Re: Re: [Rosegarden-user] Building rosegarden - liblo 0.7

 Re: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 From: - 2006-01-31 18:59:22 ```----- original Nachricht -------- Betreff: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 Gesendet: Di 31 Jan 2006 15:36:07 CET Von: "Brett McCoy" > D. Michael 'Silvan' McIntyre wrote: >=20 > > Now I get it. Math people. You're math people. You read that as 0.7 = >=3D > 0.18=20 > > and said "But wait, 7/10 is not greater than or equal to 18/100" Where= as > I,=20 > > being fundamentally number-averse, didn't read it as a float at all, an= d > just=20 > > thought "18 is greater than 7, duhhhh, what's the problem?" > >=20 > > Interesting. I apologize for being flippant about a legitimate bit of= =20 > > confusion. >=20 > I would have numbered things as 0.07... 0.18 to make that kind of thing= =20 > more obvious to everyone. :-) It's a matter of implementing comparison between 0.7 and 0.18. If you order= them "alphabetically", then 0.7 is greater than 0.18 and comes after it. I= 'm totally disappointed 'bout the usage of version numbers and so I only us= e subversion's "revisions". "One number to find them all..." But as you can= see, looking for >=3D0.7 and having 0.23 installed leads to a "no". To me it's absolutely unimportant how version numbering is done, but I do c= are about decisions made for linking. If scons sorts version numbers alphab= etically, there's something wrong with 0.7<0.23=3Dfalse. I don't need liblo= , but maybe somebody else finds it important to have linked rosegarden 'gai= nst liblo.=0A=0A=0A=0A=0A=0A=0A=0A ```

 Re: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 From: - 2006-01-31 18:59:22 ```----- original Nachricht -------- Betreff: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 Gesendet: Di 31 Jan 2006 15:36:07 CET Von: "Brett McCoy" > D. Michael 'Silvan' McIntyre wrote: >=20 > > Now I get it. Math people. You're math people. You read that as 0.7 = >=3D > 0.18=20 > > and said "But wait, 7/10 is not greater than or equal to 18/100" Where= as > I,=20 > > being fundamentally number-averse, didn't read it as a float at all, an= d > just=20 > > thought "18 is greater than 7, duhhhh, what's the problem?" > >=20 > > Interesting. I apologize for being flippant about a legitimate bit of= =20 > > confusion. >=20 > I would have numbered things as 0.07... 0.18 to make that kind of thing= =20 > more obvious to everyone. :-) It's a matter of implementing comparison between 0.7 and 0.18. If you order= them "alphabetically", then 0.7 is greater than 0.18 and comes after it. I= 'm totally disappointed 'bout the usage of version numbers and so I only us= e subversion's "revisions". "One number to find them all..." But as you can= see, looking for >=3D0.7 and having 0.23 installed leads to a "no". To me it's absolutely unimportant how version numbering is done, but I do c= are about decisions made for linking. If scons sorts version numbers alphab= etically, there's something wrong with 0.7<0.23=3Dfalse. I don't need liblo= , but maybe somebody else finds it important to have linked rosegarden 'gai= nst liblo.=0A=0A=0A=0A=0A=0A=0A=0A ```
 Re: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 From: - 2006-01-31 19:25:20 ```----- original Nachricht -------- Betreff: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 Gesendet: Di 31 Jan 2006 11:16:25 CET Von: "Chris Cannam" > As Michael says, it should be 4-1.2_cvs. Was installed into the kde-bin (/usr/kde/3.4/bin) and had it in /usr/bin to= o (gentoo package). The error message (version mismatch between 4.1.0 and 4= .1.2_cvs) is displayed in a non resizable dialog. My WM (enlightenment) doe= s not allow to resize it and there's only a single line of text visible (wo= rked around with Strg+A, Strg+C, Strg+P in text app). I had this in some ot= her Rosegarden dialog, but do not remember. Not very usable, but also no "s= how stopper". As you can see, I made it 'til the lib-conflict dialog. I did exec scons in= stall but don't know, if it finished successfully. It just said=20 scons: Building targets ... scons: `install' is up to date. scons: done building targets. (after calling scons install for a second time) and rosegarden complains ab= out "1. Maybe not successfully installed" or "2. Not installed into standar= d dir". I did not change a single thing (only copied my rosegarden-config f= rom my .kde/share/apps), only added debug=3Dall. Where did I miss the impor= tant setting (kde is installed in /usr/kde/3.4)? (Is it necessary to uninst= all the 4.1.0?) Thanks so far (maybe I'll become Rosegarden Developer at the end of that pr= ocess *g*)... >=20 >=20 > Chris >=20 >=20 > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log > files > for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat= =3D121642 > _______________________________________________ > Rosegarden-user mailing list > Rosegarden-user@... - use the link below to unsubscribe > https://lists.sourceforge.net/lists/listinfo/rosegarden-user >=20 --- original Nachricht Ende ---- =0A=0A=0A=0A=0A=0A=0A=0A ```
 Re: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 From: - 2006-01-31 19:59:06 ```----- original Nachricht -------- Betreff: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 Gesendet: Di 31 Jan 2006 11:16:25 CET Von: "Chris Cannam" > On Tuesday 31 Jan 2006 06:35, b.ohnsorg@... wrote: > > I built Rosegarden from CVS-Head and run over some message "liblo>=3D0.= 7", > > gentoo's "official" download site offers only liblo 0.18 and 0.22 >=20 scons configure debug=3D1: Checking for dssi >=3D 0.4 ... failed but: [ebuild R ] media-libs/dssi-0.9 # 'R' means that dssi would be re-insta= lled (already exists) This scons-rosegarden-gentoo is not a good combination: \$> equery b dssi # list files included in dssi ebuild [...] /usr/include/dssi.h /usr/lib /usr/lib/dssi /usr/lib/dssi/less_trivial_synth /usr/lib/dssi/less_trivial_synth.so /usr/lib/dssi/less_trivial_synth/LTS_qt /usr/lib/dssi/trivial_sampler /usr/lib/dssi/trivial_sampler.so /usr/lib/dssi/trivial_sampler/trivial_sampler_qt /usr/lib/dssi/trivial_synth.so /usr/lib/pkgconfig /usr/lib/pkgconfig/dssi.pc [...] and: \$> cat /usr/include/dssi.h | grep -i version DSSI version 0.9 as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. #define DSSI_VERSION "0.9" #define DSSI_VERSION_MAJOR 0 #define DSSI_VERSION_MINOR 9 * DSSI_API_Version int DSSI_API_Version; =0A=0A=0A=0A=0A=0A=0A=0A ```
 Re: [Rosegarden-user] Building rosegarden - liblo 0.7 From: Geoff Beasley - 2006-01-31 20:03:22 ```i had a similar issue. rosegarden had been installed previously from portage. all i did was emerge --unmerge rosegarden and all was well. hth, g.r ```
 Re: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 From: Chris Cannam - 2006-01-31 20:25:27 ```> Checking for dssi >= 0.4 ... failed If /usr/lib/pkgconfig/dssi.pc exists and lists a version more recent than 0.4, and you still get that error, then there is something horribly wrong, yes. Can you find out what it tried to test? If nothing else strace might tell you. Chris ```
 Re: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 From: - 2006-02-01 07:26:13 ```----- original Nachricht -------- Betreff: Re: [Rosegarden-user] Building rosegarden - liblo 0.7 Gesendet: Di 31 Jan 2006 23:10:38 CET Von: "D. Michael 'Silvan' McIntyre" > On Tuesday 31 January 2006 1:59 pm, b.ohnsorg@... wrote: >=20 > > It's a matter of implementing comparison between 0.7 and 0.18. If you > order > > them "alphabetically", then 0.7 is greater than 0.18 and comes after it= . >=20 > I have no idea how the comparison is implemented, but it apparently works= =20 > here. >=20 > OK, I looked. What ultimately gets run is something like this command: >=20 > \$pkg-config --modversion liblo --atleast-pkgconfig-version=3D0.7&&echo tr= ue > true >=20 > It's pkg-config making this decision, and it doesn't have a damn thing to= do >=20 > with scons or not scons. I'll bet autoconf uses pkg-config to do this jo= b=20 > too. >=20 > Thus I conclude that either=20 >=20 > a) your environment is broken >=20 > b) your environment is broken >=20 > Next question. I still do not believe, that my environment is broken. My machine's up sin= ce 2 days, 14 hours and a couple of seconds. Only change between yesterday = and today was a X-Server logout over night, a restart of ppp0 (cronjob) and= login this morning. Then some editing took place (printing the tests menti= oned above to stdout) and it worked from one moment to the other. The .pyc-= files are not under CVS-control but I think the build progress is not stabl= e. I made everything new from the start and now it's behaving totally diffe= rent. 1. scons configure debug=3D1 shows various package tests, same like yesterd= ay 2. scons compiles like yesterday, shows the same package tests again (why a= preceeding configure if it does this again and what about my debug=3D1 fro= m step 1, is it saved to some file, do I need to specify it again..?) 3. scons install now links all the .o-files togehter. Yesterday it just fin= ished with scons: done building targets. Now it shows that linking information [huge block of .o-files] Install file: "RGbuild/gui/rosegarden" as "/usr/kde/3.4/bin/rosegarden" Chmod("/usr/kde/3.4/bin/rosegarden", 0755) scons: done building targets. These lines were not present yesterday (see my mail). I uninstalled gentoo'= s rosegarden package and these lines occured. make configure, make and make= install forced the linking, copying and chmod-whatever. This build doesn't= even mention, that there's already an executable (of the wrong kind, maybe= builtin a --version switch or similar). And now rosegarden starts, shows a nice yellow rose splash screen, yellow l= ed-like buttons and nice 3D elements :-). I'll do the testing later on, g2g= 2work Thanks so far and maybe there's some way of improving this, making it more = transparent to a non-scons-familiar user (Afterall I see no advantages of s= cons, maybe I'm too far away and should take a closer look).=0A=0A=0A=0A"Je= tzt Handykosten senken mit klarmobil - 14 Ct./Min.! Hier klicken"=0Awww.kla= rmobil.de/index.html?pid=3D73025=0A ```
 Re: [Rosegarden-user] Building rosegarden - liblo 0.7 From: D. Michael 'Silvan' McIntyre - 2006-01-31 22:05:11 ```On Tuesday 31 January 2006 1:59 pm, b.ohnsorg@... wrote: > It's a matter of implementing comparison between 0.7 and 0.18. If you order > them "alphabetically", then 0.7 is greater than 0.18 and comes after it. I have no idea how the comparison is implemented, but it apparently works here. OK, I looked. What ultimately gets run is something like this command: \$pkg-config --modversion liblo --atleast-pkgconfig-version=0.7&&echo true true It's pkg-config making this decision, and it doesn't have a damn thing to do with scons or not scons. I'll bet autoconf uses pkg-config to do this job too. Thus I conclude that either a) your environment is broken b) your environment is broken Next question. -- D. Michael 'Silvan' McIntyre ---- Silvan Linux fanatic, and certified Geek; registered Linux user #243621 Author of Rosegarden Companion http://rosegarden.sourceforge.net/tutorial/ ```