New port: science/pulseview The sigrok project aims at creating a portable, cross-platform, Free/Libre/Open-Source signal analysis software suite that supports various device types, such as logic analyzers, MSOs, oscilloscopes, multimeters, LCR meters, sound level meters, thermometers, hygrometers, anemometers, light meters, DAQs, dataloggers, function generators, spectrum analyzers, power supplies, GPIB interfaces, and more. WWW: http://sigrok.org/ Fix: Patch attached with submission follows:
Doh - wrong pkg-plist - New shar attached to this email /Uffe
ping ?
Responsible Changed From-To: freebsd-ports-bugs->makc I'll take it.
Any news on this one ? If there is something wrong with this port - then please let me know.
This port have now passed two months in ports-limbo... If there is something wrong with this port I would really like to know.
Responsible Changed From-To: makc->freebsd-ports-bugs Return all PR assigned to me to the pool.
This port have now passed 3 months in ports-limbo... :-( If there is something wrong with this port I would really like to know.
ping ???
Responsible Changed From-To: freebsd-ports-bugs->jgh I'll take it.
Hi, Any news on this ? ports: devel/libsigrok, devel/libsigrokdecode, science/sigrok-cli are not much worth without this one (science/pulseview) They belong to the same suite of signal analysis software /Uffe
BTW: gave it a spin on FreeBSD 10 the source will not build with clang port needs a USE_GCC=yes I'll report back upstream /Uffe
Created attachment 145166 [details] freebsd-ports-science-pulseview.shar.txt
New shar uploaded port updated to current standards (I hope) port now builds with clang
14 months and counting, maybe this edition is the charm
Taking this out of pity.
quick feedback on pkg-plist: you don't need any of those @dirrm lines, those are all standard directories that already exist. If you were going to try an operation, it would be @dirrmtry, but that is not necessary either. Which brings it down to 2 lines, so we don't need a pkg-plist at all. I'm making that change now.
second quick hit, man page doesn't have a .gz extension. That's plain wrong, which means this wasn't poudriere tested, or even "make check-plist" tested.
quick hit #3: the standard man page directory is man/man1, not share/man/man1. We should probably patch something to make it go in the standard directory.
The BUILD_DEPENDS/RUN_DEPENDS was a mess. In reality, they should be LIB_DEPENDS (which converts to RUN_DEPENDS). These are all libraries. Oh, and pkgconfig is missing from dependency list as well. Based on this, I think I am going to have to insist on poudriere test reports (with stage-qa) for your other ports, this one isn't even close. I'm not trying to be harsh, but that's the case.
CMake Error at /usr/local/share/cmake/Modules/FindQt4.cmake:1372 (message): Found unsuitable Qt version "" from NOTFOUND, this code requires Qt 4.x But "USE_QT5= cmake gmake gui" is defined in the makefile. It looks like QT5 is not acceptable for pulseview. On a hunch, I'm going to change this to USE_QT4
Regarding qt5: I have a build here working with qt5 ??? need to go and recheck
You probably have both QT4 and QT5 on your system and pulseview is picking up the already installed QT4. that's my guess...
From trial and error, I've found QT4's moc and uic are both needed, and probably corelib (there's not QT4 core). I keep running it, hitting a config error, fixing, and repeating.
It was finally happy with "USE_QT4= corelib gui moc qmake rcc uic"
=========================================================================== ====> Running Q/A tests (stage-qa) ====> Checking for pkg-plist issues (check-plist) ===> Parsing plist ===> Checking for items in STAGEDIR missing from pkg-plist ===> Checking for directories owned by MTREEs ===> Checking for directories handled by dependencies ===> Checking for items in pkg-plist which are not in STAGEDIR ===> No pkg-plist issues found (check-plist) ====>> Checking for staging violations... done Just to be sure, I'm going to run this in Redports, 8X. I'll pass the link to the current port in a minute.
current state of port: https://redports.org/browser/jmarino/science/pulseview redports run: https://redports.org/buildarchive/20140731225117-55946/ I just noticed the patch names are non-standard (end in diff), so I'm going to rename those patches too.
mixing up platforms - I had a trunk edition of pulseview running on linux - Sorry my bad Thanks for your work - I need to go and read up on all your reports.
And yes you are right I had both QT4 and QT5 on that system
(In reply to uffe from comment #28) > And yes you are right I had both QT4 and QT5 on that system Now you have a real-life example of the value of testing ports in a clean build environment such as poudriere, tinderbox or redports (based on tinderbox). You would have caught: - The QT issues - The missing pkgconf dependency - The DEPENDS issues - the bad pkg-plist (man page didn't end in .gz) With the number of ports PRs you have open, I think it would greatly benefit everyone for you to install poudriere and test each one (and make adjustments if the tests fail) I don't know why this is marked jobs unsafe (it seems cmake ports are generally jobs safe) but for now I'll trust that it really has to be built with -j1.
A commit references this bug: Author: marino Date: Thu Jul 31 23:22:43 UTC 2014 New revision: 363683 URL: http://svnweb.freebsd.org/changeset/ports/363683 Log: Add new port: science/pulseview PR: 179854 Submitted by: uffe (uffe.org) Major QA: marino ================================================================== The sigrok project aims at creating a portable, cross-platform, Free/Libre/Open-Source signal analysis software suite that supports various device types, such as logic analyzers, MSOs, oscilloscopes, multimeters, LCR meters, sound level meters, thermometers, hygrometers, anemometers, light meters, DAQs, dataloggers, function generators, spectrum analyzers, power supplies, GPIB interfaces, and more. WWW: http://sigrok.org/ Changes: head/science/Makefile head/science/pulseview/ head/science/pulseview/Makefile head/science/pulseview/distinfo head/science/pulseview/files/ head/science/pulseview/files/patch-pv_data_signaldata.cpp head/science/pulseview/files/patch-pv_data_signaldata.h head/science/pulseview/pkg-descr
Luckily it passed 8x in Redports, so everything is fine now. Pulseview is in ports.
A commit references this bug: Author: marino Date: Sun Aug 3 16:16:22 UTC 2014 New revision: 363921 URL: http://svnweb.freebsd.org/changeset/ports/363921 Log: science/pulseview: Remove unnecessary RUN_DEPENDS Suggested by: makc PR: 179854 Changes: head/science/pulseview/Makefile