Re: [RDD] RDImport and Disk Space
On 19-11-27 12 h 16, Steve Varholy wrote: > Mark: > > Add another drive, move/copy /var/snd to it and create a symbolic link > to the new location. Done. > You can also use UnionFS to overlay a new filesystem on a new drive on top of the old filesystem. The old filesystem effectively becomes read-only as the writes are all redirected to the new filesystem. This lets you progressively move old files to the new drive, without any downtime, assuming your system supports hotplugging of new drives. Sébastien ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
[RDD] HPI drivers on Debian Jessie → ASI5020
Hello! I am seeking advice on how to install the AudioScience HPI drivers on this Jessie box, for my ASI5020 card. As of today, the latest version of the driver is 4.16.07 (released on 2016-02-08) but it seems like they dropped support for the 5020 in the 4.14.00 release (2014-11-04), although only the release notes mention it; the download page still says that 50xx are supported. [1] Browsing ASI's site, the latest "old release" I can find is version 3.14.12 [2]. It seems like all releases between 3.14.12 and 4.16.07 were removed from AudioScience's download archives. Tryphon repositories package hpklinux-source v.4.08.07, yet for some reason, the Jessie package does not include the source, just READMEs and a changelog. I have been able to find v.4.04.07 of the drivers on an old Rivendell download page [3] but this one does not build due to differences in kernel (after patching a reference to version.h: "Fatal error: kernel firmware_class module required"). I figured I could either: - Install everything on a machine with a very old kernel. > I would rather run it on a more recent system, though. - Try to locate the latest version <4.14.00 (on the off chance that: 1. they build properly 2. they do support the 50xx series) - Hack the latest compatible driver I can find so that it builds on a recent kernel > I would rather do something else (or get paid for doing it), and I am not sure I can accomplish that alone without significant effort. Any advice? Regards, Sébastien NB: I CC'd this message to supp...@audioscience.com --- 1: http://audioscience.com/internet/download/drivers/released/v4/14/drvnotes.txt 2: http://audioscience.com/internet/download/drivers/released/v3/14/12/ 3: http://www.rivendellaudio.org/rivendell/download-1.x.shtml ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Twitter RLM Status?
Correct me if I am wrong, but if one were to write a Twitter OAuth RLM, the authentication state could be saved in a local cookie, whose path would be determined e.g. by an environment variable. That cookie can then be opened at runtime. However it requires one to write such a plugin. On 30 December 2014 at 15:02, Frederick Gleason wrote: > On Dec 30, 2014, at 10:51 27, drew Roberts wrote: > > > Is it workable these days or still in a non-working state after the > OAuth change or for some other reason? > > The changes at Twitter to require OAuth means that it is no longer doable > in a stand-alone RLM. An RLM feeding some sort of server process to > preserve the OAuth state would be a viable approach. > > Cheers! > > > |--| > | Frederick F. Gleason, Jr. | Chief Developer | > | | Paravel Systems | > |--| > | A room without books is like a body without a soul. | > | -- Cicero| > |--| > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] SQL instances
On 25 October 2014 22:49, Lorne Tyndale wrote: > (unless of course, you specifically tell it to use a database called > Rivendell, but that would not be wise) > But even then, it is very unlikely that the two apps will have tables with name conflicts. Overall, it is mostly a question of performance. The database is on the same machine as the actual playout (rdairplay): Are you doing your accounting on the playout machine? Yes: Please dedicate another machine for this task. If there is no way around this, then create a new database for your billing purposes. No: Install MySQL on the machine on which you are doing the billing. The database system is on a separate machine: Create a new database for billing. Make sure you allocate enough resources so that concurrent use will not slow down the Rivendell database. Overall, this kind of setup is very common, nothing in particular should happen (e.g. corruption). Consider moving the Rivendell database back to the playout machine. If budget is tight, consider alternatives like setting up a Raspberry Pi with an external USB drive (total about $100). It could make a decent database server, ideal for low performance tasks, that you can hide anywhere convenient. Don't put the Rivendell database on it, though. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Rivendell and InnoDB (Alessio Elmi)
On 2 October 2014 19:27, James L. Stewart wrote: > So here are some stupid MySQL questions that show off my lack of knowledge > of MySQL: > > How can you tell which database engine you are running? > SHOW TABLE STATUS WHERE Name = 'table name' > Is it possible to be running one engine on some tables in a database and > another engine on other tables in the same database? (My confusing > stumbling around in MySQL sometimes leads me to think that I might be) > It totally is. Might seem confusing at first, but it makes sense once you realize that it is useful to exploit features of an engine for a specific type of data. https://dev.mysql.com/doc/refman/5.7/en/storage-engines.html It looks like they recommend not trying to convert from MyISAM to InnoDB, > but what about going the other way? > If it is okay to do, how do you do it? I am not aware that it is not recommended to convert between different engines. In any case, you have the option of exporting your data and re-importing it in the second table, however keep in mind that there are differences between engines, notably MyISAM will not enforce foreign key constraints and does not provide transaction capabilities, therefore you might hit walls when trying to import "mostly valid" data into an engine that strictly enforces constraints. Since the only recurrent problems I have with Rivendell are database > related any enlightenment could be very helpful to me. > In any case, read the MySQL docs. They might be heavy and long, but at least they are very well written (in my opinion). Here is a starting point: https://dev.mysql.com/doc/refman/5.7/en/storage-engine-setting.html -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Caching audio storage
On 1 October 2014 09:30, Alessio Elmi wrote: > Sometimes we face delay issues, from the moment we hit NEXT to the > moment it is actually played by the client. > Our present NFS server is quite old pc with a basic Raid1 2x2TB hdd in > a Fast Ethernet Lan. Hdd are Sata III but the motherboard only support > Sata I. > Hi Alessio! Sounds a lot like your disks are spinning down after a while. Verify that this is not the case: https://www.google.ca/search?q=sdparm+spindown Average seek times, even on old hard drives, are so small that it should not affect Rivendell, unless as Alban Peignier mentioned, you have concurrent access (i.e. multiple stations using the same NFS repo). There might be additional latency if the drive was not spinning at the time of request. There are a lot of parameters for NFS that can be tweaked, too. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Unable to finish compiling Rivendell 2.9 because of an unknown error
Also I just noticed that you are building the package as root. It is generally considered a Bad Idea(tm) to run the compile job as root. Only the `make install` part should be run as root, and usually it is preferred to have a sandboxed task prepare a package instead, which in turn will be installed by the package manager of your distro. On 14 August 2014 16:52, Larry Ewing wrote: > I am using OpenSUSE 13.1 64 bit on a Dell Optiplex 740 trying to compile > Rivendell 2.9 but it keeps getting stuck at lupdate: command not found. I > have searched the lists for this and have not found any similar problems. > All of my dependencies are installed and mysql is working. See error below. > > /usr/lib/qt3/bin/moc wall_clock.h -o moc_wall_clock.cpp > CXXmoc_wall_clock.o > CXXLD rdairplay > /usr/lib/qt3/bin/lupdate rdairplay.pro > /usr/lib/qt3/bin/lrelease rdairplay.pro > make[1]: Leaving directory `/root/rivendell-2.9.0/rdairplay' > Making all in rdcartslots > make[1]: Entering directory `/root/rivendell-2.9.0/rdcartslots' > CXXlocal_macros.o > CXXrdcartslots.o > /usr/lib/qt3/bin/moc rdcartslots.h -o moc_rdcartslots.cpp > CXXmoc_rdcartslots.o > CXXLD rdcartslots > lupdate rdcartslots.pro > make[1]: lupdate: Command not found > make[1]: *** [all] Error 127 > make[1]: Leaving directory `/root/rivendell-2.9.0/rdcartslots' > make: *** [all-recursive] Error 1 > suserivendell:~/rivendell-2.9.0 # > > > > Thanks, > > *Larry Ewing* > > > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev > > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Unable to finish compiling Rivendell 2.9 because of an unknown error
Try adding "/usr/lib/qt3/bin/" to your path, it seems the Makefile assumes that it is already in there. example: PATH="$PATH":/usr/lib/qt3/bin make On 14 August 2014 16:52, Larry Ewing wrote: > I am using OpenSUSE 13.1 64 bit on a Dell Optiplex 740 trying to compile > Rivendell 2.9 but it keeps getting stuck at lupdate: command not found. I > have searched the lists for this and have not found any similar problems. > All of my dependencies are installed and mysql is working. See error below. > > /usr/lib/qt3/bin/moc wall_clock.h -o moc_wall_clock.cpp > CXXmoc_wall_clock.o > CXXLD rdairplay > /usr/lib/qt3/bin/lupdate rdairplay.pro > /usr/lib/qt3/bin/lrelease rdairplay.pro > make[1]: Leaving directory `/root/rivendell-2.9.0/rdairplay' > Making all in rdcartslots > make[1]: Entering directory `/root/rivendell-2.9.0/rdcartslots' > CXXlocal_macros.o > CXXrdcartslots.o > /usr/lib/qt3/bin/moc rdcartslots.h -o moc_rdcartslots.cpp > CXXmoc_rdcartslots.o > CXXLD rdcartslots > lupdate rdcartslots.pro > make[1]: lupdate: Command not found > make[1]: *** [all] Error 127 > make[1]: Leaving directory `/root/rivendell-2.9.0/rdcartslots' > make: *** [all-recursive] Error 1 > suserivendell:~/rivendell-2.9.0 # > > > > Thanks, > > *Larry Ewing* > > > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev > > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] most of it works
Anyone, correct me if I am wrong, but the recommended setup is to have the RD PCM streams in /etc/asound.conf in order to "assign" the card to Rivendell, and configure your own .asoundrc so that it either provides "null" output as default to your apps or route it to an additional sound card. If you set Rivendell in your .asoundrc, your desktop manager (KDE, Gnome, etc.) will likely still pick up the default sound card and then "hog" it. On 13 August 2014 11:46, David Weber wrote: > I can login. I can rip cds into carts and I can import into carts > what I cannot do is hear any sound > > by the way if I put asound.conf into /etc my sound system complains that > it is keeping the sound from working so instead I have a .asoundrc but > still no sound > > David Weber dweber1941@gmail > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
[RDD] Fwd: Makefile patch suggestion
Since I did not get any feedback the first time I submitted this email, I am resubmitting it so that more people see it. The reason I am submitting this patch is so that the packaging process becomes more portable to Linux distributions other than the big fours (Debian/Ubuntu, Centos, Slackware, Gentoo). While this patch was explicitly engineered for the Arch Linux "makepkg" package building system, I believe it is of adequate portability to integrate it to the codebase and that it will probably benefit other distributions who have a similar build system. Note that one of these fixes (the "lib64" one, details in original post) needs to be factored out as I could not figure out a way to guess the conventions of a specific system (on Arch, there is no "lib64", but there is a "lib32" only on "multilib" systems, which are truly 64-bit systems that include 32-bit libraries for compatibility reasons; pure 32-bit systems will have a "lib" directory but no "lib32/64"). I would like some guidance or feedback on how to fix this once and for all. -- Forwarded message -- From: Sébastien Leblanc Date: 10 March 2014 02:35 Subject: Makefile patch suggestion To: Rivendell Users Group Hello all, I spent the last few hours tweaking the Makefiles in Rivendell so that it will compile seamlessly on Arch Linux systems. Attached is an unified diff-formatted patch to address some issues with the Makefile system. I made my changes in such a way that they should not affect other system, except for a few small details. Here is the list of modifications: ### - cae/Makefile.am Problem: caed depends on pthread, but for some reason it does not get included. Solution: add -lpthread to "caed_LDADD" variable. - configure.in: Problem: a lib path of "lib64" is forced for 64-bit systems. Some systems (e.g. Arch) keep "/usr/lib" as path for "default" architecture, and have a "lib32" directory instead for multilib installations. Solution: I am not an autotools expert and I could not devise a way to make it work for distributions that have "lib64" as a default for 64-bit systems. On Arch, the standard is lib/lib32 on 64-bit system and only "lib" on 32-bit. Other projects have defined an additional variable (LIB_SUFFIX) which is set by configure during compilation using "-DLIB_SUFFIX=64". This modification is therefore incompatible with 64-bit distros that have lib64 as the default lib path. This is a hack that should probably not be committed to repo. - lib/Makefile.am Problem: libtool can't figure out that librd.la needs to be linked against some libraries Solution: specify those libraries manually, in "librd_la_LDADD" - pam_rd/Makefile.am Problems: 1- out-of-directory packaging breaks libtool as it tries to relink the library to a non-final directory, 2- installation path does not honor prefix. On my system, the PAM libraries are in /usr/lib/security/. Solution: specify "inst_prefix_dir" [1] and add $(prefix) to "inst_libdir". - rdselect/Makefile.am Problem: multiple parallel job compilation (make -j) usually fails because it tries to "chmod" a file that was not yet installed Solution: rename target "install-exec-local" to "install-exec-hook". [2] - {icons,rlm,scripts,web/rdxport,xdg}/Makefile.am Problem: $(DESTDIR) not set, making staged installs impossible Solution: add $(DESTDIR) accordingly. Note that I did not fix the "uninstall" counterparts as I did not need to. 1: https://www.gnu.org/software/libtool/manual/html_node/Install-mode.html 2: https://www.sourceware.org/autobook/autobook/autobook_108.html ## Comments and criticism are more than welcome. -- Sébastien Leblanc -- Sébastien Leblanc diff -Naur rivendell-2.8.1-distclean/cae/Makefile.am rivendell-2.8.1-patched/cae/Makefile.am --- rivendell-2.8.1-distclean/cae/Makefile.am 2014-03-10 01:03:30.089863481 -0400 +++ rivendell-2.8.1-patched/cae/Makefile.am 2014-03-10 01:10:22.901034082 -0400 @@ -41,7 +41,8 @@ nodist_caed_SOURCES = moc_cae.cpp\ moc_cae_socket.cpp -caed_LDADD = @LIB_RDLIBS@\ +caed_LDADD = -lpthread\ + @LIB_RDLIBS@\ @LIBALSA@\ @LIBHPI@\ @LIBJACK@\ diff -Naur rivendell-2.8.1-distclean/configure.in rivendell-2.8.1-patched/configure.in --- rivendell-2.8.1-distclean/configure.in 2014-03-10 01:03:30.129866174 -0400 +++ rivendell-2.8.1-patched/configure.in 2014-03-10 01:09:46.811928529 -0400 @@ -96,8 +96,8 @@ AC_SUBST(VENDOR,$ar_gcc_distro) AC_SUBST(ARCH,$ar_gcc_arch) if test $ar_gcc_arch = x86_64 ; then - LIB_PATH=$PREFIX/lib64 - AC_SUBST(RD_LIB_PATH,lib64) + LIB_PATH=$PREFIX/lib + AC_SUBST(RD_LIB_PATH,lib) else LIB_PATH=$PREFIX/li
[RDD] Makefile patch suggestion
Hello all, I spent the last few hours tweaking the Makefiles in Rivendell so that it will compile seamlessly on Arch Linux systems. Attached is an unified diff-formatted patch to address some issues with the Makefile system. I made my changes in such a way that they should not affect other system, except for a few small details. Here is the list of modifications: ### - cae/Makefile.am Problem: caed depends on pthread, but for some reason it does not get included. Solution: add -lpthread to "caed_LDADD" variable. - configure.in: Problem: a lib path of "lib64" is forced for 64-bit systems. Some systems (e.g. Arch) keep "/usr/lib" as path for "default" architecture, and have a "lib32" directory instead for multilib installations. Solution: I am not an autotools expert and I could not devise a way to make it work for distributions that have "lib64" as a default for 64-bit systems. On Arch, the standard is lib/lib32 on 64-bit system and only "lib" on 32-bit. Other projects have defined an additional variable (LIB_SUFFIX) which is set by configure during compilation using "-DLIB_SUFFIX=64". This modification is therefore incompatible with 64-bit distros that have lib64 as the default lib path. This is a hack that should probably not be committed to repo. - lib/Makefile.am Problem: libtool can't figure out that librd.la needs to be linked against some libraries Solution: specify those libraries manually, in "librd_la_LDADD" - pam_rd/Makefile.am Problems: 1- out-of-directory packaging breaks libtool as it tries to relink the library to a non-final directory, 2- installation path does not honor prefix. On my system, the PAM libraries are in /usr/lib/security/. Solution: specify "inst_prefix_dir" [1] and add $(prefix) to "inst_libdir". - rdselect/Makefile.am Problem: multiple parallel job compilation (make -j) usually fails because it tries to "chmod" a file that was not yet installed Solution: rename target "install-exec-local" to "install-exec-hook". [2] - {icons,rlm,scripts,web/rdxport,xdg}/Makefile.am Problem: $(DESTDIR) not set, making staged installs impossible Solution: add $(DESTDIR) accordingly. Note that I did not fix the "uninstall" counterparts as I did not need to. 1: https://www.gnu.org/software/libtool/manual/html_node/Install-mode.html 2: https://www.sourceware.org/autobook/autobook/autobook_108.html ## Comments and criticism are more than welcome. -- Sébastien Leblanc diff -Naur rivendell-2.8.1-distclean/cae/Makefile.am rivendell-2.8.1-patched/cae/Makefile.am --- rivendell-2.8.1-distclean/cae/Makefile.am 2014-03-10 01:03:30.089863481 -0400 +++ rivendell-2.8.1-patched/cae/Makefile.am 2014-03-10 01:10:22.901034082 -0400 @@ -41,7 +41,8 @@ nodist_caed_SOURCES = moc_cae.cpp\ moc_cae_socket.cpp -caed_LDADD = @LIB_RDLIBS@\ +caed_LDADD = -lpthread\ + @LIB_RDLIBS@\ @LIBALSA@\ @LIBHPI@\ @LIBJACK@\ diff -Naur rivendell-2.8.1-distclean/configure.in rivendell-2.8.1-patched/configure.in --- rivendell-2.8.1-distclean/configure.in 2014-03-10 01:03:30.129866174 -0400 +++ rivendell-2.8.1-patched/configure.in 2014-03-10 01:09:46.811928529 -0400 @@ -96,8 +96,8 @@ AC_SUBST(VENDOR,$ar_gcc_distro) AC_SUBST(ARCH,$ar_gcc_arch) if test $ar_gcc_arch = x86_64 ; then - LIB_PATH=$PREFIX/lib64 - AC_SUBST(RD_LIB_PATH,lib64) + LIB_PATH=$PREFIX/lib + AC_SUBST(RD_LIB_PATH,lib) else LIB_PATH=$PREFIX/lib AC_SUBST(RD_LIB_PATH,lib) diff -Naur rivendell-2.8.1-distclean/icons/Makefile.am rivendell-2.8.1-patched/icons/Makefile.am --- rivendell-2.8.1-distclean/icons/Makefile.am 2014-03-10 01:03:30.093197038 -0400 +++ rivendell-2.8.1-patched/icons/Makefile.am 2014-03-10 00:59:05.542052023 -0400 @@ -31,18 +31,18 @@ cp redball.png $(DESTDIR)@libexecdir@ cp whiteball.png $(DESTDIR)@libexecdir@ cp progressbar.gif $(DESTDIR)@libexecdir@ - mkdir -p @prefix@/share/icons/hicolor/16x16/apps - cp rivendell-16x16.png @prefix@/share/icons/hicolor/16x16/apps/rivendell.png - mkdir -p @prefix@/share/icons/hicolor/22x22/apps - cp rivendell-22x22.png @prefix@/share/icons/hicolor/22x22/apps/rivendell.png - mkdir -p @prefix@/share/icons/hicolor/32x32/apps - cp rivendell-32x32.png @prefix@/share/icons/hicolor/32x32/apps/rivendell.png - mkdir -p @prefix@/share/icons/hicolor/48x48/apps - cp rivendell-48x48.png @prefix@/share/icons/hicolor/48x48/apps/rivendell.png - mkdir -p @prefix@/share/icons/hicolor/64x64/apps - cp rivendell-64x64.png @prefix@/share/icons/hicolor/64x64/apps/rivendell.png - mkdir -p @prefix@/share/icons/hicolor/128x128/apps - cp rivendell-128x128.png @prefix@/share/icons/hicolor/128x128/apps/rivendell.png + mkdir -p $(DESTDIR)@prefix@/share/icons/hicolor/16x16/apps + cp rivendell-16x1
Re: [RDD] Two problems in RDLogManager 2.7.0
On 18 January 2014 14:15, Luigino Bracci wrote: > Hi guys, > > Testing the new Rivendell 2.7.0 I've tried to generate a log from > RdLogManager. I'm using the template name "Production-%m%d" (it's the name > assigned by rdsvc.cpp, line 1097), but the generated log is empty. > RDLogMnager generated many errors in console: > > Invalid SQL or failed DB connection[You have an error in your SQL syntax; > check the manual that corresponds to your MySQL server version for the > right syntax to use near '-0120_LOG set ID=0,COUNT=0,TYPE=0, > SOURCE=2,START_TIME=7560,GRACE_TIM' at line 1 QMYSQL3: Unable to > execute query]: insert into Production-0120_LOG set ID=0,COUNT=0,TYPE=0, > SOURCE=2,START_TIME=7560,GRACE_TIME=0, > CART_NUMBER=10005,TIME_TYPE=0,POST_POINT="N", > TRANS_TYPE=0,EXT_START_TIME="21:00:00", > EVENT_LENGTH=90 > Database connection failed: insert into Production-0120_LOG set > ID=0,COUNT=0,TYPE=0, SOURCE=2,START_TIME=7560,GRACE_TIME=0, > CART_NUMBER=10005,TIME_TYPE=0,POST_POINT="N", > TRANS_TYPE=0,EXT_START_TIME="21:00:00", > EVENT_LENGTH=90 > Database connection restored. > _ > > Apparently the problem is in rdevent_line.cpp, > in RDEventLine::generateLog(), lines 793 and 817: It is needed to use > backquotes surrounding the name of the table, so using a dash in the name > of the template (and the SQL table) doesn't break the log generation. > > sql=QString().sprintf("insert into `%s_LOG` set > ID=%d,COUNT=%d,TYPE=%d,\ > > sql=QString().sprintf("insert into `%s_STACK` set (...) > > Once I placed this, I have no more errors and the log was generated as > expected. > > The other problem is with the clocks. I've created a new clock called > "Torta", code "CO", then I pressed "Edit grid" in RDLogManager and then > pressed "Change All" to assign the clock "Torta" to all the buttons in the > grid. > > When I close the edit grid window and I open it again, buttons from monday > to thursday (11-12) are updated with the clock "Torta", but the others > aren't. > > In the console, RDLogManager showed these errors after using the "Change > All" button: > _ > > Invalid SQL or failed DB connection[Row size too large (> 8126). Changing > some columns to TEXT or BLOB or using ROW_FORMAT=DYNAMIC or > ROW_FORMAT=COMPRESSED may help. In current row format, BLOB prefix of 768 > bytes is stored inline. QMYSQL3: Unable to execute query]: update SERVICES > set CLOCK84="Torta" where NAME="Production" > > Database connection failed: update SERVICES set CLOCK84="Torta" where > NAME="Production" > Database connection restored. > > > (This error repeats with each CLOCKnnn field from CLOCK84 to CLOCK167) > > I'm using the DB structure from a recently installed Rivendell with no > modifications, using Mysql 5.5.34 in Ubuntu Studio 12.04. The table engine > is InnoDB. > > Apparently the problem is that a row in a InnoDB table can't exceed 8126 > characters, but there are 168 CLOCK fields in every row in the SERVICES > table, each one is CHAR(64) (168 * 64 = 10752 characters). I don't > understand why this problem hasn't occurred before, maybe it's something > new in Mysql or maybe I'm doing something wrong. But this is my first time > using the "Changel all" button. > > Googling this problem, some people suggests to create the table with the > parameters ROW_FORMAT=DYNAMIC or ROW_FORMAT=COMPRESSED, or use the new > format "Barracude" available from Mysql-server 5.5.29 (see > http://stackoverflow.com/questions/15585602/change-limit-for-mysql-row-size-too-large > ). > > Another idea could be to shorten the clock's NAME field from CHAR(64) to > CHAR(32), or maybe using the clock's SHORT_NAME (it's a CHAR(8)) as primary > index, and not the clock's NAME. > > Bye, and sorry for my bad english. > I think it was a long-known, but unresolved issue that log names containing dashes (-) would fail generating. I worked around this issue by using underscores (_) instead. It is great that you tracked at least part of the problem to these lines of code. Unfortunately, many such cases exist in the Rivendell codebase, and in my opinion most of the SQL part would benefit greatly from a overhaul (which is very hard work in itself). I think that the only supported database engine for Rivendell is MyISAM, which is the engine that Rivendell should have used when creating your databases. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
[RDD] Fwd: What reliable ALSA cards are you using?
-- Forwarded message -- From: Sébastien Leblanc Date: 6 January 2014 12:44 Subject: Re: [RDD] What reliable ALSA cards are you using? To: Alan Smith The Delta 44 by M-AUDIO is a general purpose, fair quality sound card that is well supported on Linux, might want to give it a try. On 6 January 2014 12:17, Alan Smith wrote: > A new project may require me to put together 3 systems. They will be > basically similar to a sat network type setup, and only required to play > liners, IDs, and stopsets. > > The ASI cards are way too much for this type application. > > Any recommendations on some good, reliable, currently available alsa sound > cards, preferably PCI? These systems will be located in other cities with > no direct access to the machines, so the reliability aspect is the most > important. > > Thanks! > > -Alan > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
[RDD] Cut dayparting bug when using MariaDB
Hey everyone, We are running an older version of Rivendell here (2.1.1), but we had to upgrade our MySQL install to version 5.5. As we are using Arch Linux, it has been replaced by MariaDB. Upgrade was seamless except for one (major) detail: dayparting would no longer work. The only cuts that would play were EVERGREEN ones. I have tracked down the bug to an incompatible SQL line, in lib/rdcart.cpp, function RDCart::selectCut(QString *cut,const QTime &time) const (line 72 in v2.1.1, 86 in 2.6.0): select CUT_NAME,WEIGHT,LOCAL_COUNTER from CUTS where ( ( (START_DATETIME<="$1") && (END_DATETIME>="$2") ) || (START_DATETIME is null) ) && ( ( (START_DAYPART<="$3") && (END_DAYPART>="$4") || START_DAYPART is null ) ) && ($5="Y") && (CART_NUMBER=$6) && (EVERGREEN="N") && (LENGTH>0) order by LOCAL_COUNTER $1-4 are the current time (in format "-MM-dd hh:mm:ss"), $5 is the day of the week in short format (e.g. "MON" for Monday) and $6 is the cart number. The issue with this line is that MySQL would implicitly convert e.g. "2013-12-31 19:00:00" to "19:00:00" for columns START_DAYPART and END_DAYPART, because of their type (TIME). MariaDB, unless I am mistaken, does not, and therefore the times in the database are interpreted as offsets from the Unix epoch, that is comparing "17:01:00" (value in database) to "2013-12-31 17:00:00" (current time) results in a conversion to ( "1970-01-01 17:01:00" <> "2013-12-31 17:00:00" ), which means END_DAYPART would always be smaller than the current time (thus, no scheduling of that cut). Using full timestamps as values for TIME types is undocumented, therefore I have fixed it by casting the timestamp to a second string, time_str, with format "hh:mm:ss", then using it for variables $3 and $4. I have included a patch in this email. Apply with `patch -p1 < `, when in rivendell source directory. This fix should not break compatibility with MySQL, but I did not test it. Is there a bug tracker for the Rivendell project? -- Sébastien Leblanc diff -aur rivendell-2.6.0/lib/rdcart.cpp rivendell-2.6.0-timestamp/lib/rdcart.cpp --- rivendell-2.6.0/lib/rdcart.cpp 2013-12-11 13:51:47.0 -0500 +++ rivendell-2.6.0-timestamp/lib/rdcart.cpp 2013-12-30 19:26:30.0 -0500 @@ -109,6 +109,8 @@ QDate current_date=QDate::currentDate(); QString datetime_str=QDateTime(current_date,time). toString("-MM-dd hh:mm:ss"); + QString time_str=QDateTime(current_date,time). +toString("hh:mm:ss"); // if(type()==RDCart::Audio) { switch(type()) { @@ -123,8 +125,8 @@ (LENGTH>0) order by LOCAL_COUNTER", (const char *)datetime_str, (const char *)datetime_str, - (const char *)datetime_str, - (const char *)datetime_str, + (const char *)time_str, + (const char *)time_str, (const char *)RDGetShortDayNameEN(current_date.dayOfWeek()).upper(), cart_number); q=new RDSqlQuery(sql); ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Appliance Updates
On 21 November 2013 11:34, nathan lawson wrote: > Going to go with the CentOS RD Appliance (v2) instead of ubuntu for the next set of installs... To do updates of the rivendell suite is it as easy as doing an apt-get? Ive never used CentOs and really dont want to be re-compiling (im not great at it in the first place) every time an update comes out... FYI, CentOS is a carbon copy of Red Hat Entreprise Linux, which in turn is very similar to Fedora Linux (which is based on RHEL and is sponsored by Red Hat). Thus, many of the Fedora docs can be translated directly to CentOS. There is also the official RHEL docs (the CentOS docs have not yet been updated to version 6.x of CentOS): https://access.redhat.com/site/documentation/Red_Hat_Enterprise_Linux/ -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Silentjack and silence detection
On 16 November 2013 07:06, Alessio Elmi wrote: > I'm trying to experiment a bit with this tool. It's really simple and > powerful at the same time. I was wandering if there is a way to trigger an > event as soon as sound gets back. > Let's say I prepare an emergency 24h-long mp3 which is played by Silentjack > after 1 minute of silence, like after power failure of the host or a missing > log in Riv. or whatever. > But when sound gets back how to catch that event and stop playing the > emergency track? > There should be an opposite tool: when jackd output is bigger than a certain > level do simething. > Anyone? Liquidsoap has an "on_blank" function that can detect silence and then act on it. It supports two callbacks; one for silence detection and one for noise detection. It can run system commands or even output Jack audio data. You could even use it to generate a backup stream from either another Icecast mountpoint or from static files in a directory. We have it running on a web stream that we send using Icecast. On silence, it will log the event, send an email to the managers and switch to a backup source that's actually a directory filled with songs that it will play in shuffle (with crossfades), and every 40-ish seconds, it ducks the program and plays excuse messages from another directory. If you don't need it to do this, you could always send the generated Liquidsoap content to a dummy output. In our case, the main program is taken from a line-in input, and the program is encoded in both MP3 and Vorbis and streamed to a distant Icecast server. All this is running on an old 1.60GHz Pentium 4. Send me an email me if you need pointers on how to set-up Liquidsoap... -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Is Rivendell suited for my station?
On 7 November 2013 05:06, Peter van Embden wrote: > Does Rivendell have a silence-detector? If it's silent for more than a > certain time, something should start playing! Can Rivendell automatically > build an "emergency playlist" to fill a gap? Many users have used Liquidsoap to build web streaming engines, and you can tune it so that it will detect silence and play an alternative source such as a static playlist, send you an email or switch a relay to another input or anything, basically. Much cheaper (i.e. free) than these types of machines: http://www.broadcastpartners.com/website/eshop/sd-1-silence-detector-81716.html -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] ASI Soundcard
Generally better sound quality, balanced output, hardware mixer. Some models support timestretching and hardware decoding. A+++ would recommend. On 18 October 2013 16:15, Kevin Mitchell wrote: > Hi guys, > Just brought a ASI soundcard off of eBay. Within my price range. > Can anyone give any views on the ASI range of cards? They appear to be used > for broqdcast. > I am hoping for better audio output. > Regards Kevin > > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] M-Audio Delta 44 +4db Output
Regarding the decibel terminology, I have to nitpick on one point: a dB is not the same as a dBu/dBv. A straight "dB" is always a relative scale unit, and can only be used to compare things together. A dBu (or dBv) is an actual, absolute unit, that happens to use the decibel scale. So "0 dBu" is equal to the power of 1 mW going through a 600 ohm resistance. That is why gain in an amplifier (whether a speaker power amplifier or a tiny transistor) is counted in simple decibels (dB) since they compare the input to the output in the logarithmic scale (decibel scale). However, when you measure the output of a sound mixer, you may use the dBu unit as it refers to actual, physical quantities. On 17 August 2013 07:00, Cowboy wrote: > On Friday 16 August 2013 09:15:13 pm al davis wrote: >> That is misleading. > > And that is precisely the point, and the reason this question > comes up so often ! > > To shorten the story greatly, db is ALWAYS a power ratio reference > by definition. Hence, zero was defined ( correctly so ) by Ma Bell > many many years ago, when I was just a little lad, as 1 mw in > 600 ohms, *THE* reference for all things bel. > > ( side note. Ma Bell deliberately defined the range to track the response > of the human ear ( or human eye ) so as to relate directly to human > perception of sound ( or light ) ratios ) > > Some years later, somewhere across the pond, 0 was incorrectly > defined as 1 volt. Since voltage is not power, this is and was > incorrect, and so has caused much confusion ever since. > > It *is* misleading. > > As long as anyone attempts to define voltage in terms that mean power, > it will always be misleading. ( and it's now so entrenched, we can expect > this will continue in perpetuity ) > > -- > Cowboy > > http://cowboy.cwf1.com > > Clothes make the man. Naked people have little or no influence on > society. > -- Mark Twain > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Mailign list reply address
I noticed this too, for this email, for example, it defaulted to your email address instead of the list's. On 16 August 2013 11:40, Gavin Stephens wrote: > I haven't posted in a while and just noticed when I click reply it doesn't > default to the mail list. Has it always been like that? I haven't changed > email client in near a decade. > > Cheers, > Gavin. > > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] One More Problem, Can't Access Files on a NTFS Stick
On 12 July 2013 07:26, Cowboy wrote: > On Friday 12 July 2013 04:13:34 am Kevin, Natalia, Stacey and Rochelle wrote: >> I have done a search on Google and found information on how to add support >> for NTFS, I followed the instructions and rebooted, however still will not >> see files on NTFS. > > All you need is the compatible kernel module, which very well may mean > you need to recompile the kernel to add it. > You'll very likely need to specify explicitly that you want NTFS when you > mount the device. You might not be able to do this from fstab. NTFS support for Linux is most of the time provided by NTFS-3G, which is a now fully mature project. NTFS-3G runs in userspace, through FUSE, and thus requires no special kernel module to work. it features full read and write support, and many other features of NTFS. There are a few issues with advanced features of NTFS, but they are simply unsupported and will not break your filesystem. > Remember, NTFS is proprietary, and depending on version, not > compatible with NTFS ! This is a bit FUD. NTFS support nowadays is (probably) reliable. > *I* would use something else, probably VFAT. This is fair. Having an NTFS filesystem on a USB stick is asking for trouble in any case. Windows by default partitions sticks as FAT32. > That means that yes, you can add NTFS, but it is strictly experimental, > will probably change with the next version of Windows, should be > read-only, *and* carries a very high risk of destroying the filesystem > if you do decide to experiment. FUD. This is not 2005 anymore. > *I* have had good luck recovering NTFS devices, but there is a method > to the madness. > Read-only under *nix. I can then work on the files under *nix, save > the work on an ext3 system, then copy it back to the NTFS device > using a M$ OS compatible with that version of NTFS. > Finally, double-check that system using a M$ OS and M$ OS file > system tools designed specifically for that purpose. No longer required. Kevin, check for Centos NTFS-3G packages. You should find what you are looking for. Cheers! -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] Rivendell Arch Linux PKGBUILD
Scratch that last paragraph, I got it working before I sent the email, but forgot to remove that paragraph! Everything should be peachy. On 3 June 2013 21:44, Sébastien Leblanc wrote: > Hey folks! > > I updated my PKGBUILD file and submitted it to the Arch User > Repository, if anyone is interested. It is the 2.5.1 version of > Rivendell > > https://aur.archlinux.org/packages/rivendell/ > > Let me know if you are having any issues. For now, I am having trouble > with rdcartslots: it will not run at all. Maybe someone can help me > out. I have not done thorough testing, but most everything else should > work correctly. > > -- > Sébastien Leblanc -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
[RDD] Rivendell Arch Linux PKGBUILD
Hey folks! I updated my PKGBUILD file and submitted it to the Arch User Repository, if anyone is interested. It is the 2.5.1 version of Rivendell https://aur.archlinux.org/packages/rivendell/ Let me know if you are having any issues. For now, I am having trouble with rdcartslots: it will not run at all. Maybe someone can help me out. I have not done thorough testing, but most everything else should work correctly. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
Re: [RDD] rivendell database update V2.5
On 17 May 2013 22:27, Jorge Soto wrote: > It's kind of like when you are trying to open an MS access database > created with a different version. > I think Mr. Peter Claes is wondering whether he will lose data during database upgrade. > -- > On Fri, May 17, 2013 6:55 PM PDT Peter Claes wrote: > >will it harm the older system ? is this normal ? > This is the intended operation; upon upgrade of the Rivendell software, you must also upgrade the database backend. You should not lose data, but you definitely should backup / dump your database before upgrading, just in case. Use a tool such as phpMyAdmin or MySQL Administrator to save your database. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] What the h377 is...
On 16 May 2013 11:07, Alan Peterson wrote: > [...] I'm going to ask just what in tarnation Winki 3 is. [...] > It's a somewhat popular motherboard / system feature. Some manufacturers embed a Linux system either in Flash memory or in a hard drive partition that comes before the main system. This OS is a streamlined Linux system containing a web browser, an email client, a music player and some other apps. It can even access documents in the main hard drive. This gives an auxilliary system that boots real fast and can browse the web in seconds. This is mostly used as a "backup" OS, or as a fast system that you boot on the go when you have only quick things to do. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Not playing audio
The "invalid URL" error is a red herring for an improper rd-bin CGI setup. In the (I think) /usr/libexec/rd-bin you'll find CGI scripts. A local HTTP server such as Apache must be set-up to execute these files. Upon import, Rivendell communicates with the local httpd process and asks for the rdimport.cgi script (or something similar). Sorry for the lack of detail, I'm on the go right now. On 30 April 2013 18:12, Guy Curtis wrote: > I tried to import some audio with that in mind, it is saying that the url is > invalid and so skips the import, I tried using terminal rdimport and using > import function of rdlibrary and neither of them will allow me to import > audio. > > Guy Curtis > Broadcast Engineer > WAZU Peoria > > > From: Rob Landry <41001...@interpring.com> > To: Guy Curtis ; User discussion about the > Rivendell Radio Automation System > Cc: MICHAEL SMITH > Sent: Tuesday, 30 April 2013, 17:10 > > Subject: Re: [RDD] Not playing audio > > > So, now you have /var/snd. Make sure it is owned by the Rivendell user > (which is "rivendell" if you're running RD 2.0 or newer). > > Don't try to play the test tone; it's not there. Delete it and use rdlibrary > to import a .wav file or rip something off a CD. > > > Rob > > On Tue, 30 Apr 2013, Guy Curtis wrote: > >> well, I didn't have a /var/snd directory so that might have been the issue >> in and of itself. Testing now... >> >> Didn't change anything > > > > _______ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Temperature controlled jingles
Damn, just checked their terms of use (Forecast.io) and they don't allow radios, newspapers, broadcast and cable television to use their API... ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Temperature controlled jingles
Have a script that fetches the weather from some public API like forecast.io and this script will put the appropriate cart number in a known file (e.g. /var/tmp/weather). Then you have a cart which is actually a RML command that runs a shell script. This shell script will check the timestamp on /var/tmp/weather, and if it's recent enough (if there's a problem, you don't want to repeat yesterday's forecast), have the shell script add the number in that file as a cart, through rmlsend. That's lots of callbacks, I know, but I think the data freshness is worth it. Wish there was a better way to do that in Rivendell... On 9 April 2013 07:07, Daniel Roviriego wrote: > Hi. > > I Think the approach would change depending on how much earlier you ' d like > to schedule your commercial. > For instance, if it' s more than 24 hours in advance, I ' d write some code > to fill in a txt column formatted file and import that to your slots through > the traffic iport function within rivendell. > For sure people in here will have better ideas and experiences. > > Have a nice day > > Em terça-feira, 9 de abril de 2013, Peter Claes escreveu: > >> Hi guys, >> >> I've read somewhere, that there are stations that use temperature >> controlled commercial, using rivendell. >> >> I would like to have this hourly mentioned on our station. >> >> Can someone enlighten me on how i should start on this ? >> >> My guess would be a php script that Fetches the temp at for example >> Yahoo-weather. And changing a cart number in the sql database. >> >> Any help would be appreciated. >> >> Regards, >> Peter >> >> >> >> Verstuurd vanaf mijn iPhone >> ___ >> Rivendell-dev mailing list >> Rivendell-dev@lists.rivendellaudio.org >> http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev > > > > -- > Daniel Roviriego > (21) 35920701 > (21) 99561654 > > > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Best distro for Rivendell
Seriously? The one that works the best for you. A LTS distro (long-term support) might be better-suited, as you might not want to update many packages very often on a production machine, and you want only critical, stability or security-related kernel patches. So, any LTS distro, or even your preferred distro. We run Rivendell on Arch Linux here. On 8 April 2013 14:25, Guy Curtis wrote: > What is the best Linux distro for Rivendell to run on? > > Guy Curtis > Broadcast Engineer > WAZU Peoria > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Cart linking
Why not use a command cart containing two "Play next" commands? Syntax is "PX !". Can't test it, I'm not sure whether carts can be loaded fast enough if they are to be played immediately after the command runs... Also your scheduling might display that you are short on time until the carts are loaded. On 11 March 2013 08:28, Tim Camp wrote: > Just wondering if anyone knows of a way to link two carts together during > scheduling, ie when a particular cut plays, a linked cart would > automatically be loaded to follow it, under a set of circumstances. > I understand this may have to be done in a external script. > > the idea would be to load a specific backsell when it was time for a > stopset. > > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Invalid Cart...but it's not....
You could refactor your RDcatch automatic download into an external script that would be ran by RDCatch instead. This script can get today's date, add one, then get tomorrow's file. I even used to run Python scripts through Rivendell macros. Also I just thought of it, enabling auto log refresh would probably do no good because it's only the cart that gets updated, not the whole log... (I am not aware of the internals of log refresh) On 4 March 2013 15:19, Nathan Steele wrote: > BY the way offsetting in RDcatch offsets the start date as well, so > that's no good. > > Nathaniel C. Steele > Assistant Chief Engineer/Technical Director > WTRM-FM / TheCrossFM > > On 3/4/2013 2:50 PM, Nathan Steele wrote: >> That's the case, so will the cart play or not? I just found out that >> they skip them when they come up red, because it messes up the music if >> there actually is no cut in the cart. I've told them not to do that and >> see what happens. I also went up there a little while ago and looked and >> the only thing I see Invalid, actually is invalid so I don't know what's >> going on now. You get told one thing (I"m not in that early in the >> morning), and then have to figure out what's really going on. >> >> Thing is, why would it only be on monday? oh, because fridays cuts are >> expierd by that time ( they get imported with a end date 2 days after >> the start date, that's the way the programs come, it's in the cart chunk >> data.Maybe if I offset the end time by one more day >> >> Nathaniel C. Steele >> Assistant Chief Engineer/Technical Director >> WTRM-FM / TheCrossFM >> >> On 3/4/2013 2:13 PM, Sébastien Leblanc wrote: >>> Maybe the log is not refreshed when the cart is updated with today's >>> content... Is that cart playing on the same day that it is downloaded? >>> I suspect that when your log is loaded at midnight, the cart is not >>> yet available, so at that time it was invalid. >>> >>> On 4 March 2013 13:16, Nathan Steele wrote: >>>> Every monday (i'm told) we have programs that do not play (all of them). >>>> these programs are downloaded into RD by RDCATCH between 1 am and 2 am. >>>> this is done monday-friday, but only mondays seem to fail. >>>> >>>> I looked at todays log and all the programs are red saying invalid cart, >>>> but When I look at the carts in RDlibrary, I don't see what is invalid >>>> about them and it says it played once today. I will have to actually >>>> listen for the next one and see if it really plays or if the host's just >>>> think it didn't play beause it says invalid cart. butWhy does it say >>>> invalid cart anyway? I see no issues with start/end dates or dayparting, >>>> and the carts are there. >>>> >>>> Thanks, >>>> Nathan >>>> >>>> -- >>>> Nathaniel C. Steele >>>> Assistant Chief Engineer/Technical Director >>>> WTRM-FM / TheCrossFM >>>> >>>> ___ >>>> Rivendell-dev mailing list >>>> Rivendell-dev@lists.rivendellaudio.org >>>> http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev >>> >>> -- >>> Sébastien Leblanc >>> ___ >>> Rivendell-dev mailing list >>> Rivendell-dev@lists.rivendellaudio.org >>> http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev >> ___ >> Rivendell-dev mailing list >> Rivendell-dev@lists.rivendellaudio.org >> http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Invalid Cart...but it's not....
Kind of a hack, but are you able to download the carts the day before, during the evening? Are you able to enable auto log refreshes? That would surely fix your problems. However, failing so, maybe you can craft a small Rivendell macro that can add the cart right before it will play. From the wiki: > ADD NEXT > Mneumonic: PX > Syntax: > PX ! > > Insert cart in the next to play position on log machine . You would have to put that command right at the place you want to play your cart. Maybe add some padding (1s delay) as I am not too sure if it could maybe execute it too fast. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Invalid Cart...but it's not....
Maybe the log is not refreshed when the cart is updated with today's content... Is that cart playing on the same day that it is downloaded? I suspect that when your log is loaded at midnight, the cart is not yet available, so at that time it was invalid. On 4 March 2013 13:16, Nathan Steele wrote: > Every monday (i'm told) we have programs that do not play (all of them). > these programs are downloaded into RD by RDCATCH between 1 am and 2 am. > this is done monday-friday, but only mondays seem to fail. > > I looked at todays log and all the programs are red saying invalid cart, > but When I look at the carts in RDlibrary, I don't see what is invalid > about them and it says it played once today. I will have to actually > listen for the next one and see if it really plays or if the host's just > think it didn't play beause it says invalid cart. butWhy does it say > invalid cart anyway? I see no issues with start/end dates or dayparting, > and the carts are there. > > Thanks, > Nathan > > -- > Nathaniel C. Steele > Assistant Chief Engineer/Technical Director > WTRM-FM / TheCrossFM > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Jack Configuration in 2.4 - How to Autostart?
On 30 January 2013 15:30, wrote: > installation sets those 4 files with the 's' attribute (on my system FYI, the "s" attribute is officially called the sticky bit, in case you were looking for more information. Examples of programs making use of the sticky bit are su/sudo (they need root privileges to switch effective UID), mount (for user-mountable filesystems), passwd (to make system calls that allow changing passwords) and a few others. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] RDLogmanager and clocks
On 14 January 2013 10:09, Andy Brown wrote: > so I'm thinking of grabbing source and 'doing > it properly'! That's the spirit! -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] RDLogmanager and clocks
On 14 January 2013 09:54, Andy Brown wrote: > Yeah I'd tried that, the table that was created has ID as a primary key, > and its the primary key thats the issue, It's trying to add a duplicate > entry for ID: That is weird. Usually, in most database implementations for lists, the primary key is an ID that is entirely managed by the database. It is most often set as an autoincrement key, so that the client should not have to worry about counting the primary key, or avoiding collisions when others edit the same table at the same time, etc. It seems that Rivendell explicitly sets the primary key... > insert into `2013_01_15_LOG` set ID=47,COUNT=47,TYPE=0, > SOURCE=3,START_TIME=82807000,GRACE_TIME=0, >CART_NUMBER=999003,TIME_TYPE=0,POST_POINT="N", > TRANS_TYPE=1,COMMENT=""; > ERROR 1062 (23000): Duplicate entry '47' for key 'PRIMARY' > mysql> > rdlogmanager: invalid SQL or failed DB connection: insert into > `2013_01_15_LOG` set ID=8,COUNT=8,TYPE=1, > SOURCE=3,START_TIME=3607000,GRACE_TIME=-1, > CART_NUMBER=0,TIME_TYPE=0,POST_POINT="N", > TRANS_TYPE=1,COMMENT="End\ song" Watch out! Both could be different errors! The one you get by typing the command is due to the line already existing in the log. However, rdlogmanager is complaining that it either could not even connect to the MySQL server, or it got disconnected while executing the query (or else MySQL thought the query was syntactically incorrect). One thing you could do, is wipe the entire table, do some memorable operation (e.g. create some random table named as a timestamp) then generate the log with rdlogmanager (make sure it was all set up before creating that table; you don't want irrelevant queries to show up in the database logs). Afterwards, find the mysql logs with your favorite admin tool, and fetch the line containing the timestamp: now everything that follows this line is related to the log generation. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] RDLogmanager and clocks
I don't have an answer for your problem, but regarding MySQL indexes, that should not be a problem. Databases are supposed to work properly even when indexes are missing. The only drawback are performance issues, when trying to sort, or fetching a specific row in the table, based on that column (e.g. "WHERE col_name > value") UNIQUE constraints are a subclass of indexes, but these are explicitly defined at table creation (or later, for that matter), and if they are missing, you will simply risk having multiple inserts for the same values (might be bad for a Rivendell log, but MySQL will definitely not complain). I am not sure how the mysql library works this out, but I think the multiple inserts with the same ID are due to the SQL command or library failing to execute, then the log manager skips to the next one, to finally quit due to errors... This is a blind guess, but is it possible that you upgraded Rivendell without upgrading the database? ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Windows manager
Fluxbox is a proven, very lightweight window manager that is somewhat easily extensible. OTOH, we at our radio run a full-fledged XFCE4 desktop, because that machine is also used as a general purpose computer for the jockeys (browser, file explorer...). On 26 November 2012 16:21, Rivendell Steve wrote: > Whilst I'm a linux guy, I have very little to do with X windows as virtually > all my kit is headless. > > The studio playout PC is running debian wheezy. Can anyone recommend a > suitable windows manager? If it makes any difference I'm running with two > touch screen displays. > > This is only for playout, there's a seperate PC for email, etc. > > Thanks, > > Steve. > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev > -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Broadcast appliance startup question
Long startup times are often due to pretty simple defects such as a bad hostname in config, or the network is down before a daemon starts. I'm putting my bet on the hostnames in your configuration file. On 16 November 2012 16:23, Nathan Steele wrote: > Why do my broadcast appliance boxes seem to hang forever at "starting > Sendmail" it takes minutes to get past it and finish booting up...I > don't reboot often, but just wondering, when seconds count, I'll be > pulling my hair out waiting for it to boot > > Do I need to configure something? > > -- > Nathaniel C. Steele > Assistant Chief Engineer/Technical Director > WTRM-FM / TheCrossFM > > ___ > Rivendell-dev mailing list > Rivendell-dev@lists.rivendellaudio.org > http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
[RDD] Python module for Rivendell
Hi there, I wanted to let you all know that I am working on a Python module for Rivendell. For the moment, it does not do much, but I plan to add many features. It is currently in active development. You can get it on Github ( http://github.com/sebleblanc/python-rivendell ). If you encounter any problems with it, do not hesitate to fill an issue on the bug tracker ( http://github.com/sebleblanc/python-rivendell/issues ) or to ping me on the mailing list. Right now, what might be of interest is the "normalize.py" utility, which allows one to normalize carts according to loudness levels as set by EBU R128 standards. Required is libebur128, and sox if you actually want to modify the files on the hard drive. (by default it updates the "play_gain" value in the CUTS database table) Cheers! ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Newbie GPIO question
Sorry for the multiple responses... As for what you should do, you should use physical resistors (that would be my prefered solution) as software switches can sometimes, accidentally or for no apparent reason, get reset and then you get erratic behaviour. Those pull-(up|down) resistor software switches actually enable real, internal resistors inside the chips. Using physical resistors outside of the chips will not affect wether the internal ones are enabled or not, so you are better off using physical resistors. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Newbie GPIO question
On 16 August 2012 20:16, Sébastien Leblanc wrote: > then this cable acts as an antenna. Which means that this input will pickup interference as only small variations in the electrical field surrounding it will be enough to toggle the input's state. (electrons pulled by electrical fields will move through internal components of the input, "shifting the balance" from one side to another) This explains the current behaviour you are experiencing. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Newbie GPIO question
On 16 August 2012 13:43, Wayne Merricks wrote: > The manual for this card says "all I/O bits are set to a high impedance > input mode on power up and reset" and you should install pull up or pull > down resistors to get inputs to a known state. > > -- > Wayne Merricks > The Voice Asia Note: I have some background in electrical engineering High impedance input pins mean that these pins have no (actually, a very small) current flowing in or out when you connect it to something. If you leave it floating, its behaviour is only defined by the electronic design. If you physically extend the pins using a cable, then this cable acts as an antenna. You must therefore use pull-up or pull-down resistors to force a known set. A pull-up resistor connects the input to a higher voltage (+5V, e.g.) while a pull-down resistor brings it to ground (0V). Your input will still be connected to the input pin. To find the resistor values, some maths are involved. You need to use a resistor high enough so that it will not disturb your signal source, but you need a value small enough so that it does not seem as if there was nothing connected at all. For switches and simple components, you may use a simple >1K, <1M resistor. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Install error
Name of package is "libmad-devel" -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Install error
On Mon, Apr 23, 2012 at 00:13, Juan Priario wrote: > This is my config.log > > http://pastebin.com/ASGYsZmz > conftest.cc:61:17: fatal error: mad.h: No such file or directory -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Install error
Are you sure libmad has been installed? I am not too familiar with openSUSE, but I think you would need the "developer" libmad package. It includes header files for libmad. ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Trouble compiling (missing -lpthread option for CAE)
I forgot to mention that I am using Arch Linux, which follows some particular guidelines concerning the FHS (Filesystem Hierarchy Standard). Namely, packages should not add files in /usr/libexec or /usr/local. Scripts in the user repository must call "make install" with a specific $(DESTDIR) variable. In this case, it is built from another variable, that is DESTDIR="$pkgdir". What I found out however, is that some Makefile.am files are inconsistent with the rest: most have $(DESTDIR) in front of the targets for `install`, and some do not. This causes the make process to fail with errors. I had to make a couple of patches to these files and run autogen.sh to rebuild the Makefile.in files. Now the package builds, but it is missing some files (daemon scripts and /var/run/rivendell special directories) because they are still copied with "hard" names. They try to install to /etc/init.d instead of $pkgdir/etc/init.d, and I am having a hard time figuring out how the Makefile scripts are crafted... Also there is something wrong with the configure.in script: it forces the RD_LIB_PATH to lib64 on x86_64 platforms. The script should test if the variable is already defined (passed through ./configure) and not override it. Arch Linux, on x86_64, uses "lib" for 64-bit libraries and "lib32" for 32-bit. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
[RDD] Trouble compiling (missing -lpthread option for CAE)
Hello all, I had problems compiling the latest Rivendell release (2.1.3). When getting to compile 'cae', I would get this error: http://pastebin.com/dNCQNNB5 (undefined reference to symbol 'pthread_create@@GLIBC_2.2.5') I managed to make it compile by adding '-lpthread' to the command line. Then I ran make again and everything went ok. However, I am working on an installation script, and obviously, I cannot ask the user to type this command manually. I don't know enough of autotools and I would like to know if somebody could help me or give some pointers as to how I could fix that. I thought of a couple of ways: > Running ./configure, then through shell scripting, add -lpthread to the LIBS > array in cae/Makefile, then run make. Pros: quick and dirty Cons: hackish, fragile > Adding -lpthread to configure.in substitutions and then adding @PTHREAD@ to > cae/Makefile. I feel this is the right way to do it (could be wrong, I am still not sure how autotools works) but I would like to know Fred Gleason's advice on this. I figure that if it's not already there, then there must be something different about my system... I am running configure with options "--libexecdir=/srv/http/rd-bin --prefix=/usr". Thanks! -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] ASI 6012
On Thu, Dec 22, 2011 at 12:24, Nathan Steele wrote: > [root@rdbs ~]# rpm -q hpklinux --filesbypkg > hpklinux /etc/udev/rules.d/45-asihpi.rules > hpklinux /lib/firmware/asihpi/dsp5000.bin > hpklinux /lib/firmware/asihpi/dsp6200.bin > hpklinux /lib/firmware/asihpi/dsp6205.bin > hpklinux /lib/firmware/asihpi/dsp6400.bin > hpklinux /lib/firmware/asihpi/dsp6600.bin > hpklinux /lib/firmware/asihpi/dsp8700.bin > hpklinux /lib/firmware/asihpi/dsp8900.bin > hpklinux /lib/modules/2.6.18-194.26.1.el5/extra/asihpi.ko > hpklinux /usr/bin/asihpiassert > hpklinux /usr/bin/asihpiplay > hpklinux /usr/bin/asihpirec > hpklinux /usr/bin/asihpitest > hpklinux /usr/bin/asihpitune > hpklinux /usr/lib/libhpi.so > hpklinux /usr/lib/libhpi.so.9 > hpklinux /usr/lib/libhpi.so.9.0.0 > hpklinux /usr/share/doc/hpklinux-4.04.07 > hpklinux /usr/share/doc/hpklinux-4.04.07/INSTALL.hpi > hpklinux /usr/share/doc/hpklinux-4.04.07/README.hpi > hpklinux /usr/share/doc/hpklinux-4.04.07/drvnotes.txt I guess you would have to rebuild the hpklinux package since you updated the kernel in-between. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] rdlogmanager or mysql bug.
On Sun, Nov 20, 2011 at 16:15, Pierre-Luc Gauthier wrote: > Hi there, > > I do want to thank everybody on this mailing list for your help and support > and especially Fred Gleason for his wonderful software. > > Last week I've successfully compiled Rivendell 2.1.1 on Arch Linux. > Fooled around with it and started a new database from scratch. So no > database updating here, it's a bran new one. > > So I went into creating simple events and simple clocks and peppered them > into my grid leaving empty slots in the grid. > Generated some logs with rdlogmanager and got them running for a few days in > rdairplay. > So far so good. > > But, today I created many more clocks. Still everything is fine here. The > problem starts when trying to add any clocks to an empty grid. > RdLogManager -> Edit Grids -> "MyGrid" -> Edit -> "Clicked on any empty > clock" -> "Selected whatever clock" -> OK > The Clock is added in the "Grid edition" interface BUT is not added in > mysql. The CLOCKx row in the SERVICE table still displays a NULL value. > Moreover, if I close and reopen the "Grid edition" interface, the newly > added clock has disappeared. > > So I started trouble shooting the situation with RdLogManager: > I can change the clock in an already attributed slot in the grid. > But I cannot change the clock in an empty slot in the grid. > > With mysql-workbench: > I can manually change the clock in an already attributed slot in the grid. > But I cannot change the clock in an empty slot in the grid. > And if I try to do so, I get this message: > ERROR 1118: Row size too large. The maximum row size for the used table > type, not counting BLOBs, is 8126. You have to change some columns to TEXT > or BLOBs > SQL Statement: > UPDATE `Rivendell`.`SERVICES` SET `CLOCK7`='102' WHERE `NAME`='clg2' > > So that's where I am so far. > Any hints on what could resolve the issue? > > Thank you so very much. > > -- > Pierre-Luc Gauthier We apparently fixed it by changing the storage engine for SERVICES from InnoDB to MyISAM. Since Rivendell does not use advanced features of InnoDB (e.g. foreign keys) I do not think there will be an issue with this. -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
Re: [RDD] Rivendell apps unresponsive
On Wed, Nov 16, 2011 at 19:40, Wayne Merricks wrote: > > I'm not sure if this is related to your problems but I had issues when I was > using a Windows MySQL server on a linux host when I couldn't perform a DNS > lookup on the linux host. It was a custom program but it would pause for 2 > seconds or so for every DB query that was run. > > Simply turning off DNS lookup in the mysql config (or I guess adding proper > DNS records for your arch machine on your name servers might help). > > http://www.linuxask.com/questions/disable-dns-hostname-lookup-in-mysql > > Regards, > > Wayne You had it exactly. Name resolution was what I suspected first, so I thought that maybe the hostname was not set in the /etc/hosts file. We are not running on a network so everything is done on the same machine and we do not need name servers. Thank you! -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev
[RDD] Rivendell apps unresponsive
We are having some difficulties with our new radio setup. It is freshly installed, running under Arch Linux with an AudioScience card. Most of the times, when we click some buttons on some apps, Rivendell does not respond for as much as 15 seconds. You will understand that this kind of delay is unacceptable in a production environment. I have tried running ltrace against Rivendell. Results are not conclusive, as they often change, but it might give a hint on what is going on. The command I ran was: ltrace -c -C -S -o -p `pidof rdlogedit` RDLogEdit -> Action: Pressed "Voice tracker" while selecting a log. http://pastebin.com/L6s8eE4x RDLibrary -> Action: Pressed "OK" after editing a cart. 01. http://pastebin.com/dcXsCSXg 02. http://pastebin.com/0a1ipbfW 03. http://pastebin.com/ZQrY2dq6 I always ran `ltrace` only for the time it took to perform the action, then i ^C'd it as soon as the app was responsive. I suspect that it might be an issue with SQL queries taking longer than they should... Anyone having this problem? -- Sébastien Leblanc ___ Rivendell-dev mailing list Rivendell-dev@lists.rivendellaudio.org http://lists.rivendellaudio.org/mailman/listinfo/rivendell-dev