0 INFO-VAX	Sat, 11 Feb 2006	Volume 2006 : Issue 84      Contents: Alpha CPU testing equipment?% Re: Excessive paging problem - SOLVED 5 Re: Failover from SDSL to ADSL on a single Cisco 2801 5 Re: Failover from SDSL to ADSL on a single Cisco 2801 * Re: I64 Support: Stronger than is thought?* Re: I64 Support: Stronger than is thought?* Re: I64 Support: Stronger than is thought? Re: null terminated strings  RE: null terminated strings  RE: null terminated strings  Re: null terminated strings  Re: null terminated strings  Re: null terminated strings  Re: null terminated strings  Re: null terminated strings  Re: null terminated strings  Re: null terminated strings A Re: OT: Gartner wakes up company executives to X86-64 scalability A Re: OT: Gartner wakes up company executives to X86-64 scalability  Perl, DBI, and DBD::RDB & Re: Security experts say no secure OS! Source Listing Re: V7.3-2 and Joliet CDs  Re: V7.3-2 and Joliet CDs  Re: V7.3-2 and Joliet CDs  Re: V7.3-2 and Joliet CDs   F ----------------------------------------------------------------------    Date: 11 Feb 2006 10:31:32 -0800 From: "cmk" <ckaron@gmail.com>% Subject: Alpha CPU testing equipment? B Message-ID: <1139682692.241325.17650@g47g2000cwa.googlegroups.com>  	 Hi group!     + I've found something strange at scrap-shop.   & It is an unit size of an ossciloscope,. which is claimed to be an "ALPHA CPU TEST SET"'    No sign of an indentification plate.     Manufacturer unknown.  ? I'm unable to verify that such devices were ever sold publicly. 9 Google is mum about actual equipment - seems to talk only > in general about alpha's functional testing - mostly performed on complete systems.  @ I doubt this is a real thing, and don't believe it could be usedC (or usefull) in the wild. Previous owner is said to be Symbol Tech. 5 the peolpe who make bar-code, RFID scanners and such.     : Has anyone seen anything alike? How should I check if it's= the real thing? (nice to have - from enthusiast perspective).   D Maybe someone can tell a bit about Alpha's testing at manufacturing?A (or Itanic - it is of interest how these 3-D colour X-ray imaging  technologiesB which Intel bought along witch NewFocus Inc. are actually applied)     Regards, Chris    ------------------------------  % Date: Sat, 11 Feb 2006 09:58:59 +0100 + From: Bernhard Dorninger <bd_@gmxNOSPAM.at> . Subject: Re: Excessive paging problem - SOLVED. Message-ID: <1139644508.939383@news.liwest.at>   Hello again!    & For those, who might be interested....  > After experimenting with the Java cmd line args and long time ( observation, we finally found a solution  G Activating the switch -Xdynclassgc did the trick. The Java VM now runs   with a stable working set.   Cheers,  Bernhard   ------------------------------  % Date: Sat, 11 Feb 2006 01:48:15 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> > Subject: Re: Failover from SDSL to ADSL on a single Cisco 2801, Message-ID: <43ED8892.7F22BCF0@teksavvy.com>    erosbresolin@equannox.com wrote:H > We have a Cisco 2801 Router that currently routes a number of internalF > networks out via one interface onto an SDSL router and then onto theH > internet. It is configured to run NAT and there are a number of StaticH > IP mappings between given external addresses (associated with the ADSL9 > accounts) to specific servers on the internal networks.  > I > We would like to set up a new ADSL line to act as a backup in the event G > of the SDSL failing. However any new ADSL account will obviously come * > with a new set of external IP addresses. > I > Is there anyway to configure the CISCO router to allow for some form of D > automatic failover that can also handle the new IP address without& > breaking the NAT translation tables?  G This may not be the best newsgroup for this. You might want to ask in a  cisco specific newsgroup.    you can start to read \ > http://www.cisco.com/en/US/tech/tk648/tk361/technologies_white_paper09186a0080091cb9.shtml0 (description of NAT features on Cisco machines).   One thing you could do:   F Each LAN side machine has 2 interfaces sporting IPs in 2 subnets. Lets- say host 1 has IP 10.0.0.1 and IP 192.168.0.1   ; You can then map  external ip 66.33.71.27 to 10.0.0.1 , and  132.127.33.17 to 192.168.0.1  ? This way, you coudl get packets coming from link 1 to be routed ? internally to 10.0.0.1, and packets coming from link 2 to go to 9 192.168.0.1 which happen to be the same physical machine.   G Not sure how to handle failover. Perhaps CISCO has innovative solutions  in its software.  E The "real" solution is to officially become multi-homed, get your ASN F number and get your ISPs to provide BGP services to you. With that, ifG link1 goes down, traffic would automatically use link2 and use the same H IPs. But this is not a small undertaking and it needs a router with lotsG of memory to hold the world's full BGP table. Not a project you do over 
 a weekend.   ------------------------------  + Date: Sat, 11 Feb 2006 13:50:33 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)> Subject: Re: Failover from SDSL to ADSL on a single Cisco 2801$ Message-ID: <dskq38$vk3$1@online.de>  B In article <1139625446.864295.36290@o13g2000cwo.googlegroups.com>," erosbresolin@equannox.com writes:   : > Thanks for your response Phillip. Very much appreciated. > 6 > Could you please provide me with a bit more detail?   / I'm not sure if it is what you are looking for.   H The basic idea of a NAT router is that it maps all outgoing connections H to one address, so that, to the outside world, all computers behind the + NAT router appear to have the same address:   
 NAT Router ---------------------------  LAN           WAN  -----------   -------------  192.168.1.2   84.211.143.21  192.168.1.3  192.168.1.4   F The 192 addresses would be the local machines.  To the outside world,  all appear to be 84.211.143.21.   @ JF's idea was to set up a second NAT router with a different WANI address.  This could go to a different ISP or whatever.  The trick is to  H get a THIRD NAT router and connect it between the local network and the G other two NAT routers, but set it up "backwards".  In other words, the  B VMS machines will connect to its WAN port (obviously, you need an H additional hub here since there is just one WAN port) and the LAN ports H connect to the LAN ports of the first two routers.  This means that, to F the local machines, ALL incoming connections will appear to come from ! the WAN port of the third router.   I NAT (and especially PAT, if you are using that as well) is a performance  H hit, since all packets need to be dis- and re-assembled.  However, this E is a cheap solution to avoid one ISP being a single point of failure.    See   ? http://groups.google.de/group/comp.os.vms/browse_thread/thread/ F 1943627ead5db1e6/0f6183a92e978783?q=%2BNAT+%2Brouter++%2BWAN+%2BMezei& rnum=1#0f6183a92e978783    ------------------------------  % Date: Sat, 11 Feb 2006 02:43:34 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 3 Subject: Re: I64 Support: Stronger than is thought? , Message-ID: <43ED9585.BAA73768@teksavvy.com>   Bill Todd wrote: > I > Well, possibly stronger than *you* thought, if you hadn't been aware of 2 > the companies nominally offering Itanic systems.  H The issue with those IA64 systems from different companies is that whileF they may share the CPU, they are not compatible. VMS only runs on HP'sH machines. And the japanese companies who do sell some IA64 things abuild( highly proprietary specialised machines.   ------------------------------    Date: 11 Feb 2006 09:04:50 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) 3 Subject: Re: I64 Support: Stronger than is thought? 3 Message-ID: <5PEnAa4HQuhr@eisner.encompasserve.org>   \ In article <43ED9585.BAA73768@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes: > Bill Todd wrote: >>  J >> Well, possibly stronger than *you* thought, if you hadn't been aware of3 >> the companies nominally offering Itanic systems.  > J > The issue with those IA64 systems from different companies is that whileH > they may share the CPU, they are not compatible. VMS only runs on HP's > machines.    How do you know that ?  C How many different machines are there on which you have tried VMS ?   C The last I heard was that Itanium system infrastructure (similar to E console code) is rather uniform and that the major limitation is that F VMS is only _supported_ on HP machines.  Perhaps the hobbyist licenses9 allow use on any machine, since they are not _supported_.   B This seems no different from Alpha, where there were a few non-DEC% machines built for Unix and the like.    ------------------------------  % Date: Sat, 11 Feb 2006 11:08:50 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>3 Subject: Re: I64 Support: Stronger than is thought? + Message-ID: <43EE1A21.8CB12489@comcast.net>    Bill Todd wrote: >  > [snip]E > So I find the *decrease* in support for Itanic (Sun bailed out very G > early, IBM stuck around until the handwriting on the wall was clearer J > but then didn't hesitate to leave, and others - even including HP - haveF > showed increasing interest in x86-64 for high-end use recently) moreJ > significant than the fact that the rest of those corporations who boughtG > into Itanic early-on have not yet completely given up on it.  Perhaps H > you weren't aware that they've been on board all along, and that's the > reason you were impressed.  H I was more impressed with them STILL being on-board with I64 in the faceD of some high-profile defections. Then again, IBM set their precedentD with PCs: they established the market, but did not really embrace itE with gusto until it started running away with the "boodle". THEN they * realized (too late) what they had started.  ? Can't say as I64 will do the same. It may just be an incredibly B expensive version of the next Alpha. However, with others on boardF besides HP, at least it will have more market presence than Alpha, and, that may be good in the long run. We'll see.   --   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/   ------------------------------  % Date: Sat, 11 Feb 2006 02:01:58 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> $ Subject: Re: null terminated strings, Message-ID: <43ED8BC8.6B225724@teksavvy.com>   Dave Froble wrote:C > I know of no other language that handles arrays in such a manner. D > Having used a number of assembly languages, I understand what it'sI > doing.  What I don't understand is why such a thing was included in any * > language that's more than an assembler.     D Not all "arrays" are declared as arrays. You can declare a structureC (record format) and then dynamically alloacte a buffer big eough to H contain a dynamically calculated number of instances of that record. YouG can then use pointers to access individual elements within that buffer.   C C also allows you to pass this pointer to a called subroutine which G expected an argument of that structure type and the subroutine can then  access that array.  H Also, it allows you to pass not just the base element of the array , butC also the address of the nth element of the array and the subroutine F treats that nth element as element 0. (eg: skip processing the first 5G records by passing the address of the 6th record instead of that of the  first record).  H The pointer artihmetic becomes useful in structures because the compilerC automatically calculates the structure size for you for any pointer  arithmetic you do.  D While VMS has very limited support for multybvyte characters, others@ OSes do and that allows the "char" to become a 2 byte entity andG programs that scan a character array then automatically advance 2 bytes / at a time when the pointer in incremented by 1.    ------------------------------  % Date: Sat, 11 Feb 2006 18:48:56 +1100 6 From: "O'Brien Paddy" <Paddy.O'Brien@transgrid.com.au>$ Subject: RE: null terminated stringsX Message-ID: <0A7046B0A95F2B41B3712F0C5FD1CDC307E670@ex-tg2-pr.corporate.transgrid.local>   -----Original Message-----. From: Dave Froble [mailto:davef@tsoft-inc.com]' Sent: Friday, February 10, 2006 6:18 PM  To: Info-VAX@Mvb.Saic.Com $ Subject: Re: null terminated strings     JF Mezei wrote:  > Dave Froble wrote: >=20H >>I think it's totally relavent.  Range checking is a restriction, and IF >>would not want to be without it.  While it may have been left out ofD >>some older stuff, there's no valid reason to leave it out of a HLL >>today. >=20 >=20 >=20H > Consider the following scenario. You have multiple arrays of differentJ > sizes. Think item lists where the last item is a 0 to indicate it is the > last item.=20  >=20G > You pass this to a subroutine. The subroutine has no wait to know how J > many items are in the array, especially since the same subroutine can beG > passed different size arrays during the same invocation of the image.    Wrong!  J If the address of the array descriptor is passed, then any array passed=20- to the subprogram brings with it it's bounds.   F You get such usefulness when you have a good calling standard and a=20 decent language.   *****   4 I believe you are wrong, but with some reservations.  L The passing mechanism is a bit more complex.  Descriptors are normally pass=L ed for length of declared Character (Fortran) variables (not sure of other =H languages), but I believe the VMS standard only used to allow for this).  L Since Fortran 90/95, there does have to be a mechanism to pass array sizes,=L  if the routine has an INTERFACE block.  Prior to that, I think most langua=L ges passed the location of the first element in the array.  Look at VAX mac=0 ro code for how it accepts arrays or characters.  L John R., Does Fortran set up a descriptor, or have you created a separate c=( alling standard for this f90/95 feature?   Regards, Paddy=20     G *********************************************************************** ; Please consider the environment before printing this email.   C "This electronic message and any attachments may contain privileged @ and confidential information intended only for the use of the=20D addressees named above.  If you are not the intended recipient of=20C this email, please delete the message and any attachment and advise D the sender.  You are hereby notified that any use, dissemination,=207 distribution, reproduction of this email is prohibited.   C If you have received the email in error, please notify TransGrid=20 C immediately.  Any views expressed in this email are those of the=20 ? individual sender except where the sender expressly and with=20 C authority states them to be the views of TransGrid.  TransGrid uses > virus-scanning software but excludes any liability for viruses contained in any attachment.  < Please note the email address for TransGrid personnel is now$ firstname.lastname@transgrid.com.au"  G ***********************************************************************    ------------------------------  % Date: Sat, 11 Feb 2006 19:19:47 +1100 6 From: "O'Brien Paddy" <Paddy.O'Brien@transgrid.com.au>$ Subject: RE: null terminated stringsX Message-ID: <0A7046B0A95F2B41B3712F0C5FD1CDC307E672@ex-tg2-pr.corporate.transgrid.local>   -----Original Message-----2 From: Karsten Nyblad [mailto:nospam@nospam.nospam]* Sent: Saturday, February 11, 2006 12:53 AM To: Info-VAX@Mvb.Saic.Com $ Subject: Re: null terminated strings     Dave Froble wrote:G > I'm in the minority on this issue.  I'm aware of that.  Most of my=20 I > career has been writing new stuff, and most of the maintenance is on=20 J > stuff I wrote.  Not declaring variables hasn't been a large problem for= =20 K > me.  That said, I'll agree that it's safer, and better in many instances.  >=20J > When using local temporary variables, it is nice to not have to declare= =20  > such.   H Even a programmer with a poor typewriting does not use that much time=20F typing the code his writes.  The programmer uses much more time for=20I reading, debugging, and understanding the code.  Thus languages should=20 G make code readable, easy to understand, and as many errors should be=20 K caught before debugging.  Say your finger slips, such that you get a new=20 J implicitly declare variable.  It will take you much longer time finding=20C that bug with a debugger than if it is reported by the compiler.=20 F Further, you will have to use more time for proofreading your code,=20J because you have to look for faulty typings.  At last it will not be as=20F easy to maintain, because later programmers will have more problems=208 finding out how and for what purpose variables are used.  H I am sure that the extra work of typing declarations is an investment=20J that is paid back before the code has gone into production, and is paid=201 back many times if the code has to be maintained.    *****   L I agree with your sentiments, and definitely explicit type declarations do =L help.  I find, though, that I like to put declarations into alphabetical or=L der.  I am a Fortran (mainly) programmer from the early '60's and I still u=L nderstand implicit type names.  I find it easier to verify whether J is an =@ integer or a real if I can quickly look at the declaration list.  L The worst thing that I found was when an ex-boss hired a student to add dec=L larations to an old program/routines.  One variable, N-something, was not d=L eclared, so the student assumed it was an integer from the implicit Fortran=L  rules.  It had never been, it was supposed to be a real but never declared=L  as such.  Now that the routine now explicitly typed it, we took longer to =L find that bug because we took the declaration on faith.  Beware just adding=H  declarations in a language like Fortran that has implicit declarations.  L And in Fortran 90, no routine should be CONTAINS unless both have been chec=, ked with IMPLICIT NONE.  Beware inheritance.   Regards, Paddy    G *********************************************************************** ; Please consider the environment before printing this email.   C "This electronic message and any attachments may contain privileged @ and confidential information intended only for the use of the=20D addressees named above.  If you are not the intended recipient of=20C this email, please delete the message and any attachment and advise D the sender.  You are hereby notified that any use, dissemination,=207 distribution, reproduction of this email is prohibited.   C If you have received the email in error, please notify TransGrid=20 C immediately.  Any views expressed in this email are those of the=20 ? individual sender except where the sender expressly and with=20 C authority states them to be the views of TransGrid.  TransGrid uses > virus-scanning software but excludes any liability for viruses contained in any attachment.  < Please note the email address for TransGrid personnel is now$ firstname.lastname@transgrid.com.au"  G ***********************************************************************    ------------------------------    Date: 11 Feb 2006 02:41:19 -06002 From: "Dave Weatherall" <djw-nothere@nospam.nohow>$ Subject: Re: null terminated strings? Message-ID: <DTiotGxQ0bj6-pn2-7mwqgCOMoYwH@dave2_os2.home.ours>   D On Sat, 11 Feb 2006 03:34:57 UTC, Dave Froble <davef@tsoft-inc.com>  wrote:   > Larry Kilgallen wrote: > > Karsten Nyblad wrote:  > >  > > N > >>Let a pointer in C point to an element in an array.  Then the pointer +   N > >>1 will point to the next element in the array.  (Assuming that the type   N > >>of the pointer matches the element type of the array.)  But I wonder how  ( > >>many people use that feature today.  > >  > > ` > > In article <43ED3339.DBAE181E@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes: > >  > >>Larry Kilgallen wrote: > >>J > >>>Whereas I was wondering how many will remember how it works when they > >>>write such code.  > >> > >>, > >>Pointer arithmetic in C is well defined. > >  > > G > > But if Karsten is correct that not many people use the feature, the @ > > fact that it is well defined will not help them remember it. > I > Earlier in this thread, at least I think so, the claim was made that C  < > was designed to have the capabilities of PDP-11 assembler. > D > I know of no other language that handles arrays in such a manner. E > Having used a number of assembly languages, I understand what it's  J > doing.  What I don't understand is why such a thing was included in any J > language that's more than an assembler.  I've been thinking rather hard K > on the technique, and I challenge anyone to provide a need for such that  + > isn't more easily done in another manner.  > I > What it suggests to me is the capability of looking at some part of an  K > array element, not necessarily including the first byte/bit in the array.    DaveE             consider the autoincrement/decrement addressing modes on  F pdp and VAX. The value of the index register was incremented according, to the size of the otem being operated on :-  , 	movb	(r5)+, r0		; r5 bumped up one location 	movw	(r5)+, r0		; r5 bumped 2 	movl	(r5)+, r0		; r5 bumped 4  B This is the behaviour that the C pointer arithmetic is following, . methinks. It allowed the compiler to translate   	p++     or      p = p +1   ) to the correct autoincrement instruction.   A Actually if you look at the assembler code generated by many VAX  = compilers you might find the same constructs. Certainly ISTR  F VAX-Fortran generating them for x = y DO loops. Once you get to arraysD of structures that are not 'natural' the compiler need to generate  . ADD instructions using sizeof(array element).   E In the 80s, I took on responsibility for some Macro-11 that was full    of statements along the lines of   	cmp	(r0)+,(r0)+   which was just to avoid   
 	add	#4,r0  F It saved a word ISTR. I wasn't sure the two memory fetches were worth F it and put the ADD back.  (not to mention the possibility of an accessE violation from (r0) when moved to VMS). To be fair the code had been  F wriiten originally for a small memory pdp (16K?) inder DOS-11. By the F time I picked it up it was on RSX (32K non-overlayed tasks) looking to VMS.  = I may have mis-read your point. If so I apologise in advance.    --   Cheers - Dave W.   ------------------------------    Date: 11 Feb 2006 02:41:21 -06002 From: "Dave Weatherall" <djw-nothere@nospam.nohow>$ Subject: Re: null terminated strings? Message-ID: <DTiotGxQ0bj6-pn2-vVTkwM7rX7XW@dave2_os2.home.ours>   1 On Fri, 10 Feb 2006 18:17:39 UTC, Karsten Nyblad   <nospam@nospam.nospam> wrote:    > Dave Froble wrote: > > Karsten Nyblad wrote:  > >  > >> Dave Froble wrote:  > >>I > >>> I'm in the minority on this issue.  I'm aware of that.  Most of my  K > >>> career has been writing new stuff, and most of the maintenance is on  J > >>> stuff I wrote.  Not declaring variables hasn't been a large problem I > >>> for me.  That said, I'll agree that it's safer, and better in many   > >>> instances. > >>> F > >>> When using local temporary variables, it is nice to not have to  > >>> declare such.  > >> > >> > >>K > >> Even a programmer with a poor typewriting does not use that much time  I > >> typing the code his writes.  The programmer uses much more time for  L > >> reading, debugging, and understanding the code.  Thus languages should J > >> make code readable, easy to understand, and as many errors should be J > >> caught before debugging.  Say your finger slips, such that you get a I > >> new implicitly declare variable.  It will take you much longer time  N > >> finding that bug with a debugger than if it is reported by the compiler.  > >  > > H > > Well, it can be caught by the compiler.  I have utilities that will M > > parse the /LIST output of the compiler, and report on variables that are  L > > referenced only once.  I can then check out such, and determine whether L > > it's a typo, or something brought in with some boiler-plate sections of 	 > > code.  > H > Well, then you had better not make the same typo twice.  Being bad at H > spelling I could easily make the same spelling error twice.  I have a D > hunch that writing and using that utility has taken you a sizable > > fraction of the time you have save not writing declarations. > I > >> Further, you will have to use more time for proofreading your code,  2 > >> because you have to look for faulty typings.  > >  > > 5 > > Proof reading your code is a bad thing ??????????  > E > Certainly not, but proofreading to find errors that could be found  G > automatically is a waste of resources.  And you can look for only so  + > many different types of errors at a time.  > J > >> At last it will not be as easy to maintain, because later programmersT > > will have more problems finding out how and for what purpose variables are used. > >  > > M > > When very descriptive variable names are used, this is less of a problem. J >  > Further, declaring an obscure name for a variable is no help to laterJ > > programmers.  This is a red herring.  The declaration of a variable is; >  > of no help in determining what a variable is used for.  > J > All programmers have their own way of writing very descriptive variable K > names.  Unfortunately what is very descriptive to one programmer may not  G > be it to an other.  Thus the variables and their use should still be  J > documented.  The declaration is an appropiate place for a comment doing  > that.  > H > And if all variables are declared then you know that you can find the H > declaration.  You will not search for a variable's declartion that is G > not there or worse:  Search for a declaration that is there, but you  9 > fail to find it and think the variable is not declared.   A Strangley enough if you're used to assembler then you're used to  7 declaring / typing your variables / memory locations...   F IMPLICIT NONE has probably saved me a lot of trouble. It's not easily A quantified. Forgetting it has caused me not much volume-wise but  ( enough time-wise. (The proverbial typo).    --   Cheers - Dave W.   ------------------------------  % Date: Sat, 11 Feb 2006 10:41:02 +0100 + From: Karsten Nyblad <nospam@nospam.nospam> $ Subject: Re: null terminated strings= Message-ID: <43edb127$0$78285$157c6196@dreader1.cybercity.dk>    Larry Kilgallen wrote:m > In article <43ecef27$0$67263$157c6196@dreader2.cybercity.dk>, Karsten Nyblad <nospam@nospam.nospam> writes:  >  >>Larry Kilgallen wrote: >>n >>>In article <43eca195$0$67264$157c6196@dreader2.cybercity.dk>, Karsten Nyblad <nospam@nospam.nospam> writes: >>>  >>>  >>> J >>>>Besides both Ada and Pascal are languages where pointers are used for J >>>>implementing data structures when arrays are not appropriate.  So Ada 2 >>>>and Pascal do not differ from C on that topic. >>>  >>> J >>>No, Ada Access Values differ considerably from C Pointers, in that theyK >>>do not allow pointer arithmetic.  One does not add 1 to a pointer to get I >>>the next byte (or is it longword).  In that regard, C is not much more & >>>than a glorified assembly language. >>J >>Let a pointer in C point to an element in an array.  Then the pointer + J >>1 will point to the next element in the array.  (Assuming that the type K >>of the pointer matches the element type of the array.)  But I wonder how  % >>many people use that feature today.  >  > G > Whereas I was wondering how many will remember how it works when they  > write such code.  I Yes, I have seen programmers having a hard time with the construct *p++,  5 where they thought it meant (*p)++.  It means *(p++).    ------------------------------    Date: 11 Feb 2006 08:59:25 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) $ Subject: Re: null terminated strings3 Message-ID: <HLYVU$SXHYhO@eisner.encompasserve.org>   \ In article <43ED8BC8.6B225724@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes: > Dave Froble wrote:D >> I know of no other language that handles arrays in such a manner.E >> Having used a number of assembly languages, I understand what it's J >> doing.  What I don't understand is why such a thing was included in any+ >> language that's more than an assembler.   >  > F > Not all "arrays" are declared as arrays. You can declare a structureE > (record format) and then dynamically alloacte a buffer big eough to J > contain a dynamically calculated number of instances of that record. YouI > can then use pointers to access individual elements within that buffer.   C In Ada that dynamic allocation is automatic (after type definition) @ and type safe.  But it is declared as an array, because it is an array (of that record type).  E > C also allows you to pass this pointer to a called subroutine which I > expected an argument of that structure type and the subroutine can then  > access that array.  F In Ada, you pass the array as a whole (variable length) and the bounds& of the array are automatically passed.  J > Also, it allows you to pass not just the base element of the array , butE > also the address of the nth element of the array and the subroutine H > treats that nth element as element 0. (eg: skip processing the first 5I > records by passing the address of the 6th record instead of that of the  > first record).  F In Ada you pass a slice (as discussed before) specifying the first and- last elements of the sub-array to b3e passed.   J > The pointer artihmetic becomes useful in structures because the compilerE > automatically calculates the structure size for you for any pointer  > arithmetic you do.  G To someone who has seen Ada or Pascal, that is a very clumsy way to get F that effect.  I don't know PL/I that well, so it is Tom's problem that he is so shy :-)   ------------------------------  % Date: Sat, 11 Feb 2006 07:21:24 -0800 # From: "Tom Linden" <tom@kednos.com> $ Subject: Re: null terminated strings( Message-ID: <ops4td9ycgzgicya@hyrrokkin>  H On 11 Feb 2006 08:59:25 -0600, Larry Kilgallen <Kilgallen@SpamCop.net>   wrote:  9 > In article <43ED8BC8.6B225724@teksavvy.com>, JF Mezei   ( > <jfmezei.spamnot@teksavvy.com> writes: >> Dave Froble wrote: E >>> I know of no other language that handles arrays in such a manner. F >>> Having used a number of assembly languages, I understand what it'sK >>> doing.  What I don't understand is why such a thing was included in any + >>> language that's more than an assembler.  >> >>G >> Not all "arrays" are declared as arrays. You can declare a structure F >> (record format) and then dynamically alloacte a buffer big eough toK >> contain a dynamically calculated number of instances of that record. You J >> can then use pointers to access individual elements within that buffer. > E > In Ada that dynamic allocation is automatic (after type definition) B > and type safe.  But it is declared as an array, because it is an > array (of that record type). > F >> C also allows you to pass this pointer to a called subroutine whichJ >> expected an argument of that structure type and the subroutine can then >> access that array.  > H > In Ada, you pass the array as a whole (variable length) and the bounds( > of the array are automatically passed. > K >> Also, it allows you to pass not just the base element of the array , but F >> also the address of the nth element of the array and the subroutineI >> treats that nth element as element 0. (eg: skip processing the first 5 J >> records by passing the address of the 6th record instead of that of the >> first record).  > H > In Ada you pass a slice (as discussed before) specifying the first and/ > last elements of the sub-array to b3e passed.  > K >> The pointer artihmetic becomes useful in structures because the compiler F >> automatically calculates the structure size for you for any pointer >> arithmetic you do.  > I > To someone who has seen Ada or Pascal, that is a very clumsy way to get H > that effect.  I don't know PL/I that well, so it is Tom's problem that > he is so shy :-)  K It is not that I am shy, we have these endless discussions leading to ennui K on programming techniques in C, when the real problem is that C shouldn't    beI used for any serious programming project.  It is too primitive, makes you J reinvent too many wheels and is not well suited for writing reliable code. And the syntax is ugly.    ------------------------------  % Date: Sat, 11 Feb 2006 12:54:57 -0500 ' From: Dave Froble <davef@tsoft-inc.com> $ Subject: Re: null terminated strings/ Message-ID: <FtednfqeK61TuXPeRVn-tg@libcom.com>    JF Mezei wrote:  > Dave Froble wrote: > C >>I know of no other language that handles arrays in such a manner. D >>Having used a number of assembly languages, I understand what it'sI >>doing.  What I don't understand is why such a thing was included in any * >>language that's more than an assembler.  >  >  > F > Not all "arrays" are declared as arrays. You can declare a structureE > (record format) and then dynamically alloacte a buffer big eough to J > contain a dynamically calculated number of instances of that record. YouI > can then use pointers to access individual elements within that buffer.   	 Hmmm ....    You mean something like:   Define a structure:   
 Record ABC	 	Word	Idx  	Long	Count  	String	Desc=8
 End Record  ) Then declare some array of the structure:    Dim ABC Struc(99)   1 Then access some part of an element in the array:    	Print Struc(7)::Count  C While I've seen syntax I like a bit better, (Visual Basic uses '.'  I instead of '::'), I didn't need any special type of index.  Just a plain   old numeric data type.  A And everything you've mentioned doesn't need a special data type.   E > C also allows you to pass this pointer to a called subroutine which I > expected an argument of that structure type and the subroutine can then  > access that array. > J > Also, it allows you to pass not just the base element of the array , butE > also the address of the nth element of the array and the subroutine H > treats that nth element as element 0. (eg: skip processing the first 5I > records by passing the address of the 6th record instead of that of the  > first record). > J > The pointer artihmetic becomes useful in structures because the compilerE > automatically calculates the structure size for you for any pointer  > arithmetic you do. > F > While VMS has very limited support for multybvyte characters, othersB > OSes do and that allows the "char" to become a 2 byte entity andI > programs that scan a character array then automatically advance 2 bytes 1 > at a time when the pointer in incremented by 1.   I I need to quite doing this.  Just because I don't like many things about  , C isn't enough reason to continue this rant.   --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  % Date: Sat, 11 Feb 2006 13:03:35 -0500 ' From: Dave Froble <davef@tsoft-inc.com> $ Subject: Re: null terminated strings9 Message-ID: <jdmdnRY145lOu3PenZ2dnUVZ_t6dnZ2d@libcom.com>    Dave Weatherall wrote:F > On Sat, 11 Feb 2006 03:34:57 UTC, Dave Froble <davef@tsoft-inc.com>  > wrote: >  >  >>Larry Kilgallen wrote: >> >>>Karsten Nyblad wrote: >>>  >>>  >>> N >>>>Let a pointer in C point to an element in an array.  Then the pointer +   N >>>>1 will point to the next element in the array.  (Assuming that the type   N >>>>of the pointer matches the element type of the array.)  But I wonder how  ( >>>>many people use that feature today.  >>>  >>> _ >>>In article <43ED3339.DBAE181E@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes:  >>>  >>>  >>>>Larry Kilgallen wrote: >>>> >>>>J >>>>>Whereas I was wondering how many will remember how it works when they >>>>>write such code.  >>>> >>>>, >>>>Pointer arithmetic in C is well defined. >>>  >>> F >>>But if Karsten is correct that not many people use the feature, the? >>>fact that it is well defined will not help them remember it.  >>I >>Earlier in this thread, at least I think so, the claim was made that C  < >>was designed to have the capabilities of PDP-11 assembler. >>D >>I know of no other language that handles arrays in such a manner. E >>Having used a number of assembly languages, I understand what it's  J >>doing.  What I don't understand is why such a thing was included in any J >>language that's more than an assembler.  I've been thinking rather hard K >>on the technique, and I challenge anyone to provide a need for such that  + >>isn't more easily done in another manner.  >>I >>What it suggests to me is the capability of looking at some part of an  K >>array element, not necessarily including the first byte/bit in the array.  >  >  > DaveG >             consider the autoincrement/decrement addressing modes on  H > pdp and VAX. The value of the index register was incremented according. > to the size of the otem being operated on :- > . > 	movb	(r5)+, r0		; r5 bumped up one location  > 	movw	(r5)+, r0		; r5 bumped 2  > 	movl	(r5)+, r0		; r5 bumped 4 > D > This is the behaviour that the C pointer arithmetic is following, 0 > methinks. It allowed the compiler to translate >  > 	p++     or      p = p +1  > + > to the correct autoincrement instruction.  > C > Actually if you look at the assembler code generated by many VAX  ? > compilers you might find the same constructs. Certainly ISTR  H > VAX-Fortran generating them for x = y DO loops. Once you get to arraysF > of structures that are not 'natural' the compiler need to generate  0 > ADD instructions using sizeof(array element).  > G > In the 80s, I took on responsibility for some Macro-11 that was full  " > of statements along the lines of >  > 	cmp	(r0)+,(r0)+ >  > which was just to avoid  >  > 	add	#4,r0 > H > It saved a word ISTR. I wasn't sure the two memory fetches were worth H > it and put the ADD back.  (not to mention the possibility of an accessG > violation from (r0) when moved to VMS). To be fair the code had been  H > wriiten originally for a small memory pdp (16K?) inder DOS-11. By the H > time I picked it up it was on RSX (32K non-overlayed tasks) looking to > VMS. > ? > I may have mis-read your point. If so I apologise in advance.  >   I You read it correctly.  The point is, why does a language need a special  H pointer data type when the compiler knows the structure, even if it's a F user defined structure.  As Larry, Tom, and I have been saying, other F languages do not need this, they automatically handle the offsets.  A H normal numeric data type is sufficient for indicating an element in the 
 structure.  - C offers too many opportunities for mistakes.    --  4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  # Date: Sat, 11 Feb 2006 13:42:02 GMT ( From: Alan Greig <greigaln@netscape.net>J Subject: Re: OT: Gartner wakes up company executives to X86-64 scalability> Message-ID: <KQlHf.277508$vl2.74655@fe2.news.blueyonder.co.uk>   David J Dachtera wrote:    >  > J > Search the IBM website for "xSeries 460". Found some usable info. there.  H And I've read it all but I'm more interested in what Gartner are saying 
 in this case.    --  
 Alan Greig   ------------------------------  % Date: Sat, 11 Feb 2006 11:00:44 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>J Subject: Re: OT: Gartner wakes up company executives to X86-64 scalability+ Message-ID: <43EE183C.E712333D@comcast.net>    Alan Greig wrote:  >  > David J Dachtera wrote:  >  > >  > > L > > Search the IBM website for "xSeries 460". Found some usable info. there. > I > And I've read it all but I'm more interested in what Gartner are saying  > in this case.   D Rather disgusting to think about actually paying for such dreck, eh?   --   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/   ------------------------------   Date: 11 Feb 2006 17:42:20 GMT From: healyzh@aracnet.com   Subject: Perl, DBI, and DBD::RDB, Message-ID: <dsl7ls02j63@enews4.newsguy.com>  J Based on Alan Winston's book on OpenVMS Web Servers, there is a version ofJ Perl for OpenVMS that already has the DBI installed.  It doesn't appear toL be in Perl V5.6-1 or Perl 5.8-6 from the HP website.  Does anyone know where to find this version of Perl?   L Or does anyone know where to find a version of DBI that will build with PerlB 5.8-6 under OpenVMS 7.3-2?  My attemps to build DBI 1.50 have been questionable :^(    3 tests and 45 subtests skipped.H Failed 45/49 test scripts, 8.16% okay. 1387/1896 subtests failed, 26.85% okay.   I The end result is that I want to be able to use the DBD::RDB perl module.    	Zane    ------------------------------  # Date: Sat, 11 Feb 2006 13:25:16 GMT " From:   VAXman-  @SendSpamHere.ORG/ Subject: Re: Security experts say no secure OS! 0 Message-ID: <00A5126D.268444C0@SendSpamHere.ORG>  _ In article <PoudnQH8FqG6vHDenZ2dnUVZ_tidnZ2d@bresnan.com>, GreyCloud <mist@cumulus.com> writes:  >  >  >bob@instantwhip.com wrote:  > I >> these guys obviously don't know about OpenVMS, or in IBMs case, do not 
 >> want toH >> acknowledge it ... put o boot on VMS, sure, but where decides what it >> can do ... what
 >> idiots ...  >>  x >> http://www.informationweek.com/news/showArticle.jhtml;jsessionid=L1WPBLPNLDMS2QSNDBCSKHSCJUMEKJVN?articleID=179103249 >>   > H >Funny that the journalists always say "... experts ..." but never give % >out names.  Very weak on their part.   M ex -- adj: out of fashion, demode or outmoded, antique, old-fashioned, passe. ? spurt -- n: the occurrence of a sudden discharge (as of liquid)   * Thus, ex-spurts are has-been wankers.            --  K 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: 11 Feb 2006 01:07:00 -0800 From: shofu_au@yahoo.com.au  Subject: Source Listing C Message-ID: <1139648820.194537.107400@f14g2000cwb.googlegroups.com>   	 Hi Group,   G Just about to upgrade from Open VMS 6.2 Alpha to Open VMS 8.2 on Alpha.   (DS25s from 8400s)   = I have the source code listings for 6.2.0 circa 1995/6 on CD.   E Will I need to buy the entire kit or is the update set the way to go?   8 If I just buy the upgrade can I use my existing licence?   How many CDs is 8.2 Alpha on?   B Any idea of the price for the upgrade kits?  I am sitting down....   Thanks   Mark   ------------------------------  # Date: Sat, 11 Feb 2006 13:30:56 GMT " From:   VAXman-  @SendSpamHere.ORG" Subject: Re: V7.3-2 and Joliet CDs0 Message-ID: <00A5126D.F13A75C5@SendSpamHere.ORG>  ` In article <43ED6394.C951D3E3@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes: >  > " >VAXman-, @SendSpamHere.ORG wrote: >>  f >> In article <C9fo01pUHKV7@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes: >> > >> >c >> >In article <43EBFB69.CC050B8@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes:  >> >N >> >> Discovered something interesting the other night. I did MOUNT/OVER=ID ofK >> >> a Joliet format CD I'd burned (Easy CD Creator V3.5) that evening and D >> >> V7.3-2 MOUNTed it as ISO-9660. Only the "short" filenames were7 >> >> "visible", but the files were at least available.  >> >K >> >ISO 9660 requires that files be accessible by standard names as well as J >> >by other naming methods.  Of course Joliet is not exactly a compatibleJ >> >extension as authorized by ISO 9660, but it does allow compliance with >> >that one point.  >>  G >> I've tried mounting some of these but if the short name has a ~, VMS  >> can't find it on the CD.  > : >Hhmmm... just tried extended parse style, too. No schmae. >  >Drat...  H Yup.  There doesn't seem to be a VMS way around such CDs.  I've also hadH problems with them on OS X -- not the same type of problems but problemsH that still have me pining for the day when Billzebub and his vile hoard H at the Redmond Academy of Teenaged Software Neophytes Emitting Substand-- ard Technologies will begin to use STANDARDS!    --  K 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: 11 Feb 2006 09:10:04 -0600- From: Kilgallen@SpamCop.net (Larry Kilgallen) " Subject: Re: V7.3-2 and Joliet CDs3 Message-ID: <Kx44HeZERofB@eisner.encompasserve.org>   U In article <00A5126D.F13A75C5@SendSpamHere.ORG>,   VAXman-  @SendSpamHere.ORG writes: b > In article <43ED6394.C951D3E3@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes: >> >># >>VAXman-, @SendSpamHere.ORG wrote:   H >>> I've tried mounting some of these but if the short name has a ~, VMS >>> can't find it on the CD. >>; >>Hhmmm... just tried extended parse style, too. No schmae.  >>	 >>Drat...  > J > Yup.  There doesn't seem to be a VMS way around such CDs.  I've also hadJ > problems with them on OS X -- not the same type of problems but problemsJ > that still have me pining for the day when Billzebub and his vile hoard J > at the Redmond Academy of Teenaged Software Neophytes Emitting Substand-/ > ard Technologies will begin to use STANDARDS!   M I am convinced it is possible to read such CDROMs if the Joliet specification 6 is the only deviation from the ISO-9660 specification.  N But by using a tilde character in the filename of the primary (8.3) hierarchy,H the MAKER OF THE FORMATTING TOOL has violated ISO-9660.  Unless the toolI used to make the CDROM came from Microsoft, the only part of this problem M one can blame on Microsoft is setting a bad example for standards compliance.    ------------------------------  # Date: Sat, 11 Feb 2006 17:04:09 GMT " From:   VAXman-  @SendSpamHere.ORG" Subject: Re: V7.3-2 and Joliet CDs0 Message-ID: <00A5128B.BA87F50B@SendSpamHere.ORG>  c In article <Kx44HeZERofB@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes:  >  > V >In article <00A5126D.F13A75C5@SendSpamHere.ORG>,   VAXman-  @SendSpamHere.ORG writes:c >> In article <43ED6394.C951D3E3@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes:  >>>  >>> $ >>>VAXman-, @SendSpamHere.ORG wrote: > I >>>> I've tried mounting some of these but if the short name has a ~, VMS  >>>> can't find it on the CD.  >>> < >>>Hhmmm... just tried extended parse style, too. No schmae. >>> 
 >>>Drat... >>  K >> Yup.  There doesn't seem to be a VMS way around such CDs.  I've also had K >> problems with them on OS X -- not the same type of problems but problems K >> that still have me pining for the day when Billzebub and his vile hoard  K >> at the Redmond Academy of Teenaged Software Neophytes Emitting Substand- 0 >> ard Technologies will begin to use STANDARDS! > N >I am convinced it is possible to read such CDROMs if the Joliet specification7 >is the only deviation from the ISO-9660 specification.  > O >But by using a tilde character in the filename of the primary (8.3) hierarchy, I >the MAKER OF THE FORMATTING TOOL has violated ISO-9660.  Unless the tool J >used to make the CDROM came from Microsoft, the only part of this problemN >one can blame on Microsoft is setting a bad example for standards compliance.  M "Microsoft is setting a bad example for standards compliance."  News at 11...   O Don't bother watching, we've seen this headline every day for the entire exist- . ence of that shady Washington state operation.     --  K 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: Sat, 11 Feb 2006 17:47:11 GMT % From: Roger Ivie <rivie@ridgenet.net> " Subject: Re: V7.3-2 and Joliet CDs3 Message-ID: <slrndus8ov.cl0.rivie@stench.no.domain>   = On 2006-02-11, Larry Kilgallen <Kilgallen@SpamCop.net> wrote: P > But by using a tilde character in the filename of the primary (8.3) hierarchy,J > the MAKER OF THE FORMATTING TOOL has violated ISO-9660.  Unless the toolK > used to make the CDROM came from Microsoft, the only part of this problem O > one can blame on Microsoft is setting a bad example for standards compliance.   D On Windows, I would expect the formatting tool to use the short name? generated by Windows when the file was created as the 8.3 name. > Microsoft, in its infinite wisdom, uses tildes in those names.  > If I'm working in XP on a FAT32 volume and create a file namedH feldergarb.txt, XP creates an 8.3 file named felder~1.txt decorated withH FAT32 name extensions that specify the long name as feldergarb.txt; thatF is, the Windows operating system has created two names for the file. ID would expect a Windows CD formatting program to use the Windows file names for that file.  @ The naming scheme is the same for other filesystems supported byI Windows, but I don't know whether on (say) an NTFS file system the short  K name is decorated with name extensions or the long name is decorated with a  short name property.  C By the way, WordStar (REAL WordStar; 3.3, 4.0; I'm not certain what F happened in the 5.x timeframe) also doesn't like tildes in file names. Very annoying. --  
 roger ivie rivie@ridgenet.net   ------------------------------   End of INFO-VAX 2006.084 ************************