1 INFO-VAX	Thu, 21 Apr 2005	Volume 2005 : Issue 222       Contents: Re: 'touch' command under VMS? Apache, mysql, phpMyAdmin  Re: Apache, mysql, phpMyAdmin & Re: Apache, mysql, phpMyAdmin - solved& Re: Apache, mysql, phpMyAdmin - solved( Re: Could a PC do this - Don't think so.( Re: Could a PC do this - Don't think so.( Re: Could a PC do this - Don't think so. Re: Could a PC do this?  Re: Could a PC do this?  Re: Could a PC do this?  Re: Could a PC do this?  Re: Could a PC do this?  Re: Could a PC do this?  Re: Could a PC do this?  Re: Could a PC do this?  Re: Could a PC do this?  Re: Could a PC do this?  DNBE1, state: CALLING ???  Re: HP SSL for OpenVMS Legato NetWorker Re: Legato NetWorker Re: Legato NetWorker Re: Legato NetWorker Re: Meine geilen Bilder A Re: Problem running x-windows client (exceed) over VPN connection A Re: Problem running x-windows client (exceed) over VPN connection A Re: Problem running x-windows client (exceed) over VPN connection A Re: Problem running x-windows client (exceed) over VPN connection A Re: Problem running x-windows client (exceed) over VPN connection A Re: Problem running x-windows client (exceed) over VPN connection A Re: Problem running x-windows client (exceed) over VPN connection , Re: Relative file size changes with copying?, Re: Relative file size changes with copying?, Re: Relative file size changes with copying?, Re: Relative file size changes with copying?, Re: Relative file size changes with copying?, Re: Relative file size changes with copying?, Re: Relative file size changes with copying?* RTOS options, was: Re: Could a PC do this?/ [OpenVMS VAX V7.3] Updating with VAXRPC02_073 ?   F ----------------------------------------------------------------------  % Date: Thu, 21 Apr 2005 08:38:37 -0500 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>' Subject: Re: 'touch' command under VMS? D Message-ID: <craigberry-9545FB.08383621042005@news.isp.giganews.com>  ; In article <6d3fd$42665882$82a1f19e$5559@news1.tudelft.nl>, )  JOUKJ <joukj@hrem.stm.tudelft.nl> wrote:    > Neil Cherry wrote:E > > Is there an equivalent command to Unix's touch command under VMS? I > > I need this for tftp where we need to touch the file before we put it  > > there. Thanks   A And yet one more to throw in the hopper.  Assuming you have perl   defined as a foreign command:   5 $ touch == "''perl' -e ""utime(undef,undef,@ARGV);"""    ------------------------------  % Date: Thu, 21 Apr 2005 13:49:25 +0100 0 From: Chris Sharman <chris.sharman@sorry.nospam>" Subject: Apache, mysql, phpMyAdmin4 Message-ID: <d487gl$d7m$1$8302bc10@news.demon.co.uk>  5 Installed all the latest versions on Alpha VMS 7.3-1. H All appears to be working, but I can't get phpMyAdmin to login properly.  I I've got mysql installed on a test machine with passwords unset (ie open  I access), and phpMyAdmin (& Issue-Tracker) work there (phpMyAdmin puts up  ? a standard user/password browser dialog, and accepts anything).   G On a different machine, I've set the mysql root password (as described  8 in http://www.pi-net.dyndns.org/anonymous/kits/#mysql ).  F The local command line mysql client works fine (with password), but I E can't access the mysql server from phpMyAdmin, either from the local  / webserver, or from the original test webserver.   J I'm not really clear whether it's an Apache, mysql, or phpMyAdmin problem.  I I've read Alan Winston's book, but I'm still unclear on how it all hangs  . together - there isn't much on authentication.  H I don't really need anything complicated at this stage: one functioning E username/password pair will suffice, so long as it's not open access. F I don't really want to use vms authentication, because I don't really E want all of the vms logins on the box to have access to the database.   3 Can anyone point me in the right direction please ?    Thanks,  Chris    ------------------------------  # Date: Thu, 21 Apr 2005 13:39:20 GMT . From: "Jerry Alan Braga" <jabraga@flanagan.ca>& Subject: Re: Apache, mysql, phpMyAdmin+ Message-ID: <c2O9e.40366$yV3.7549@clgrps12>   A Depending on the verions of mySQL vs PHP you may need to use the  / OLD_PASSWORD('') function to connect correctly.   5 I had the same problem with PHP 4.3.? and mySQL 4.1.x   I It seems that since mySQL 4.1 they have changed the way they do password  M encryption vs the way PHP encrypts the password to match.  PHP 5.x does this   correctly again.  9 1) edit the my.cnf and add this into the [mysqld] section   G #Use old password encryption method (needed for 4.0 and older clients).  old_password=1   2) restart mysql server   / 3) go into mySQL and update the user account to    mySQL> use mysql; H mySQL> update user set password = old_password('PASS') where user = '?'; mySQL> commit; mySQL> flush privileges; mySQL> exit;   4)  try phpMyAdmin again  > "Chris Sharman" <chris.sharman@sorry.nospam> wrote in message . news:d487gl$d7m$1$8302bc10@news.demon.co.uk...7 > Installed all the latest versions on Alpha VMS 7.3-1. J > All appears to be working, but I can't get phpMyAdmin to login properly. > K > I've got mysql installed on a test machine with passwords unset (ie open  M > access), and phpMyAdmin (& Issue-Tracker) work there (phpMyAdmin puts up a  ? > standard user/password browser dialog, and accepts anything).  > L > On a different machine, I've set the mysql root password (as described in 7 > http://www.pi-net.dyndns.org/anonymous/kits/#mysql ).  > H > The local command line mysql client works fine (with password), but I G > can't access the mysql server from phpMyAdmin, either from the local  1 > webserver, or from the original test webserver.  > L > I'm not really clear whether it's an Apache, mysql, or phpMyAdmin problem. > K > I've read Alan Winston's book, but I'm still unclear on how it all hangs  0 > together - there isn't much on authentication. > J > I don't really need anything complicated at this stage: one functioning G > username/password pair will suffice, so long as it's not open access. M > I don't really want to use vms authentication, because I don't really want  B > all of the vms logins on the box to have access to the database. > 5 > Can anyone point me in the right direction please ?  > 	 > Thanks,  > Chris    ------------------------------  % Date: Thu, 21 Apr 2005 14:04:31 +0100 0 From: Chris Sharman <chris.sharman@sorry.nospam>/ Subject: Re: Apache, mysql, phpMyAdmin - solved 4 Message-ID: <d488cv$ea0$1$8302bc10@news.demon.co.uk>   Found it in a mysql forum.I Apparently I'm using a php version linked against an older mysql client,  F so I have to use the OLD_PASSWORD function for calculating passwords,  not the PASSWORD function.0 (Alternatives would be rebuilding php, etc etc).   Thanks Chris    ------------------------------  # Date: Thu, 21 Apr 2005 13:56:31 GMT ! From: Nigel Barker <nigel@hp.com> / Subject: Re: Apache, mysql, phpMyAdmin - solved 8 Message-ID: <k4cf61leghgaabvqv97j6leg0dk39p4m5j@4ax.com>  N On Thu, 21 Apr 2005 14:04:31 +0100, Chris Sharman <chris.sharman@sorry.nospam> wrote:   >Found it in a mysql forum. J >Apparently I'm using a php version linked against an older mysql client, G >so I have to use the OLD_PASSWORD function for calculating passwords,   >not the PASSWORD function. 1 >(Alternatives would be rebuilding php, etc etc).  >  >Thanks  >Chris  P I encountered exactly the same thing a couple of weeks ago. Had me scratching myN head for a bit. Glad that you found the answer by yourself though. It's all in the documentation of course:-)   -- Nigel Barker Live from the sunny Cote d'Azur    ------------------------------  % Date: Wed, 20 Apr 2005 22:48:24 -0700 ( From: Jeff Cameron <roktsci@comcast.net>1 Subject: Re: Could a PC do this - Don't think so. / Message-ID: <BE8C8CB8.C4B9%roktsci@comcast.net>   G I've written a great deal of specialized real-time telemetry processing J systems using both PC's and OpenVMS for AXP. The requirements dictated theI type of platform I needed to use. On the VMS platform, my chosen language L was DEC FORTRAN, while on the PC, C and C++ were much more effective for me.  C On the PC, it was certainly much more difficult to implement double L buffering that it was on VMS, but not impossible by any means. Certainly theL philosophy of the double buffering I/O paradigm is considerably different on a PC than in OpenVMS.   H Using all the facilities in VMS like event flags, ASTs, Distributed lockJ manager, and VMS multi-threading allowed me to build complex multi-processJ solutions with minimal effort and most important little debugging and high reliability.  E On the PC side, I quite often ran into major problems like the kernel * hanging or crashing and the infamous BSOD.  L One of my telemetry processing systems implemented on VMS processed multipleH asynchronous parallel data streams which were multiplexed with data fromG many sources, including flight control, aeronautic sensors, GPS, Radar, J multi-spectral imaging and various non-imaging sensors. The data had to beL de-multiplexed, calibrated into engineering units and presented to a host ofF multiple signal processing algorithms running as separate asynchronousL processes. Then the results had to be re-aligned temporally and fused into aJ presentable information for the eager audience. I did this on a single 190D Mhz CPU Alpha server 2100. The system processed and recorded all theI important aspects of over 12 hours of telemetry from a Predator UAV. This A even included the performance of Automatic Target Recognition and I multi-sensor fusion algorithms. Most of the processing was done under the E context of ASTs. While it was running, I could use the MONITOR SYSTEM C command to see that the whole system was taking up less than 50% of  available CPU.  G No way could I do this on a PC. In fact I did build a similar, but much J smaller scale telemetry processing system that did much less, and I had toK use 5 separate PCs each with a CPU processor with greater clock speeds than  my single 190 Mhz Alpha 2100.   H In conclusion. If you have written a VMS solution in FORTRAN using EventB flags and AST's, don't count on easily migrating the code to a PC.  
 Best of luck.  Jeff   On 4/20/05 8:43 AM, in article6 lMu9e.18006$5F3.2741@news-server.bigpond.net.au, "Jon" <jwatmuff@bigpond.com> wrote:   K > This message is posted in the hope there are some who can offer advice on M > the basis of broader knowledge of BOTH "DEC" hardware/ VMS Operating System C > and Personal Computers. I appreciate there is a lot of background K > information here, but the answer to the question raised at the end is not 
 > trivial. > I > Over the deacde of the 1990's I developed a set of about 200 high level M > routines written in VMS FORTRAN for automated control of a small-scale wind E > tunnel experiment using a microVAX II. At the lowest level, devices I > performing high speed DAQ, were controlled using QIO's, AST's and other K > System Services submitted by separate independent processes, e.g. up to 1 E > MHz 16-bit double-buffered DAQ using a DRQ3B parallel interface. By L > dynamically raising (e.g. above OS with level 17) and lowering the processG > priority, the system could effectively operate for short periods as a A > real-time system, yet alternate buffers could also be processed F > simultaneously in real-time. Even though the CPU has the highest DMAN > priority, control of the bus was relinquished for enough clock cycles duringL > the computations (presumably for internal CPU operations) so that the dataM > flow was not impeded. I envisaged this to be like firing bullets (the data) M > through a propeller (computations). Previously obtained files could even be K > independently processed post-experimentally on the same system etc. while  > all this was going on. > M > The higher level processes controlled the sequence of events using clusters L > of event flags to communicate with lower level processes and the arrays ofJ > data and control variables were shared using "global" common blocks. TheI > higher level routines were layered into higher and higher levels in the J > sense that the very highest level routines could perform substantial andJ > complete subsets of fairly complex tasks, such as frequent and automaticM > calibration of probes etc. In essence large data sets on spatially dense 3D K > grids could be obtained on a point-by-point basis by running continuously I > for days on end. The longest experiment ran without a hitch for 30 days K > continuously without scheduled manual intervention and chalked up 24 days I > CPU time. The configuration ended up being extremely powerful since new D > experiments could be designed and implemented with minimal effort. > G > The application was eventually and relatively easily and successfully K > transferred to an Alpha PWS 500au running OpenVMS using the same external L > instrumentation, but using PCI equivalents of the Q-Bus hardware, e.g. theK > DCI-1100 replacement for the DRQ3B from The Logical Co. This PCI card and G > the OpenVMS driver is still available today ... but at a cost of $2K.  > M > My question concerns whether this type of functionality could be reproduced K > on a PC, since this would allow thousands of lines of existing high level  > FORTRAN code to be reused. > K > Asking this question to the PC community is pointless in many ways, since M > they tend to think of themselves sophisticated, with point-and-click visual M > programming and consider VMS to be old-fashioned, etc. We all know that VMS K > command procedures can be used to process thousands of files - a daunting N > and in fact impossibly tedious to perform manually using the point-and-clickH > approach. In summary, we are talking about a vast difference in scale. > N > I have experimented enough to know that shared memory works with PC FORTRAN,N > although in a somewhat limited form. But what about the equivalents of EventL > Flags, QIO's and ASTs. I am aware of some similarities between VMS and NT,K > e.g. VMS AST equivalent with NT APC. However, writing device drivers from M > scratch defeats the objective. Alternatively, it might be possible to avoid N > these issues all together by using PC DAQ cards with their own internal dataK > buffers. However, this approach would require linking with the C code DAQ F > drivers for control of the data sequences. etc. under a FORTRAN mainK > program. Most of the information I've found describes how to call FORTRAN L > DLL's from the DAQ C code. There appears to be nothing about linking DAQ CF > code device drivers into a main application written in FORTRAN.  AnyN > comments, suggestions and/or prior experiences would be greatly appreciated. >  >    ------------------------------  # Date: Thu, 21 Apr 2005 14:38:58 GMT " From: "Jon" <jwatmuff@bigpond.com>1 Subject: Re: Could a PC do this - Don't think so. = Message-ID: <6WO9e.19097$5F3.2305@news-server.bigpond.net.au>    JeffH Thanks for providing your experiences. They just confirm how incredibly K powerful the facilities provided in VMS are for real-time data processing,  I and as you say, with minimal effort. The power of VMS for real-time data  L acquisition is not widely appreciated, even among VMS aficianados. The fact J that a single 190 MHz Alpha can perform a more demanding data acquisition K task than 5 PC's and still be using only 50% CPU must surely be considered  K amazing. As far as reliability is concerned, I tried but could never break  K the system. For example, the critical component in the instrumentation was  I an old Tustin A/D built in the 1980's which only had an 8K internal FIFO  K buffer, together with an overflow bit that was checked after each transfer  I of up to 2e5 samples. The Tustin was clocked externally, and not once in  M probably billions of buffer transfers over a 10-year period did the internal  L Tustin FIFO buffer overflow, despite the CPU priority blocking transfers to M the memory bus during real-time processing. It was this that made me realize  M that a few bytes of data could sneak through from the Tustin during the free  K clock cycles when control of the memory bus was relinquished after the CPU  H completed transfer to internal registers and was performing an internal F arithmetic operation - the "bullets passing between the blades of the J propeller". Remember that these 1MHz 16-bit data rates were on a (?3MHz?) J Q-Bus. As a matter of interest, the process would become CPU bound on the J MicroVAX II at these rates. However, the Alpha 500au PWS based system was I just awesome. I still have an Alpha 500au PWS Open VMS system in my home   office in among the new PCs.  H Thanks also to the others and those who sent email. The difficulties of E re-establishing the system go beyond the relatively high cost of the  J DCI-1100 interface since a Tustin A/D would also be required. We bought a G few new ones at NASA in the early 1990's at $16K each - beautiful hand  I crafted instruments - but now hopelessly obsolete. What a shame that all  H this tremendous robust functionality - designed by engineers for use by J engineers - has gone by the way-side in favor of point-and-click LabView. K However the PC hardware is now so cheap that I will probably buy something  A to play with. But I'll keep the old Alpha 500au for a while yet.     ------------------------------  # Date: Thu, 21 Apr 2005 14:56:22 GMT " From: "Jon" <jwatmuff@bigpond.com>1 Subject: Re: Could a PC do this - Don't think so. > Message-ID: <qaP9e.19104$5F3.18259@news-server.bigpond.net.au>   Just found the following9 http://www.aplabs.com/Integrated_Systems/Tustin_Elec.html  The saga may yet live on.  Jon    ------------------------------  + Date: Thu, 21 Apr 2005 10:52:45 +0000 (UTC)  From: david20@alpha2.mdx.ac.uk  Subject: Re: Could a PC do this?) Message-ID: <d480lt$pft$1@news.mdx.ac.uk>   d In article <4266D565.18877.2617BDF3@localhost>, "Stanley F. Quayle" <squayle@insight.rr.com> writes:H >> The application was eventually and relatively easily and successfullyC >> transferred to an Alpha PWS 500au running OpenVMS using the same C >> external instrumentation, but using PCI equivalents of the Q-Bus I >> hardware, e.g. the DCI-1100 replacement for the DRQ3B from The Logical H >> Co. This PCI card and the OpenVMS driver is still available today ... >> but at a cost of $2K. >>  C >> My question concerns whether this type of functionality could be C >> reproduced on a PC, since this would allow thousands of lines of 1 >> existing high level FORTRAN code to be reused.  > ; >[Shameless Plug (tm) Alert -- I am a CHARON-VAX reseller.]  > E >CHARON-VAX/XM for Windows could control the DCI-1100.  In fact, The  G >Logical Company is the preferred partner for CHARON-VAX I/O devices.   + >Some development work would be required.    > ? >Your highly-VMS-specific code would work unchanged in the VMS  B >environment provided by CHARON-VAX.  It would emulate a VAX, but D >you've already indicated that it's not a problem.  And the virtual , >VAX would be far faster than a MicroVAX II. > ' >Details on CHARON-VAX is available at:  > ' >  http://www.stanq.com/charon-vax.html  >   J Wouldn't the fact that  CHARON-VAX/XM for Windows is running under windows7 compromise it's ability to run real-time applications ?     
 David Webb Security team leader CCSS Middlesex University           >--Stan Quayle >Quayle Consulting Inc.  >  >---------- . >Stanley F. Quayle, P.E. N8SQ  +1 614-868-13634 >8572 North Spring Ct., Pickerington, OH  43147  USA1 >stan-at-stanq-dot-com       http://www.stanq.com  >  >  >    ------------------------------  % Date: Thu, 21 Apr 2005 08:00:49 -0400 2 From: "Stanley F. Quayle" <squayle@insight.rr.com>  Subject: Re: Could a PC do this?/ Message-ID: <42675DB1.17499.282C1DA3@localhost>   8 On 21 Apr 2005 at 10:52, david20@alpha2.mdx.ac.uk wrote:D > Wouldn't the fact that  CHARON-VAX/XM for Windows is running underA > windows compromise it's ability to run real-time applications ?   : Nope.  In fact, there are lots of military and industrial - applications running on CHARON-VAX right now.   D For a supported configuration, CHARON-VAX requires a dual-processor > PC.  One processor emulates, the VAX.  The other does network B functions and any Windows "housekeeping".  While you can get away F with only one processor for number crunching, a real time application 3 like this would definitely require both processors.   
 --Stan Quayle  Quayle Consulting Inc.  
 ----------- Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363 3 8572 North Spring Ct., Pickerington, OH  43147  USA 0 stan-at-stanq-dot-com       http://www.stanq.com   ------------------------------    Date: 21 Apr 2005 07:28:05 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)   Subject: Re: Could a PC do this?3 Message-ID: <orsA9K+otgUq@eisner.encompasserve.org>   b In article <lMu9e.18006$5F3.2741@news-server.bigpond.net.au>, "Jon" <jwatmuff@bigpond.com> writes: > H > The application was eventually and relatively easily and successfully L > transferred to an Alpha PWS 500au running OpenVMS using the same external M > instrumentation, but using PCI equivalents of the Q-Bus hardware, e.g. the  L > DCI-1100 replacement for the DRQ3B from The Logical Co. This PCI card and G > the OpenVMS driver is still available today ... but at a cost of $2K.  >   E    OBTW, I'd consider $2K cheap compared to the cost of other options H    you might consider.  And do consider that the slowest, cheapest Alpha-    you can get off eBay will handle the load.   G    And, yes, there are real-time OS available for PC hardware.  Look at *    the total cost, including porting cost.  =    IMHO it'll be hard to beat a good used Apha on total cost.    ------------------------------    Date: 21 Apr 2005 07:32:27 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)   Subject: Re: Could a PC do this?3 Message-ID: <$jhlOvlNPjox@eisner.encompasserve.org>   J In article <d480lt$pft$1@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes: > L > Wouldn't the fact that  CHARON-VAX/XM for Windows is running under windows9 > compromise it's ability to run real-time applications ?   C    There have been claims made that you can pretty much get Windows B    out of the way.  My issue wih this is that all of our real timeF    applications access disks.  I don't think Charon-VAX writes to it'sB    simulated disks without going through the Windows file system. G    Although Dave Cutler used his usual design for the low levels of the I    WNT I/O system, I don't trust that the entire Windows file system will $    be deterministic and low latency.   ------------------------------  # Date: Thu, 21 Apr 2005 12:37:39 GMT * From: "FredK" <fred.nospam@nospam.dec.com>  Subject: Re: Could a PC do this?2 Message-ID: <n8N9e.4291$x03.4280@news.cpqcorp.net>  H "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message- news:$jhlOvlNPjox@eisner.encompasserve.org... L > In article <d480lt$pft$1@news.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes: > > F > > Wouldn't the fact that  CHARON-VAX/XM for Windows is running under windows ; > > compromise it's ability to run real-time applications ?  > E >    There have been claims made that you can pretty much get Windows D >    out of the way.  My issue wih this is that all of our real timeH >    applications access disks.  I don't think Charon-VAX writes to it'sC >    simulated disks without going through the Windows file system. I >    Although Dave Cutler used his usual design for the low levels of the K >    WNT I/O system, I don't trust that the entire Windows file system will & >    be deterministic and low latency. >   L The NT file system is heavily cached, and uses a lazy writer... which is whyG when you shut down windows you see that message about saving file data.    ------------------------------  % Date: Thu, 21 Apr 2005 15:03:12 +0200 ! From: Soterro <soterroatyahoocom>   Subject: Re: Could a PC do this?9 Message-ID: <4267a3c4$0$314$4d4ef98e@read.news.ch.uu.net>    Dave Froble wrote:  > If you read the original post: > G > The application was eventually and relatively easily and successfully K > transferred to an Alpha PWS 500au running OpenVMS using the same external L > instrumentation, but using PCI equivalents of the Q-Bus hardware, e.g. theK > DCI-1100 replacement for the DRQ3B from The Logical Co. This PCI card and G > the OpenVMS driver is still available today ... but at a cost of $2K.  > I > So, it appears that "anything else that supports VMS" isn't the issue.  " > The issue is the interface card. > F > What might really be cost prohibitive is re-writing the application.  F I think this is the key statement. The application runs also on Alpha F and depends only on the tools available to be ported to Itanium - can = AEST be of quick help here? I know almost nothing about it...   H A piece of hardware for 2K might seem costly _for a piece of hardware_, H but it's only a piece in the big schema. Charon VAX is not free either, E a new dual processor PC plus a W2003 OS license are also not exactly  E nothing. Or the new development environment and new _trustworthy_ PC  ' (not quite grandma's browsing station).   C Then think about porting the application: paid time for the people  I involved, learning curves for everybody, researching possible technology  I replacements, redesign according to the new environment, implement, test  G and retest, come back to unsatisfactory solutions chosen for replacing  H this or that technology, will all that be free? I'm sure porting can be F exciting for the programmers, but I can't see any other justification G for going the porting way. It looks like saving a penny today (the 2k)  - versus spending ton'o'money the next year(s).   I Unless the application is 20 lines of code, I'd stay on the existing VMS  I line. Please notice, I didn't even start thinking whether the new Win OS  E can do that or not (just as a theoretical point, I'd vote against it   anyway).   S    ------------------------------  % Date: Thu, 21 Apr 2005 09:17:33 -0400 2 From: "Stanley F. Quayle" <squayle@insight.rr.com>  Subject: Re: Could a PC do this?. Message-ID: <42676FAD.8106.28726048@localhost>  @ "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote inI > There have been claims made that you can pretty much get Windows out of G > the way.  My issue wih this is that all of our real time applications H > access disks.  I don't think Charon-VAX writes to it's simulated disks1 > without going through the Windows file system.    > You can access physical SCSI disks by addressing the SCSI bus 	 directly.     % On 21 Apr 2005 at 12:37, FredK wrote: G > The NT file system is heavily cached, and uses a lazy writer... which E > is why when you shut down windows you see that message about saving  > file data.  E I've seen entire disks cached on systems with enough physical memory.   
 --Stan Quayle  Quayle Consulting Inc.  
 ----------- Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363 3 8572 North Spring Ct., Pickerington, OH  43147  USA 0 stan-at-stanq-dot-com       http://www.stanq.com   ------------------------------  % Date: Thu, 21 Apr 2005 06:58:36 -0700 # From: "Tom Linden" <tom@kednos.com>   Subject: Re: Could a PC do this?( Message-ID: <opspk43yq7zgicya@hyrrokkin>  , On 21 Apr 2005 07:28:05 -0500, Bob Koehler  0 <koehler@eisner.nospam.encompasserve.org> wrote:  G > In article <lMu9e.18006$5F3.2741@news-server.bigpond.net.au>, "Jon"     > <jwatmuff@bigpond.com> writes: >>H >> The application was eventually and relatively easily and successfullyE >> transferred to an Alpha PWS 500au running OpenVMS using the same    >> external K >> instrumentation, but using PCI equivalents of the Q-Bus hardware, e.g.    >> theJ >> DCI-1100 replacement for the DRQ3B from The Logical Co. This PCI card   >> andH >> the OpenVMS driver is still available today ... but at a cost of $2K. >> > G >    OBTW, I'd consider $2K cheap compared to the cost of other options J >    you might consider.  And do consider that the slowest, cheapest Alpha/ >    you can get off eBay will handle the load.  > I >    And, yes, there are real-time OS available for PC hardware.  Look at , >    the total cost, including porting cost. > ? >    IMHO it'll be hard to beat a good used Apha on total cost.  > H It wasn't clear from the OP what was wrong with the VAX.  That is also a  possibility, pick up some VAXen.   ------------------------------    Date: 21 Apr 2005 12:32:25 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)   Subject: Re: Could a PC do this?3 Message-ID: <N4BMN8V+7WGm@eisner.encompasserve.org>   _ In article <n8N9e.4291$x03.4280@news.cpqcorp.net>, "FredK" <fred.nospam@nospam.dec.com> writes:  > N > The NT file system is heavily cached, and uses a lazy writer... which is whyI > when you shut down windows you see that message about saving file data.   G    I'd be worried about it delaying some of my interrupts when it deals     with updating the cache.    ------------------------------    Date: 21 Apr 2005 12:33:33 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)   Subject: Re: Could a PC do this?3 Message-ID: <k4CcvRdURFU8@eisner.encompasserve.org>   c In article <42676FAD.8106.28726048@localhost>, "Stanley F. Quayle" <squayle@insight.rr.com> writes: B > "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote inJ >> There have been claims made that you can pretty much get Windows out ofH >> the way.  My issue wih this is that all of our real time applicationsI >> access disks.  I don't think Charon-VAX writes to it's simulated disks 2 >> without going through the Windows file system.  > @ > You can access physical SCSI disks by addressing the SCSI bus  > directly.   C    With Files-11 on it?  How do I keep Windows from overwriting the     file system header?   ------------------------------    Date: 21 Apr 2005 08:00:41 -0700  From: bone_sam@hotmail.com (sam)" Subject: DNBE1, state: CALLING ???< Message-ID: <db8b79bb.0504210700.a3ce32a@posting.google.com>  F I have just replaced a DNBE-1 card with a new one and can now see this$ state. The firware is correct on it.D Does anybody know what the calling state means? and how to get it to the state below it?   , - Device PT37A (type: DNBE1, state: CALLING)A     - Device PT37B (type: DNBE1, boot: V4, firmware: V2.4, state:  RUNNING)   Cheers   ------------------------------  # Date: Thu, 21 Apr 2005 13:28:24 GMT ! From: Nigel Barker <nigel@hp.com>  Subject: Re: HP SSL for OpenVMS 8 Message-ID: <leaf61pe3dlv8d1neno5maovge87gjdqrt@4ax.com>  M On 20 Apr 2005 15:16:55 -0700, "vichoty@gmail.com" <vichoty@gmail.com> wrote:    >Hi, > H >We have a module that has coded against the RSA SSL-C APIs, which shareH >the exact same interface as openssl except for the certificate handling >capabilities. > D >We are considering to use HP SSL for the OpenVMS platform and wouldH >like to know if anyone knows if HP SSL provides just the same interface >as openssl or possibly more?  > A >As far as I can see, APIs like below cannont be found in HP SSL:  >X509_STORE_CTX_get_ex_data() % >SSLCERT_STORE_CTX_get_current_cert()  >SSLCERT_to_binary() >  >Thanks  >Victor   5 HP SSL for OpenVMS _is_ OpenSSL. From the web site at 7 http://h71000.www7.hp.com/openvms/products/ssl/ssl.html   N "HP SSL Version 1.2 for OpenVMS is based on OpenSSL 0.9.7d and includes all of- the latest security updates from OpenSSL.org"    -- Nigel Barker Live from the sunny Cote d'Azur    ------------------------------  # Date: Thu, 21 Apr 2005 07:33:38 GMT + From: Jack Patteeuw <jjpatteeuw@nospam.net>  Subject: Legato NetWorker C Message-ID: <mHI9e.10886$go4.9346@newsread2.news.atl.earthlink.net>   I I understand an evaluation version of this is now being distributed with  K recent version of VMS.  Has anyone tried it, or is using it in production ?   F Personally I have been using it for years on Unix (Tru64, Solaris and H Linux) and, in that environment, it is fantastic !  It handles multiple E input streams (disks) going to multiple output streams (tapes).  You  E must trust it's "cataloging" because you really don't know what file  H from what system is written to what tape !  (Catalogs are backed up and   can be recovered or re-created.)  G One feature I like about Networker over ABS and MDMS is, if there is a  G failure on tape drive the rest of the save completes to another drive !     
 Jack Patteeuw 4 "Coming back to VMS from the dark side of the force"  B "A PC with WinDoze is an application.  I use mine to make toast !"  ) "A goal, without a plan, is only a dream; ,   A dream, can become a goal, with a plan !"   ------------------------------  % Date: Thu, 21 Apr 2005 05:44:10 -0700q( From: Jeff Cameron <roktsci@comcast.net> Subject: Re: Legato NetWorker / Message-ID: <BE8CEE2A.C4E9%roktsci@comcast.net>   B We (MTI [www.mti.com]) sell install and provide support for Legato3 NetWorker. We are an authorized EMC/Legato Partner.o  H You have pretty much hit the nail on the head. It does all the wonderfulL things you said. One other thing is that it performs much better than nativeF VMS backup, primarily because of it's multi-streaming and multiplexingJ ability. It is also strong in providing individual file recovery. It's oneF weakness is in Disaster Recovery. While it can be used in DR scenario,L Legato recommends that you still make independent VMS backups of your systemE disk and the disk where Networked is installed. This speeds up you DRiJ operation. Unfortunately NetWorker cannot track the tape or tapes that you! write your Native VMS backups on.   G MTI has it's own Heterogeneous platform Enterprise backup software, andbH while it it certainly is not nearly as fast as Networker, it does recordJ backups using native VMS BACKUP (it also can use native unix TAR, cpio and dump).  K I have many customers who have used and continue to use our backup software I (Called Oasis Netbackup, not to be confused with Veritas' NetBackup [see:eI http://support.mti.com/MTICare/OpenVMS/NetBackup/index.htm). A handful ofwL them have decided to migrate to Legato Networker so that they have a uniform7 backup method across their entire computing Enterprise.h  L Most of them have come back to our Netbackup product because it VMS centric.G Our Netbackup software does have one additional drawback being that therH backup/storage node must be a VMS platform running our TapeControl mediaE management software. But all in all, we are considerably cheaper than>
 Networker.  K Don't get me wrong. Networker for VMS is a great well performing enterprisedL backup solution, and we fully support it and in many cases recommend it overB our Netbackup product. Different customers have different uses and? requirements, and that dictates which product is best for them.    Jeff Cameron On 4/21/05 12:33 AM, in articlegF mHI9e.10886$go4.9346@newsread2.news.atl.earthlink.net, "Jack Patteeuw" <jjpatteeuw@nospam.net> wrote:  J > I understand an evaluation version of this is now being distributed withM > recent version of VMS.  Has anyone tried it, or is using it in production ?M > G > Personally I have been using it for years on Unix (Tru64, Solaris and I > Linux) and, in that environment, it is fantastic !  It handles multiplecF > input streams (disks) going to multiple output streams (tapes).  YouF > must trust it's "cataloging" because you really don't know what fileI > from what system is written to what tape !  (Catalogs are backed up andb" > can be recovered or re-created.) > H > One feature I like about Networker over ABS and MDMS is, if there is aI > failure on tape drive the rest of the save completes to another drive !m >  >  > Jack Patteeuwk6 > "Coming back to VMS from the dark side of the force" > D > "A PC with WinDoze is an application.  I use mine to make toast !" > + > "A goal, without a plan, is only a dream;e. >   A dream, can become a goal, with a plan !"   ------------------------------  % Date: Thu, 21 Apr 2005 16:07:49 +0200n( From: "Rudolf Wingert" <win@fom.fgan.de> Subject: Re: Legato NetWorkere3 Message-ID: <000201c5467b$8023d1c0$994614ac@wat153>    Hello,  B My basic question is: is the Legato NetWorker for OpenVMS a serverE license or only a client license, so that I must have any Do Nix witha UNIX or Windows server?h   Best regards R. Wingert    ------------------------------  % Date: Thu, 21 Apr 2005 10:36:15 -0400s From: norm.raphael@metso.com Subject: Re: Legato NetWorker Q Message-ID: <OF918B7D42.1D4D4888-ON85256FEA.00501CDF-85256FEA.00501AF0@metso.com>:  C "Rudolf Wingert" <win@fom.fgan.de> wrote on 04/21/2005 10:07:49 AM:    > Hello, >nD > My basic question is: is the Legato NetWorker for OpenVMS a serverG > license or only a client license, so that I must have any Do Nix with. > UNIX or Windows server?e  3 Actually it is a storage-node license, but yes, youd4 need the NetWorker Serve to be on a non-OpenVMS box.  ; You do not want this solution in OpenVMS-only environments.e   >h > Best regards R. Wingertt >[   ------------------------------  % Date: Thu, 21 Apr 2005 16:02:13 +0200 " From: "Ariane" <ariane@freenet.de>  Subject: Re: Meine geilen Bilder: Message-ID: <4267b17d$2$17971$6d4158fb@reader-1.xsnews.nl>  " Hi, hier sind meine geilen Bilder! My nude Pics!!!A  http://www.geile-tipps.info/go/    -- Posted by News Bulk Poster Unregistered version   ------------------------------    Date: 21 Apr 2005 00:29:17 -0700$ From: juno10000@hotmail.com (Triger)J Subject: Re: Problem running x-windows client (exceed) over VPN connection= Message-ID: <d969c33c.0504202329.4c9ba471@posting.google.com>o  C Ok, guys I was experimenting by your answers and here is situation:   " WHEN LAPTOP IS CONNECTED IN LAN:  E 1. "set display...decw$clock..." routine works fine, clock appears on 
 laptop screenaE 2. Telnet from Alpha server to laptop works fine (when I start telnetc server on laptop, of course)  ! WHEN LAPTOP IS CONNECTED WIA WPN:w  ; 1. "set display....@decw$clock..." doesn't work, got error: ,         "x toolkit error: can't open display0          %dwt-f-nomsg, Message number 03AB8204 "  3 2. Telnet from Alpha server to laptop doesn't work  .        "Invalid or unknown host <ip address> "  E Obviosly, problem could be that Alpha "cannot see" laptop over VPN ?? 
 Am I right ??n  0 And Soterro asked for complete list of messages:@ When I try to run x-windows exceed application I got window with header:e  ) "Status:    Connection closed(read error) !  start method:   REXEC (port 512)r  host:    <alpha ip>  user id:  <username> 8  command: <@username_script.com_<laptop ip address>>   "  5 And in window body i have messages mentioned earlier:n2        %DWT-F-NOMSG, Message error number 03AB82047        %TRACE-F-TRACEBACK , symbolic stack dumb followsh<                            ................                "    # Any advice is highly appreciated...i
 Best regards;s   ------------------------------   Date: 21 Apr 2005 12:27:55 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)J Subject: Re: Problem running x-windows client (exceed) over VPN connection, Message-ID: <3cpkibF6pfa01U1@individual.net>  = In article <d969c33c.0504202329.4c9ba471@posting.google.com>,w' 	juno10000@hotmail.com (Triger) writes:rE > Ok, guys I was experimenting by your answers and here is situation:- > $ > WHEN LAPTOP IS CONNECTED IN LAN:  G > 1. "set display...decw$clock..." routine works fine, clock appears ona > laptop screenAG > 2. Telnet from Alpha server to laptop works fine (when I start telnet  > server on laptop, of course) > # > WHEN LAPTOP IS CONNECTED WIA WPN:n > = > 1. "set display....@decw$clock..." doesn't work, got error:n. >         "x toolkit error: can't open display2 >          %dwt-f-nomsg, Message number 03AB8204 " > 5 > 2. Telnet from Alpha server to laptop doesn't work  0 >        "Invalid or unknown host <ip address> " > G > Obviosly, problem could be that Alpha "cannot see" laptop over VPN ??e > Am I right ??r  * Is it perhaps possible your VPN is NATing?  B What with NAT and rules on Firewalls I don't control, I have givenA up trying to run X remotely outside the enterprise and have opted-1 for VNC instead.  Is a version available for VMS?5   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   P   ------------------------------  # Date: Thu, 21 Apr 2005 12:35:09 GMTI* From: "FredK" <fred.nospam@nospam.dec.com>J Subject: Re: Problem running x-windows client (exceed) over VPN connection2 Message-ID: <16N9e.4290$c53.3243@news.cpqcorp.net>  5 "Bill Gunshannon" <bill@cs.uofs.edu> wrote in message & news:3cpkibF6pfa01U1@individual.net...? > In article <d969c33c.0504202329.4c9ba471@posting.google.com>,s( > juno10000@hotmail.com (Triger) writes:  D > What with NAT and rules on Firewalls I don't control, I have givenC > up trying to run X remotely outside the enterprise and have optedm3 > for VNC instead.  Is a version available for VMS?  >t  ! There is a client, but no server.    ------------------------------   Date: 21 Apr 2005 13:10:37 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)J Subject: Re: Problem running x-windows client (exceed) over VPN connection, Message-ID: <3cpn2dF6qq467U1@individual.net>  2 In article <16N9e.4290$c53.3243@news.cpqcorp.net>,- 	"FredK" <fred.nospam@nospam.dec.com> writes:t > 7 > "Bill Gunshannon" <bill@cs.uofs.edu> wrote in messagen( > news:3cpkibF6pfa01U1@individual.net...@ >> In article <d969c33c.0504202329.4c9ba471@posting.google.com>,) >> juno10000@hotmail.com (Triger) writes:t > E >> What with NAT and rules on Firewalls I don't control, I have given D >> up trying to run X remotely outside the enterprise and have opted4 >> for VNC instead.  Is a version available for VMS? >> > # > There is a client, but no server.e  dB Well, another option would be to run (shudder!) a windows box with@ VNC Server and an Xserver (Cygwin has a nice one and it's free).B Let the windows box access X 0n the VMS machine and use VNC to get at the desktop remotely.   bill x  f   -- WJ Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>   r   ------------------------------  % Date: Thu, 21 Apr 2005 16:10:44 +0200i, From: Albrecht Schlosser <ajs856@tiscali.de>J Subject: Re: Problem running x-windows client (exceed) over VPN connection, Message-ID: <g9c84d.1ag.ln@news.hus-soft.de>   Bill Gunshannon wrote:4 > In article <16N9e.4290$c53.3243@news.cpqcorp.net>,/ > 	"FredK" <fred.nospam@nospam.dec.com> writes:$ > 7 >>"Bill Gunshannon" <bill@cs.uofs.edu> wrote in message.( >>news:3cpkibF6pfa01U1@individual.net... >>@ >>>In article <d969c33c.0504202329.4c9ba471@posting.google.com>,) >>>juno10000@hotmail.com (Triger) writes:O >>E >>>What with NAT and rules on Firewalls I don't control, I have givenbD >>>up trying to run X remotely outside the enterprise and have opted4 >>>for VNC instead.  Is a version available for VMS? >>>e >># >>There is a client, but no server.g >  >   D > Well, another option would be to run (shudder!) a windows box withB > VNC Server and an Xserver (Cygwin has a nice one and it's free).D > Let the windows box access X 0n the VMS machine and use VNC to get > at the desktop remotely.  P Or use ssh to connect to the server and do port forwarding for the display port P   through the ssh connection. Should work, but I didn't do it (yet), so I can't Q give more details. Putty is a source for a PC ssh client with some documentation eC about port forwarding, and there is also something in the VMS docs.l   Albrecht   ------------------------------  % Date: Thu, 21 Apr 2005 17:29:59 +0200e! From: Soterro <soterroatyahoocom>yJ Subject: Re: Problem running x-windows client (exceed) over VPN connection9 Message-ID: <4267c62b$0$299$4d4ef98e@read.news.ch.uu.net>a   Albrecht Schlosser wrote:pE > Or use ssh to connect to the server and do port forwarding for the  I > display port  through the ssh connection. Should work, but I didn't do eI > it (yet), so I can't give more details. Putty is a source for a PC ssh lJ > client with some documentation about port forwarding, and there is also  > something in the VMS docs.  > There was also a thread on this some time ago, um, here it is: http://tinyurl.com/8ue66    (also the long URL just in case) http://groups.google.com/groups?hl=la&lr=&threadm=6cCdnfr-0puu74XdRVn-sw%40comcast.com&rnum=1&prev=/groups%3Fhl%3Dla%26lr%3D%26scoring%3Dd%26q%3D%2522X11%2Bvia%2BSSH%2Bfrom%2BOpenVMS%2Bto%2BWin2000%2522%26btnG%3DQuaere   Sh   ------------------------------   Date: 21 Apr 2005 15:53:01 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)J Subject: Re: Problem running x-windows client (exceed) over VPN connection, Message-ID: <3cq0itF6k82qfU1@individual.net>  , In article <g9c84d.1ag.ln@news.hus-soft.de>,/ 	Albrecht Schlosser <ajs856@tiscali.de> writes:  > Bill Gunshannon wrote:5 >> In article <16N9e.4290$c53.3243@news.cpqcorp.net>,e0 >> 	"FredK" <fred.nospam@nospam.dec.com> writes: >> h8 >>>"Bill Gunshannon" <bill@cs.uofs.edu> wrote in message) >>>news:3cpkibF6pfa01U1@individual.net...n >>>eA >>>>In article <d969c33c.0504202329.4c9ba471@posting.google.com>, * >>>>juno10000@hotmail.com (Triger) writes: >>> F >>>>What with NAT and rules on Firewalls I don't control, I have givenE >>>>up trying to run X remotely outside the enterprise and have opted 5 >>>>for VNC instead.  Is a version available for VMS?c >>>> >>>r$ >>>There is a client, but no server. >> u >>  E >> Well, another option would be to run (shudder!) a windows box witheC >> VNC Server and an Xserver (Cygwin has a nice one and it's free). E >> Let the windows box access X 0n the VMS machine and use VNC to get  >> at the desktop remotely.h > R > Or use ssh to connect to the server and do port forwarding for the display port R >   through the ssh connection. Should work, but I didn't do it (yet), so I can't S > give more details. Putty is a source for a PC ssh client with some documentation cE > about port forwarding, and there is also something in the VMS docs.  >   D I didn't suggest that because I seem to remember a conversation hereC that mentioned port forwarding with ssh not being a vailable in all H cases (perhaps dependant on what IP stack you are running?).  Of course,C if SSH and portforwarding are available they are the best (and mostH secure) asnwer to the problem.   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Thu, 21 Apr 2005 11:38:46 +0100f* From: Peter Bell <peter@bellfamily.org.uk>5 Subject: Re: Relative file size changes with copying?a8 Message-ID: <6579085f4d.peter@iyonix.earley.fourcom.com>  1 In message <aWu9e.4216$bs2.1162@news.cpqcorp.net> 3           "Hein" <hein.nomail@hp.nomail.com> wrote:b  9 > "Peter Bell" <peter@bellfamily.org.uk> wrote in message-4 > news:ba97895e4d.peter@iyonix.earley.fourcom.com...L > > We have a minor inconvenience with copying relative files under VMS 6.2. > >gK > > We find that if we copy a file which is relative, fixed length records,p > :iK > > Why doesn't the used size get preserved when the file is copied, rathere; > > than increasing the used size up to the allocated size?s > D > I just tried using COPY on VMS V7.1 and the EOF mark was preserved >  > $ dir/size=all *.rel  > TMP.REL;1                  9/9 > $ dismount mda0s > $ init mda0 hein /clus=100 > $ moun mda0 hein$ > %MOUNT-I-MOUNTED, HEIN mounted ... > $ cre/dir mda0:[hein]C > $ copy/log tmp.rel mda0:G > %COPY-S-COPIED, U$1:[HEIN]TMP.REL;1 copied to MDA0:[HEIN]TMP.REL;1 (9b	 > blocks)  > $ dir/size=all mda0:" > TMP.REL;1                  9/100 > $ anal/rms mda0:tmp.rel  > RMS FILE ATTRIBUTES % >         File Organization: relativeu > : 8 >          Blocks Allocated: 100, Default Extend Size: 0 >         Bucket Size: 1 >  FIXED PROLOGn > :c >         End-of-File VBN: 10  >         Prolog Version: 1f  F Okay - thanks for that.  However, if you look at my example below, youE will see that the used size doesn't increase on the first copy to theb8 device, but on a subsequent copy within the same device.   ALPHA1> ANAL/RMS DB:ABC.DAT;1a# DKA0:[ACSDEV.DEVELOP.DATA]ABC.DAT;1d :n RMS FILE ATTRIBUTESy  #         File Organization: relative  :i6         Blocks Allocated: 4251, Default Extend Size: 1         Bucket Size: 18m   FIXED PROLOG :r         End-of-File VBN: 4252t         Prolog Version: 1s   ALPHA1> dir db:abc.dat/siz:all  $ Directory DKA0:[ACSDEV.DEVELOP.DATA]  ! ABC.DAT;1               4251/4251e  " Total of 1 file, 4251/4251 blocks.& ALPHA1> cop db:abc.dat dka300:[acslog]* ALPHA1> dir dka300:[acslog]abc.dat/siz:all   Directory DKA300:[ACSLOG]e  ! ABC.DAT;1               4251/4255n  " Total of 1 file, 4251/4255 blocks.2 ALPHA1> cop dka300:[acslog]abc.dat dka300:[acslog]* ALPHA1> dir dka300:[acslog]abc.dat/siz:all   Directory DKA300:[ACSLOG]t  ! ABC.DAT;2               4255/4255t! ABC.DAT;1               4251/4255o  # Total of 2 files, 8506/8510 blocks.s& ALPHA1> cop dka300:[acslog]abc.dat db: ALPHA1> dir db:abc.dat/siz:all  $ Directory DKA0:[ACSDEV.DEVELOP.DATA]  ! ABC.DAT;2               4255/4257d! ABC.DAT;1               4251/4251   # Total of 2 files, 8506/8508 blocks.a! ALPHA1> cop db:abc.dat db:abc.datp ALPHA1> dir db:abc.dat/siz:all  $ Directory DKA0:[ACSDEV.DEVELOP.DATA]  ! ABC.DAT;3               4257/4257a! ABC.DAT;2               4255/4257 ! ABC.DAT;1               4251/4251-  % Total of 3 files, 12763/12765 blocks.p ALPHA1> ANAL/RMS DB:ABC.DATJ# DKA0:[ACSDEV.DEVELOP.DATA]ABC.DAT;3R :C RMS FILE ATTRIBUTESo  #         File Organization: relative4 :36         Blocks Allocated: 4257, Default Extend Size: 1         Bucket Size: 18.   FIXED PROLOG :e         End-of-File VBN: 4252m         Prolog Version: 1-  3 > The behaviour may have changed with VMS versions.1  > Perhaps, but your experiment didn't explore the area where I'm experiencing this oddity.g  ! > What tool did you use for copy.    Straightforward COPY command.B  > > You may find that BACKUP preserves EOF, where COPY does not? > I > Finally, The EOF marker is largely meaningless in the context of an RMS>K > relative file: empty and deleted cells will be automatically skipped overaD > during  sequential read operations.  (This particular behaviour isB > documented in  the _OpenVMS Guide to File Applications_ manual.)  F Absolutely, but it's just confusing when the file size changes betweenG what should be identical files.  As Bill says, this can be awkward.  InfH our case, the cluster sizes are not prime and this particular file stopsF extending when it reaches 4260 blocks, with our current mix of 3 and 5 cluster sizes.   > Hope this helps some,   E Some, yes.  But I'd still like to know why it's happening and, betterDG still, find a way of preventing it from happening.  It would be nice ifHG my file retained its size as 4251/4255 when copied on the device with a F cluster size of 5 - it could then revert to 4251/4251 when copied back& to the device with cluster size of 3.    -- a$ Peter Bell - peter@bellfamily.org.uk   ------------------------------  % Date: Thu, 21 Apr 2005 13:11:51 +0100a- From: John Laird <nospam@laird-towers.org.uk> 5 Subject: Re: Relative file size changes with copying?h8 Message-ID: <6p5f61dpog5n2h284lkuoaubm2tguvsn87@4ax.com>  H On Thu, 21 Apr 2005 11:38:46 +0100, Peter Bell <peter@bellfamily.org.uk> wrote:  2 >In message <aWu9e.4216$bs2.1162@news.cpqcorp.net>4 >          "Hein" <hein.nomail@hp.nomail.com> wrote: >> >> Hope this helps some, >SF >Some, yes.  But I'd still like to know why it's happening and, betterH >still, find a way of preventing it from happening.  It would be nice ifH >my file retained its size as 4251/4255 when copied on the device with aG >cluster size of 5 - it could then revert to 4251/4251 when copied backu' >to the device with cluster size of 3. D  J Hein also said "The EOF marker is largely meaningless in the context of anL RMS relative file".  That is the explanation, and it also applies to indexedF files.  COPY and BACKUP must therefore faithfully copy all *allocated*H blocks, as they may well contain valid data, and thus where the two diskK cluster sizes are not simple multiples of one another, the file will always,K tend to grow.  Yours stopped at a multiple of 15, for example, as expected.   G Look at it another way - if I gave you a non-sequential file showing atyJ 4251/4255 blocks which *did* have data in blocks 4252-4255 and you managedJ to lose that data when copying to another disk, I'd be more upset than you are !    -- e4 If the good die young, then I'm gonna live forever.    Mail john rather than nospam...h   ------------------------------  % Date: Thu, 21 Apr 2005 13:46:38 +0100p* From: Peter Bell <peter@bellfamily.org.uk>5 Subject: Re: Relative file size changes with copying?e8 Message-ID: <4e2e145f4d.peter@iyonix.earley.fourcom.com>  7 In message <6p5f61dpog5n2h284lkuoaubm2tguvsn87@4ax.com>e8           John Laird <nospam@laird-towers.org.uk> wrote:  J > On Thu, 21 Apr 2005 11:38:46 +0100, Peter Bell <peter@bellfamily.org.uk> > wrote: > 4 > >In message <aWu9e.4216$bs2.1162@news.cpqcorp.net>6 > >          "Hein" <hein.nomail@hp.nomail.com> wrote: > >> > >> Hope this helps some, > >tH > >Some, yes.  But I'd still like to know why it's happening and, betterJ > >still, find a way of preventing it from happening.  It would be nice ifJ > >my file retained its size as 4251/4255 when copied on the device with aI > >cluster size of 5 - it could then revert to 4251/4251 when copied backc) > >to the device with cluster size of 3.   > L > Hein also said "The EOF marker is largely meaningless in the context of anN > RMS relative file".  That is the explanation, and it also applies to indexedH > files.  COPY and BACKUP must therefore faithfully copy all *allocated*J > blocks, as they may well contain valid data, and thus where the two diskM > cluster sizes are not simple multiples of one another, the file will alwaysR > tend to grow.t  @ Thanks for that - I do now understand.  It seems that VMS/RMS is# protecting against errant programs.a   -- l$ Peter Bell - peter@bellfamily.org.uk   ------------------------------    Date: 21 Apr 2005 07:51:54 -0500 From: briggs@encompasserve.org5 Subject: Re: Relative file size changes with copying?a3 Message-ID: <70a4HhDzENg7@eisner.encompasserve.org>h  h In article <s_WdndBvL8yQYvvfRVn-1A@metrocastcablevision.com>, Bill Todd <billtodd@metrocast.net> writes:I > Well, the behavior described would still be kind of awkward if you for lG > some reason needed to move the file back and forth between two disks eI > whose cluster sizes were prime relative to each other:  the file would y& > grow a bit with every new migration.  B All pairs of numbers, including those that are co-prime have least common multiples.   B It's just that when they're co-prime, the least common multiple is& simply the product of the two numbers.   	John Briggs   ------------------------------  % Date: Thu, 21 Apr 2005 14:01:47 +0100e- From: John Laird <nospam@laird-towers.org.uk> 5 Subject: Re: Relative file size changes with copying?28 Message-ID: <is8f61tsbmlfrlut2udlogfujs7oa340vj@4ax.com>  H On Thu, 21 Apr 2005 13:46:38 +0100, Peter Bell <peter@bellfamily.org.uk> wrote:  A >Thanks for that - I do now understand.  It seems that VMS/RMS is $ >protecting against errant programs.  J The errant "program" being RMS itself ;-)  When a file is properly closed,J the EOF should be correct.  However, abnormal image rundowns tend to leaveK it behind, as it were, even although RMS will be involved in the rundown byt way of flushing buffers, etc.h   -- e2 Tomorrow has been cancelled until further notice.    Mail john rather than nospam...l   ------------------------------  % Date: Thu, 21 Apr 2005 11:47:03 -0400v( From: Bill Todd <billtodd@metrocast.net>5 Subject: Re: Relative file size changes with copying?g= Message-ID: <xZadncBtJJdlV_rfRVn-jg@metrocastcablevision.com>    briggs@encompasserve.org wrote:0j > In article <s_WdndBvL8yQYvvfRVn-1A@metrocastcablevision.com>, Bill Todd <billtodd@metrocast.net> writes: > I >>Well, the behavior described would still be kind of awkward if you for :G >>some reason needed to move the file back and forth between two disks uI >>whose cluster sizes were prime relative to each other:  the file would t& >>grow a bit with every new migration. >  > D > All pairs of numbers, including those that are co-prime have least > common multiples.  > D > It's just that when they're co-prime, the least common multiple is( > simply the product of the two numbers.  E Ah - just wasn't really awake, I guess.  And for that matter a least  I common multiple would exist for any number of disks among which the file r moved.   - bill   ------------------------------  % Date: Thu, 21 Apr 2005 12:00:36 -0400h( From: Bill Todd <billtodd@metrocast.net>5 Subject: Re: Relative file size changes with copying? = Message-ID: <CoCdnTzArK-4U_rfRVn-gA@metrocastcablevision.com>    John Laird wrote:tJ > On Thu, 21 Apr 2005 13:46:38 +0100, Peter Bell <peter@bellfamily.org.uk> > wrote: >  > B >>Thanks for that - I do now understand.  It seems that VMS/RMS is% >>protecting against errant programs.e >  > L > The errant "program" being RMS itself ;-)  When a file is properly closed,L > the EOF should be correct.  However, abnormal image rundowns tend to leaveM > it behind, as it were, even although RMS will be involved in the rundown byc > way of flushing buffers, etc.   H That sounds more like a bug in RMS, then:  if it has the opportunity to G flush dirty buffers, surely it could update EOF as well if appropriate.b  D That's not necessarily a problem for relative files, since the only C effect that a subsequent straight-forward file-copy operation by a  G program less tolerant of such errors than COPY seems to be would be to tD leave unwritten record slots which the original application had not G considered important enough to have flushed to disk in the first place tF (since *surely* on such a flush RMS will update EOF if appropriate as I part of the operation before returning - right?).  With an indexed file, aD however, such a naive copy after such a failure to update EOF could D leave the file structurally corrupt, since blocks within the copied G portion could then be pointing to garbage in the uncopied portion (and cI I'm not sure that some sequence of truncate-then-extend-again operations  H on a sequential file couldn't leave you with, say, half a record at the H new copy's end - unless RMS forces a new EOF to disk on truncate, which B under the circumstances had better be the case, but then does RMS G actually check that truncating a sequential file honors any applicable  E record boundaries anyway?  if not, a user could screw up such things  G anyway, though in that case it wouldn't really be as much RMS's fault).    - bill   ------------------------------    Date: 21 Apr 2005 07:36:05 -0500B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley)3 Subject: RTOS options, was: Re: Could a PC do this? 3 Message-ID: <QlO18gQlSRDW@eisner.encompasserve.org>N  R In article <OumdnTzhPcHdNvvfRVn-3g@igs.net>, "John Smith" <a@nonymous.com> writes:H > You could probably do it on a 'PC' if you used QNX or one of the other0 > 'real-time' operating systems such as VxWorks. > I > You might also have better luck finding a wider range of controller D/A E > cards that already have drivers for one of these operating systems.c > K > I can't offer any current thoughts on which to use - last time I used QNXn > was about 12-15 years ago. > 
 > www.qnx.com: > http://www.windriver.com/e >  > L > Does anybody buy VMS to do this sort of thing any longer, even though it's > pretty good at it? >   H I agree that for people with existing VMS realtime applications there isI no urgent need to move to a non-VMS solution, but the problem is that formH people creating new realtime systems today, the fact that VMS is "prettyI good at it" is pretty much irrelevent, especially if you don't have a VMS2 background.r  E First, the hardware in a x86/ARM/whatever based system is very cheap,eD even if you need an embedded solution instead of a RTOS running on a desktop PC.   J For example, people are building x86 based embedded systems using Mini-ITXL (See http://www.mini-itx.com ). If you prefer a non-x86 environment, you canJ get embedded ARM boards from http://www.embeddedarm.com for under 200 USD.> (But there's no one in the UK that stocks this ARM board. :-()  G Second, there are a wide range of RTOS's available, especially for x86.5I A number of these RTOS are available for free, and there's even a cutdowncO QNX version that you can use for free, provided your project is non-commercial. N [And as we all know, if you start using a OS as a student or for personal use,I there's an increased chance that you will choose it in a work environmentl3 once you move into a job that requires such tools.]o  / Here's something that I wrote a few months ago:   I |  There are a number of free RTOS options available already, so one morew: |  probably will not make much difference. A few examples: |  eI |  a) In the Linux area, you have RTLinux, which is implemented as a hard O |  realtime kernel running below Linux (it runs Linux as a task). A major issuerL |  here is that the Linux device drivers run in the non-realtime part of theL |  system, so if you need realtime device drivers for your project, you have4 |  to implement them again under the RTLinux kernel. |  aJ |  As an example, there appears to be no real-time USB support in RTLinux,D |  (If I am wrong about that, I would _really_ appreciate a pointer) |  'C |  b) RedHat has a RTOS implemented from the ground up called eCos.t |  h+ |  c) OAR Corp produce a RTOS called RTEMS.e |  yO |  d) If you are a hobbyist, and hence your project is non-commercial, then QNX N |  make available a cut down version of QNX (called QNX NC) for free use. NoteM |  that if you want to build embedded systems with it, you have to go back tohA |  QNX NC 6.1, as QNX removed the embedded tools from QNX NC 6.2.d  J I actually had to choose an RTOS recently for an embedded hobbyist projectK that I am working on, and I ended up choosing RTEMS. (I chose it because itiI has Ada support, was designed to be used in embedded systems, and runs onh x86 hardware.)   Simon.   -- 3B Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       7 Microsoft: The Standard Oil Company of the 21st century-   ------------------------------    Date: 21 Apr 2005 07:54:33 +01006 From: eplan@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)8 Subject: [OpenVMS VAX V7.3] Updating with VAXRPC02_073 ?, Message-ID: <42675c39$1@news.langstoeger.at>  D I just saw an rather old (Aug 2004) ECO kit for OpenVMS VAX V7.3 andG noticed that I haven't checked my VAX for a while and indeed did miss atI couple of ECOs (eg. DECnet-Plus V7.3 ECO4). So after browsing the ReleaseeM Notes I thought I give it a try and install them and alas VAXRPC02_073 claimsl    ;         You are attempting to install an old version of ther<         remedial kit. Should you choose to proceed with thisH         installation, prior fixes in images replaced by this kit will be
         lost.S   * Do you want to continue [NO]?E  , Does anybody already know the story behind ?  M Release Notes state one needs it for DCE V3.1 (I still have V3.0) and for RPCmM (I remember having installed an RPC ECO for my Alphas some weeks/months ago). O I haven't installed DCE or RPC related ECOs before (since VAX/VMS V7.3 upgrade) N and only mention of RPC in ECO logfiles I find is in TCPIP and TCPware so far.8 So what the heck is it claiming to incorrectly replace ?   TIAp   -- @ Peter "EPLAN" LANGSTOEGERt% Network and OpenVMS system specialistg E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------   End of INFO-VAX 2005.222 ************************ion ?   F Personally I have been using it for years on Unix (Tru64, Solaris and H Linux) and, in that environment, it is fantastic !  It handles multiple E input streams (disks) 0 Port 8.218 at Host 83.146.44.251 accepted. <<< RETR moria_500.76eZ >>> 150 IMAGE retrieve of /disk$misc/decus/vax_games/moria_500.76 (15646 bytes) started.; >>> 226 Transfer completed.  14592 (8) bytes transferred., <<< PORT 83,146,44,251,8,2194 >>> 200 Port 8.219 at Host 83.146.44.251 accepted. <<< RETR moria_500.77eZ >>> 150 IMAGE retrieve of /disk$misc/decus/vax_games/moria_500.77 (15642 bytes) started.; >>> 226 Transfer completed.  14538 (8) bytes transferred., <<< PORT 83,146,44