@eisner.nospam.encompasserve.org]+ > Sent: Wednesday, October 22, 2003 3:04 PMt > To: Info-VAX@Mvb.Saic.Com  > Subject: Re: Product problem >  > N > In article <PBEDIAGOKEDCKCLPJKAFMEMNCDAA.hvanderw@comcast.net>, "Hank Vander& > Waal" <hvanderw@comcast.net> writes: > > >>But if I do a directory of sys$common:[sysexe] it is there ! >  > 3 >    directory/full to make sure it's really there.  >  >  >   B You CAN get to and work with data on DRA0 - a RAID disk, correct? 1 Device mounts and does I/O correctly, in general?   B Yes, there are several things that can happen to corrupt the PCSI : databases on the disk.  What exact errors do you get from:   $ PRODUCT SHOW PRODUCT         ande $ PRODUCT SHOW HISTORY   bobM   ------------------------------  % Date: Fri, 24 Oct 2003 14:24:43 -0400e* From: JF Mezei <jfmezei.spamnot@istop.com>& Subject: Re: Spam from Hewlett-Packard) Message-ID: <3F996E66.3CB7139D@istop.com>0  K If the HP employee went over your head directly to your president, then the K response should come from your president and also go over the HP employee'su* head and talk from president to president.    9 Your president should state 2 things to the HP president:SL 1- ask HP to ensure HP employees don't bother him directly with petty stuff.  H 2- specify that the president trusts his employees/consultants , perhapsK noting that you are well known in the VMS community with tons of experiencerK and well respected and that if you state that you have been abused by an HPoL employee not respecting your request to be taken out of a mass mailing list,H then the president fully supports you and not HP and that if HP does notJ immediatly correct the problem and if the HP president does not immediatlyI apologize directly to  the employee (you), that he (the president of youraN copmpany) will remember this event when the time comes to evaluate upcoming IT purchase decisions.    ------------------------------  # Date: Fri, 24 Oct 2003 23:23:46 GMTr# From: "John Smith" <a@nonymous.com>y& Subject: Re: Spam from Hewlett-PackardJ Message-ID: <6wimb.281192$ko%.162011@news04.bloor.is.net.cable.rogers.com>   Alan,e  1 Ask for the same injunction granted in this case.g  ! =================================   L http://story.news.yahoo.com/news?tmpl=story&cid=569&ncid=738&e=1&u=/nm/20031 024/tc_nm/tech_spam_lawsuit_dc    , California Wins Its First Anti-Spam Judgment October 24, 2003   By Barbara Grady  H SAN FRANCISCO (Reuters) - California won its first anti-spam judgment onE Friday when a court fined a marketing firm $2 million for sending out G millions of unsolicited e-mails telling people how to spam, the state'sl attorney general said.  J Attorney General Bill Lockyer brought the case against PW Marketing of LosK Angeles County and its owners Paul Willis and Claudia Griffin in 2002 undereJ a 1998 state anti-spam law. The law was strengthened last month to make it easier to sue spammers.   K Lockyer's spokesman Tom Dressler said while this case was decided under theXI original statute, *the attorney general's office expects in the future it6< will be easier to try cases under the updated, tougher law.*  J PW Marketing and Willis and Griffin were charged with sending out millionsK of e-mails advertising "how to" guides on spamming and long lists of e-mailF
 addresses.  B The judgment, which Lockyer said will be the model for future spamE injunctions, forbids PW Marketing from sending unsolicited commercial E e-mail, accessing computers that belong to other people without theiroH permission and disguising its identity by sending e-mails that appear to# originate from a different address.   K *The injunction also forbids Willis and Griffin for 10 years from owning orw9 managing any business that advertises over the Internet.*c *(HP - take note)*  K The tougher measures in the new statute include allowing individuals to suetJ spammers and collect damages of up to $1,000 per e-mail. Another provisionH forbids sending unsolicited e-mail advertisements unless recipients give) prior permission to receive such e-mails.o  L The old law made it illegal to send to recipients who had specified they didK not want to receive e-mail advertising. It also required senders to providelG a phone number or valid e-mail address for opting out on each e-mail --tE something the company did not do, the attorney general's office said.    ------------------------------  % Date: Fri, 24 Oct 2003 19:25:55 -0700 3 From: Alan Frisbie <Usenet01REMOVE@Flying-Disk.com>s& Subject: Re: Spam from Hewlett-Packard, Message-ID: <3F99DF33.30004@Flying-Disk.com>   John Smith wrote:  > Alan Frisbie wrote:s > 6 >> Sadly, this problem has taken a turn for the worse. >>: >> Apparently, Mr. Swartwood decided to get revenge on me.= >> Yesterday afternoon, he called my client, Nelson Nameplated; >> Company (where I had been getting the spams), and left ae7 >> voice-mail message for the president of the company.e >>A >> He falsely claimed that I had sent bogus and harassing e-mailss4 >> to him.   He did not offer any proof or examples.  9 [NOTE: After a request from Nelson, he has yet to do so.]V  > >> I have never sent any e-mail of any kind to him, nor have I@ >> ever asked anyone to do so.   I think it would be wrong, justC >> as I think it is wrong for him to make such claims to my client.t >>@ >> I am deeply saddened that a Hewlett-Packard employee would doD >> such a thing, especially someone in a position of responsibility.0 >> The company is greatly diminished in my eyes.  J > Retain a lawyer to sue the bastard and HP for defamation of character...  > sue in both CA and TX.3  > I would prefer to avoid any legal action, however warranted it? might be.   My goal is to convince HP to change their practicesl7 so that they cannot be used as a tool to harass others.   ? I would be perfectly satisfied by a letter from Dan Swartwood's = manager, apologizing for the behavior of his employee, and ane< agreement to use only confirmed (closed-loop) opt-in for all? their mailing lists.   Needless to say, this would require thatf@ all existing mailing lists be confirmed before using them again.   He can always contact me at:   Alan E. Frisbieh Flying Disk Systems, Inc.  4759 Round Top Drive Los Angeles, CA 90065e  ? I hope that some concerned HP employee will make sure this getsd to the right person.  . Alan                  Remove "REMOVE" to reply   ------------------------------  % Date: Fri, 24 Oct 2003 22:39:12 -0400n* From: JF Mezei <jfmezei.spamnot@istop.com>& Subject: Re: Spam from Hewlett-Packard) Message-ID: <3F99E22A.2FAC7C04@istop.com>r   Alan Frisbie wrote: A > I would be perfectly satisfied by a letter from Dan Swartwood'sa8 > manager, apologizing for the behavior of his employee,    K If the culprit went over your head directly to your president, wasting youreJ president's time, then it is your president that needs to complain to HP'sK president. HP's president will pass it down to whatever level she thinks is  applicable.   9 You can draft the letter and have your president sign it.v   ------------------------------  # Date: Fri, 24 Oct 2003 22:07:41 GMT # From: "John Smith" <a@nonymous.com>e4 Subject: Swan Song For The Alpha Processor - TechWebJ Message-ID: <Nohmb.31842$3f.24733@twister01.bloor.is.net.cable.rogers.com>  2 http://www.techweb.com/wire/story/TWB20031023S0014  ! Swan Song For The Alpha Processor-! October 24, 2003 (11:45 a.m. EST) ! By W. David Gardner, TechWeb Newsm  I The Alpha processor -- the 64-bit machine on life support for much of itsDG tormented lifetime -- has finally been placed on death row. Its presentmH proprietor -- the Hewlett Packard Co. -- said the final iteration of the Alpha began shipping this week. J "When it was introduced by Digital Equipment Corp. in November of 1992, itL was the fastest chip on the planet," said Terry Shannon, a computer industryH consultant who watched the processor grow from conception at the DigitalJ Equipment Corp. to its adolescence at the Compaq Computer Co. to its finalJ resting place at HP. "The new server still goes like a bat out of hell. IfK you have to have an end of the line, it's better to go out with a bang than  with a whimper."  K Although this week's shipment of the AlphaServer GS1280 marks the beginningl@ of the end for the Alpha -- it will be the final major processorK improvement -- HP will support the machine for years and provide additionalaH hardware and software updates. "The new products," said Rich Marcello, aL senior vice president at HP, "...are designed to meet the computing needs ofC AlphaServer customers while ensuring a smooth transition to Itaniumt@ architecture-based HP integrity servers on their own timetable."  F Nobody in charge of the processor over its long history seemed to knowJ exactly what to do with it or how to market it very effectively. There areK those who think it played a central role in the demise and disappearance of J the Digital Equipment Corp. first and later Compaq. And then there was theK machine's role in the murky proceedings between Digital and Microsoft. WhenmG that music ended, Microsoft ended up with Windows NT and with Digital'shL legendary OS software designer Dave Cutler while Digital began fading out of the computer industry for good.t  E Shannon estimates there are 400,000 Alpha systems in use; taking intohJ consideration terminals and timesharing applications that translates to 10H million people "touched by" the processor. He said its user base is veryE active and, as a typical example, noted that there were more than 400qF attendees at an Alpha users group he addressed in the Netherlands last8 summer. Shannon's consultancy is located in Albuquerque.  K What became the Alpha's workhorse OS -- VMS -- remains an important part ofsH the Digital and Alpha legacy and is likely to play a role in the ItaniumG family. Shannon noted the Alpha didn't have a full-blown OS when it wascK announced and it wasn't until VMS was ported over from Digital's VAX family3F several months later that the 64-bit processor got its first major OS.G Shannon said VMS is enjoying a revived second wind these days primarilylL because it is so secure and robust and security-minded users are looking for2 such software. VMS is being ported to the Itanium.  K Shannon, who sold 100,000 copies of a best-selling book on VMS, said the OSeL could be "very important" for HP's Itanium particularly for users who need a "rock solid" OS that is secure.n  E In a statement, HP said the GS1280 is the most powerful system in theeL AlphaServer line. The EV7-based line has been enhanced with a doubled memoryL capacity reaching as much as eight gigabytes per central processing unit andK a new memory channel interconnect adapter that lowers the cost of clustered I systems. An optional input/output drawer scheme quadruples I/O rack spacer	 capacity.V  G HP's GS1280 contains 64 of the 64-bit Alphas. The EV7-based AlphaServermL family runs OpenVMA and Tru64 Unix operations systems. HP's strategy for theG Alpha is to move users to its 64-bit Itanium family. The HP program for B moving Alpha users is called the AlphaServer Retain Trust program.  J The Alpha has crossed paths with other major computer manufacturers. AlphaI Processor, a joint venture of Compaq and Samsung was discontinued severalhA months ago. Sun has made a run at the Alpha user base and IBM hasmD manufactured the Alpha chips in its Vermont semiconductor operation.  H When the Alpha was announced Digital promised to keep the machine in theI technology forefront maintaining twice the speed of competing processors.eE But the machine often had problems with operating system software. NoeJ effective OS was developed in time for the first users. Digital's attemptsJ to develop alternative operating systems failed and VMS was ported over toK the Alpha from the VAX family. At Digital, Dave Cutler was writing softwarewK for 64-bit machines. He later developed software for Microsoft on a DigitalwC Alpha. At some point, Digital and Microsoft entered into protracted L negotiations and deliberations involving 64-bit machines. Finally, MicrosoftE agreed to provide software for the Alpha but later the software giantg# withdrew its support for the Alpha.r  J "The Alpha never made it to the desktop," said Shannon. "No desktop equals+ no volume. And no volume equals no income."    ------------------------------  # Date: Sat, 25 Oct 2003 00:30:35 GMTt5 From: rdeininger@mindspringdot.com (Robert Deininger)t8 Subject: Re: Swan Song For The Alpha Processor - TechWebL Message-ID: <rdeininger-2410032040430001@user-105n8f9.dialup.mindspring.com>  J In article <Nohmb.31842$3f.24733@twister01.bloor.is.net.cable.rogers.com>,$ "John Smith" <a@nonymous.com> wrote:  3 >http://www.techweb.com/wire/story/TWB20031023S0014. >m" >Swan Song For The Alpha Processor" >October 24, 2003 (11:45 a.m. EST)" >By W. David Gardner, TechWeb News >y ...t  L >Although this week's shipment of the AlphaServer GS1280 marks the beginningA >of the end for the Alpha -- it will be the final major processor L >improvement -- HP will support the machine for years and provide additional  >hardware and software updates.   G Well they got that part wrong.  The CPU speed-up is still to come.  ThefI recent system announcement was about more memory, higher processor count,h and IO enhancements.   ------------------------------  % Date: Fri, 24 Oct 2003 21:39:46 +0200l- From: Marc Van Dyck <marc.vandyck@brutele.be>a# Subject: Re: Time server protocole?s: Message-ID: <mesnews.c5137d3a.d855b47d.164.584@brutele.be>   "H Vlems" a pens trs fort :r/ > "Dirk Munk" <munk@home.nl> schreef in berichtp. > news:bnb01q$425$1@news1.tilbu1.nb.home.nl... >> Didier Morandi wrote: >>> Helli Everybodo, >>>  >> rN >> 1. connect a supported time receiver to a serial port of your system. This N >> radio receiver will use a precise time signal from a powerfull transmitter J >> and your VAX can sync its clock on that signal. (DCF77 is a well known M >> German time provider with a transmitter in Frankfurt. But there is also a  N >> French station). If you do this your system can also act as a timeprovider  >> for NTP and DTSS. >> r > Dirk,n >tK > could you elaborate on this option? What kind of equipment do you need to   > run your own stratum 1 server? >s > Hans  F A radio clock receiver with a serial port (made by many manufacturers)D and a small program (I once made one in Pascal that I could retrieveF if you're interested) to read the port (serial port on a LAT or TELNETC terminal server is OK) and feed the time to DTSS so that it acts as.E a DTSS time provider. Took me a week to do it, including the program,sC altough I didn't write any code (other than DCL) in the 10 previousu years.   Marc.    -- -
 Marc Van Dyck-   ------------------------------  % Date: Fri, 24 Oct 2003 23:40:52 +0200l From: Dirk Munk <munk@home.nl># Subject: Re: Time server protocole?e2 Message-ID: <bnc6cc$kf4$1@news1.tilbu1.nb.home.nl>   H Vlems wrote:/ > "Dirk Munk" <munk@home.nl> schreef in berichte. > news:bnb01q$425$1@news1.tilbu1.nb.home.nl... >  >>Didier Morandi wrote:  >> >>>Helli Everybodo,s >>>i >>L >>1. connect a supported time receiver to a serial port of your system. This >  > radiot > J >>receiver will use a precise time signal from a powerfull transmitter and >  > your > K >>VAX can sync its clock on that signal. (DCF77 is a well known German time F >>provider with a transmitter in Frankfurt. But there is also a French >  > station).r > G >>If you do this your system can also act as a timeprovider for NTP and  >  > DTSS.f >  > Dirk,t > K > could you elaborate on this option? What kind of equipment do you need to.  > run your own stratum 1 server? >  > Hans >  Good question.M Appendix B of the DECnet-Plus DECdts Management Guide has a listing of these SN receivers. It also claims that there should be a DTSS$provider_acts.c example . program in sys$common:[syshelp.examples.dtss].  N However I can't find the file or the directory. A long time ago it was there. P I'm sure about that because I used this source to build a executable for a Hopf 7 Clock connected to a serial port of my VAX workstation.   M Five receivers are mentioned in this manual, but since the manual dates from 1J 1996 it is questionable if any of those receivers are still on the market.   ------------------------------  % Date: Fri, 24 Oct 2003 18:27:53 -0400D  From: John Santos <JOHN@egh.com># Subject: Re: Time server protocole? 4 Message-ID: <1031024182601.403B-100000@Ives.egh.com>  % On Fri, 24 Oct 2003, Dirk Munk wrote:u   > H Vlems wrote:1 > > "Dirk Munk" <munk@home.nl> schreef in bericht30 > > news:bnb01q$425$1@news1.tilbu1.nb.home.nl... > >  > >>Didier Morandi wrote:h > >> > >>>Helli Everybodo,  > >>>  > >>N > >>1. connect a supported time receiver to a serial port of your system. This > > 	 > > radioo > > L > >>receiver will use a precise time signal from a powerfull transmitter and > >  > > your > > M > >>VAX can sync its clock on that signal. (DCF77 is a well known German timeeH > >>provider with a transmitter in Frankfurt. But there is also a French > > 
 > > station).- > > I > >>If you do this your system can also act as a timeprovider for NTP andb > > 	 > > DTSS.j > > 	 > > Dirk,I > > M > > could you elaborate on this option? What kind of equipment do you need to0" > > run your own stratum 1 server? > >  > > Hans > >  > Good question.O > Appendix B of the DECnet-Plus DECdts Management Guide has a listing of these lP > receivers. It also claims that there should be a DTSS$provider_acts.c example 0 > program in sys$common:[syshelp.examples.dtss].  % ??? sys$common:[syshlp.examples.dtss]i   (No "e" in syshlp...)h   It's on my system.  V7.3-1.a   > P > However I can't find the file or the directory. A long time ago it was there. R > I'm sure about that because I used this source to build a executable for a Hopf 9 > Clock connected to a serial port of my VAX workstation.e > O > Five receivers are mentioned in this manual, but since the manual dates from :L > 1996 it is questionable if any of those receivers are still on the market. >  >  >  >    -- . John Santosi Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   Date: 24 Oct 2003 18:16 CDTB. From: carl@nospam.gerg.tamu.edu (Carl Perkins)# Subject: Re: Time server protocole?o4 Message-ID: <24OCT200318160725@nospam.gerg.tamu.edu>  & Didier Morandi <no@spam.com> writes... }Mike Rechtman wrote:oG }> Didier: What stack are you using? Setting up NTP on a TCPIP V5.3 is  H }> fairly immediate - enable and start in TCPIP$CONFIG, copy a template G }> file to *.CONF and edit -  if you have connectivity to a NTP server o, }> (i.e. is your firewall open on port 123?) } P }I tell you, one post, three answers, two pages of RTFM, a few DCL commands and 6 }you know the first name of my uncle. VMS, I love you. } 	 }BUT !...d } 4 }DTL02> ty SYS$SPECIFIC:[TCPIP$NTP]TCPIP$NTP_RUN.LOG' }24 Oct 14:39:55  ntpd version = 3-5.91uO }24 Oct 14:39:55  tickadj = 97, tick = 976, tvu_maxslew = 99231, est. hz = 1024-& }24 Oct 14:39:55  precision = 976 usecQ }24 Oct 14:39:55  drift file 'SYS$SPECIFIC:[TCPIP$NTP]TCPIP$NTP.DRIFT' does not eu  }xist: no such file or directory3 }24 Oct 14:39:55  will attempt to create drift filemN }24 Oct 14:39:55  read drift of 0 from SYS$SPECIFIC:[TCPIP$NTP]TCPIP$NTP.DRIFT; }24 Oct 14:44:48  synchronized to 145.238.110.68, stratum=2s1 }24 Oct 14:44:49  time reset (slew) 90.183853 secm& }24 Oct 14:44:49  synchronization lostL }24 Oct 14:44:51  Previous time adjustment incomplete; residual 0.008198 secL }24 Oct 14:44:52  Previous time adjustment incomplete; residual 0.003611 secL }24 Oct 14:44:53  Previous time adjustment incomplete; residual 0.003513 secL }24 Oct 14:44:54  Previous time adjustment incomplete; residual 0.003318 sec }.../..  }(many lines ommited)  }  }Where is the HOWTO??????? }  }D.   F I would guess that the variation in timing in talking to the server isD larger than is optimal (perhaps an overloaded branch of the internetA between you and it). This could cause many more adjustments to bet" attempted than are really needed.   F You might have better luck just by picking a different server. You canF probably specify multiple servers in your configuration file, in whichF case it will measure delays and the degree of randomness in the delaysF for each server and pick the best server to use (the system I use as aB local server switches between the systems it uses as servers a few times per day, usually).  H I'm using Multinet's XNTP rather than the one from TCPIP Services. SinceG the 4th of June (the start of the current log) it has logged a total ofmE 145 messages on this node, 174 on another (both client only), and 693tE on the node is use as a local server for a few systems which gets itstF time from servers that are just a couple of network hops away. This isD clearly much less than you are getting. A message per second is just rediculous.)  H There may be parameters you can set in the configuration file to improve the situation.  : It is also possible that your system's clock drifts a lot.  H A possibility... Does the message rate slow down when you log off? ThereI used to be a problem (fixed by an ECO, the Graphics V5 ECO on VMS V7.2-1)nG that caused the ELSA Gloria Synergy cards to induce clock drift when itiH was in use. Are you using that card? Before I applied the ECO to fix it,E I would get a lot of messages - I think it was a dozen or so messagessF per hour when I was logged in but not doing much, and a few per minute1 when doing things that updated the display a lot.o   --- Carl   ------------------------------  # Date: Fri, 24 Oct 2003 23:32:08 GMTh- From: hoffman@xdelta.hp.nospam (Hoff Hoffman)w# Subject: Re: Time server protocole?e2 Message-ID: <YDimb.7793$Mg7.2178@news.cpqcorp.net>  W In article <1031024182601.403B-100000@Ives.egh.com>, John Santos <JOHN@egh.com> writes:r& :On Fri, 24 Oct 2003, Dirk Munk wrote: :i :> H Vlems wrote:e ..    +   I am not sure who wrote what here, but...h    P :> Appendix B of the DECnet-Plus DECdts Management Guide has a listing of these Q :> receivers. It also claims that there should be a DTSS$provider_acts.c example o1 :> program in sys$common:[syshelp.examples.dtss].i :c& :??? sys$common:[syshlp.examples.dtss] :4 :(No "e" in syshlp...) :r :It's on my system.  V7.3-1.    @   Please take the time to acquire and to read the OpenVMS FAQ.    H   This particular DTSS$NTP_PROVIDER.C module is very specifically called<   out in the FAQ, as its availability is version-specific.    I   Details on general OpenVMS time-keeping, time servers, daylight savingseJ   time (DST), the tool to switch between standard time and daylight time, G   version-specific TDF issues, etc., are all discussed and/or otherwisecB   referenced within the OpenVMS FAQ.  URL references are included.  %     http://www.hp.com/go/openvms/faq/   p  N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.coms   ------------------------------  # Date: Fri, 24 Oct 2003 17:54:34 GMTe" From:   VAXman-  @SendSpamHere.ORGM Subject: Re: Tru64 unix Layeredproducts won't install on my OpenVMS machines!d0 Message-ID: <00A27DBE.D29A652F@SendSpamHere.ORG>  j In article <7DgeYt1QkEUV@eisner.encompasserve.org>, kuhrt@nospammy.encompasserve.org (Marty Kuhrt) writes:d >In article <nf9mb.7746$zF6.5549@news.cpqcorp.net>, Mark Schafer <mark.schafer@nospamhp.com> writes:H >> I'm checking with DSPP on this issue.  On the surface, it looks like J >> they shipped the UNIX Q4 kit to the wrong list.  The OpenVMS Q4 update  >> isn't due yet.s >> N >> Sorry for the inconvenience.n >> o >> Mark. >> w >e? >I checked with my DSPP rep and she said that that is the case,V@ >the Tru64 kit went to the wrong list.  The VMS kit will be out  >in November (IIRC).   >o< >Also, the old DSPP (ISV/ASAP/ETC) 800 number now goes to a ? >"call our hot talk line" message.  I didn't call the suggesteds@ >800 number they mentioned in the message, but might it be a new> >way to get developers interest?  "oooooh, VMS developers are  >soooo HOT, baby!"  H What!  You mean DSPP dosn't stand for Dirty Sex Pornography Program? ;)    -- rL VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)COM             5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------  % Date: Fri, 24 Oct 2003 13:40:21 -0500e( From: Wayne Sewell <wayne@tachysoft.com>M Subject: Re: Tru64 unix Layeredproducts won't install on my OpenVMS machines!n0 Message-ID: <00A27DBC.D56E9521.10@tachysoft.com>  / >From: Mark Schafer <mark.schafer@nospamhp.com>n/ >Organization: Partner Technical Services - DRCrN >Subject: Re: Tru64 unix Layeredproducts won't install on my OpenVMS machines! >hF >I'm checking with DSPP on this issue.  On the surface, it looks like H >they shipped the UNIX Q4 kit to the wrong list.  The OpenVMS Q4 update  >isn't due yet.. >p >Sorry for the inconvenience.: >e    L I got a reponse to the message I left on the DSPP phone line.  They say theyO are aware of the problem, and it was an adminstrative error as we had surmised..5 The vms kit doesn't ship until the end of november.     D So the only real inconvenince is to fedex, who transported a packageM needlessly, and to the trash boys, who have to haul it to the landfill (whichmL should be the ultimate destination for all eunuchs software).  I doubt fedexO minds too much, since they got paid.  Don't know the feelings of the trash boysaM on the matter.  Maybe they will give the eunuchs CDs a place of prominence on & the truck, right next to the dog shit.   WaynetO =============================================================================== N Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html   oO ===============================================================================kH Randolph Duke (in Trading Places): "Mother always said you were greedy."1    Mortimer Duke: "She meant it as a compliment!"t   ------------------------------  % Date: Fri, 24 Oct 2003 15:43:05 -0700s# From: "Tom Linden" <tom@kednos.com>- Subject: VAX emulators9 Message-ID: <CIEJLCMNHNNDLLOOGNJIKEDHIFAA.tom@kednos.com>n  : HP's web site contains part numbers for transfering OS and; layered product license to CHARON-VAX.  There has been some-: discussion in the past about other emulators like simh and; eVAX.  I don't think these are in a commercial form, but if-< they were, and you wanted to use one would the same transfer  license as for CHARON-VAX apply?  / The HP web site doesn't make any mention of it.l> http://h71000.www7.hp.com/openvms/sri-charon-vax-emulator.html --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.530 / Virus Database: 325 - Release Date: 10/22/2003    ------------------------------    Date: 24 Oct 2003 16:10:18 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen)m; Subject: Re: VAXscan now available on OpenVMS Alpha systems 3 Message-ID: <nZC0iv8Y3heG@eisner.encompasserve.org>e  W In article <3f99383f$0$251$636a55ce@news.free.fr>, Didier Morandi <no@spam.com> writes:r1 > (from http://h71000.www7.hp.com/news/scan.html)e > 0 > VAXscan now available on OpenVMS Alpha systems > N > VAXscan is a block-structured programming language that runs in the OpenVMS S > Alpha and OpenVMS VAX environments to build tools to manipulate text strings and a
 > text files.   B But on Alpha, there is none of the SCAN-specific debugger support,? so one must _build_ (all but the simplest) tools on VAX and use  Alpha only to _run_ them.o   ------------------------------  % Date: Fri, 24 Oct 2003 21:07:45 +0200 " From: Didier Morandi <no@spam.com>7 Subject: Re: [survey] why do you still use VAX systems?h2 Message-ID: <3f997881$0$228$636a55ce@news.free.fr>  $ One of Tom Linden's Customers wrote:  3 >>>>Have you thought about the CHARON VAX emulator?. >>>aH >>>Yes that was considered but the Sales group did not think we would be5 >>>happy with the performance compared with our 7840.g  F Let me laugh. A further announce to come on this subject will take by  surprize more than one...-   D.   ------------------------------   End of INFO-VAX 2003.591 ************************penvms/faq3 > --------------------------- pure personal opiniong >---------------------------F >        Hoff (Stephen) Hoffman   OpenVMS Engineering    sĜJ	SY;Jk4?h66* KXCW6hP"'PD<e::4O7a9h砡'%_tߢUYe	ԉI2/U4qSt	Q++:L]j|f%[T֟Zl߯qBI;KO&&t[԰X.gQc=SL5iJgb߲LLbWR$RH7	ĞH"=H|[|R&7-~Xt]NpZZq0i9"qn8t4eQTZ_o,!soheAdm[ӕNs][HxA6+qS
e]GXV)7R^׶,ڸ_뤾6۩Z@:4:"0!r$z/qsG}xݮw;۲^@0Jf*Ѐk#	tP-wg<ji0ǅh5S=7H;hAHA[
Od(rt{oƻUTF6sտ{3 w>i6
$|z;<tҽr=Ѣ0=q)/9A.v3Tռv=LcJsezIN_=,KzA)ݼad,B_$N4__zCmQst(uo99ʟFeVkx
_L*mS0O{OSWYYXZЋ';rA;LyvI7qݔt33Q
fCF8+(u?u<w\~ϾyΘET44wX9,Tih\ɮ|opPR/$toyQ4/ٝ*qE-W?pab+9vfMN틭`NhH:EG͊^iVtgOf`i.-nQ#?Wfk_B5Nfǵ(^d)bW3:x|WyrGnez=HY/W?ryLequZ,YױvҮ[Ե5<v:2Vn<y,l'\&ƚ&a^+KauXXr;432Eeӕ*|YI>VV9St-z{g&D|H/b@B0(IDyV$ |E
	v߭{811NiU	6&mb57!ee=MhJ-5Y6
4	zג20g_1udn5Lhc<$jH@VާV!NC`[|T;oQwՁ6ٓEzyM;o%ϬAt'L@Sqd}N'qNJ,<ip4QVW~dLHf&Q\r-?@~&ф$k+]amwгv%ں`ZQNTZ	@ 2{{{7I<9}s%H`)'[-O47ZMF%My!]r8g?C^֤40Jw*zc/tQvA5;ſvA.P/~
A+7AQL=s2Uw
F_U,#~|W;ͷ|@5OY%4=ϩ-;
1L-+k&De
q,0]3G0'$ޜ0hyU7>%
4`QYZI'BusQv\\bn1gHHSpX)㟊cXˀVKhK K(y[}=YTiпv1m+={M:Y#j)t++]7LWkRW
"\Qn9Z

{1db©%z E^LQ$`mo'4>vN(%-aK3=bhN\?F|T:9Opʬn+.Z+\y{4?uW_7~_{yrDm+h;yS
aUwiaX	{/
y(ݺ^.BM30IIg!˲{Q2Yo%M>SEUT=dTj/-2|Vݴ)ѣ!طByN'Pl|njGD2EȧM醧1USp?fb?5͓iOaoJMZ84E7&keoLE5dU^[M{tX	)2'ʙ\5>C~VR^cxrF*hx\\lB
ͤk04&k	KQrx~)H4`3r3\S!Mr9Dk$
3u]̇(1nru?"㺘MWNf]lg21t	qA9	jw0QQ5F0_PO߀ ̀s݌_:Ѡ oе5pf+c5RG>Ca5!$ƿ@#3w[4=%
~Πb^k1;173^:nC5:̸q(9Y`FtkLg9r5Z/;0\98<ޠ?Ȏ"Mz_*o#dRzo߮z!^z,uWrENYŏʨ,;:ӧYYg*Vc*۶p;*jI5\쵼|._zB~Xp$d /TlUδÚ	pO8kUr'_0O'[Dss*5|mQ$;.X@!@D7x2$"]#"_KGFdL1{㆚b42scYFYJ(vQ*Ǣ=`otQ:03ˤ$ҁz@G,̑Ռ=~B>-xtu4=9ZV@0hu24Sh&A(E$rÕy<]C_2?*Ow癞all|' J(~l0A32{pg֡vFo42@5Dbr~\B '{Xe{s	({G?CDKs yrjW D4YB4Y''l!+KiMf<$^w l"3ٮ(EAJ%gz2 0cvz&& -UJY*!N&(C]'5: