Re: odd permissions on mounted disk
On Fri, 4 Nov 2011, Andrew Z wrote: ... [az@server media]$ ls -l ./ total 2 dr-xr-xr-x 4 4294967295 4294967295 136 Oct 29 23:43 dvd Well 4294967295 is (2^32)-1 so for some reason either those are the uid/gid on the real device or something is not passing the info through to the mount... I have no idea why that then prevents access to the files underneath, unless there are places where only the owner is allowed access (e.g mode 700 directories or similar)... Can you alter the mount (or fstab since you have it mentioned there) to pass suitable uid=...,gid=... options? (see the mount man page entry for udf for lists of valid options)... On sl5 we normally rely on hal to do the actual mounts which for most users is done via the gnome volume manager - and we tweak the gconf settings for it to pass suitable options to hal (if they are not correct already). I believe that things are different in sl6. -- /\ | "Computers are different from telephones. Computers do not ring." | | -- A. Tanenbaum, "Computer Networks", p. 32 | -| | Jon Peatfield, _Computer_ Officer, DAMTP, University of Cambridge | | Mail: jp...@damtp.cam.ac.uk Web: http://www.damtp.cam.ac.uk/ | \/
Re: vlc-1.1.11-72.el5.i386.rpm
On Sat, Nov 5, 2011 at 3:48 PM, Yasha Karant wrote: > Thank you. I am (well) aware of the power of VLC, particularly compared > with a number of other offerings/applications, both open systems and > proprietary for fee. Although I would very much like to comment upon the > issues you have raised concerning the reading and use of legally purchased > and owned for-personal-use DVDs across region codes, as this list repeatedly > has stated that it is strictly technology for technicians/technologists and > devoid of any intellectual issues or discussions -- including societal > issues as required under the ACM code of ethics, I shall not take the bait. Well, it wasn't bait, I was pointing out the need for caution, not trying to make a moral judgment, and explaining why you won't find vlc's components in our favorite upstream vendor's open source or free software licensed source repositories. > I would very much appreciate it if you would provide the tools, syntax, > etc., for the needed mock build environment, as I presume that the resulting > containment is "safer" than allowing ATrpms to have free reign (and free > update requests) for a production end-user stable EL workstation. The mock package itself is available from EPEL. By default, it points to (that other rebuild from our favorite upstream vendor) repositories, which need re-arrangement to point to somewhat different names of Scientific Linux channels. Let me spend a bit of time setting it up for atrpms access. I'll post those in a separate thread. I had some issues installing it under SL 5.7, due to package version conflicts with EPEL. Under SL 6.1 on x86_64, vlc pulls in no fewer than 43 packages from atrpms. I don't think you want to take on building that youself without atrpms. Maintaing your own dependency tree will get pretty old, pretty fast.
Re: Need Fortran advice
John--you should note that gcc3.4 (same c compiler as in Sci. Linux 4) is also available in Sci. Linux 5, there is a compatible g77 as part of that. That should be able to compile anything that you compiled before. Steve On Sat, 5 Nov 2011, John Schaad wrote: FORTRAN -- For many years I have helped support, have used, and continue to use a package of software written mostly in FORTRAN for a DEC VAX Mainframe running the VMS operating system. Later, I helped develop a Graphic User Interface written mostly in C and X-Windows, and it was ported to run on DEC VAXstations (VMS), DECstations (DEC-UNIX - ULTRIX), and the PC Platform under Linux. The original port to the PC platform with Linux was with Slackware, and later Redhat 5.1 through 6.2, 7.2, and 9.0. A batch-only version was ported to run on MS-Windows using Cygnus/Cygwin. In recent years, however, we have found the most reliable Linux Platform to use is Scientific Linux -- presently the 4.X series. I have attached a PDF file with links to the IPF home page, source code, and manuals, along with some of the history, in case anyone is interested in exploring it further. In the interest of continuing development in parallel with the later versions of Scientific Linux, I have tried to compile and install the IPF software on SL-5.X and also SL-6.X. I found that with the switch to "gfortran" as the supported FORTRAN compiler, the FORTRAN code would no longer compile. It appears that there are major differences between gfortran and the FORTRAN versions used in SL-4.X and earlier versions, and it looks like we would have to re-write a lot of our FORTRAN Code, based on the compiler error messages that are displayed.That was a frustrating discovery! I am glad to see that Scientific Linux 4.X will continue to be supported for a while, at least for the near future. Does anyone have experience, and possibly some suggestions, for porting earlier FORTRAN code to be usable with the gfortran compiler? John Schaad - BPA Transmission Services http://transmission.bpa.gov/ NOTE: As a "cost cutting measure", BPA officially switched to a commercial, 3rd party, "off-the-shelf" powerflow software package, that actually took many years following that decision to be improved and upgraded sufficiently to the point where it could come close to the capability of IPF for the percentage of "successfully solved" power flow base cases. At present, continued support of IPF is on a volunteer basis by some of the original BPA developers and current users. Many utilities around the world still use IPF, due to its quality of system solutions, minimal cost, and ease of use. It is public domain software, with certain license restrictions, as described in the license on the IPF website. On Mon, Oct 31, 2011 at 6:58 PM, Todd And Margo Chester < toddandma...@gmail.com> wrote: Hi Guys, I have an extremely intelligent retired college professor for a customer. He writes doctor level text books on structural engineering. He is a sweet old duffer with an I.Q. I would guess at 200. Part of what he does in his text books is give Fortran code to go along with the extremely complicated equations in his books. He has asked me to put him together a new computer (I figure Fedora Core 15 x64 bit). I do not know squat about Fortran compilers. Question: is there a good Fortran environment I should put on his new computer? Bear in mind when you answer, that if my customer likes your suggestion, doctoral engineering students will be required to use it (tormented with it). Many thanks, -T -- -- Steven C. Timm, Ph.D (630) 840-8525 t...@fnal.gov http://home.fnal.gov/~timm/ Fermilab Computing Division, Scientific Computing Facilities, Grid Facilities Department, FermiGrid Services Group, Group Leader. Lead of FermiCloud project.
Re: odd permissions on mounted disk
On Sat, Nov 5, 2011 at 10:17 AM, Andrew Z wrote: > Akemi, thank you for response, but I'm mounting udf, not nfs. > Not sure where to look next... Sorry, I was indeed thinking only of NFS. What you are seeing is indeed odd. Looks as if the system is not finding the UID/GID and thus using that particular number. Akemi
Re: Need Fortran advice
I sometimes need to use gfortran compiler switch -std=legacy for some older codes. Jim. On 05/11/2011, at 3:57 PM, John Schaad wrote: > FORTRAN -- For many years I have helped support, have used, and continue to > use a package of software written mostly in FORTRAN for a DEC VAX Mainframe > running the VMS operating system. Later, I helped develop a Graphic User > Interface written mostly in C and X-Windows, and it was ported to run on DEC > VAXstations (VMS), DECstations (DEC-UNIX - ULTRIX), and the PC Platform under > Linux. The original port to the PC platform with Linux was with Slackware, > and later Redhat 5.1 through 6.2, 7.2, and 9.0. A batch-only version was > ported to run on MS-Windows using Cygnus/Cygwin. In recent years, however, > we have found the most reliable Linux Platform to use is Scientific Linux -- > presently the 4.X series. > > I have attached a PDF file with links to the IPF home page, source code, and > manuals, along with some of the history, in case anyone is interested in > exploring it further. > > In the interest of continuing development in parallel with the later versions > of Scientific Linux, I have tried to compile and install the IPF software on > SL-5.X and also SL-6.X. I found that with the switch to "gfortran" as the > supported FORTRAN compiler, the FORTRAN code would no longer compile. It > appears that there are major differences between gfortran and the FORTRAN > versions used in SL-4.X and earlier versions, and it looks like we would have > to re-write a lot of our FORTRAN Code, based on the compiler error messages > that are displayed.That was a frustrating discovery! I am glad to see > that Scientific Linux 4.X will continue to be supported for a while, at least > for the near future. > > Does anyone have experience, and possibly some suggestions, for porting > earlier FORTRAN code to be usable with the gfortran compiler? > > John Schaad - > BPA Transmission Services > http://transmission.bpa.gov/ > > NOTE: As a "cost cutting measure", BPA officially switched to a commercial, > 3rd party, "off-the-shelf" powerflow software package, that actually took > many years following that decision to be improved and upgraded sufficiently > to the point where it could come close to the capability of IPF for the > percentage of "successfully solved" power flow base cases. At present, > continued support of IPF is on a volunteer basis by some of the original BPA > developers and current users. Many utilities around the world still use IPF, > due to its quality of system solutions, minimal cost, and ease of use. It is > public domain software, with certain license restrictions, as described in > the license on the IPF website. > > > On Mon, Oct 31, 2011 at 6:58 PM, Todd And Margo Chester > wrote: > Hi Guys, > > I have an extremely intelligent retired college professor for > a customer. He writes doctor level text books on structural > engineering. He is a sweet old duffer with an I.Q. I would guess > at 200. > > Part of what he does in his text books is give Fortran code to > go along with the extremely complicated equations in his books. > He has asked me to put him together a new computer (I figure > Fedora Core 15 x64 bit). > > I do not know squat about Fortran compilers. > > Question: is there a good Fortran environment I should put on > his new computer? Bear in mind when you answer, that if my customer > likes your suggestion, doctoral engineering students will be required > to use it (tormented with it). > > Many thanks, > -T > >
Re: vlc-1.1.11-72.el5.i386.rpm
On 11/05/2011 11:57 AM, Nico Kadel-Garcia wrote: On Sat, Nov 5, 2011 at 2:02 AM, Yasha Karant wrote: I am attempting to install vlc-1.1.11-72.el5.i386.rpm from http://packages.atrpms.net/dist/el5/vlc/ There is a long list of dependencies missing, appended below. Is there a way to specify to get just these packages from ATrpms without "damaging" the underlying production EL 5.7 system (e.g., replacing SL packages by other packages from ATrpms that then cause problems and instability with EL 5.7)? Is there another add-on EL 5 RPM repository that requires fewer additions to stock EL 5.7 but still has a vlc 1.x version? First warning: vlc supports MPEG formats and playing DVD's. There are various patents and reverse engineering legal issues which interfere with open source and especially genuine freeware licensing or deployment of such software. This is why such tools are not in our favorite upstream vendor's codeline, nor will they be, unless such cumbersome licensing can be resolved. Since VLC can typically deal with DVD's, there's the whole libdvdcss sawsuit history that makes it unavailable for our favorite upstream vendor's core distributions. So it's *not* going to work fully without such non-Scientific-Linux provided components, unless our friends at Scientific Linux were to take that on. I don't see a point to that when atrpms and the other repositories are doiing such a good job. That said, if you're in a legal position to use these patented software tools, and you suspect some of the dependencies are extraneous, you can use "mock" to try building the SRPM in a clean SL 5.7 environment. I can send you, or the group, my /etc/mock files for using a local repo for precisely this sort of work. It's much, much, much more efficient to use "mock" from local repositories than reaching out to external mirrors, and different configs to work with JPackage or RPMforge repos as well. This keeps from cluttering your working system with a lot of libraries and dependencies that may cause other adventures. But Yasha, VLC is a powerful and flexible multimedia player, It *needs* access to those libraries in order to manage even half of the different formats and encodings it manages. Thank you. I am (well) aware of the power of VLC, particularly compared with a number of other offerings/applications, both open systems and proprietary for fee. Although I would very much like to comment upon the issues you have raised concerning the reading and use of legally purchased and owned for-personal-use DVDs across region codes, as this list repeatedly has stated that it is strictly technology for technicians/technologists and devoid of any intellectual issues or discussions -- including societal issues as required under the ACM code of ethics, I shall not take the bait. I would very much appreciate it if you would provide the tools, syntax, etc., for the needed mock build environment, as I presume that the resulting containment is "safer" than allowing ATrpms to have free reign (and free update requests) for a production end-user stable EL workstation. Yasha Karant
Re: vlc-1.1.11-72.el5.i386.rpm
On Sat, Nov 5, 2011 at 2:02 AM, Yasha Karant wrote: > I am attempting to install vlc-1.1.11-72.el5.i386.rpm from > http://packages.atrpms.net/dist/el5/vlc/ > > There is a long list of dependencies missing, appended below. Is there a > way to specify to get just these packages from ATrpms without "damaging" the > underlying production EL 5.7 system (e.g., replacing SL packages by other > packages from ATrpms that then cause problems and instability with EL 5.7)? > Is there another add-on EL 5 RPM repository that requires fewer additions > to stock EL 5.7 but still has a vlc 1.x version? First warning: vlc supports MPEG formats and playing DVD's. There are various patents and reverse engineering legal issues which interfere with open source and especially genuine freeware licensing or deployment of such software. This is why such tools are not in our favorite upstream vendor's codeline, nor will they be, unless such cumbersome licensing can be resolved. Since VLC can typically deal with DVD's, there's the whole libdvdcss sawsuit history that makes it unavailable for our favorite upstream vendor's core distributions. So it's *not* going to work fully without such non-Scientific-Linux provided components, unless our friends at Scientific Linux were to take that on. I don't see a point to that when atrpms and the other repositories are doiing such a good job. That said, if you're in a legal position to use these patented software tools, and you suspect some of the dependencies are extraneous, you can use "mock" to try building the SRPM in a clean SL 5.7 environment. I can send you, or the group, my /etc/mock files for using a local repo for precisely this sort of work. It's much, much, much more efficient to use "mock" from local repositories than reaching out to external mirrors, and different configs to work with JPackage or RPMforge repos as well. This keeps from cluttering your working system with a lot of libraries and dependencies that may cause other adventures. But Yasha, VLC is a powerful and flexible multimedia player, It *needs* access to those libraries in order to manage even half of the different formats and encodings it manages.
Re: odd permissions on mounted disk
Akemi, thank you for response, but I'm mounting udf, not nfs. Not sure where to look next... -- Sent from my Android phone with K-9 Mail. Please excuse my brevity. Akemi Yagi wrote: On Fri, Nov 4, 2011 at 7:59 PM, Andrew Z wrote: > i think it's sort of stupid question, but after the latest kernel update i > observed the following behavior. > total 2 > dr-xr-xr-x 4 4294967295 4294967295 136 Oct 29 23:43 dvd > > I'm not sure i ever saw the permissions set like this. > > Any ideas are welcome. This is due to the difference between NFSv3 and NFSv4. NFSv3 uses UID/GID whereas NFSv4 uses user@domainname. For more details, see for example, http://www.novell.com/support/viewContent.do?externalId=7005060&sliceId=1 http://serverfault.com/questions/98741/files-mounted-over-nfsv4-are-owned-by-4294967294-uids-and-gids-match Akemi
Re: odd permissions on mounted disk
On Fri, Nov 4, 2011 at 7:59 PM, Andrew Z wrote: > i think it's sort of stupid question, but after the latest kernel update i > observed the following behavior. > total 2 > dr-xr-xr-x 4 4294967295 4294967295 136 Oct 29 23:43 dvd > > I'm not sure i ever saw the permissions set like this. > > Any ideas are welcome. This is due to the difference between NFSv3 and NFSv4. NFSv3 uses UID/GID whereas NFSv4 uses user@domainname. For more details, see for example, http://www.novell.com/support/viewContent.do?externalId=7005060&sliceId=1 http://serverfault.com/questions/98741/files-mounted-over-nfsv4-are-owned-by-4294967294-uids-and-gids-match Akemi
