1 INFO-VAX	Sun, 18 Jun 2006	Volume 2006 : Issue 336       Contents:0 Re: Just Ask The Vendor: Can OpenVMS Scale Down?0 Re: Just Ask The Vendor: Can OpenVMS Scale Down?0 Re: Just Ask The Vendor: Can OpenVMS Scale Down?0 Re: Just Ask The Vendor: Can OpenVMS Scale Down?P Re: Q: (long) DCL symbols not appearing in Show Symbol output [SLS V2.9H] V2.9H]  F ----------------------------------------------------------------------  # Date: Sat, 17 Jun 2006 19:01:40 GMT + From: "Villy Madsen" <Villy.Madsen@shaw.ca> 9 Subject: Re: Just Ask The Vendor: Can OpenVMS Scale Down? , Message-ID: <okYkg.52572$iF6.40963@pd7tw2no>  L And don't forget that the price of the ERP solution, installed tweeked etc &J in service - is often (grossly) understated.  Then, if it's built around aD large expensive backend product (db) then you have another source of2 on-going cost that keeps going up and up and up...  J I remember one discussion - where data from some smaller AS/400s was beingI sent to the ERP for consolidation.  The ERP provider wanted a license for I every seat on the AS/400 that was used to generate the data being sent up J for consolidation.  He wanted it, I don't know whether he got it or not...   Villy    ------------------------------  % Date: Sat, 17 Jun 2006 17:55:08 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 9 Subject: Re: Just Ask The Vendor: Can OpenVMS Scale Down? , Message-ID: <44947A35.903272C9@teksavvy.com>   Dave Froble wrote:C > The other side of this is that a company that is a leader in it's H > industry is such a leader for some reasons.  Usually it's because theyB > understand the industry better, and do a better job than others.  C Consider a maker of chocolate bars. Their specialty/expertise is in F making chocolate, designing the manufacturing process so that they canB magically insert the caramel inside its chocolate bars, as well as1 mounting a marketing campaign to sell those bars.   E Do they need to write an email system in house ? Nop. They can use an = off-the-shelf system whcih fits their needs quite well.  As a D multinational chocolate corporation, do they need to write their ownE payroll system ? They would be better off having seperate systems for B each country they operate in and buy locally produced payroll that2 fulfills localised practices for payroll, tax etc.  @ As long as an operation doesn't actually touch the production orA marketing of its core product, it can be "outsourced" to either a M shrinkwrapped software or really outsourced as a externally provided service.     @ Now, lets take ERP. This is more ticky. If you have a perishableD commodity whose price varies a lot, then you might want to have veryD specialised software that looks at the price of milk on a day by dayC basis. And when milk is cheaper, you get more delivered and ramp up S production of chocolate bars, and slow it down on days when milk is more expensive.   E For a non perishable product like cocoa, you may want to stockpile it S when it is cheap to buy so that you don't have to buy it when it is more expensive.   ? But suppose chocolate bars have seasonal variations and you are G approaching a lull period. That automated system would have to know not G to start stockpiling stuff even if the price is low because it won't be  needed for a long time.   E If you are making platic trinkets and have a fixed price contract for D plastic, then it is much easier to get an ERP system that meets yourK needs. You just match sales to production quantities and order accordingly.   D If you look at aircraft manufacturing, it gets more complex. When an< airline orders an airplane, it is highly customized, and theF manufactrurer promises a delivery date sometime in the future. BetweenD now and then, the computer has to keep track of that order, and makeE sure that it orders every custom part for that aircraft in time to be H delivered on the day that they will be working on that specific aircraftG for that specific customer (in fact it gets worse, because the aircraft F is assembled in different locations, brought to a main assembly place,G then moved to a finishing hangar where engines, cockpit electronics etc H are put, then moved to another place wherte the aircraft interios, seatsD etc are installed, so the computer needs to make sure that the rightE parts are delivered at the right time for the stage of assembly where H the part is installed, as well as delivered to the right location). SomeE parts, like the engines, may take months in lead time to order, while  others may take days.   B The question then becomes: how much money/effort is it to fight toF reshape SAP  to meet your needs, versus continuing to upgrade your ownG in-house software that was designed with those industry specific needs.   D SAP tells you that you need to reshape your company to fit SAP.  But? there are cases where this is very difficult if not impossible.   H The problem becomes that when a package becomes "trendy", then many justC want to implement it "to look cool" without really thinking through K about whether this package can really be customized to your company or not.     N >  When they adapt their organization to a generic piece of software, they canG > lose that advantage they had and become just another average company.   H But there are many areas of a corporation when a business process can beD streamlined even though manufacturing aspects are very specific. TheE trick is to know which part of a company can be made "vanilla" to fit H SAP and be more productive, and which part cannot be made to fit SAP and< SAP has to be made to fit the specific needs of the company.    G > An example of standardization is giving up the advantages of Alpha to / > use the generic itanic.  Where has this gone?   6 Actually, th initial promise of moving VMS to industry? standard/commodity hardware was not all bad. Yes, you loose the E performance advantage of Alpha. But you also gain the access to lower G priced hardware abnd ride on the markleting bandwagon that the industry 3 standard commodity and competitive environment has.   C The problem is that they didn't choose a platform that is "industry F standard" nor "commodity" nor "competitive" and one which doesn't evenF have a real performance edge. We we end up with a worse situation then@ with Alpha. At least Alpha was recognised as having real edge in+ performamce and/or design and capabilities.   G Had Digital played their cards right, they would have been able to make H Alpha "industry standard" either in 1995 when Windows 95 was introduced,F or in the 1998-1999 time frame when everyone and anyone was buying newG PCs to be Y2K compatible. Had that IA64 thing been ready for prime time E back then, Intel may have succeeded in switching the whole world from A 8086 to IA64. Alpha was ready for prime time, but nobody tried to  leverage that.  F Look at Compaq, instead of throwing in the towel, why didn't they justG give AMD rights to manufacture Alphas ? Get Sony to adopt Alpha for its E game consoles ? There was potential there. But nobody at the helm was C interested in pursiuing a vision of a fully leveraged potential for  Alpha.      C > company continuing to produce the itanic.  A leading edge company = > decided to become just another follower with no advantages.   E When you have a cancer on a limb, it is sometimes too late to save it F and you are better off cutting it off  and make the best of therest of@ the body. At this point in time,  VMS has battered disease after7 diseased related to its platform and lack of marketing.   G What is best right now is to lay the platform question to rest once and H for all by porting it to the 8086. It may not be as elegant as Alpha, itD may not be as performing  as as fancy as ALpha, but it is a platformG whose existenace is not in question. It is also the "core" platform for L HP, not some necessary evil platform that HP is stuck with due to contracts.  F By going to the 8086, VMS may lose performance edge it had with Alpha,D but it gains an assured future, it gains access to marketing, and itC gains access to fully scalable machines from laptop to data centre, G something which had purposefully been prevented by the various onwer of : VMS. And with Alpha dead, there really is no other option.  E I used to disagree with Mr Dachtera's desire to see VMS ported to the G 8086. But I have come to realise that by now, it is not only desirable, F but imperative to move VMS to it. The door opened when the 8086 became 64 bit capable.   H There is more talk in the press about the failure of IA64 than there wasH prior to June 25 2001 about Alpha's possible demise. VMS is worse off on6 IA64 today than it was on Alpha prior to June 25 2001.   ------------------------------  % Date: Sat, 17 Jun 2006 18:00:35 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 9 Subject: Re: Just Ask The Vendor: Can OpenVMS Scale Down? , Message-ID: <44947B7B.7D5D0A0A@teksavvy.com>  C > >> adapt the organization to the ERP system and not the other way 
 > >> around.      H I find it interesting to compare this to All-in-1.  All-In-1 was in many= cases simimilar to EVE: a baseline implementation meant to be G customised. In fact, customers were encouraged to customised ALLIN1 and E were given all the tools to do so.  They wanted you to integrate your B aopps to it, and write apps in ALLIN1. (which allowed mail enabled applciations etc etc).  D Interestingly, few clients really saw the potential of that and manyG just ran a valilla uncustomised ALLIN1 (interestinly, most DEC officesa ? round the world are guylty of running an uncustomized version).     H So it is interesting to see SAP being sold with a opposite philosophy ofF reducing customisations and adapting he company to the package instead of the other way around.   ------------------------------  % Date: Sat, 17 Jun 2006 23:10:29 -0400 ' From: Dave Froble <davef@tsoft-inc.com> 9 Subject: Re: Just Ask The Vendor: Can OpenVMS Scale Down? 9 Message-ID: <v62dnTpb1qaJXgnZnZ2dneKdnZydnZ2d@libcom.com>    JF Mezei wrote:  > Dave Froble wrote:D >> The other side of this is that a company that is a leader in it'sI >> industry is such a leader for some reasons.  Usually it's because they C >> understand the industry better, and do a better job than others.  > E > Consider a maker of chocolate bars. Their specialty/expertise is in H > making chocolate, designing the manufacturing process so that they canD > magically insert the caramel inside its chocolate bars, as well as3 > mounting a marketing campaign to sell those bars.  > G > Do they need to write an email system in house ? Nop. They can use an ? > off-the-shelf system whcih fits their needs quite well.  As a F > multinational chocolate corporation, do they need to write their ownG > payroll system ? They would be better off having seperate systems for D > each country they operate in and buy locally produced payroll that4 > fulfills localised practices for payroll, tax etc.  @ Some things, and payroll is a special case, actually need to be I 'standard'.  Small and medium sized companies are better off just paying  H a bank or ADP or such to do the payroll.  That avoids multiple problems.  B > As long as an operation doesn't actually touch the production orC > marketing of its core product, it can be "outsourced" to either a O > shrinkwrapped software or really outsourced as a externally provided service.  >  > B > Now, lets take ERP. This is more ticky. If you have a perishableF > commodity whose price varies a lot, then you might want to have veryF > specialised software that looks at the price of milk on a day by dayE > basis. And when milk is cheaper, you get more delivered and ramp up U > production of chocolate bars, and slow it down on days when milk is more expensive.  > G > For a non perishable product like cocoa, you may want to stockpile it U > when it is cheap to buy so that you don't have to buy it when it is more expensive.  > A > But suppose chocolate bars have seasonal variations and you are I > approaching a lull period. That automated system would have to know not I > to start stockpiling stuff even if the price is low because it won't be  > needed for a long time.   H Now you're touching on some key issues.  A very good and specific sales I analysis system is needed to insure you produce enough, and not too much.   G A former customer was a sporting goods distributor.  Consider baseball  H equipment.  Consider just how seasonal such can be.  An old friend once H looked at the company and marveled at their capability to do well in an I environment where nine (9) months of the year the company had a negative  
 cash flow.  D Now consider purchasing gloves from a supplier in the Orient.  Lead H times can approach six (6) months.  You better have a real good idea of H what will be the hot sellers this year, and just how many you can sell. E   Don't order enough, your customers (stores) get upset because they  F cannot get enough product.  Order too much, and end up destroying the  excess.   E A custom sales analysis system was designed to allow the customer to  G track bookings (made in Sept for delivery in Jan-Mar) by week, current  B year vs last year vs 2 years ago.  The ordering of long lead time H products became VERY effective.  That's called 'doing it better'.  That / makes you more profitable than the competitors.   G > If you are making platic trinkets and have a fixed price contract for F > plastic, then it is much easier to get an ERP system that meets yourM > needs. You just match sales to production quantities and order accordingly.  > F > If you look at aircraft manufacturing, it gets more complex. When an> > airline orders an airplane, it is highly customized, and theH > manufactrurer promises a delivery date sometime in the future. BetweenF > now and then, the computer has to keep track of that order, and makeG > sure that it orders every custom part for that aircraft in time to be J > delivered on the day that they will be working on that specific aircraftI > for that specific customer (in fact it gets worse, because the aircraft H > is assembled in different locations, brought to a main assembly place,I > then moved to a finishing hangar where engines, cockpit electronics etc J > are put, then moved to another place wherte the aircraft interios, seatsF > etc are installed, so the computer needs to make sure that the rightG > parts are delivered at the right time for the stage of assembly where J > the part is installed, as well as delivered to the right location). SomeG > parts, like the engines, may take months in lead time to order, while  > others may take days.  > D > The question then becomes: how much money/effort is it to fight toH > reshape SAP  to meet your needs, versus continuing to upgrade your ownI > in-house software that was designed with those industry specific needs.    Using $250/hr SAP consultants.  F > SAP tells you that you need to reshape your company to fit SAP.  ButA > there are cases where this is very difficult if not impossible.    The road to hell.   J > The problem becomes that when a package becomes "trendy", then many justE > want to implement it "to look cool" without really thinking through M > about whether this package can really be customized to your company or not.   F Beancounters don't understand, and will listen to other beancounters, F the large accounting firms, who also sell the SAP consulting services.  O >>  When they adapt their organization to a generic piece of software, they can H >> lose that advantage they had and become just another average company. > J > But there are many areas of a corporation when a business process can beF > streamlined even though manufacturing aspects are very specific. TheG > trick is to know which part of a company can be made "vanilla" to fit J > SAP and be more productive, and which part cannot be made to fit SAP and> > SAP has to be made to fit the specific needs of the company. >  > H >> An example of standardization is giving up the advantages of Alpha to0 >> use the generic itanic.  Where has this gone?  B And now stay tuned for another adventure in JF never, never, land.  8 > Actually, th initial promise of moving VMS to industryA > standard/commodity hardware was not all bad. Yes, you loose the G > performance advantage of Alpha. But you also gain the access to lower I > priced hardware abnd ride on the markleting bandwagon that the industry 5 > standard commodity and competitive environment has.  > E > The problem is that they didn't choose a platform that is "industry H > standard" nor "commodity" nor "competitive" and one which doesn't evenH > have a real performance edge. We we end up with a worse situation thenB > with Alpha. At least Alpha was recognised as having real edge in- > performamce and/or design and capabilities.  > I > Had Digital played their cards right, they would have been able to make J > Alpha "industry standard" either in 1995 when Windows 95 was introduced,H > or in the 1998-1999 time frame when everyone and anyone was buying newI > PCs to be Y2K compatible. Had that IA64 thing been ready for prime time G > back then, Intel may have succeeded in switching the whole world from C > 8086 to IA64. Alpha was ready for prime time, but nobody tried to  > leverage that. > H > Look at Compaq, instead of throwing in the towel, why didn't they justI > give AMD rights to manufacture Alphas ? Get Sony to adopt Alpha for its G > game consoles ? There was potential there. But nobody at the helm was E > interested in pursiuing a vision of a fully leveraged potential for 	 > Alpha.   >  > D >> company continuing to produce the itanic.  A leading edge company> >> decided to become just another follower with no advantages. > G > When you have a cancer on a limb, it is sometimes too late to save it H > and you are better off cutting it off  and make the best of therest ofB > the body. At this point in time,  VMS has battered disease after9 > diseased related to its platform and lack of marketing.  > I > What is best right now is to lay the platform question to rest once and J > for all by porting it to the 8086. It may not be as elegant as Alpha, itF > may not be as performing  as as fancy as ALpha, but it is a platformI > whose existenace is not in question. It is also the "core" platform for N > HP, not some necessary evil platform that HP is stuck with due to contracts. > H > By going to the 8086, VMS may lose performance edge it had with Alpha,F > but it gains an assured future, it gains access to marketing, and itE > gains access to fully scalable machines from laptop to data centre, I > something which had purposefully been prevented by the various onwer of < > VMS. And with Alpha dead, there really is no other option. > G > I used to disagree with Mr Dachtera's desire to see VMS ported to the I > 8086. But I have come to realise that by now, it is not only desirable, H > but imperative to move VMS to it. The door opened when the 8086 became > 64 bit capable.  > J > There is more talk in the press about the failure of IA64 than there wasJ > prior to June 25 2001 about Alpha's possible demise. VMS is worse off on8 > IA64 today than it was on Alpha prior to June 25 2001.     --  4 David Froble                       Tel: 724-529-0450> Dave Froble Enterprises, Inc.      E-Mail: davef@tsoft-inc.com DFE Ultralights, Inc.  170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  % Date: Sat, 17 Jun 2006 22:54:47 -0500 6 From: "David J. Dachtera" <djesys.no@spam.comcast.net>Y Subject: Re: Q: (long) DCL symbols not appearing in Show Symbol output [SLS V2.9H] V2.9H] 0 Message-ID: <4494CE87.F8FBE095@spam.comcast.net>   Ken Fairfield wrote: > C > Having worked with VMS for more than 20 years, I really thought I ? > understood DCL.  But here's a puzzle I just can't understand!  > @ > In the context of SLS's POST_PROCESS_EACH.COM (and similar forH > PRE_PROCESS_FIRST, PRE_PROCESS_EACH and POST_PROCESS_LAST, SLS V2.9H),J > I was trying to determine what *symbols* are available to that procedureH > (since the documentation in Section 5.3.3.1 & 5.3.3.2 of the "Guide toE > Backup and Restore Operations" is known to be incorrect).  Near the : > beginning of my procedure, I inserted the following DCL: > E > $Write Sys$Output "''F$Fao("!30*% In POST_PROCESS_EACH !30*%!/ ")'"  > $ Show Symbol *  > $ Show Symbol /Local  *  > $ Show Symbol /Global *  > $ Show Symbol /Local  status > $ Show Symbol /Global status! > $ Show Symbol /Local  statusmsg ! > $ Show Symbol /Global statusmsg  > $ Show Symbol first_reel > $ Show Symbol last_reel  > $ Show Symbol first  > $ Show Symbol last > $ Show Symbol status > $ Show Symbol statusmsg E > $Write Sys$Output "''F$Fao("!/ !30*% In POST_PROCESS_EACH !30*%")'"  > C > There's a lot of redundancy here as I was trying "many" different F > approaches.  Note that the documentation claims that all of 'first',G > 'last', 'status' and 'statusmsg', are supposed to be available to all I > four procedures.  'first_reel' and 'last_reel' are *not* documented but / > are, in fact, available in POST_PROCESS_EACH.  > H > I discovered three things that I cannot explain (see the output at the > bottom of this post):  > E >     1) *Neither* 'first_reel' nor 'last_reel' are present in any of B >        the wildcard Show Symbol commands.  They are present on a6 >        Show Symbol of those names.  How can that be? > C >     2) 'status' is not present on any of the wildcard Show Symbol D >        commands, nor on "Show Symbol [ /Local | /Global ] status".E >        However, the local symbol 'status' is present (see output of C >        the next to last Show Symbol command below).  How can that 
 >        be??  > F >     3) 'statusmsg' does not appear in *any* Show Symbol output, evenB >        the last.  However, it is present. For example, I can do: > ; >          $ Write Sys$Output "Final status = ''statusmsg'"  > E >        and have, e.g, "Final status = BACKUP-W-ACCONFLICT," written H >        to the log file, and this after testing $ If statusmsg.Nes."" . >        How can this be???  > C > I have tried to reproduce some of these effects interactively, by G > defining local and global symbols of the same name but with different I > values, and by varying "symbol assign (:=)" versus "string assign (=)", 9 > but nothing unexpected, nothing like the above, occurs.  > F > Can anyone explain what is happening?  Is there a 3rd, or 4th, classA > of symbols I've never heard of that are hidden (sometimes) from % > Show Symbol??  Have I gone crazy???   8 No - you're fine. Just one of the "finer" points of DCL.  8 What you're seeing is symbol defined at lesser "depths".  E Here's a little example I just whipped up on the little Alpha here at  home...    DJAS01::DDACHTERA$ type one.com  $ @two.com 1 2 3 4 5 6 7 8   $ exit DJAS01::DDACHTERA$ type two.com 
 $ set noon
 $ show sym p%  $ del/sym p1 $ del/sym p2 $ del/sym p3 $ del/sym p4 $ del/sym p5 $ del/sym p6 $ del/sym p7 $ del/sym p8
 $ show sym p% 
 $ show sym p1  $ say "P1 = """, p1, """"  $ exit DJAS01::DDACHTERA$ set ver' DJAS01::DDACHTERA$ @one a b c d e f g h  $ @two.com 1 2 3 4 5 6 7 8  
 $ set noon
 $ show sym p% 
   P1 = "1"
   P2 = "2"
   P3 = "3"
   P4 = "4"
   P5 = "5"
   P6 = "6"
   P7 = "7"
   P8 = "8" $ del/sym p1 $ del/sym p2 $ del/sym p3 $ del/sym p4 $ del/sym p5 $ del/sym p6 $ del/sym p7 $ del/sym p8
 $ show sym p% = %DCL-W-UNDSYM, undefined symbol - check validity and spelling 
 $ show sym p1 
   P1 = "A" $ say "P1 = """, p1, """"  P1 = "A" $ exit $ exit DJAS01::DDACHTERA$    H I might even add that to my DCL programming hands-on lab (which has been accepted for Houston).  C P1, at the end of TWO.COM, is a symbol local to depth one(1). It is F available, but "invisible" , if you will, to a wildcarded SHOW SYMBOL.F Note that when P1 if referenced explicitly in a SHOW SYM[BOL] command,G it is displayed as a local symbol - just not designated as local to the  current procedure depth.  . Guy - any thoughts here?  Suggestion for this:   $ SHOW SYMBOL/FULL P1    P1 = "A" [1/2]  ) ...where the numbers in the brackets are:    [symbol-depth/current-depth]  @ SET SYMBOL/SCOPE can be used to adjust this behavior (left as an exercise for the reader).    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  & Unofficial OpenVMS Marketing Home Page! http://www.djesys.com/vms/market/   ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/   ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/   ------------------------------   End of INFO-VAX 2006.336 ************************