1 INFO-VAX	Sat, 24 Mar 2001	Volume 2001 : Issue 165       Contents: "green bar" in DCPS  RE: "green bar" in DCPS  Re: "green bar" in DCPS  Re: "green bar" in DCPS  Re: "green bar" in DCPS  Re: "green bar" in DCPS ) BASIC 1.4 install causing a startup error C Re: Boring the pants off the COV (was RE: Dumbing Down VMS with...)  Re: compaction "nabled"? dsn  RE: dsn  Re: dsn P Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS     EducationalProgramP Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS     EducationalProgramP Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS     EducationalProgramP Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS     EducationalProgramB re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS EducatioM Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS EducationalProgram) P Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducationalProgram) ProP Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducationalProgram) ProP Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducationalProgram) Pro Re: listserv programP Re: New qualifiers for some DCL commands (was: Re: Dumbing Down VMS withUNIX Ele# Re: POSIX Streams, File Permissions # Re: POSIX Streams, File Permissions 9 Source listing do NOT coincide with the operating system. = Re: Source listing do NOT coincide with the operating system. = Re: Source listing do NOT coincide with the operating system.  Upgrading to alphavms 7.2-1  Re: Upgrading to alphavms 7.2-1  Re: Upgrading to alphavms 7.2-1  Re: Upgrading to alphavms 7.2-1  Re: [DCL] minute of the day  Re: [DCL] minute of the day  Re: [DCL] minute of the day   F ----------------------------------------------------------------------  % Date: Fri, 23 Mar 2001 14:04:54 -0500 / From: "Webb, William W" <wwebb1@email.usps.gov>  Subject: "green bar" in DCPSK Message-ID: <D46FE9B132FB9B44AEC242A96E4AB75001925A6E@rlghncst625.usps.gov>   ( Sometimes our Printserver 32 prints with' line numbering and shaded bars that are  reminiscent of green bar paper.   ' This seems to occur most often when the ' parameter /PARAM=(SIDES=2) is used, but  it's by no means consistent.  ' Does anybody know exactly how this gets 
 triggered?  1 (Paging Paul Anderson... paging Paul Anderson...)  ;^)    ==============================* William W. Webb, EDS, c/o USPS DSSC/OSS/MS. 4924 Green Road Raleigh, NC 27616 919 874 3043      ------------------------------  % Date: Fri, 23 Mar 2001 14:16:26 -0600 * From: WILLIAM WEBB <WWEBB1@email.usps.gov>  Subject: RE: "green bar" in DCPS- Message-ID: <0033000019548808000002L082*@MHS>   # =0ASorry, VMS 7.2-1 Alpha, DCPS 1.7    > -----Original Message-----1 > From: Info-VAX-Request@Mvb.Saic.Com at INTERNET & > Sent: Friday, March 23, 2001 2:11 PM8 > To: Webb, William W; Info-VAX@Mvb.Saic.Com at INTERNET > Subject: "green bar" in DCPS >  > * > Sometimes our Printserver 32 prints with) > line numbering and shaded bars that are ! > reminiscent of green bar paper.  > ) > This seems to occur most often when the - > parameter /PARAM=3D(SIDES=3D2) is used, but  > it's by no means consistent. > ) > Does anybody know exactly how this gets  > triggered? > 3 > (Paging Paul Anderson... paging Paul Anderson...)  > ;^)  > H > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D , > William W. Webb, EDS, c/o USPS DSSC/OSS/MS0 > 4924 Green Road Raleigh, NC 27616 919 874 3043 >=   ------------------------------    Date: 23 Mar 2001 14:27:05 -05003 From: malmberg@encompasserve.org (John E. Malmberg)   Subject: Re: "green bar" in DCPS3 Message-ID: <AQNs4pwVdVGu@eisner.encompasserve.org>   K In article <D46FE9B132FB9B44AEC242A96E4AB75001925A6E@rlghncst625.usps.gov>, 8 "Webb, William W" <wwebb1@email.usps.government> writes:  * > Sometimes our Printserver 32 prints with) > line numbering and shaded bars that are ! > reminiscent of green bar paper.   + If you like that, for plain text files use:    /param=(data_type=list)   ) > This seems to occur most often when the ) > parameter /PARAM=(SIDES=2) is used, but  > it's by no means consistent.  , Yes it is consistent, if you know the rules.  ) > Does anybody know exactly how this gets  > triggered?   $help print_parameter data_type   G This will give you a start.  The default list of file extensions should @ be in the DCPS manual somewhere, and IIRC: you can customize it.  B Note: DCPS_OPEN license required for this functionality currently.   -John  wb8tyw@qsl.network Personal Opinion Only    ------------------------------  # Date: Fri, 23 Mar 2001 19:36:48 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)   Subject: Re: "green bar" in DCPS0 Message-ID: <009F972F.27E44C90@SendSpamHere.ORG>  } In article <D46FE9B132FB9B44AEC242A96E4AB75001925A6E@rlghncst625.usps.gov>, "Webb, William W" <wwebb1@email.usps.gov> writes: ) >Sometimes our Printserver 32 prints with ( >line numbering and shaded bars that are  >reminiscent of green bar paper. > ( >This seems to occur most often when the( >parameter /PARAM=(SIDES=2) is used, but >it's by no means consistent.  > ( >Does anybody know exactly how this gets >triggered?  > 2 >(Paging Paul Anderson... paging Paul Anderson...) >;^) >  >============================== + >William W. Webb, EDS, c/o USPS DSSC/OSS/MS / >4924 Green Road Raleigh, NC 27616 919 874 3043  >   J Look in the manual for the discussion of DCPS$FILE_EXTENSION_DATA_TYPE.DAT  H Depending upon the file extension of the file being sent to the printer,G this file determines who to format that output.   Usually, any language 2 source files will be printed with the "green bar".  G FYI, if you have a color laser, create a setup file with the following:   ' /setgray { .6 exch .6 setrgbcolor } def   I and place it in your DCPS setup library (I call it GREY2GREEN).  Then, if I you issue a PRINT/SETUP=GREY2GREEN/.... <file.ext>, the grey bars are now  green. --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              O city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  # Date: Fri, 23 Mar 2001 19:39:42 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)  Subject: Re: "green bar" in DCPS1 Message-ID: <2ENu6.111$fB6.3965@news.cpqcorp.net>   } In article <D46FE9B132FB9B44AEC242A96E4AB75001925A6E@rlghncst625.usps.gov>, "Webb, William W" <wwebb1@email.usps.gov> writes: I :Sometimes our Printserver 32 prints with line numbering and shaded bars  + :that are reminiscent of green bar paper... 3 :Does anybody know exactly how this gets triggered?   H   This behaviour is based on the file extension, or by explicit request "   via PRINT /PARAM=DATA_TYPE=LIST.  F   To customize the processing of file extensions within DCPS, use the    information available in:   9     SYS$LIBRARY:DCPS$FILE_EXTENSION_DATA_TYPE.DAT_DEFAULT   #   to create your own site-specific:   1     SYS$LIBRARY:DCPS$FILE_EXTENSION_DATA_TYPE.DAT      D   On a related note, the separator tray logical name feature that isI   available with DCPS-PLUS is handy.  With a printer supporting multiple  K   paper trays, you can define this logical name to the (smallest-capacity)  J   tray and fill it with colored paper, and the print separator pages will 2   be printed on the colored paper from that tray.   8     $ DEFINE/SYSTEM/EXEC DCPS$queuename_SEPARATOR_TRAY n  =   Where n is the (numeric) tray containing the colored paper.   N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 23 Mar 2001 14:43:16 -0500 0 From: Paul Anderson <paul.r.anderson@compaq.com>  Subject: Re: "green bar" in DCPSK Message-ID: <paul.r.anderson-A3965D.14431623032001@gazette.loc1.tandem.com>   3 In article <AQNs4pwVdVGu@eisner.encompasserve.org>, 5  malmberg@encompasserve.org (John E. Malmberg) wrote:   M > In article <D46FE9B132FB9B44AEC242A96E4AB75001925A6E@rlghncst625.usps.gov>, : > "Webb, William W" <wwebb1@email.usps.government> writes: > E > Sometimes our Printserver 32 prints with line numbering and shaded  / > bars that are reminiscent of green bar paper.  > B > This will give you a start. The default list of file extensions F > should be in the DCPS manual somewhere, and IIRC: you can customize  > it.    The file to look for is D SYS$LIBRARY:DCPS$FILE_EXTENSION_DATA_TYPE.DAT_DEFAULT.  If you make % changes, copy the file to .DAT first.    Paul   --  ,    Paul Anderson, OpenVMS Engineering (DCPS),    Compaq Computer Corporation, Littleton MA   ------------------------------  # Date: Fri, 23 Mar 2001 23:30:58 GMT + From: rjordan@mars.mcs.net (Richard Jordan) 2 Subject: BASIC 1.4 install causing a startup error3 Message-ID: <S0Ru6.2613$JN5.78384@news.goodnet.com>   H We have a new DS10 being configured for a customer.  600MHz, 256MB, dualE 9GB drives on a KZPBA controller, etc.  We have a personal use BASIC  C license  assigned to the development account name (not SYSTEM), and C used SYSTEM to install BASIC 1.4 from the March 2001 condist.  Full G privs were enabled, andd there were no errors on install except for the ? IVP (which failed because SYSTEM is not licensed to use BASIC).   L Since that install, we get the following error very early in system startup:  A  %INSTALL-I-NONSHRADR, image installed ignoring '/SHARE=ADDRESS'  8      DISK$ALPHASYS:<SYS0.SYSCOMMON.SYSLIB>DEC$BASRTL.EXEI  -INSTALL-E-NOTSHRSYMVEC, symbol vector is not in NONSHRADR image section   F There is a mention of the INSTALL-I-NONSHRADR message coming up duringH the install of BASIC, with the fix of makign sure all privs are enabled,B but I haven't been able to locate any reference to it happening inD system startup (checked Google/deja, VMS Wizard, and DSNlink).  I'llA put a call in on Monday, but was hoping someone here might have a  quicker answer.   2 I have done the following to try and correct this:  B      Relicense BASIC to SYSTEM and reinstall (IVP passed, no help)D      Checked ECOs newer than installed for relevant info (underway).  8 Following products installed (plus VMSINSTAL kits below)  < ----------------------------------- ----------- ------------5 PRODUCT                             KIT TYPE    STATE < ----------------------------------- ----------- ------------9 DEC AXPVMS DECNET_PHASE_IV V7.2-1   Full LP     Installed 9 DEC AXPVMS DWMOTIF V1.2-5           Full LP     Installed 9 DEC AXPVMS NS_NAV_EXPORT V3.3       Full LP     Installed 9 DEC AXPVMS OPENVMS V7.2-1           Platform    Installed 9 DEC AXPVMS TCPIP V5.1-15            Full LP     Installed 9 DEC AXPVMS VMS V7.2-1               Oper System Installed < ----------------------------------- ----------- ------------    A Following ECOs installed (done a few weeks ago, I know that there ? are a updates available, some of which are planned for install)    DEC AXPVMS VMS721_CPU2208 V1.0 DEC AXPVMS VMS721_DDTM V1.0  DEC AXPVMS VMS721_DQCONFIG V2.0  DEC AXPVMS VMS721_DRIVER V1.0  DEC AXPVMS VMS721_F11X V2.0   DEC AXPVMS VMS721_FIBRECHAN V3.0 DEC AXPVMS VMS721_IPC V1.0 DEC AXPVMS VMS721_LAN V1.0 DEC AXPVMS VMS721_P350 V1.0  DEC AXPVMS VMS721_PCSI V1.0  DEC AXPVMS VMS721_SCSI V1.0   DEC AXPVMS VMS721_SHADOWING V3.0 DEC AXPVMS VMS721_SYS V8.0 DEC AXPVMS VMS721_UPDATE V1.0   @ And the following VMSINSTAL kits (preinstalls, I only did BASIC)   OPEN3DA          |  4.9  ALPDWEUR01_U5    |  1.2  DIAA             |  3.3  FIIVA01          |  7.2  BASIC            |  1.4   & Thanks much for any thoughts or ideas.   Rich Jordan  rjordan@mcs.net    ------------------------------  % Date: Fri, 23 Mar 2001 23:58:25 +0000 ) From: Christof Brass <brass@infopuls.com> L Subject: Re: Boring the pants off the COV (was RE: Dumbing Down VMS with...), Message-ID: <3ABBE321.CC14CABA@infopuls.com>   Brian Wheeler wrote: > . > In article <3ABB4C28.E5316DD1@infopuls.com>,5 >         Christof Brass <brass@infopuls.com> writes:  > C > > To be clear: what the majority thinks or does is no measure for C > > quality. It could be the extinction of some minorities. In that A > > I could reference to the butt head (Brian Wheeler) in this NG B > > accusing me of beeing like the Nazis against the Jews. I wouldD > > have used a more drastic word to denote this stupid ashole but IA > > don't know any suitable words which will pass the four letter  > > words filters. > G > Actually, I was merely accusing you of being irrational, and ignoring H > the facts when they didn't suit you.  The Nazi reference was to invoke0 > Godel's law and shut down this tedious thread. >  > Brian   ; Tedious is that I'm accused of irrationality like communist = regimes did without anything rational to disprove the accused  person. 4 If you read what came out so far of real facts it is< interesting. See Hoff's, Pauls Repacholi's, Mark Berryman's,= Christopher Smith's and some other's posts (sorry if I forgot 7 one important contributor). Honestly based on the facts * presented UNIX crap isn't in a good shape.   ------------------------------    Date: 24 Mar 2001 04:24:19 +0800, From: Paul Repacholi <prep@prep.synonet.com>! Subject: Re: compaction "nabled"? - Message-ID: <87ofusnsi4.fsf@prep.synonet.com>   / koehler@encompasserve.org (Bob Koehler) writes:   E > Has this come up before (I thought I might have seen it, but Google & > and Wizard searches didn't find it)?   > VMS 7.2-1, DEC 3000 600S > J > Magtape MUMBLE$MKB100:, device type TLZ06, is online, allocated, mountedM >     foreign, record-oriented device, file-oriented device, error logging is I >     enabled, device is busy, controller supports compaction (compaction ( >     nabled), device supports fastskip. >     ^^^^^^ >  > Somebody run short a byte?  ; Good to see you are doing your thing for the E-con-ome Bob.   , Or it is all part of the e bubble popping...   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  # Date: Fri, 23 Mar 2001 20:47:17 GMT & From: "john nixon" <jnixon@cfl.rr.com> Subject: dsn: Message-ID: <pDOu6.122$086.126495@typhoon.tampabay.rr.com>  @ I tried to download an ECO over DSN, but I got an error message:  , %DSN-E-CSCNOREC, requested record not found,  K I wasn't sure exactly which record wasn't found, so I sent the dsn customer F assistance desk (dsn%cad) a mail message asking for help.  The message$ bounced due to :     you guessed it?  + %DSN-E-CSCNOREC, requested record not found   G I think I can remember my access number, but I can't remember the phone ) number to call.  Can anyone help me here?   H By the way, I am trying to find all ECOs that I need prior to installingE Fibre Channel on my VMS 7.2-1 GS140.  I have most of the ECOs already K applied, but I never installed any fibre channel ECOs.  I sent in a request J to DSN a few weeks ago asking for advice on which ECOs to apply in advanceH of adding fibre channel, but the answer said none specifically for fibreK channel, but he recommended   vms72_hardware_v0100  (which I think has been L put on hold, but I dont think we need anyway.  Can anyone help me with these
 two probs?   thanks,     John    ------------------------------  % Date: Fri, 23 Mar 2001 21:13:05 +0000 5 From: "Steeples, Oliver" <Oliver.Steeples@compaq.com>  Subject: RE: dsnN Message-ID: <F498D199EDB12D468CD2C66680D308018B1326@reoexc04.emea.cpqcorp.net>  J For the DSN prob you will have to contact your local Compaq office as they7 should be able to sort it out.  As for the fibre patch: % dec-axpvms-vms721_fibre_scsi-v0400--4    Available from: 3 ftp://ftp.service.digital.com/public/vms/axp/v7.2-1   D This is a must if fibre-channel is being used, needs sys and pcsi as	 pre-req's    Regards, 	Oliver        -----Original Message-----+ From: john nixon [mailto:jnixon@cfl.rr.com] $ Sent: Friday, March 23, 2001 8:47 PM To: Info-VAX@Mvb.Saic.ComV Subject: dsn    @ I tried to download an ECO over DSN, but I got an error message:  , %DSN-E-CSCNOREC, requested record not found,  K I wasn't sure exactly which record wasn't found, so I sent the dsn customer F assistance desk (dsn%cad) a mail message asking for help.  The message$ bounced due to :     you guessed it?  + %DSN-E-CSCNOREC, requested record not found   G I think I can remember my access number, but I can't remember the phonec) number to call.  Can anyone help me here?N  H By the way, I am trying to find all ECOs that I need prior to installingE Fibre Channel on my VMS 7.2-1 GS140.  I have most of the ECOs already K applied, but I never installed any fibre channel ECOs.  I sent in a requestbJ to DSN a few weeks ago asking for advice on which ECOs to apply in advanceH of adding fibre channel, but the answer said none specifically for fibreK channel, but he recommended   vms72_hardware_v0100  (which I think has beenuL put on hold, but I dont think we need anyway.  Can anyone help me with these
 two probs?   thanks,     Johne   ------------------------------  % Date: Fri, 23 Mar 2001 22:33:03 +0100 7 From: Martin Zinser <zinser@sysdev.deutsche-boerse.com>M Subject: Re: dsn2 Message-ID: <3ABBC10F.FAB1A44A@sysdev.exchange.de>   Hello!   john nixon wrote:s > J > By the way, I am trying to find all ECOs that I need prior to installingG > Fibre Channel on my VMS 7.2-1 GS140.  I have most of the ECOs already M > applied, but I never installed any fibre channel ECOs.  I sent in a request L > to DSN a few weeks ago asking for advice on which ECOs to apply in advanceJ > of adding fibre channel, but the answer said none specifically for fibreM > channel, but he recommended   vms72_hardware_v0100  (which I think has been-N > put on hold, but I dont think we need anyway.  Can anyone help me with these > two probs?  B Check out http://www.openvms.compaq.com/openvms/fibre/. This lists	 relevant 9E information along with appropriate ECOs. You might want to check the hE Fibre channel firmware fix too (We were one of the sites experiencingp this problem, not really funny :-(n   				Greetings, MartinR   ------------------------------  % Date: Fri, 23 Mar 2001 12:21:38 -0800o0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>Y Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS     EducationalProgram., Message-ID: <3ABB3FD2.21850A12@Mvb.Saic.Com>  E I've been out of touch with news for a couple of days so I'm a little F bit behind in responding to comments to my posting, pleae bear with me< (I've still got over 500 Info-VAX messages to wade through). ...0   Brian Wheeler wrote: > . > In article <3AB62A8A.4233172D@mvb.saic.com>,< >         Mark Berryman <Mark.Berryman@Mvb.Saic.Com> writes: > > Brian Wheeler wrote: > >> . > >> . > >> .R > >> Unix is as easy to administer.  You have not shown any cases where it is not. > >C > > Case 1:  > >iI > > I just received a report on one of my VMS systems of an unrecoverablerI > > error on a disk block.  Information from the report gave me the drivenK > > and block number in question, the application that generated the error,s$ > > and the file with the bad block. > >lL > > A short time ago I discovered in the syslog of a Solaris system a reportD > > of 3 bad blocks.  Device and block numbers were given.  NO otherG > > information available.  Contacted Sun support, it turns out SolarisSL > > provides no mechanism for telling which file is using a given block, norJ > > is it capable of being configured to tell me which application or file  > > triggers a bad block report. > >mH > > How do I find out which file(s) have bad data in them so they can be > > corrected? > >  > K > Under linux (using ext2) its possible (and quite easy) using debugfs, andW- > assuming the bad block in question is 5323:w >  > # debugfs /dev/hda1r6 > debugfs 1.18, 11-Nov-1999 for EXT2 FS 0.5b, 95/08/09 > debugfs:  icheck 5323n > Block   Inode number > 5323    30 > debugfs:  ncheck 30l > Inode   Pathname > 30      /vmlinux-2.2.17-14
 > debugfs: > H > So its not a Unix issue, per se, but a Solaris specific issue.  Though> > there's no technical reason why it can't be done on Solaris.  E There is a technical reason, Solaris hasn't support icheck for a longrG time and no replacement was ever given.  As I said in my post, even Sun 1 Support said "sorry, there is no way to do this".o  E But you have proven one of the points of this thread - when comparingrH the difference of managing VMS and managing Unix the first question that must be asked is: Which Unix?-   > > Case 2:- > > I > > An application, as part of its normal operations has put about 10,000SK > > files in a small directory tree.  I now need to clean out about half of9I > > them.  All permutations using wildcards fail (the string generated by72 > > the shell is too long and the command aborts). > >b  > > How do I remove these files? > M > Using bash (though I'm other shells have very similar functionality), I didm	 > a test:  >  > $ ls | wc -l6 >   10002              <-- # of files in the directory > $ ls *' > bash: /bin/ls: Argument list too longs > $ for n in *; do rm $n; done > $ ls | wc -l	 >       0i  D Please note that I stated that I only wanted to delete *some* of theC files, not all of them.  That said, while providing a means for the F administrator to write code to make his/her life easier is certainly aE good thing, *requiring* the administrator to write code to accomplishTE the job at all hardly qualifies as "ease of administration", wouldn'tn you say?   > > Case 3:s > >tJ > > A Solaris system I administer started giving very poor performance.  IF > > tracked it down to one user whose program was taking all available > > memory.e > >oI > > How did I find this out (or, perhaps a better question, how would youD > > find this out)?c > >t" > > What can I do to prevent this? > >N > M > The script (or shell) which starts the program needs to use ulimit to limitnK > memory usage.  Sample output of "ulimit -a" shows the default settings on 
 > solaris:' > core file size (blocks)     unlimitedn' > data seg size (kbytes)      unlimited-' > file size (blocks)          unlimited ! > open files                  256a  > pipe size (512 bytes)       10" > stack size (kbytes)         8192' > cpu time (seconds)          unlimitedc" > max user processes          2901' > virtual memory (kbytes)     unlimited6 > O > Once the limit is lowered by the superuser for a process, it cannot be raisedt2 > by a normal process (though, it can be lowered).  G Ah yes, ulimit.  A feature implemented in the shell, and only supportedyH by certain shells (csh uses the limit command, for example).  Each shellF also uses a different syntax even when they use the same command name.  @ Since this is implemented by the shell, this also means that theA sysadmin must write a separate system-wide login profile for each_B different shell supported on the system and then must come up withG his/her own method of how to implement and maintain a database of whicht% values are to be given to which user.u  $ Again, this is "ease of management"?   > > Case 4:a > >tI > > One user has an application he is working on that needs to be able too > > open a privileged TCP port.y > >fK > > How do I give him the ability to do so without giving him access to theo > > entire system? > M > Ok, this is trickier.  I got 75% of them :)  Seriously, though, I think youhL > can do it with the capability patches to linux, but I've never had to deal
 > with it.  B You can do anything with linux, it comes in source form and can be	 modified.s  H Since I work in an environment that does not consider linux as valid forB the enterprise please provide an answer that will work on Solaris, HP/UX, AIX, Tru64, etc.i  K > Out of curiosity, is there an easy way to delete a directory structure on 9 > vms without having to run "DELETE *.*;*" over and over?.     Certainly, a single command:  " $ delete [...]*.*;*,;*,;*,;*,;*,;*  D Add as many ",;*" to the end of the command as you like depending on< potential directory depth, it doesn't hurt to have too many.  
 Mark.Berrymanh Mark.Berryman@Mvb.Saic.Com   ------------------------------  % Date: Fri, 23 Mar 2001 12:29:42 -0800 0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>Y Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS     EducationalPrograma, Message-ID: <3ABB41B6.2AA2F5E3@Mvb.Saic.Com>   Bill Gunshannon wrote: > 5 > In article <9987av$n5a$1@flotsam.uits.indiana.edu>,n/ >  bdwheele@indiana.edu (Brian Wheeler) writes:t > |> > |> > Case 2: > |> >L > |> > An application, as part of its normal operations has put about 10,000N > |> > files in a small directory tree.  I now need to clean out about half ofL > |> > them.  All permutations using wildcards fail (the string generated by5 > |> > the shell is too long and the command aborts).2 > |> ># > |> > How do I remove these files?h > |>P > |> Using bash (though I'm other shells have very similar functionality), I did > |> a test: > |> > |> $ ls | wc -lv9 > |>   10002              <-- # of files in the directoryB > |> $ ls *r* > |> bash: /bin/ls: Argument list too long! > |> $ for n in *; do rm $n; doneg > |> $ ls | wc -le > |>       0 > G > I read this as a case where he did not want to just delete everythingnD > in the directory.  If that were the case, it is trivial to do in a > shell independant manner.f >  $ ls | wc -lT7 >    10002              <-- # of files in the directory " >  $ find . -type f -exec rm {} \; >  $ ls | wc -lu >    0  D A good answer, and one I've used on occasion.  However, find suffersF from what I consider to be a fatal flaw - it cannot select files basedH on creation date, only various modification dates.  I need to be able toH select files based on both a wildcard pattern and a creation date ealier than a given date.  C What would be your solution in this case?  (and remember, we're notn8 talking "can it be done" but "how easy is it to do it".)  eF > One of the strengths of the software tools concept (on which much ofI > unix is based) is that there is usually more than one way to accomplishl4 > a task and a well stocked toolbox is a good thing.  F The problem I have with this is that I get tired of having to take outD that "well stocked toolbox" and building my own solution each time IG encounter an issue.  It would be nice if the answer was already there. -3 Flexibility is great but it can be carried too far.   * You may, of course, feel very differently.  
 Mark Berryman- Mark.Berryman@Mvb.Saic.Com   ------------------------------  % Date: Fri, 23 Mar 2001 13:05:39 -0800V0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>Y Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS     EducationalProgramo, Message-ID: <3ABB4A23.1EE00BE0@Mvb.Saic.Com>   Bill Gunshannon wrote: > . > In article <3AB62A8A.4233172D@Mvb.Saic.Com>,5 >  Mark Berryman <Mark.Berryman@Mvb.Saic.Com> writes:- > |> Brian Wheeler wrote:l > |> > . > |> > . > |> > .T > |> > Unix is as easy to administer.  You have not shown any cases where it is not. > |> > |> Case 1: > |>J > |> I just received a report on one of my VMS systems of an unrecoverableJ > |> error on a disk block.  Information from the report gave me the driveL > |> and block number in question, the application that generated the error,% > |> and the file with the bad block.E > |>M > |> A short time ago I discovered in the syslog of a Solaris system a report E > |> of 3 bad blocks.  Device and block numbers were given.  NO other H > |> information available.  Contacted Sun support, it turns out SolarisM > |> provides no mechanism for telling which file is using a given block, nor K > |> is it capable of being configured to tell me which application or filea! > |> triggers a bad block report.  > |>I > |> How do I find out which file(s) have bad data in them so they can bem > |> corrected?< > C > Never had to do it, but I am pretty sure ther eis a way to do it.DG > I would expect that most administrators don't learn how to do obscure E > tasks till the time comes when they actually have to, regardless of,	 > the OS.   D I don't consider finding a file with a bad block obscure, on non-VMSD systems I have to do it several times a year (YMMV).  You say you'reD pretty sure there is a way to do it when I explicitly indicated thatH even Sun Support said there wasn't a way do it.  I'm looking for someoneH who actually knows a way to do this so we can compare it to VMS since we* are in an "ease of management" discussion.   > |> > |> Case 2: > |>J > |> An application, as part of its normal operations has put about 10,000L > |> files in a small directory tree.  I now need to clean out about half ofJ > |> them.  All permutations using wildcards fail (the string generated by3 > |> the shell is too long and the command aborts).a > |>! > |> How do I remove these files?n > C > Insufficient data provided.  How about giveing me the command yousE > would use on VMS.  You claim that "All permutations using wildcards.D > fail".  If you can't use wildcards in VMS, how do you specify themE > without explicitly listing all the files, which unix can do just as  > well.b > @ > Provide an accurate description of a real-world case and I can > probably answer this one.   F Ok, I will clarify and provide a specific example.  (I originally leftE the description somewhat vague because I was looking for *any* answerrE that could address the problem of what you do on unix when a wildcard 1 selects too many files to fit on a command line).e  E I did not say that the wildcards failed on VMS, I said they failed on,B Unix because the shell tried to create a command line that was too1 long.  The wildcard operation worked fine on VMS.r   Here's a real-world case:d  C In a small directory tree containing over 10,000 files - delete allpE files matching the wildcard zy%*yz.% (must start with the letters zy, H must have at least 3 letters after zy and before the period and the lastH two of those letters must be yz, must have exactly 1 character after theF period) and the file must have been created before 1-jan-2001.  (Don't4 ask me, I didn't design the file-naming convention).  D It does not matter if the directories themselves are deleted or not,2 they will be reused if present and created if not.  F The command I use on VMS is: $delete/before=1-jan-2001 [...]zy%*yz.%;*  8 The method on Unix that is compartively just as easy is?   > |> > |> Case 3: > |>K > |> A Solaris system I administer started giving very poor performance.  ItG > |> tracked it down to one user whose program was taking all availablea > |> memory. > H > Big deal.  I saw this on a Univac Mainframe 20 years ago.  The correctI > answer is to smack the programmer inthe head and tell him to stop beingyI > stupid, but you probably want a technical solution rather than a social  > one. >  > |>J > |> How did I find this out (or, perhaps a better question, how would you > |> find this out)? > E > One of the first things I do when I detect a performance problem ishF > run "top".  Now, before you say this isn't "standard" unix (whateverD > that is) "top is just a program that formats kernel data in a moreG > user friendly way.  All the same information is available from commoneG > utilities.  It would immediately show who the hog is and what programe > he was running.e  D Which utility will show me how much memory a user has and break thisD down to how much is private to his heap, and how much is shared with other processes?   >  > |># > |> What can I do to prevent this?a > < > Pretty much the same way you would on VMS. Process Limits.  @ Again, how to you manage the process limits and how easy is that management compared to VMS?@   > |> > |> Case 4: > |>J > |> One user has an application he is working on that needs to be able to  > |> open a privileged TCP port. > |>L > |> How do I give him the ability to do so without giving him access to the > |> entire system?i > G > Again, too vague to answer.  While developing I would guess you stickuD > him on a development machine where he can't do any harm.  Once theC > application is ready for production, it is easy to start a single H > process with the priveledges necessary to do this. (Hint: news servers( > do it all the time to access port 119)  G Vague? I see nothing vague about it.  The question was "how do you give F a user the ability to open a privileged port without giving him access to the entire system?".e  G Putting each developer on his own system is not a practical answer (noto in my case anway).    F > But, considering that there is no TCPIP stack that is a part of VMS,2 > it seems that this is impossible under VMS.  :-)  E Every VMS system I've bought has come with an IP stack and there is af1 trivial answer to my question with that IP stack.   H It is just as trivial with any other of the IP stacks available for VMS.  , I think you're trying to dodge the question.  
 Mark Berrymanb   ------------------------------  % Date: Fri, 23 Mar 2001 16:35:21 -0500e5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com>cY Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS     EducationalProgramb1 Message-ID: <slPu6.115$fB6.4077@news.cpqcorp.net>   / Yes.  Hard Links, Soft Links, and Mount Points.s        Bob Koehler wrote in message ...@ >In article <4Py5Os95U9V=AErH3PPNb55kFc8D@4ax.com>, David Beatty, <David.Beatty@sasSPAMITUPTHEZZZ.com> writes: >>@ >> I rarely use aliases that much, so I don't know the answer to9 >> the following question: can alias pount across mountedo9 >> devices?  Unix will allow this -- probably because then> >> implementation of the Unix file systems pretty much require >> that ability. >> >aF >Aliases and hard links won't, but it looks like soft links are comingA >with COE and that's what UNIX uses to point across mount points.n >iI >I wonder if we're getting mount points with COE?  Or just simulating allt, >device names in / like the porting library? >tG >----------------------------------------------------------------------n@ >Bob Koehler                     | Computer Sciences Corporation> >NASA GSFC Flight Software       | Federal Sector, Civil GroupF >                                | please remove ".aspm" when replying   ------------------------------  + Date: Sat, 24 Mar 2001 00:18:46 +0100 (CET)a: From: "Gotfryd Smolik, VMS lists" <gotfryd@stanpol.com.pl>K Subject: re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS Educatio J Message-ID: <Pine.LNX.4.21.0103240002200.12725-100000@irys.stanpol.com.pl>  ' On Thu, 22 Mar 2001, Nigel Arnot wrote:.  B +> Question: Can you pipe a list of filenames into DELETE on VMS?? [...] O +Rhetorical? Ans: No, because piping isn't the "VMS way". On VMS, you typicallyt2 +write (say) DELETE.TMP containing lines formatted +$ DELETE file1e +$ DELETE file2  u +...$ +and them invoke it with @DELETE.TMP  ? ...noting the fact, that (with all DCL limitation) you can have   file *list* on delete command:e# $ DELETE X.1,TST.BIN,*.OLD,*.COM;-1   ?  The second meaning of "list" and point to "piping" may suggesth2 a file list - and a little command procedure with: $ READ/END=u  and DELETE the_read_name   I +or you write a procedure containing F$SEARCH and F$FILE_ATTR (which haveoF +no direct Unix equivalents, find being the closest) to locate delete : +targets and then feed those directly to DELETE commands.   J  Erm... f$file suggest more than a "piped list" can do :), IMHO of course.  J +I'm not saying that this is better any more than I'm saying that the UnixG +approach is. I manage both VMS and Linux, and there are times I wish I H +had BASH and friends on VMS, also times I wish I had DCL and friends on +Linux.   6 FYI: a limited DCL free product is available on Linux.  K +One advantage of the VMS way is debugging. Creating the DELETE.TMP file is  +a harmless operation.  
  IMHO not.$ 1. Create the file list, b.ex. with:F $ DIR/NOHE/NOTRAI/SEL=SIZ=MIN=100/SIN=-0:30 [-...*OLD*]/OUT=DELETE.TMP 2. $ EDIT/EDT DELETE.TMPu *CI [PF1 PF3 RETURN RIGHT"." RETURN $ DELETE SPACE ENTER PF1 value PF1 ENTER]  (and modify 1. line+exit)F  A  Och, any prefferred way to change all record separator to record ( separator plus "$ DELETE " is allowed :)  7 + You can then TYPE it and check that the list of files E +is sensible before invoking it. In contrast, if you feed find outputhH +directly to rm, you are doing the equivalent of looking down the barrel +of a loaded gun.o    -:)  Regards - Gotfryd   -- rE =====================================================================,F $ ON F$ERROR("LANGUAGE","ENGLISH","IN_MESSAGE").GT.F$ERROR("NORMAL") - 		THEN EXCUSE/OBJECT=MEp. $!                        GS@stanpol.zabrze.plE =====================================================================b   ------------------------------  % Date: Sat, 24 Mar 2001 00:12:45 +0000 ) From: Christof Brass <brass@infopuls.com>0V Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMS EducationalProgram), Message-ID: <3ABBE67D.99E24172@infopuls.com>   Brian Wheeler wrote: > - > In article <3ABA9A86.E75BFE9@infopuls.com>, 5 >         Christof Brass <brass@infopuls.com> writes:h > > Brian Wheeler wrote: > >>1 > >> In article <3AB925E1.1C7BA213@infopuls.com>, 8 > >>         Christof Brass <brass@infopuls.com> writes: > >> > Brian Wheeler wrote:h > >> >>: > >m
 > > [SNIP] > >s > >> >2 > >> >> > WINE requires some sort of Windows also. > >> >>eP > >> >> No, it does not.  It is a reimplementation of windows.  Original windowsN > >> >> code is only required if you're emulating a windows binary....and only2 > >> >> then if the WINE libraries are incomplete. > >> >G > >> > Yup. Learned that in between. What do you mean with "emulating arF > >> > windows binary"? Are talking about a genuine M$ tool that comes! > >> > with Windoze like the CLI?I > >>R > >> On x86 machines you can run a windows .exe file as if it were native, becauseL > >> of environment emulation routines (mainly LDT and GDT tinkering) in theT > >> kernel.  The wine version of the dlls will be loaded unless they are overriddenI > >> (by the user) to use the native windows ones, or if there is no winer& > >> equivalent (such as vbrunx00.dll) > > C > > Sorry, don't understand that. Is WINE part of the Linux kernel?n > N > No, it uses facilities in the linux kernel to do things which most processesQ > do not:  setup LDTs.  They're useless to anything _but_ emulating the structure 1 > of another environment, such as DOS or Windows.t > A > > What does it mean "as if it were native"? Can you install thecB > > original Windoze DLLs or leave the WINE versions of these DLLs
 > > in place?  > L > Either way.  The wine software loads the windows.exe into memory, and doesK > fixups for the dlls.  For each of the dlls required (or requested later),uM > wine makes a choice:  use the 'real' dll, or use the wine-supplied dll.  ByhP > default it uses the wine-supplied, but can be overridden.  So, for all intentsK > and purposes, the windows .exe file is being run as a native application:hH > there is no emulation, other than in initial setup and library fixups.  > Exactly what I thought how it should work. I misunderstood the@ English "as if it were native" as expressing it weren't native -? native as beeing executed without emulation. That's why I don'tu2 like the use of the word "emulation" further down.  % > >> >> > AFAIK from a sound PR eventoP > >> >> > organised by SUN to promote Solaris/386 you even have to buy a WindowsN > >> >> > licence - though I'm not sure about this because this was about 1990O > >> >> > or even earlier. One of the best technical PR events I ever met. OnlyyS > >> >> > a few SUN people but highly qualified, sound answers, very good concepts.v > >> >> > Alas UNIX crap ... > >> >>o2 > >> >> WABI required a windows license, not Wine. > >> >B > >> > Okay then - as far as I understood the concept is the same:G > >> > offering the Windoze API on the host platform. Maybe the licenceo< > >> > requirement comes from a contract between SUN and M$. > >>R > >> WABI does binary emulation _only_.  Wine does binary emulation _and_ provides: > >> a source API to provide building native applications. > >eA > > What is a source API? If I have the header files of Windoze IhD > > can write an application for WABI because all other Windoze appsB > > that run under WABI are written that way. What is specifically9 > > needed for building native ("native" wrt what?) apps?  > I > Applications which run under WABI are _strictly_ x86 windows EXE files.:M > There is no other choice.  That is what it emulates, and if you wish to runa4 > something under it, you must generate that format.  @ This is a special use of the word "emulate" but it might well be correct.  L > Wine, on the other hand, will _allow_ those files to be "emulated" (thoughP > its only an environment emulation, the code is run natively on the processor),M > but also provides winelib, which is a native (I.E. linux ELF shared object) O > library which allows you to recompile your 'windows' source code into a linux>- > ELF executable with the same functionality.) > ? > So, if you had "hello.c" which was windows source, you could:s > 
 > Under wabi: C >         compile with a windows compiler and then run the x86 .exe  > 
 > Under wine: C >         compile with a windows compiler and then run the x86 .exerP > or      compile with a native compiler (i.e. gcc) and run the linux executable > 	 > Get it?- > Brian   ; Got it completely. It is plausible and in congruence what Is; thought. I didn't know and didn't think about the option toa; generate a Linux (or Solaris which isn't possible) binary - @ probably because these techniques have been developed to offer a@ way to run programs on Linux/Solaris of which the sources aren't around.o: Very good explanation anyway. The concept is convincing. I7 wonder why it isn't wider spread. VMware is widely used @ (although offering a lot of different possibilities) for running Win apps on Linux. Thanks.e   ------------------------------  % Date: Fri, 23 Mar 2001 23:36:59 +0000n) From: Christof Brass <brass@infopuls.com>oY Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducationalProgram) Pro + Message-ID: <3ABBDE1B.98453C3@infopuls.com>o   Paul Repacholi wrote:w > - > Christof Brass <brass@infopuls.com> writes:a > F > > Will this also remove all directories? I remember sometimes havingG > > the directories left. I normally use backup to move files to retaini > > the modification date. > K > No. But you do not always want to nuke the directories and go through the G > overhead of re-creating and extending them. The delete get rid of thet0 > directories VERY fast once the files are gone. >  > --> > Paul Repacholi                               1 Crescent Rd.,9 > +61 (08) 9257-1001                           Kalamunda.:B >                                              West Australia 60760 > Raw, Cooked or Well-done, it's all half baked.  8 Do you mean the "$ DELETE [...]*.*;*"? On my system this> complains about deleting a directory which isn't empty because there is a directory in it.c   ------------------------------  % Date: Fri, 23 Mar 2001 23:41:54 +0000W) From: Christof Brass <brass@infopuls.com>.Y Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducationalProgram) Pro>, Message-ID: <3ABBDF42.2646B6AB@infopuls.com>   Mark Berryman wrote: >  > Bill Gunshannon wrote: > >r7 > > In article <9987av$n5a$1@flotsam.uits.indiana.edu>,t1 > >  bdwheele@indiana.edu (Brian Wheeler) writes:p > > |> > > |> > Case 2: > > |> >N > > |> > An application, as part of its normal operations has put about 10,000P > > |> > files in a small directory tree.  I now need to clean out about half ofN > > |> > them.  All permutations using wildcards fail (the string generated by7 > > |> > the shell is too long and the command aborts).f > > |> >% > > |> > How do I remove these files?  > > |>R > > |> Using bash (though I'm other shells have very similar functionality), I did > > |> a test: > > |> > > |> $ ls | wc -li; > > |>   10002              <-- # of files in the directory 
 > > |> $ ls * , > > |> bash: /bin/ls: Argument list too long# > > |> $ for n in *; do rm $n; done  > > |> $ ls | wc -ln > > |>       0 > > I > > I read this as a case where he did not want to just delete everything F > > in the directory.  If that were the case, it is trivial to do in a > > shell independant manner.t > >  $ ls | wc -l 9 > >    10002              <-- # of files in the directoryh$ > >  $ find . -type f -exec rm {} \; > >  $ ls | wc -lv > >    0 > F > A good answer, and one I've used on occasion.  However, find suffersH > from what I consider to be a fatal flaw - it cannot select files basedJ > on creation date, only various modification dates.  I need to be able toJ > select files based on both a wildcard pattern and a creation date ealier > than a given date. > E > What would be your solution in this case?  (and remember, we're notl: > talking "can it be done" but "how easy is it to do it".) > H > > One of the strengths of the software tools concept (on which much ofK > > unix is based) is that there is usually more than one way to accomplish 6 > > a task and a well stocked toolbox is a good thing. > H > The problem I have with this is that I get tired of having to take outF > that "well stocked toolbox" and building my own solution each time IH > encounter an issue.  It would be nice if the answer was already there.5 > Flexibility is great but it can be carried too far.y > , > You may, of course, feel very differently. >  > Mark Berrymano > Mark.Berryman@Mvb.Saic.Com  = While I like flexibility where it is appropriate I agree thatu; wrt known problems and objects this need for flexibility inn? simple tasks turns out to be a consequence of the fact that theV: implementors didn't think through. "ls", "rm" and "mv" are@ excellent examples as they all three deal with files. The number< and type of file attributes is fixed. Why not offer a simple; mechanism to select and sort according to these attributes?    ------------------------------  % Date: Fri, 23 Mar 2001 23:52:57 +0000t) From: Christof Brass <brass@infopuls.com>eY Subject: Re: Dumbing Down VMS with UNIX Elements? (was Re: OpenVMSEducationalProgram) Prot, Message-ID: <3ABBE1D9.988F9A5C@infopuls.com>   Mark Berryman wrote: > G > I've been out of touch with news for a couple of days so I'm a little H > bit behind in responding to comments to my posting, pleae bear with me> > (I've still got over 500 Info-VAX messages to wade through).  # Congratulations to your first shot!z  
 [BIG SNIP]   > Certainly, a single command: > $ > $ delete [...]*.*;*,;*,;*,;*,;*,;* > F > Add as many ",;*" to the end of the command as you like depending on> > potential directory depth, it doesn't hurt to have too many.   This is the hit!   >  > Mark.Berrymanl > Mark.Berryman@Mvb.Saic.Com  : I admire your efficiency in destroying the "UNIX is better- because it's simpler and more flexible" tale.    ------------------------------  % Date: Sat, 24 Mar 2001 01:26:49 -0500 ' From: Jim Becker <jbecker@ui.urban.org>s Subject: Re: listserv programt, Message-ID: <3ABC3E29.7FB628F5@ui.urban.org>   Dave Pampreen wrote: > N > Is there a listserv program which runs on Alpha OpenVMS whcih users can view& > old messages (archives) via the web? >  > Dave  D LISTSERV (produced by L-Soft: http://www.lsoft.com/) runs on OpenVMSD and other platforms. It supports searchable list archives with a web
 interface.   --
 Jim Becker+ The Urban Institute (http://www.urban.org/)g' Encompass (http://www.encompassus.org/) 5 ESILUG (http://eisner.encompasserve.org/lugs/esilug/)r   ------------------------------  % Date: Sat, 24 Mar 2001 00:02:03 +0000>) From: Christof Brass <brass@infopuls.com> Y Subject: Re: New qualifiers for some DCL commands (was: Re: Dumbing Down VMS withUNIX Elee, Message-ID: <3ABBE3FB.9403C071@infopuls.com>   Carl Perkins wrote:  > 5 > jordan@lisa.gemair.com (Jordan Henderson) writes...hD > }Oh?  Delete offers the "necessary selection criteria"?  Let's sayD > }I wanted to perform the very commonly requested deleting files in? > }reverse order?  How do I do that with delete?  In Unix it's:t > }w > }       ls -r1 | xargs rm  > @ > DELETE, PURGE, DIRECTORY, and probably some others could use a > /REVERSE qualifier.l > < > Even better might be a more comprehensive /SORT qualifier. > @ > With a /SORT qualifer you could have things along the lines of > $ DIR/SORT=KEY=(SIZE,ASCEND) > or" > $ DIR/SORT=KEY=(CREATED,DESCEND) > or > $ DIR/SORT=SPEC=myspec.spc > or, of course,8 > $ DELETE/SORT=(KEY=(NAME,DESCEND),KEY=(TYPE,DESCEND),-" >     KEY=(VERSION,DESCEND)) *.*;*A > (perhaps a key called "filespec" or something along those linesoE > to get the full spec including name, type, and version in one key).t > F > After all, there are callable sorting routunes are there not? AddingF > sort functionality should therefore, in theory, not be all that hardF > (which is easy to say since I'm not the one who would be adding it). > @ > }Or, say I want to delete empty .tmp files, which is something > }I've wanted to do.e
 > }That's: > } @ > }       ls -s1 *.tmp |  awk '/^ *0 / { print $2 };' | xargs rm > }0J > }So, I must be missing the options on DEL that provides an way to do the > }above easily. > }-Jordan Henderson > E > One of the good things people mention about DCL is it's regularity.fC > To help promote this, and because it would be useful, I'd like to C > suggest that DELETE and PURGE and possibly some others could also9D > use the /SELECT= qualifier that was added to the DIRECTORY command/ > sometime around VMS V6.1. You could then do a % > $ DELETE/SELECT=SIZE=MAX=0 *.TMP;*.o > 
 > --- Carl  @ Very good suggestions! Exactly what is in line with the VMS way.= Instead of postponing the sorting and selecting to other moreU= complicated commands it's done where it can be done best. The > mentioned commands all know exactly with which data they deal.   ------------------------------  % Date: Fri, 23 Mar 2001 23:20:16 +0000 ) From: Christof Brass <brass@infopuls.com> , Subject: Re: POSIX Streams, File Permissions, Message-ID: <3ABBDA30.110E9DAF@infopuls.com>   Dave Weatherall wrote: > C > On Fri, 23 Mar 2001 01:00:18, Christof Brass <brass@infopuls.com>  > wrote: > & > > Shane.F.Smith@Healthnet.com wrote: > > > L > > > Did I understand right? Christof just asked for a Unix feature on VMS? > > >" > > > Shanee > > > J > > > Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> on > > > 03/22/2001 02:57:16 AM > > >s! > > > To:   Info-VAX@Mvb.Saic.Com:	 > > > cc:g > > >M3 > > > Subject:  Re: POSIX Streams, File Permissionss > > > 1 > > > Christof Brass <brass@infopuls.com> writes:p > > >pG > > > > And I like to have the possibility of case sensitive file namespD > > > > but the change is only half way. It only remembers the firstF > > > > writing wrt case. It doesn't allow to have two files named "a"E > > > > and "A". I would have liked another option to make it as some  > > > > other OSs wrt case.o > > >iL > > > AFAIK, what you describe is ODS-5 extended file name parsing, which is > > > case-lN > > > preserving and case-blind. What COE gives you is real, stupid Unix/Posix > > > case-tP > > > sensitive file naming, i.e., "a" and "A" are different files. (This is one > > > of@ > > > the few things MS's newer file systems got right, IMNSHO.) > > >i > > >      Jan > >IB > > Not specifically. I only want to have the option to be able toB > > exactly use the names as they are wrt case and be different if@ > > only one character is different. That UNIX crap may have the% > > same feature is of no importance.t > >nC > > But why "real, stupid" ("Unix/Posix case-sensitive file name")?c> > > Is it stupid to make a difference between "MAC" and "Mac"? > E > Chris it does sound a little incongruous for someone who dislikes C G > and Unix to imply that they consider case sensitivity in names (ala Ce > and Unix) a 'good' thing :-) > B > I don't think it a good thing but then again I do Fortran  for a	 > living.C > F > Which prompts the thought, if Unix commands are terse to save typingE > why is it that you  _need_  two fingers to type commands? (answer -RF > you do only need one - you use the Caps Lock and not Shift Lock...). >  > Cheers - Dave.  > Dislike is a very friendly word and a strong understatement to say the least.; Again: I like case sensitivity in filenames sometimes but Ie@ don't like all sorts of characters in filenames. But what I like8 is not important because only the case sensitivity has a? technical background. E.g. there are programming languages thatn; take Mac and MAC as beeing different. To have the option to=. adjust this per directory may be very usefull.   ------------------------------  % Date: Sat, 24 Mar 2001 00:18:06 +0000	) From: Christof Brass <brass@infopuls.com> , Subject: Re: POSIX Streams, File Permissions, Message-ID: <3ABBE7BE.FA3BF31E@infopuls.com>   Carl Perkins wrote:- > ' > Shane.F.Smith@Healthnet.com writes...:I > }Did I understand right? Christof just asked for a Unix feature on VMS?i > }w > }Shane > F > Not only did he ask for one, he asked for one of the stupidest ones. > F > Case sensitivity is completely and utterly moronic (not only in fileH > names, but in programming languages and every other place it happens). > E > The COE thingy is, of course, requiring such a feature. Fortunatelyf6 > it will only happen if you deliberately activate it. > 
 > --- Carl >   ' Yes it have to be activated explicitly. > But why is it stupid to make a difference between Mac and MAC?? Especially in a programming language in which we try to expressn; our algorithmical thoughts, objects a.s.o.. If we have thisq> disctinction in our natural scripture why not carrying forward> to our PLs and filesystems? I would rather restrict the use of> special characters which are used in natural language texts to< separate words/sentences or to express certain grammatically" important meanings like questions.   ------------------------------  # Date: Fri, 23 Mar 2001 19:16:37 GMTi= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)(B Subject: Source listing do NOT coincide with the operating system.0 Message-ID: <009F972C.5602D25E@SendSpamHere.ORG>  / I've already posed this same question to CSC...o  H I installed a ZLXp-L2 video card which has been sitting on the shelf andH booted V7.2-1 (a virgin install too).  VMS bugchecks with a BUG$_DECWIN-G DOWS bugcheck.  So, I started looking into the crash.  The crash was ineH SYS$GUBDRIVER so I pulled up the source listings and started looking.  IG found discrepancies between the code and the source listings.  Strange, I thought I.  So, I looked at the date on the .EXE -- 7-APR-1999.  The datenI on the .LIS and .MAP files is 29-MAY-1999.  Excuse me?  More than a month  and a half time difference???   G I tend to NOT install patches because I don't have listings which coin- E cide with the patched images but now I'm finding that even the virgin:H O/S code does not coincide with the listings.  Can somebody in engineer- ing explain this mess?   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMg            iO city, n., 1. a place where trees are cut down and streets are named after them.y   ------------------------------  # Date: Fri, 23 Mar 2001 19:29:46 GMTe= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)eF Subject: Re: Source listing do NOT coincide with the operating system.0 Message-ID: <009F972E.2C9EAF42@SendSpamHere.ORG>  p In article <009F972C.5602D25E@SendSpamHere.ORG>, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:0 >I've already posed this same question to CSC... > I >I installed a ZLXp-L2 video card which has been sitting on the shelf andtI >booted V7.2-1 (a virgin install too).  VMS bugchecks with a BUG$_DECWIN-lH >DOWS bugcheck.  So, I started looking into the crash.  The crash was inI >SYS$GUBDRIVER so I pulled up the source listings and started looking.  I H >found discrepancies between the code and the source listings.  Strange,J >thought I.  So, I looked at the date on the .EXE -- 7-APR-1999.  The dateJ >on the .LIS and .MAP files is 29-MAY-1999.  Excuse me?  More than a month >and a half time difference??? s >nH >I tend to NOT install patches because I don't have listings which coin-F >cide with the patched images but now I'm finding that even the virginI >O/S code does not coincide with the listings.  Can somebody in engineer-i >ing explain this mess?u >r >-- P >VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM >           P >city, n., 1. a place where trees are cut down and streets are named after them.    E FYI, WIS came back and told me a cannot log a service request.  Guessh* I'll just have to hope for an answer here.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              O city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  % Date: Fri, 23 Mar 2001 21:39:51 +0000 + From: "antonio.carlini" <arcarlini@iee.org>eF Subject: Re: Source listing do NOT coincide with the operating system.' Message-ID: <3ABBC2A7.C09AF636@iee.org>>  & "Brian Schenkenberger, VAXman-" wrote:K > thought I.  So, I looked at the date on the .EXE -- 7-APR-1999.  The date K > on the .LIS and .MAP files is 29-MAY-1999.  Excuse me?  More than a monthe > and a half time difference???   , Obviously built on an Alpha --- nothing else/ could produce a .EXE so quickly that it travelsv *back* in time :-)  3 Seriously I have no idea what actually happened ... 2 it looks like someone manually "rolled back" to an earlier image. u  . What do you get for date and id from ANA/IMAGE, and how does that correspond to what you see in the .MAP file?    Antonioy   -- e   ---------------a- Antonio Carlini             arcarlini@iee.orgm   ------------------------------  % Date: Fri, 23 Mar 2001 13:36:01 -0600t% From: "steve smith" <ssmith@lbpc.com>I$ Subject: Upgrading to alphavms 7.2-1& Message-ID: <gBNu6.44$n%4.1460@client>  I I'm planning to upgrade from alphavms 7.1-2 to 7.2-1.  Some brain cell is I ringing in my head, about run time libraries, being overwritten for basic L and other languages. Will I have to reinstall the languages after I upgrade, and install the 20 patches.V  6 Then again this brain cell might doing the death dance   -- Steve Smitha Manager Technical Services Information Technology Law Bulletin Publishing Co. 
 (312)644-7067r (312)644-4355 Faxe ndelise@LBPC.com http://www.lawbulletin.com   ------------------------------  % Date: Fri, 23 Mar 2001 14:52:58 -0500a- From: John Reagan <reagan@hiyall.zko.dec.com>t( Subject: Re: Upgrading to alphavms 7.2-11 Message-ID: <3ABB634A.F40C2B8@hiyall.zko.dec.com>h   steve smith wrote: > K > I'm planning to upgrade from alphavms 7.1-2 to 7.2-1.  Some brain cell is K > ringing in my head, about run time libraries, being overwritten for basic N > and other languages. Will I have to reinstall the languages after I upgrade, > and install the 20 patches.r  E Some of the compilers (Pascal for one, perhaps BASIC) often ship RTLs?H independently of VMS and do not use the PCSI installer.  In those cases,E its quite possible that upgrading to V7.2-1 will regress the language-F RTLs.  Now, V7.2-1 is quite new so your chances of regression might beG smaller.  For example, I know in Pascal's case, the RTL included inside:G the compiler kit is in fact the V7.2-1 RTL so no regression will really.G occur (but you didn't know that until I told you or until you found thea section in my release notes).t  G As for PCSI-based kits and patches, PCSI should take care of everything> for you.   -- : John Reagan  Compaq Pascal Project Leader   ------------------------------  % Date: Fri, 23 Mar 2001 15:57:12 -0500i  From: norm.raphael@jamesbury.com( Subject: Re: Upgrading to alphavms 7.2-14 Message-ID: <C2256A18.007242AF.00@jklh21.valmet.com>  4 I'm not that comfortable with PCSI OpenVMS upgrades. How does PCSI find the patches? ( How does it know which patches to apply?F Is this process now really automagic with PCSI-based install/upgrades?          3 reagan@hiyall.zko.dec.com on 03/23/2001 02:52:58 PM    To:   Info-VAX@mvb.saic.com  cc:t) Subject:  Re: Upgrading to alphavms 7.2-1          [Snip]  G As for PCSI-based kits and patches, PCSI should take care of everything  for you.   -- John Reagan  Compaq Pascal Project Leader   ------------------------------  # Date: Fri, 23 Mar 2001 22:36:51 GMTe2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)( Subject: Re: Upgrading to alphavms 7.2-11 Message-ID: <7eQu6.116$fB6.4165@news.cpqcorp.net>$  W In article <C2256A18.007242AF.00@jklh21.valmet.com>, norm.raphael@jamesbury.com writes:I : 5 :I'm not that comfortable with PCSI OpenVMS upgrades.   L   They're fast, they're easy, and they're easier than VMSINSTAL/KITINSTAL...  4   Once you've done a few, you'll really like PCSI.    K   If you've created few VMSINSTAL KITINSTAL procedures, you'll also really hL   like PCSI -- well, once you learn to split off the kit configuration task !   from the kit installation task.e    :How does PCSI find the patches?  E   PCSI maintains data on all objects that PCSI has installed, and allmE   objects that have been registered with PCSI.  It also maintains theoD   information necessary for the the ECO kits that apply to products.  ) :How does it know which patches to apply?g  C   With V7.1-2 and later, each object has its own generation number.iE   PCSI simply uses the most recent.  This tends to avoid regressions.   G :Is this process now really automagic with PCSI-based install/upgrades?i  
   Yes, it is.   '   Here is what I don't like about PCSI:.  F     Bugs: The PCSI kit that is available for PCSI fixes several subtleC     bugs within PCSI, provides some baseline support for newer PCSIo$     kits, and should be installed.    G     Problem: if you delete parts of the PCSI database, PCSI can forget oF     about components or otherwise become confused.  Fix: don't delete      files you don't own.  C     Problem: PCSI does not provide a version-regression capability. D     Fix: if you want to back out an OpenVMS ECO kit that uses PCSI, 8     you have to start with the base product and work up.  G   The above should not be infered as a dislike of PCSI itself nor that DD   there are particular problems with it nor that I recommend againstG   it.  PCSI works very well, and I make regular use of it.  I can (and tI   do) encourage products to use it, and I use PCSI for packages I create.tG   And I definitely find it superior to VMSINSTAL -- VMSINSTAL itself istI   nicely designed, but the environment and the flexibility that VMSINSTALgH   provides to the KITINSTAL makes for some real complexity and for some B   very hairy installation kits and some weird installation bugs...  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------    Date: 24 Mar 2001 03:03:03 +0800, From: Paul Repacholi <prep@prep.synonet.com>$ Subject: Re: [DCL] minute of the day- Message-ID: <87wv9gnw9k.fsf@prep.synonet.com>t  " Paul Sture <paul@sture.ch> writes:  F > > >The computer room staff nearly went nuts trying to figure out whyA > > >jobs were not generating any output, yet there were no errorwC > > >messages.  After a couple of dozen jobs went by, the simulatorcD > > >would drop out and return control to the real operating system,E > > >which made it just that much harder to diagnose the problem.  Of.E > > >course, if it saw one of *my* jobs, it would drop out instantly.-A > > >:-) Admittedly, this was a fairly primitive operating system C > > >(this was in 1971 or so) and rather easy to simulate, but whatr= > > >the hey.  It was a lot of laughs for not that much code.e  ? > There was a story many years ago that the operators on an ICLeE > (GEORGE ?) system were amazed to see a Happy New Year banner appear F > on the console at the start of the new year. The customer was prettyC > upset when they worked out the amount of cpu per year that simplee4 > date test soaked up - IIRC it was quite expensive.  F The best I have come across was a program called CRAMCOPY. Most of theE time, it would swallow the command, print it on the line printer, andtB copy the decks. Or it would print the command end to end reversed.  C The prize, was for it to start the very first part of the copy, and(E then proceed to type out some number of verses of Eskimo Nell. And itO6 had I think 126 or so verses hidden in its happy guts.  C Just hitting halt was not a good option, as you could end up with a*B destroyed pack. Plus it had a heap of anti tamper tricks built in.B One of the programers found the dispatch to the type out code, and= patched it out. The decks where copyied in interleaved order;e# 0-1-2-3-4 -> 1-0-3-2-5 ... Oh dear!i  E Eventually, some one found a spot where a jump disarmed the mischief. A Sad thing is, it was several orders an magnitude better code thanI' anything else that run of those things.s   -- i< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------   Date: 23 Mar 2001 18:07:28 CDT= From: wayne@tachysoft.xxx.412538.killspam.015a (Wayne Sewell) $ Subject: Re: [DCL] minute of the day. Message-ID: <xcj8HDMvddB3@tachxxsoftxxconsult>  ` In article <he4mbtoo2otfh7usbjrm46ssvm44tqapid@4ax.com>, Alan Greig <a.greig@virgin.net> writes:G > On 22 Mar 2001 20:56:56 CDT, wayne@tachysoft.xxx.412538.killspam.015a  > (Wayne Sewell) wrote:E  O >>Admittedly, this was a fairly primitive operating system (this was in 1971 oreP >>so) and rather easy to simulate, but what the hey.  It was a lot of laughs for >>not that much code.a > H > Of course now no distinction is made between a bit  of fun and genuineF > destructive damage. You'd probably get 5 years behind bars for doing
 > that today.u    H Well, you can kind of understand that attitude.  Nobody ever seems to doL harmless pranks any more.  Destruction, chaos, and theft is the norm.  These% hacker/cracker types are a nasty lot.t  N Of course, even back in 71, those guys would probably have been rather annoyedO with me if they had known I was the perpetrator.  Fortunately, they never had a F clue what was happening, much less who did it.  They thought it was anE intermittent bug in the operating system, even though that was a rare.  occurrence in the pre-billy era.     -- eO ===============================================================================cM Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxxi: http://www.tachysoft.xxx/www/tachyon.html and wayne.html  K change .xxx to .com in addresses above, assuming you are not a spambot  :-)tO =============================================================================== O Dean Wormer to Flounder: "Fat, drunk, and stupid is no way to go through life."    ------------------------------  % Date: Fri, 23 Mar 2001 18:06:17 -0800a! From: Shane.F.Smith@Healthnet.coms$ Subject: Re: [DCL] minute of the dayD Message-ID: <OF32645EA5.E0AF6136-ON88256A19.000B1EDE@foundation.com>  I You can still get away with a lot in England, largely because Brits don'ttI call in the law as readily. I once blew up a statistics lecturer in front E of an auditorium full of students....... (Before you ask, yes, he did " deserve it, no I didn't hurt him.)   Shanea          E wayne@tachysoft.xxx.412538.killspam.015a (Wayne Sewell) on 03/23/2001t 03:07:28 PM    To:   Info-VAX@Mvb.Saic.Comi cc:i  % Subject:  Re: [DCL] minute of the day     C In article <he4mbtoo2otfh7usbjrm46ssvm44tqapid@4ax.com>, Alan Greig  <a.greig@virgin.net> writes:G > On 22 Mar 2001 20:56:56 CDT, wayne@tachysoft.xxx.412538.killspam.015as > (Wayne Sewell) wrote:   G >>Admittedly, this was a fairly primitive operating system (this was inn 1971 ordE >>so) and rather easy to simulate, but what the hey.  It was a lot ofc
 laughs for >>not that much code.  > H > Of course now no distinction is made between a bit  of fun and genuineF > destructive damage. You'd probably get 5 years behind bars for doing
 > that today.     H Well, you can kind of understand that attitude.  Nobody ever seems to doE harmless pranks any more.  Destruction, chaos, and theft is the norm.i Theses% hacker/cracker types are a nasty lot.a  F Of course, even back in 71, those guys would probably have been rather annoyedrI with me if they had known I was the perpetrator.  Fortunately, they neverg had aeF clue what was happening, much less who did it.  They thought it was anE intermittent bug in the operating system, even though that was a rare   occurrence in the pre-billy era.     --O ===============================================================================B  8 Wayne Sewell, Tachyon Software Consulting  (281)812-0738 wayne@tachysoft.xxxt8 http://www.tachysoft.xxx/www/tachyon.html and wayne.htmlK change .xxx to .com in addresses above, assuming you are not a spambot  :-)hO ===============================================================================u  H Dean Wormer to Flounder: "Fat, drunk, and stupid is no way to go through life."   ------------------------------   End of INFO-VAX 2001.165 ************************