1 INFO-VAX	Mon, 28 Apr 2003	Volume 2003 : Issue 233       Contents:= Re: How Alpha will save Itanium - must reading for Bill Todd! = Re: How Alpha will save Itanium - must reading for Bill Todd! = RE: How Alpha will save Itanium - must reading for Bill Todd! = RE: How Alpha will save Itanium - must reading for Bill Todd! = Re: How Alpha will save Itanium - must reading for Bill Todd! = Re: How Alpha will save Itanium - must reading for Bill Todd! = Re: How Alpha will save Itanium - must reading for Bill Todd! = RE: How Alpha will save Itanium - must reading for Bill Todd! = Re: How Alpha will save Itanium - must reading for Bill Todd! = Re: How Alpha will save Itanium - must reading for Bill Todd! H Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyH Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyH Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyH Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyH Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyH Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyH Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopoly" Re: Setting MTU for TCPIP services& X-windows: adding a widget to a systemG Re: [DECnet-Plus V7.3-1 ECO2] What has happened to the DECNET_VERSION ? G Re: [DECnet-Plus V7.3-1 ECO2] What has happened to the DECNET_VERSION ?  [hp OpenVMS] SSL on VAX ?  Re: [hp OpenVMS] SSL on VAX ? 7 Re: [SYSMAN, TCPIP] Unprintable Chars switches terminal   F ----------------------------------------------------------------------  % Date: Sun, 27 Apr 2003 14:44:34 -0400 * From: "Bill Todd" <billtodd@metrocast.net>F Subject: Re: How Alpha will save Itanium - must reading for Bill Todd!2 Message-ID: <kt-cnZ6SJOGJuDGjXTWcoA@metrocast.net>  2 "Main, Kerry" <Kerry.Main@hp.com> wrote in messageL news:BE56C50EA024184DAF48F0B9A47F5CF4040ECFF4@kaoexc01.americas.cpqcorp.net. .. > Ahhh Bill,  H > Still resorting to slimy tactics like pulling private conference notes > into public forums eh?   You asked for it, you got it:    quote:  F I never stated *anywhere* [emphasis mine] that that an IPF design withJ "Alpha influenced" features would be *required* for VMS to run when it was first ported to IPF.  L Go for it - do your google stuff and try and find a post from me that statedK it was a *requirement* that had to be there by 2004 for the initial OpenVMS  release.  
 end quote.  F So I indeed went for it, and I certainly found it.  And now you'd likeJ people to ignore your many statements that so clearly back up the originalE assertion which I made (and you flatly denied) because *some* of them I (definitely not all - see the c.o.v. quotes) weren't made to the world at " large but only to a smaller group?  J > Ah well, others should make note of this great example of your integrity= > when considering membership for future private discussions.   J What people would be well advised to remember is the fusilade of lies thatJ Compaq issued to accompany the announcement of the Alphacide and the vigorK with which people like you promoted them.  Anyone with half a brain and the = inclination to use it could see that there was no way that an K 'Alpha-influenced' (let alone 'Alpha-Inside') Itanic was going to appear in K anything like the VMS porting time-frame, but that didn't stop you and your I ilk from foisting that crap on anyone who would listen (and many who were K loath to believe that Compaq and you were that slimy were willing to lend a  sympathetic ear).   L As I noted previously, any claim you had to privacy ceased at the point whenJ you began actively selling that Compaq party line to our group rather thanJ simply acting as an individual participant.  Act like a whore, get treated% like one - in private, and in public.   E > Question - when you joined that forum that included a number of the F > current active cov readers, did you or did you not agree to keep all& > conversations in that forum private?  I Well, I actually do not remember such a formal commitment on my part, but G I'll happily accept any *evidence* you might be able to find, Google or F otherwise (do be specific in quoting:  I keep virtually all of my sent" emails, so I'll be able to check).   ...   F > Having stated this, there is nothing in the attached which quotes meD > saying an Alpha inflence was *required* for OpenVMS to ship in the > initial releases of OpenVMS.  I You really are a shameless liar, Kerry:  the evidence is not yet 12 hours 6 old, and people have longer attention spans than that.  L Let's see:  Aside from the *many* statements to the effect that the VMS portD *would be* to an Alpha-influenced Itanic, I not only found an actualG instance where you referred to the 'Alpha features ... required to make  OpenVMS, NSK and@ Tru64 work with its IA64-2' but another where you said 'the IA64H architecture will be updated to enable Tru64, OpenVMS and NSK to use it'J ('to enable' clearly indicating that they could *not* use it without thoseB updates) and a third (this one a c.o.v. post) where you said 'IA64G technologies will be integrated and upgraded with Alpha technologies to I allow these OS's to be ported' ('to allow' carrying the same meaning that + 'to enable' carried in the previous quote).   I So wriggle and weasel though you may (I won't bother honoring the rest of I your crap today with a detailed refutation:  people really don't need any I help to see it for what it is, and you for what you are), your statements  were very clear.   - bill   ------------------------------  % Date: Sun, 27 Apr 2003 16:31:20 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>F Subject: Re: How Alpha will save Itanium - must reading for Bill Todd!) Message-ID: <3EAC3E16.B0E9DBD7@istop.com>   $ re: Revealing private conversations.  I Mr Main, while I do not support the revealing of private conversations, I K understand where Mr Todd is coming from.  When one denies ever saying X and J challenges the other to prove otherwise, and in a very insulting way, this: greatly motivates the other to find and produce the proof.  K Had the initial denial not been made, Mr Todd would not have needed to look J for the proof to respond to your challenge. Perhaps all Mr Tood could haveL done is publicly state that he staill had copies of private emails where youK stated X and be willing to produce them publicly with Mr Main's permission.   M Saying different things in private and in public is never a really good thing + because those things often catch up to you.   J Now, I had left that private forum well before the Alphacide, so I was notK privy to those conversations. Yet I still recall many statements (from many J people) about current IA64s not being able to take on the job of replacing5 Alphas, especially the larger Wildfire class systems.   H So I am not sure Mr Todd would have needed to go to private conversationR forums to find the proof. However, now that the harm is done, both have been hurt.  N One because his public sayings differ from his private ones (so what are we toM make of its future public statements ?), and the other for revealing contents  of private letters.   L The fact that Hp/Compaq felt so strongly a need to justify the Alphacide andN glorify their IA64 decision is a good indication that internally, they knew itK was a technological mistake and they know it was a very unpopular decision. M Moving to a new unproven platform at a time when there are rumours that Intel 6 won't keep it for very long is never a very good idea.  L If Digital employees were honest with us and agreed that IA64 is an inferiorJ platform and that the decision was taken purly because of the realtionshipK with Intel, then at least we'd have more respect for their statements about + IA64 because they woudl appear more honest.   N But anytime some ex Digital emopoloyee who used to extoll the vertues of AlphaM and downgrade IA64 to an interesting , expensive, delayed experiment, but now N claims IA64 is going to be industry standard low cost etc etc, then one really= has to wonder which side of the person was telling the truth.   L And this is the biggest problem with HP today. We have no confidence in whatK they are saying because we see so many examples of "we never said that", or R "yes, we have strong commitment to VMS, but expect VMS users to migrate to HP-UX".   ------------------------------  % Date: Sun, 27 Apr 2003 21:01:10 -0400 ' From: "Main, Kerry" <Kerry.Main@hp.com> F Subject: RE: How Alpha will save Itanium - must reading for Bill Todd!T Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF4040ECFF7@kaoexc01.americas.cpqcorp.net>   Bill,    You really are a case.=20   H >>> I not only found an actual instance where you referred to the 'AlphaB features ... required to make OpenVMS, NSK and Tru64 work with itsA IA64-2' but another where you said 'the IA64 architecture will be 4 updated to enable Tru64, OpenVMS and NSK to use it>>  G Having re-read that I should have stated "any features required to make H OpenVMS, NSK and Tru64 work with the IA64 would be added.." I will grantE you the Alpha features word was not a good choice for that particular E quote. However, as lock step and additional features surrounding this C capability was added to IPF for the NSK offering, this statement is  certainly true.=20  F Fortunately, to the best of my knowledge, there has been no additional/ HW features required by OpenVMS to do the port.   6 Here is my original quote that you picked out earlier:? "however, given that the EV8 team is going to work on the IA64+ G (whatever flavour) that VMS, Tru64 and NSK runs on, it is not really  a @ big stretch to surmise that a good amount of EV8 technology (VMS= friendly?) will show up in subsequent generations of IA64.<<<   E There is nothing wrong with this statement. I stated it was not a big G stretch and it certainly is not to assume that some of the new features D will be good for OpenVMS. And given that it is the EV8 team activelyG participating (EV8 team lead was put in charge of Intel future designs) D the subsequent IPF designs, my statement "about a good amount of EV8B technology will show in subsequent versions of IPF" is correct.=20  8 Note that I did not say *designs* - rather *technology*.  D Also reference: (a bit dated, but relatively close to the time these' quotes in the newsgroup were happening) ) http://news.com.com/2100-1001-826527.html $ "Itanium to take on Alpha influence"  ? With respect to the privacy issue - you asked me the following:   C >>> Well, I actually do not remember such a formal commitment on my G part, but I'll happily accept any *evidence* you might be able to find,  >>>   E Here is the actual quote from the first email to all involved in that B private (emphasize private) newsgroup - unlike you I value privacyC committments and have excluded the authors name for confidentiality  reasons:=20   = First sentence: (AND billtodd@foo.mv.com is on the TO: list.)   C "I am writing to invite you to join a private discussion about ..."   D Subsequent paragraph from that first email which discusses the group rules:  ( "What do I want from you at this point ?' --------------------------------------- E Two important things, the first of which is confidentiality.  I don't 8 want the whole world knowing of our discussion .... "=20  * So, you asked for it and I gave it to you.  F Bottom line is that you obviously do not agree with many of the eventsB that have taken place in the last 18 months  and have decided thatE spewing venom and trying to trash anyone who has an alternate view on G any of these events in a newsgroup like cov will somehow have an impact B on what happened in the past or what will happen in the future.=20  < Well, all I can say is that your strategy is not working.=20   Regards   
 Kerry Main Senior Consultant  Hewlett-Packard (Canada) Co.! Consulting & Integration Services  Voice: 613-592-4660  Fax   : 613-591-4477 Email: kerryDOTmain@hpDOTcom-     (remove the DOT's and replace with "."'s)  OpenVMS DCL - the original .COM      > -----Original Message-----4 > From: Bill Todd [mailto:billtodd@metrocast.net]=20 > Sent: April 27, 2003 2:45 PM > To: Info-VAX@Mvb.Saic.Com H > Subject: Re: How Alpha will save Itanium - must reading for Bill Todd! >=20 >=20 >=204 > "Main, Kerry" <Kerry.Main@hp.com> wrote in message@ > news:BE56C50EA024184DAF48F0B9A47F5CF4040ECFF4@kaoexc01.america > s.cpqcorp.net. > .. > > Ahhh Bill, >=20< > > Still resorting to slimy tactics like pulling private=20 > conference notes > > into public forums eh? >=20 > You asked for it, you got it:  >=20 > quote: >=20H > I never stated *anywhere* [emphasis mine] that that an IPF design with? > "Alpha influenced" features would be *required* for VMS to=20  > run when it was  > first ported to IPF. >=20B > Go for it - do your google stuff and try and find a post from=20 > me that stated@ > it was a *requirement* that had to be there by 2004 for the=20 > initial OpenVMS 
 > release. >=20 > end quote. >=20H > So I indeed went for it, and I certainly found it.  And now you'd likeB > people to ignore your many statements that so clearly back up=20 > the originalG > assertion which I made (and you flatly denied) because *some* of them A > (definitely not all - see the c.o.v. quotes) weren't made to=20  > the world at$ > large but only to a smaller group? >=20@ > > Ah well, others should make note of this great example of=20 > your integrity? > > when considering membership for future private discussions.  >=20B > What people would be well advised to remember is the fusilade=20 > of lies thatA > Compaq issued to accompany the announcement of the Alphacide=20  > and the vigor B > with which people like you promoted them.  Anyone with half a=20 > brain and the ? > inclination to use it could see that there was no way that an = > 'Alpha-influenced' (let alone 'Alpha-Inside') Itanic was=20  > going to appear in> > anything like the VMS porting time-frame, but that didn't=20 > stop you and your @ > ilk from foisting that crap on anyone who would listen (and=20 > many who were > > loath to believe that Compaq and you were that slimy were=20 > willing to lend a  > sympathetic ear).  >=20B > As I noted previously, any claim you had to privacy ceased at=20 > the point when= > you began actively selling that Compaq party line to our=20  > group rather than < > simply acting as an individual participant.  Act like a=20 > whore, get treated' > like one - in private, and in public.  >=20G > > Question - when you joined that forum that included a number of the H > > current active cov readers, did you or did you not agree to keep all( > > conversations in that forum private? >=20A > Well, I actually do not remember such a formal commitment on=20  > my part, butB > I'll happily accept any *evidence* you might be able to find,=20 > Google or H > otherwise (do be specific in quoting:  I keep virtually all of my sent$ > emails, so I'll be able to check). >=20 > ...  >=20H > > Having stated this, there is nothing in the attached which quotes meF > > saying an Alpha inflence was *required* for OpenVMS to ship in the  > > initial releases of OpenVMS. >=20A > You really are a shameless liar, Kerry:  the evidence is not=20  > yet 12 hours8 > old, and people have longer attention spans than that. >=20? > Let's see:  Aside from the *many* statements to the effect=20  > that the VMS port F > *would be* to an Alpha-influenced Itanic, I not only found an actual; > instance where you referred to the 'Alpha features ...=20  > required to make > OpenVMS, NSK andB > Tru64 work with its IA64-2' but another where you said 'the IA64B > architecture will be updated to enable Tru64, OpenVMS and NSK=20 > to use it'A > ('to enable' clearly indicating that they could *not* use it=20  > without those D > updates) and a third (this one a c.o.v. post) where you said 'IA64< > technologies will be integrated and upgraded with Alpha=20 > technologies to A > allow these OS's to be ported' ('to allow' carrying the same=20  > meaning that- > 'to enable' carried in the previous quote).  >=20B > So wriggle and weasel though you may (I won't bother honoring=20
 > the rest of ? > your crap today with a detailed refutation:  people really=20  > don't need any> > help to see it for what it is, and you for what you are),=20 > your statements  > were very clear. >=20 > - bill >=20 >=20 >=20 >=20   ------------------------------  % Date: Sun, 27 Apr 2003 21:52:02 -0400 ' From: "Main, Kerry" <Kerry.Main@hp.com> F Subject: RE: How Alpha will save Itanium - must reading for Bill Todd!T Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF4040ECFF8@kaoexc01.americas.cpqcorp.net>   JF,    Lets not lose focus -=20  ? While anyone's words can be micro-analyzed and looked at with a E microscope after the fact, the reality is that I did not say anything G intentionally misleading or "tow any company line" in private or on cov * as Bill has alluded on numerous occasions.  F >>> Saying different things in private and in public is never a really9 good thing because those things often catch up to you. >>   F That is certainly true and is why I never said anything in the private+ newsgroup that could come back and bite me.   C It does amaze me though that, with all of your banking and security B background, that you think so little of someone breaking a privacyD commitment simply because they want to further their own agenda in a" newsgroup as Bill continues to do.  G >>> Yet I still recall many statements (from many people) about current G IA64s not being able to take on the job of replacing Alphas, especially % the larger Wildfire class systems.<<<   H Absolutely. I stated it in the private newsgroup and I have stated it inD cov a number of times - IMHO, (read my opinion only) the initial IPFH systems that are released with official OpenVMS support in 2005 will notF beat the latest and biggest Alpha systems available when this happens. So what if this is true?  F Its no different than when the initial Alpha systems were released andG some of the big VAX's were still faster at the time. However, over time G the Alpha systems began to pull away from the big VAX systems. And so I C expect that the bigger IPF based systems will pull away from bigger  Alpha's over time as well.=20   G > But anytime some ex Digital emopoloyee who used to extoll the vertues B of Alpha and downgrade IA64 to an interesting , expensive, delayedD experiment, but now claims IA64 is going to be industry standard lowH cost etc etc, then one really has to wonder which side of the person was telling the truth.>>  B I agree - who is that you are talking about?  Certainly not me.=20  F Before the announcement, Intel did not have the Alpha EV8 designers orE the compiler engineers they now have. Now they do. So, does that mean G IPF future versions have a much better chance of long term success than  before the announcement?=20    Absolutely.   H However, do I still feel Alpha is the best platform today and likely for the next couple of years?=20   Absolutely.=20  E Would I have preferred if Digital had got its collective act together C and made Alpha the standard for desktop and server technologies?=20   C Sure, but for some reason the Digital BOD felt they did not need to , consult me about any decisions they made.=20  
 Imagine that.   % Things change and time marches on.=20    Regards   
 Kerry Main Senior Consultant  Hewlett-Packard (Canada) Co.! Consulting & Integration Services  Voice: 613-592-4660  Fax   : 613-591-4477 Email: kerryDOTmain@hpDOTcom-     (remove the DOT's and replace with "."'s)  OpenVMS DCL - the original .COM      > -----Original Message-----6 > From: JF Mezei [mailto:jfmezei.spamnot@istop.com]=20 > Sent: April 27, 2003 4:31 PM > To: Info-VAX@Mvb.Saic.Com H > Subject: Re: How Alpha will save Itanium - must reading for Bill Todd! >=20 >=20& > re: Revealing private conversations. >=20= > Mr Main, while I do not support the revealing of private=20  > conversations, I> > understand where Mr Todd is coming from.  When one denies=20 > ever saying X and ; > challenges the other to prove otherwise, and in a very=20  > insulting way, this < > greatly motivates the other to find and produce the proof. >=20A > Had the initial denial not been made, Mr Todd would not have=20  > needed to look? > for the proof to respond to your challenge. Perhaps all Mr=20  > Tood could have @ > done is publicly state that he staill had copies of private=20 > emails where you= > stated X and be willing to produce them publicly with Mr=20  > Main's permission. >=20@ > Saying different things in private and in public is never a=20 > really good thing - > because those things often catch up to you.  >=20B > Now, I had left that private forum well before the Alphacide,=20 > so I was not: > privy to those conversations. Yet I still recall many=20 > statements (from many B > people) about current IA64s not being able to take on the job=20 > of replacing7 > Alphas, especially the larger Wildfire class systems.  >=20@ > So I am not sure Mr Todd would have needed to go to private=20 > conversationB > forums to find the proof. However, now that the harm is done,=20 > both have been hurt. >=20@ > One because his public sayings differ from his private ones=20 > (so what are we to? > make of its future public statements ?), and the other for=20  > revealing contents > of private letters.  >=20? > The fact that Hp/Compaq felt so strongly a need to justify=20  > the Alphacide and : > glorify their IA64 decision is a good indication that=20 > internally, they knew it< > was a technological mistake and they know it was a very=20 > unpopular decision. ? > Moving to a new unproven platform at a time when there are=20  > rumours that Intel8 > won't keep it for very long is never a very good idea. >=20B > If Digital employees were honest with us and agreed that IA64=20 > is an inferiorB > platform and that the decision was taken purly because of the=20 > realtionship? > with Intel, then at least we'd have more respect for their=20  > statements about- > IA64 because they woudl appear more honest.  >=20B > But anytime some ex Digital emopoloyee who used to extoll the=20 > vertues of Alpha> > and downgrade IA64 to an interesting , expensive, delayed=20 > experiment, but now > > claims IA64 is going to be industry standard low cost etc=20 > etc, then one really? > has to wonder which side of the person was telling the truth.  >=20> > And this is the biggest problem with HP today. We have no=20 > confidence in whatA > they are saying because we see so many examples of "we never=20  > said that", orA > "yes, we have strong commitment to VMS, but expect VMS users=20  > to migrate to HP-UX".  >=20   ------------------------------  % Date: Sun, 27 Apr 2003 22:09:08 -0400 * From: "Bill Todd" <billtodd@metrocast.net>F Subject: Re: How Alpha will save Itanium - must reading for Bill Todd!2 Message-ID: <bGedneE197bXEDGjXTWcrg@metrocast.net>  2 "Main, Kerry" <Kerry.Main@hp.com> wrote in messageL news:BE56C50EA024184DAF48F0B9A47F5CF4040ECFF7@kaoexc01.americas.cpqcorp.net. .. > Bill,  >  > You really are a case.  2 And you really are a liar.  Persistent at it, too.  H >>> I not only found an actual instance where you referred to the 'AlphaB features ... required to make OpenVMS, NSK and Tru64 work with itsA IA64-2' but another where you said 'the IA64 architecture will be 4 updated to enable Tru64, OpenVMS and NSK to use it>>  I > Having re-read that I should have stated "any features required to make = > OpenVMS, NSK and Tru64 work with the IA64 would be added.."   J But that's not what you stated:  I quoted what you stated, and three of myJ citations (plus many more that express the same thought without explicitlyI stating that changes would be *required*) clearly reflect the assertion I F made and that you categorically denied (and are still trying to deny).   ...   8 > Here is my original quote that you picked out earlier:A > "however, given that the EV8 team is going to work on the IA64+ I > (whatever flavour) that VMS, Tru64 and NSK runs on, it is not really  a B > big stretch to surmise that a good amount of EV8 technology (VMS? > friendly?) will show up in subsequent generations of IA64.<<<   H No, that is not your original quote, but a rather late one (7/10/01).  IL presented 6 earlier examples, three of which *explicitly* indicated that VMS@ et al. would *require* an Alpha-enhanced Itanic for their ports.  B The quote you cite immediately above is one of the many additionalI *supporting* citations I came up with:  it clearly states that the Itanic I which VMS et al. will run on will be the *result* of work by the EV8 team L (which is also patently false), but stops short of noting explicitly that itH will be *required* for VMS et al. to run on (as 3 of the other citations do).   ...   A > With respect to the privacy issue - you asked me the following:  > C >>> Well, I actually do not remember such a formal commitment on my G part, but I'll happily accept any *evidence* you might be able to find,  >>>  > G > Here is the actual quote from the first email to all involved in that D > private (emphasize private) newsgroup - unlike you I value privacyE > committments and have excluded the authors name for confidentiality 
 > reasons: > ? > First sentence: (AND billtodd@foo.mv.com is on the TO: list.)  > E > "I am writing to invite you to join a private discussion about ..."  > F > Subsequent paragraph from that first email which discusses the group > rules: > * > "What do I want from you at this point ?) > --------------------------------------- G > Two important things, the first of which is confidentiality.  I don't 7 > want the whole world knowing of our discussion .... "  > , > So, you asked for it and I gave it to you.  C No, that is not what I asked for.  The interchange went as follows:   G > > Question - when you joined that forum that included a number of the H > > current active cov readers, did you or did you not agree to keep all( > > conversations in that forum private? > K > Well, I actually do not remember such a formal commitment on my part, but I > I'll happily accept any *evidence* you might be able to find, Google or H > otherwise (do be specific in quoting:  I keep virtually all of my sent$ > emails, so I'll be able to check).  F While the group's originator expressed a desire for confidentiality (aG desire which my actions after the Alphacide may or may not have been in I conflict with:  I'd have to see the entire original document and possibly J also ask the originator to have a definite opinion on that matter), what IJ asked for was some evidence of your assertion that I had "agree[d] to keepK all conversations in that forum private" (a somewhat different and decidely J more specific restriction, coupled with the assertion that it had not only+ been expressed but also actively accepted).   E The fact that we eventually *did* come out of the closet after having I entered into conversations with Rick Marcello et al. (and long before the J Alphacide occurred) proves that the originator's statement which you quoteL above was not blanket in scope and everlasting in time.  And I'll note againJ that your active and diligent promotion of Compaq's lies immediately afterH the Alphacide made it clear that your participation in the group from atJ least that point onward was as a Compaq whore rather than as an interestedI individual to whom normal courtesies of confidentiality might apply.  But H I'll still invite you to present any evidence that I actually broke someI kind of formal commitment by making your statements public (even though I J agree with your repeated assertions that you "did not state anything thereK that could not have been stated on cov" and thus find your sense of outrage & just a bit self-serving in any event).   - bill   ------------------------------  % Date: Sun, 27 Apr 2003 22:21:39 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>F Subject: Re: How Alpha will save Itanium - must reading for Bill Todd!) Message-ID: <3EAC901C.84B1E657@istop.com>    "Main, Kerry" wrote:H > Fortunately, to the best of my knowledge, there has been no additional1 > HW features required by OpenVMS to do the port.   B Could HP build a wildfire/marvel class machine with VMS running inL Galaxy/cluster configs with the current iteration of IA64 ? If not, why not N and what will be done to IA64 to allow HP to eventually build replacements for marvels/Wildfires ?    ------------------------------  % Date: Sun, 27 Apr 2003 23:03:23 -0400 * From: "Bill Todd" <billtodd@metrocast.net>F Subject: Re: How Alpha will save Itanium - must reading for Bill Todd!2 Message-ID: <AkqdnV1ONvVjBDGjXTWcpw@metrocast.net>  7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in message # news:3EAC901C.84B1E657@istop.com...  > "Main, Kerry" wrote:J > > Fortunately, to the best of my knowledge, there has been no additional3 > > HW features required by OpenVMS to do the port.  > D > Could HP build a wildfire/marvel class machine with VMS running inI > Galaxy/cluster configs with the current iteration of IA64 ? If not, why  not   K Of course not:  Itanic lacks the on-chip glue that provides Marvel with its J exceptional memory bandwidth and latency.  SGI's Itanic2 Altix systems areK the closest any Itanic platform comes to Marvel-class performance:  they're L a lot better than the Itanic systems HP has to offer, but still no match forI Marvel (though of course since Alpha development has largely ceased later K Itanic systems may start to attain by brute force what they can't attain by  architecture).  L > and what will be done to IA64 to allow HP to eventually build replacements for  > marvels/Wildfires ?   L The first possibility is whatever the EV8 team is working on for 2006-7, but. no relevant details about that are yet public.   - bill   ------------------------------  % Date: Mon, 28 Apr 2003 00:10:50 -0400 ' From: "Main, Kerry" <Kerry.Main@hp.com> F Subject: RE: How Alpha will save Itanium - must reading for Bill Todd!T Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF4040ECFFC@kaoexc01.americas.cpqcorp.net>   Bill,   C >>> what I asked for was some evidence of your assertion that I had F "agree[d] to keep all conversations in that forum private" (a somewhatB different and decidely more specific restriction, coupled with the? assertion that it had not only been expressed but also actively 
 accepted).<<<   F What gobblie goop - a lawyer would have a great time with all that.=20  H A list moderator sends a note to all participants that says here are theG rules. The general etiquette is that if you don't agree, then you don't ) participate. You participated. Period.=20   2 Why not just admit you made a mistake and move on?  F Anyway, name calling, venom spewing and mud throwing is your specialty) and I must admit you are good at that.=20   G However, since that is not one of the area's I want to specialize in, I E will refrain from further personal comments and let others draw their F own conclusions as to what constitutes integrity and respect even if aD formal document is not in place as you seem to think is required for online communications.  E I don't have time to respond to the micro-analysis of every word from G private emails and how it might be interpreted (or misinterpreted as in , your case) but lets take a few examples -=20    Quote #1 from you in your email:C "The above was your characterization of the port customers would be > making with VMS - i.e., a port to some combined 'IA64 + Alpha'> architecture rather than to the existing Itanic architecture."   My response #1: G Given most OpenVMS Customers will be porting in the 2006-2008 timeframe G when the Montecito and Tandlewood based systems are available (the ones E with "Alpha influence"), this is an absolutely true statement. Do you C think that most mission critical Customers are going to jump on the , initial IPF / OpenVMS systems that come out?   Quote #2 from me in your email: G "So, will IA64-2 more closely resemble IA64 or Alpha EV8? The answer is H likely somewhere in between, but it will almost certainly be a different9 chip architecture than what is available today as "IA64".    You Stated in response: B "So while you don't say a new chip will be 'required' for VMS, youE clearly state that the chip it will be ported to will not be the same E IA64 architecture known at that time, with the clear implication thatAA the target of the port will contain significant Alpha influence."l  @ My response #2: Thank you for at least admitting I did not stateD *required*. As to it not being the same architecture, there are manyH changes between the Merced, McKinley, Madison, Montecito etc in the sameG manner as there are many architecture differences between EV4, EV5, EV6fD and EV7 Alpha architectures. I only stated that the IPF-2 (Mckinley)? would be different than the initial Merced and it certainly is.   E Anyway, I've made my point, but you are certainly free to continue toe/ state your views(?) as it's a public newsgroup.B   Regards   
 Kerry Main Senior Consultant  Hewlett-Packard (Canada) Co.! Consulting & Integration Services- Voice: 613-592-4660s Fax   : 613-591-4477 Email: kerryDOTmain@hpDOTcom-     (remove the DOT's and replace with "."'s)a OpenVMS DCL - the original .COMo  ! ..snipped to save a few tree's ..    ------------------------------  % Date: Mon, 28 Apr 2003 00:13:50 -0400a* From: JF Mezei <jfmezei.spamnot@istop.com>F Subject: Re: How Alpha will save Itanium - must reading for Bill Todd!) Message-ID: <3EACAA7D.5B73139B@istop.com>    Bill Todd wrote:M > Of course not:  Itanic lacks the on-chip glue that provides Marvel with itsaL > exceptional memory bandwidth and latency.  SGI's Itanic2 Altix systems areE > the closest any Itanic platform comes to Marvel-class performance:    J Here is where I get lost. If current IA64 chips produced by Intel lack theJ hardware to allow wildfire class machines (or lack the better glue to make@ marvel), how come SGI can make "Altix" systems that come close ?  M But did EV6x chips used to build the original Wildfires have the on-chip glue>M ? Or is that an EV7-only thing ?  If they were able to build Wildfire systemsgM with capabilities to allow VMS to run in a Galaxy environment without on-chipmJ glue, couldn't they make IS64 systems today with similar capabilities ? OrC does IA64 still lack what it takes to scale systems to that level ?t  N I am not talking about matching performance. Nobody expects performance out ofI IA64. But if IA64 can compensate with an increased number of CPUs then iteE might not be so bad. On the other hand, if one must wait for a futureeI generation of IA64 things before they are capable of being assembled intoyL machines with sufficient numbers of CPUs, then IA64 is clearly not yet ready$ for prime time enterprise computing.   ------------------------------  % Date: Mon, 28 Apr 2003 00:41:23 -0400a* From: JF Mezei <jfmezei.spamnot@istop.com>F Subject: Re: How Alpha will save Itanium - must reading for Bill Todd!) Message-ID: <3EACB0EF.D335FF76@istop.com>y   "Main, Kerry" wrote:E > What gobblie goop - a lawyer would have a great time with all that.t   It's Gobble dee gook ...  J > A list moderator sends a note to all participants that says here are theI > rules. The general etiquette is that if you don't agree, then you don'tu( > participate. You participated. Period.  K What about Freedom of Information Act ? :-) :-) Imagine if we could requestnN Capellas make public all the emails between him and Carly. Imagine if we couldE peer into the emails between Gorham, Marcello, Stallard , Carly etc ?   L We'd have a much better understanding of what is really going on, and the HPI employees would think twice about screwing with VMS customers beecause we H would eventually find out. It is called transparency and accountability.  ? One final thing. While it is true that the group did have a nonyH disclosure/confidential nature, there is a difference between disclosingF material pertaining to the purpose of the group (revealing  details ofD conversations with marcello etc) and revealing general stuff such asI preference for dark chocolate or technical discussion about IA64 based ont publicly available information.    ------------------------------  % Date: Sun, 27 Apr 2003 16:06:39 -0400o* From: JF Mezei <jfmezei.spamnot@istop.com>Q Subject: Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopoly ) Message-ID: <3EAC384F.60750BDA@istop.com>m   jmfbahciv@aol.com wrote:@ > To become an expert, one does need to spend years and years in> > an area.  Programming these days spends so much time getting, > swapped in and out, nothing gets improved.  K Well said. Just as your skills on oen technology become valuable, the presseK decides that something else has become the de-facto standard fad-of-the-day / and all of a sudden, your skills are worthless.s  K And, as the press declares that new technology to be the de-facto standard,lL companies will hire everyone and anyone who pretends to have experience with that technology.  K A young guy I knew got hired as a not-junior NOTES administrator/programmer J after having toyed with NOTES (lotus, not DEC) for a couple of months as aH USER, and this is the type of guy who didn't really know how to organiseN fields in a database (for instance, he wanted to put text fields that describeM a flag instead of putting a 1 or 0 in a single character field with one fieldt	 per flag.e  K NOTES was superceded by JAVA (which I guess is still fairly "soupe-du-jour" 1 these days). Is NOTES still very much in demand ?t   ------------------------------  % Date: Sun, 27 Apr 2003 22:49:41 +0200a* From: Morten Reistad <mrr@reistad.priv.no>Q Subject: Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopoly / Message-ID: <5pfh8b.bmt.ln@via.reistad.priv.no>l  " According to  <jmfbahciv@aol.com>:1 >In article <j7k88b.4ta1.ln@via.reistad.priv.no>,s/ >   Morten Reistad <mrr@reistad.priv.no> wrote:-$ >>According to  <jmfbahciv@aol.com>:0 >>>In article <b8644r$11p8$1@godfrey.mcc.ac.uk>,/ >>>   Geoff Lane <zzassgl@zoe.mcc.ac.uk> wrote: 6 >>>>In alt.folklore.computers jmfbahciv@aol.com wrote: [snip] >>>9D >>>What was even more awesome...every single one of those guys could >>>do this.  >>; >>The best Open Source people still do this. But code todaye; >>needs to scale a couple of magnitudes further than in the / >>hayday of operating systems around 1972-1985.  >>> >>It has just fallen to the wayside of commercial programming. >d? >I think I'm going to respectfully disagree :-).  I worked witho= >a guy who could do the same thing with data base management c= >systems.  I never did watch him enough to understand how he e> >thought but he did the ka-chunk processing at that level.  HeC >was also one of the very few who was able to do ka-chunking at the B >exec level, user level, compiler level and data base level.  FromB >my experience (which is not extensive) these types are very rare.  ? Some people can do this. My claim is that there is not a singlet@ corporation that has the ability to explicitly screen for them, D find them, hire them and care for them; not least, have a documented@ career path in place. They are exceptionally few in number, evenL compared to some vital, but similarly scarce people like (good) actuarians, = structural engineers, meterologists, metal fatigue analysts,  ? and technically capable multi-language translators. These othere; groups of people may seem esotheric, but I know of specifice: career paths and skill evaluations for them in regular use by companies I know.    ; >>It seems the corporate world hasn't managed to handle thee >>world of programming at all. >d; >If people don't hone their skills over decades on the samea? >flavor, there isn't going to be experts in any computing area.e  > Yes, the good ones do. But a person with an engineering degree< and firm knowledge of four languages is about as scarce, but? is instantly recognized as an asset by the companies that care.a< This person will be honed carefully with skill expansion and productivity use.c  9 Not so for a programmers that has proven to be brilliant.   C >Look at how science does their biz.  It is expected to stay withinr@ >an area of expertise.  An astronomy advisor doesn't do a career> >review of a grad student and fail him/her because the student< >didn't get majors in biology and chemistry and physics and ? >economics and business management and art and  basket weaving. B >On the contrary, the kid would get told to pick one and ONLY one. >And even that's too general.i >H? >To become an expert, one does need to spend years and years ine= >an area.  Programming these days spends so much time gettingv+ >swapped in and out, nothing gets improved.e  ; I agree, there are very few drivers to improve programming.a9 The new languages are boring, and rehanshes of old stuff.K6 Very few new things in the methods department. But the; Open Source movement is embracing almost all the new stuff.19 It may be a necessary step to break open the field to new % ideas. Sort of a necessary recession.   = >>These teams tend to follow the sociological rules laid downhF >>by Brooks in his "Man Month" books (there is a revised one out now).! >>The PHBs never understand this.  >d: >Some do.  Stock markets don't.  There is no such thing as: >long term investment, where long term means more than one >year.  5 You don't have to be particularly long term to follow46 the Brooks models. The BSD cycle is 4 months, and that3 is within the 15-month possible mindspan of a stockt market.a  = >>All the major teams in Open Source tend to follow a similari& >>model, with either a "Boss-Sidekick" >h? >Nope.  Boss-Umbrella.  If there ain't an umbrella the project t >is doomed.e  > I was thinking of a structure used a lot by Norway's military,> "Boss-NK", where the NK stands for "2.nd in command". The boss= is to stay at a strategic or possibly tactical mindset, whiles8 the NK is to have an operation or tactical focus. The NK> does a lot of umbrella stuff; but is also the sparring partner? of the boss, and sbould be able to take on the boss's job in anh	 instant. b  > Therefore the name "Sidekick". It is about more than being the	 umbrella.t    >> .. or a tight 3-4 person team9 >>on top, and tight communication with 5-10 other people.a >>= >>This is exactly the "surgical team" Brooks suggested in the  >>"Man Month". p >>1 >>When this model breaks down, the project forks.d >g: >I didn't see projects forking.  I did see projects taking7 >double the time.  No matter how large the group, theren: >are only three workers.  If you have a really good group,D >the worker role gets picked up by another person when someone needs	 >a break.   B Brooks says 4 workers, 1 central and 3 "neighbours". Each of theseC can have another 3 "neighbours", but they only get to do supporting : stuff, because the core person is too busy to talk to them
 directly.   > >>We have seen this in BSD, which is now forked in three, with8 >>some possible further fragmentation. XFree86 is in the< >>process of forking right now, but Linus has so far managed= >>to aviod a major fork in the OS that almost bears his name.  >s> >That's what happens when the project control isn't "owned" by= >one entity who is NOT an individual with serious prima donna 9 >problems (it always seems to boil down to NIH syndrome).   @ There are no lack of "ownership" among the Open Source projects.9 The major Open Source forks has been over issues that hass: had a recognizable mix of personal and ideological issues.  : E:g. OpenBSD, which wants to stay on the "pure and narrow"= path, and stay fully secure, fully reviewed and scrutinized; t8 at the explicit expense of being even a possible desktop9 challenge to Windows because of the lack of optimizationsa: and device drivers. And cudos to the team for pulling this7 off. There was a biggish fight when they forked off theh main BSD stream.  = Linus Thorvalds is doing this very well too; but his model ist: to have several "Umbrella/Sidekick" persons, one for everyA major field.This seems a novel structure in software development.e> He may have an advantage my having his name on the front cover though.e  7 And, this is passing the corporate world completely by.C  5 RedHat did try, but lacked the momentum and number of37 quality people. IBM has been almost invisible; and thatO5 is odd considering their level of corporate exposure.p  4 Except for the odd attempt from SGI I see noone else+ even trying to care for elite programmers. m   -- mrr   ------------------------------  % Date: Sun, 27 Apr 2003 17:48:43 -0400j* From: JF Mezei <jfmezei.spamnot@istop.com>Q Subject: Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopoly ) Message-ID: <3EAC5034.4BC8666D@istop.com>A   Morten Reistad wrote:a= > I agree, there are very few drivers to improve programming. ; > The new languages are boring, and rehanshes of old stuff.n0 > Very few new things in the methods department.  K This is another area that I disliked. Some companies paid big bucks to someeD large consulting firms to implement a "programming model" which is aG philosophy for programming, testing etc etc.  Once they have paid theirdD millions for stuff that any good manager should have been able to doM themselves, they are not about to relinquish those "methods" which are forcedyN upon programmers and which may render the most brilliant programmer into a notX very productive one because he has to spend so much time doing paperwork and management.  L It is a bit like ISO9000 which doesn't dictate HWO you do things, but ratherL how you document them. It doesn't improve your product or your productivity,F it only ensures that your product (and how you make it) is documented.  K The problem isn't so much the documentation, but an excess in documentationsM and reduced freedoms to innovate. So your fancy programmer who might have theuK ability to greatly innovate and bring new ideas is held down by establishedeS procedures which force everyone to go by the book and not differentiate themselves.f  N Of course, letting your programmers loose is also dangerous. (Sue, how tight aI grip do you keep on your engineers ?) A proper balance is what is needed.    ------------------------------  % Date: Sun, 27 Apr 2003 21:59:24 -0400d: From: Charles Shannon Hendrix <shannon@news.widomaker.com>Q Subject: Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyt. Message-ID: <st1i8b.e8m.ln@escape.shannon.net>  D In article <b8ghsn$fc8$1@bob.news.rcn.net>, jmfbahciv@aol.com wrote:  @ > I think I'm going to respectfully disagree :-).  I worked with> > a guy who could do the same thing with data base management > > systems.  I never did watch him enough to understand how he ? > thought but he did the ka-chunk processing at that level.  HeiD > was also one of the very few who was able to do ka-chunking at theC > exec level, user level, compiler level and data base level.  From C > my experience (which is not extensive) these types are very rare.   D The problem is that the systems you are talking about were simple in comparison to today's monsters.l  F There are more parts on my graphics card than in an entire PDP-10, andD more code in on my "desktop" than any PDP-10 ever had, several times over.   8 This is without even getting into modern server systems.  D It was much easier back then to be a bit god, because there were far! fewer bits to keep in one's head.i  H That's one reason I argue that we've gotten overcomplicated.  Much of itE we need of course, but most we probably don't.  We are also well past G our ability to maintain code at a consistent and good level of quality.   D But that has less to do with people today not being as good and moreF to do with it being so very much harder to have the "bit god" level of system knowledge.   D Expectations are also tremendously higher, life is more complicated,+ and taxes are higher.  Yes, it counts... :)t  < > If people don't hone their skills over decades on the same@ > flavor, there isn't going to be experts in any computing area.  7 I don't think it takes decades, but it does take years.   9 We move to new things far faster than we can master them.l  ? Change is good, but only if it accomplishes something worth theo associated costs.h  C > On the contrary, the kid would get told to pick one and ONLY one.  > And even that's too general.  F Well, science long ago expanded so much it had to move to specialists.  F However, it is well known that this is a dead end strategy.  It causesB all kinds of communication problems, among other things.  Too many specialists is bad.   ? One proposed solution was to train everyone in two fields.  The C paper I read said that just two skills for each person, when groupssE communication, greatly increase the flow of information and thus work  gets done faster.-  @ > To become an expert, one does need to spend years and years in> > an area.  Programming these days spends so much time getting, > swapped in and out, nothing gets improved.  D Systems get yanked out from under you very fast, you often move to aD new project before finishing your current one, and some projects are* practically perpetual with no end in site.  B It's frustrating too because everyone wants to be able to point at  something and say, "I did that."  = >>All the major teams in Open Source tend to follow a similarr& >>model, with either a "Boss-Sidekick" > @ > Nope.  Boss-Umbrella.  If there ain't an umbrella the project  > is doomed.  * You need to learn a lot about open source.  B There are a number of very active, very good programs that have no	 umbrella.a  H It's nice for usre, but part of the point of OS is that it isn't needed.  ; > I didn't see projects forking.  I did see projects takingr8 > double the time.  No matter how large the group, there; > are only three workers.  If you have a really good group,dE > the worker role gets picked up by another person when someone needsu
 > a break.  * Depends on the project and the leadership.  > >>We have seen this in BSD, which is now forked in three, with8 >>some possible further fragmentation. XFree86 is in the< >>process of forking right now, but Linus has so far managed= >>to aviod a major fork in the OS that almost bears his name.   . Had not heard about XFree forking.  Got a URL?  ? > That's what happens when the project control isn't "owned" byt
 > one entity    G Sometimes, but obviously its pretty rare.  Try looking at some projectse, before suggesting that this is what happens.  F I have news for you to: projects in commercial ventures fork too, evenG within a department.  I have been in some vicious project forks before.l3 It's ugly on the inside of those nice glass towers.   G > who is NOT an individual with serious prima donna problems (it alwaysf& > seems to boil down to NIH syndrome).  0 You wouldn't be talking about OpenBSD would you?   ------------------------------  % Date: Sun, 27 Apr 2003 21:40:40 -0400a: From: Charles Shannon Hendrix <shannon@news.widomaker.com>Q Subject: Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyt. Message-ID: <oq0i8b.e8m.ln@escape.shannon.net>  B In article <3EAB866F.1010703@beagle-ears.com>, Lars Poulsen wrote:  A > a line item with a title like "FCC Mandated Charge". The chargenC > in question is not mandated by the FCC (although it is explicitly E > ALLOWED by the FCC. The amount is determined at its sole discretionr   ALLOWED == MANDATEDa  E There is no such thing as "allowed" or "discretion" when dealing with  companies like this.  > If the FCC believed that this would ever be optional, they are% delusional and need psychiatric help.    ------------------------------    Date: 27 Apr 2003 18:33:02 -0700$ From: Marco S Hyman <marc@snafu.org>Q Subject: Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopoly + Message-ID: <x7el3nh0dt.fsf@hana.snafu.org>t  , JF Mezei <jfmezei.spamnot@istop.com> writes:  N > It is a bit like ISO9000 which doesn't dictate HWO you do things, but ratherN > how you document them. It doesn't improve your product or your productivity,H > it only ensures that your product (and how you make it) is documented.  J Not quite (unless it's changed since 1995 or so).   Certification requiresI that you documment your processes and that you follow your documentation. J Your documentation may state that a software design step is to eat lots ofF pizza and beer while brainstorming ideas.   That's fine as long as you follow the documentation :-)  M > The problem isn't so much the documentation, but an excess in documentationuK > and reduced freedoms to innovate. So your fancy programmer who might have. > the   A You can always change the documentation, providing you follow the 1 documented procedure on documentation changes :-)c  G The problem lots of companies have (had?) with gaining certification istD that they create reams of documentation specifying how things shouldG be in a perfect world.   Wrong!   That's a prescription for failing the.K audit.   I think some managers get embarassed puting down actual procedurese in black and white.e  A The one time I worked for a company that attempted (and received)hD certification a litany was drilled into our heads: document what you do -- do what you document.   F The result was the formalization of a lot of ad hoc procedures.   ThisD was good.  Once the procedure were in black and white for all to see they became easier to modify.c   // marcs   ------------------------------  % Date: Sun, 27 Apr 2003 22:02:16 -0400 : From: Charles Shannon Hendrix <shannon@news.widomaker.com>Q Subject: Re: IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM monopolyn. Message-ID: <832i8b.e8m.ln@escape.shannon.net>  E In article <5pfh8b.bmt.ln@via.reistad.priv.no>, Morten Reistad wrote:   6 > Except for the odd attempt from SGI I see noone else- > even trying to care for elite programmers. o  > From what I have seen, they don't even want you to become one.  B I have many times pushed very hard and gotten quite good, and they0 either cut me out of projects, or burned me out.  G The fact is, a year of intensive but not insane coding can get you at atH stage where you are very proeductive.  The problem is that few companiesC avoid burning those people out, don't encourage others to get therea@ (part of the burnout problem), and they don't reward you for it.   ------------------------------  % Date: Mon, 28 Apr 2003 00:04:38 +0200h From: Dirk Munk <munk@home.nl>+ Subject: Re: Setting MTU for TCPIP servicesb2 Message-ID: <b8hkdf$m3e$1@news1.tilbu1.nb.home.nl>   JF Mezei wrote:i   > N > The new ISP recommends 1452 ! Perhaps because of hig taxes in Canada, I have; > to leave more room for the tax to fit inside the 1500 :-)  >  > I >>I haven't actually noticed any problems with this.  Well, perhaps theree/ >>are some unnoticed problems due to this.  :-|s >  > M > I haven't read up on how MTU works. But if you send a 1500 byte packet, the=O > router may have to split it into 2 packets (one large and one very small witheH > some padding to fit minimum packet lengths) and that wastes bandwidth. > M > Consider a heavy duty FTP transfer: If I send at 1452 (or whatever the real-K > value is), then all my packets are full. If I send at 1500, then half theeO > packets are full, and the other half are just little stubs to make up for thed > overflow of 1500 into 1452.G  Q True, this is called packet fragmentation. Certain TCPIP packages (not TCPIP for 5F VMS it seems) support a ping command with a 'no-fragment' bit set. In H combination with a ping packetsize, this allows you to test the maximum / packetsize that you can use with a remote host.c     > N > Come to think of it, perhaps the 1452 is designed to reduce any padding whenN > there are packets that are split in two since the second packet will have 48T > bytes instead of 8, so it need not be padded as much to reach minimum packet size. > # > Anyone can confirm or deny this ?s  + Another topic, but also applicable to ADSL:r  Q Just today I read an article about ADSL software (for Windooz of course). As you yP know cable and ADSL connections have asymetric speeds, a high download speed in O combination with a relatively low upload speed. One of those software packages hO used a interesting trick. The ACK messages (= upload traffic) for the download eQ traffic are send with a higher priority (internally in the TCPIP stack) then the rQ normal upload traffic. The result is a far better download performance (500% !!)  N when the upload link is also heavily loaded. Now that would be a nice feature  for TCPIP on VMS .......   ------------------------------  % Date: Sun, 27 Apr 2003 17:51:30 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>/ Subject: X-windows: adding a widget to a systeml) Message-ID: <3EAC50DB.F787D8CA@istop.com>n  5 OK, I am getting my hands dirty with X-windows/Motif.d  M I have found source code for a widget that allows numbers to be entered (withn- decimal etc). How does one make it available:w 	1- to a program 	2- to the UIL compiler ?c  F Is it possible to install it so that it is available by default to allM applications on a system, or must those additional widgets be linked directlyl with each application ?    ------------------------------  # Date: Sun, 27 Apr 2003 18:45:27 GMTt6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)P Subject: Re: [DECnet-Plus V7.3-1 ECO2] What has happened to the DECNET_VERSION ?3 Message-ID: <bzVqa.43924$v62.475005@news.chello.at>t  t In article <3eabed66$0$49113$e4fe514c@news.xs4all.nl>, Bart Zorn <B.Zorn@xs4all.nospam.nl> writes: Paul Sture wrote:o >> In article <ltPqa.35409$v62.366397@news.chello.at>, peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) writes: X >>>In article <PlTqU03tqXwA@elias.decus.ch>, p_sture@elias.decus.ch (Paul Sture) writes: >>>  >>>	$ ncl sho impo
 >>>	Node 0( >>>	at 2003-04-27-13:45:54.580+02:00Iinf >>>t >>>	Characteristicsf >>>e+ >>>	    Implementation                    =  >>>	       { >>>	          [h" >>>	          Name = OpenVMS AXP ," >>>	          Version = "V7.3-1  " >>>	          ] ,f >>>	          [n5 >>>	          Name = Compaq DECnet-Plus for OpenVMS ,u> >>>	          Version = "V7.3-1 ECO01 16-SEP-2002 10:33:00.25" >>>	          ]t >>>	       }  5 Sorry for the confusion, I forgot to reboot after thes3 DNVOSI ECO2 installation (on this computer only ;-)b  I >> %PCSI-I-OBJSKP, file [SYSEXE]DNS$SERVER.EXE pertains to an option thate& >was not selected; file update skipped >> >> ^^^^^^??????? was ist das?  > D >DECdns is the naming service for DECnet-Plus. Up to VMS V7.3 DECdnsD >server was only available for VAX (or on Tru64 Unix). Starting withG >V7.3-1 it is on Alpha, too. If you install or upgrade DECnet-Plus, andoH >you accept the default configuration options, DECdns server will not be >installed.n  K As we also saw on the last DECUS germany symposium, most of us VMS manglersnD didn't even notice the appearance of DECdns on Alpha with V7.3-1 ;-)I And to nitpick, DECdns is not limited to DECnet. I had DECmcc registeringe things in DECdns also...  D >Most sites use the Local name database in stead of DECdns. However,@ >DECdns was one of big advantages of DECnet-Plus over DECnet IV, >especially for large networks.l  I And before it became DECdns, there was DNS V1 and it was a predecessor ofa! what X.500 became. A directory...t  0 >> $ write sys$output f$getsyi("decnet_version") >> 00050500n  3 So, you reproduced the error I ranted upon. Thanks.h  2 >>>>$  $ write sys$output f$gets("decnet_version") >>>>00050E04 >>>q4 >>>Das ist anscheinend DECnet-Plus V7.3-1 ohne ECOs. >>	 >> Genau.l  H Sorry for the little german in this discussion. It was my fault (posting( a text which was intended to be a PM)...    K So, to rephrase my statement: DNVOSI ECO2 brought us a bogus DECNET_VERSIONm) Please someone with a support contract...i   -- t Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialistl E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------   Date: 27 Apr 03 21:42:52 +0200) From: p_sture@elias.decus.ch (Paul Sture)gP Subject: Re: [DECnet-Plus V7.3-1 ECO2] What has happened to the DECNET_VERSION ?) Message-ID: <eprLJypBE6hc@elias.decus.ch>   l In article <bzVqa.43924$v62.475005@news.chello.at>, peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) writes: > 1 >>> $ write sys$output f$getsyi("decnet_version")i >>> 00050500 > 5 > So, you reproduced the error I ranted upon. Thanks.m >   . No problem. You made me look at my ECO levels.   3 >>>>>$  $ write sys$output f$gets("decnet_version")u
 >>>>>00050E04  >>>>5 >>>>Das ist anscheinend DECnet-Plus V7.3-1 ohne ECOs.c >>>m
 >>> Genau. > J > Sorry for the little german in this discussion. It was my fault (posting* > a text which was intended to be a PM)...  D Not a problem here, but I was concerned that someone with the answer might miss it :-)-   >  > M > So, to rephrase my statement: DNVOSI ECO2 brought us a bogus DECNET_VERSIONc+ > Please someone with a support contract...a >   G I will look to see what we have at work, but not before Tuesday for me.s   --  
 Paul Sture   ------------------------------  # Date: Sun, 27 Apr 2003 19:52:13 GMTt6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)" Subject: [hp OpenVMS] SSL on VAX ?3 Message-ID: <NxWqa.45016$v62.492636@news.chello.at>M  G Recently I listed SSL on the list of differences between VAX and Alpha,tH because DEC/COMPAQ/HPQ doesn't provide SSL for the VAX. But, now I found in SYS$STARTUP:SSL$STARTUP.COM   $       ARCH = "VAX"7 $       IF F$GETSYI("CPU") .GE. 128 THEN ARCH = "ALPHA"o   ando  M $       DEFINE/NOLOG/SYSTEM/EXEC        SSL$EXE         SSL$ROOT:['ARCH'_EXE]c  O so it seems (at least for me), there are plans for making SSL available on VAX.c: Is this true ? Does anyone have any insight ? Time plans ?   TIA   L PS: And we should hint the engineers that above code won't work with I64 ;-) --   Peter "EPLAN" LANGSTOEGERt% Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Sun, 27 Apr 2003 17:49:39 -04000* From: JF Mezei <jfmezei.spamnot@istop.com>& Subject: Re: [hp OpenVMS] SSL on VAX ?) Message-ID: <3EAC506C.A808187B@istop.com>0    Peter 'EPLAN' LANGSTOEGER wrote: > I > Recently I listed SSL on the list of differences between VAX and Alpha,oJ > because DEC/COMPAQ/HPQ doesn't provide SSL for the VAX. But, now I found  > in SYS$STARTUP:SSL$STARTUP.COM  + openSLL has been available on vax for ages.r   ------------------------------  % Date: Sun, 27 Apr 2003 15:59:24 -0400r* From: JF Mezei <jfmezei.spamnot@istop.com>@ Subject: Re: [SYSMAN, TCPIP] Unprintable Chars switches terminal) Message-ID: <3EAC369C.25DD94ED@istop.com>    Paul Sture wrote:tH > A faster method of resetting is to go into either EVE or EDT in screen > mode, and exit.a  M $SET TERM/INQUIRE brings the terminal back from graphics mode to normal mode.   ) (or on DECTERM COMMANDS -> RESET TERMINALt  O > > Where does the garbage (^N) come from ? From TCPIP display or from SYSMAN ?   M I think that the engineers owe us an explanation of why TCPIP SHO CONF COMM'syM output gets scrambled  when done from SYSMAN.  What sort of IO is TCPIP doingoE that screws up so many things ? (eg: what do we need to avoid doing).s   ------------------------------   End of INFO-VAX 2003.233 ************************ and respect even if aD formal document is not in place as you seem to think is required for online communications.  E I don't have time to respond to the micro-analysis of every word from G private emails and how it might be interpreted (or misinterpreted as in , your case) but lets take a few examples -=20    Quote #1 from you in your xH?&/ONئ* wݳM^xi0?0кw *>3Gr\,r.ր(e|0L8 ,f?WlAi*X$dӰqSԹ5ϵ;58	V}q b61oghoPK    r?"  #    fts.cIM VMSV       V5.5 IM$ VFAB P     B % @    m!Ц IM VFHC ,     ; 
|R%|    IM" VDAT ,     % ndşn  IM VPRO X     'b1@  0       IM VRDT      =R nYsG\\G׍EBH[$v0 :IPK%.jĞ]jgW23luqٮe|lmY_X%msj2;vbl'
|N#dǬ~4L%Cxcof-X\!:YݷGVIZ(vm=l3}<cm>8EHـۯDH2+AmsA}*;６2~=, b)qT֎
uqV	?Rf"9!Wy$3>f/\9\S
٩`w#:;$
A'SCJ湽2ejAJ8f`y)Eqg'<T,
Xb&@J'ʲ7E8#$xxks|(dAdr&
Qa"(h['c,KDkɒvGuD`$d047W-ȫfggjd"<VD-
ۋ䑱$hmP2
6Ѝ"(uh#uN08