Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
The Razor-Qt Thread
View unanswered posts
View posts from last 24 hours
View posts from last 7 days

Goto page Previous  1, 2, 3, 4, 5, 6  Next  
Reply to topic    Gentoo Forums Forum Index Desktop Environments
View previous topic :: View next topic  
Author Message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Fri Mar 28, 2014 12:26 pm    Post subject: Reply with quote

TomWij wrote:
Sounds like the ebuild misses some dependencies;

I thought ebuild was Gentoo specific? I had to get that for debian.. though I'm considering more and more going for the dev version and help out, at least in the testing dept. But I can't find a way to contact developers there.. and I looked for Feathercoin around here to no avail =/
Quote:
Those are available as dev-libs/boost (dunno which version it needs though) and media-gfx/qrencode.

Yep, that turned out to be a problem(sorry, I did not remember the version at the time :( ), how do I select a particular version for download? I tried the colon ( : ), dots and dashes -- between the name and the version -- also to no avail.

Most recent was emerged as 1.52.0-r6, while the daemon still complains about 1.49.0. Qrencode worked though (thx! :wink: ) but now it complains about a libminiupnpc.so.5.. tried getting net-libs/miniupnpc (v1.8) but nope.
That is why I'm considering going for 9999 (over 9 thousand :lol: ) if it doesn't turn out to be a dead project... :(
Quote:
9999 is a live VCS version; this usually pulls source code from the source repository from upstream, such that you get the latest development code.

I see... but how would I convince emerge to do that, I suppose I must pass it the github location?
Quote:
Available as dev-qt/qtgui:4 in the Portage tree which sounds like another missing dependency.

Yes.. didn't seem to make a difference though.. why would they list this as the sole dep (in the readme) and yet it doesn 't include everything above :? boost would be understandable for it's used by the daemon in the background (separate executable)... the rest seems to be graphics / qt stuff...
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Mar 29, 2014 2:26 pm    Post subject: Reply with quote

MarcoMarin wrote:
I thought ebuild was Gentoo specific? I had to get that for debian..

Yes, ebuild is Gentoo specific afaik. This is confusing. Are you running Gentoo or Debian?
MarcoMarin wrote:
how do I select a particular version for download?

A specific version is `emerge =category/name-version` (eg. `emerge =media-gfx/qrencode-3.3.1`); for more syntax, see "Atom Prefix Operators" in `man 5 ebuild`. Note that the usage of < or > needs you to surround the entire atom with quotes, as to avoid the shell from seeing it as redirection.
MarcoMarin wrote:
but now it complains about a libminiupnpc.so.5.. tried getting net-libs/miniupnpc (v1.8) but nope.

That needs =net-libs/miniunpnc-1.5 which is no longer in the portage tree; to get that back, you could obtain it from http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/net-libs/miniupnpc/?hideattic=0 and put it in a local overlay after which you can emerge it again.
MarcoMarin wrote:
I see... but how would I convince emerge to do that, I suppose I must pass it the github location?

That location is listed in the 9999 ebuild, in EGIT_REPO_URI.
MarcoMarin wrote:
Yes.. didn't seem to make a difference though.. why would they list this as the sole dep (in the readme) and yet it doesn 't include everything above :? boost would be understandable for it's used by the daemon in the background (separate executable)... the rest seems to be graphics / qt stuff...

You might not need it; in other words, it might be an optional dependency for an optional graphical front-end.
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Mon Mar 31, 2014 12:42 pm    Post subject: Reply with quote

TomWij wrote:

Yes, ebuild is Gentoo specific afaik. This is confusing. Are you running Gentoo or Debian?

Sorry, I was referring to the executable I got, it was a .deb package. I don't think I'd have the privilege to bug people here was I not on gentoo.. :-)
Quote:
Note that the usage of < or > needs you to surround the entire atom with quotes, as to avoid the shell from seeing it as redirection.

thanks, that should prevent some major headaches. :-)
Quote:
That needs =net-libs/miniunpnc-1.5 which is no longer in the portage tree; to get that back, you could obtain it from

hmm.. I think this is another hint for me to compile the app for myself, I suppose it would use the most recent ones instead of requiring whatever it was compiled with?
Quote:
That location is listed in the 9999 ebuild, in EGIT_REPO_URI.

Sorry for the n00bishness, I do not understand that statement. Where can I list the "9999 ebuild" to see it? is that referring to some variable/field the github project page?
Quote:
You might not need it; in other words, it might be an optional dependency for an optional graphical front-end.

I'm ever more inclined to help the development directly... :? While I learn to contact the people there (if there is any), I'm trying to set the environment working, using the litecoin-qt project which is already in portage. They've got the same set-up and though I can merge the wallet without problems, I'm having difficulty making my GPU work with openGL.. :?
https://forums.gentoo.org/viewtopic-p-7527118.html#7527118
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Mon Mar 31, 2014 5:18 pm    Post subject: Reply with quote

MarcoMarin wrote:
I suppose it would use the most recent ones instead of requiring whatever it was compiled with?

Most of the time, yes; maybe it occasionally needs an older one.
MarcoMarin wrote:
Where can I list the "9999 ebuild" to see it? is that referring to some variable/field the github project page?

It would be an ebuild file from Gentoo which could be find at /usr/portage/${category}/${name}/${name}-9999.ebuild and there will be a line with EGIT_REPO_URI that is the URI of the Git repository.
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Mon Mar 31, 2014 9:01 pm    Post subject: Reply with quote

but as I said, I can't find it in portage with the emerge command, unless there is some flag to access 9999 versions? (to prevent people getting unstable stuff inadvertently)

Here, just to confirm it exists (or existed)...

https://github.com/no-hope/gentoo-repo/tree/master/net-p2p/feathercoin-qt

says 1 contributor... 11 months ago. :( doesn't sound promising... perhaps after I set litecoin-qt working and the video card mining, I could ask the litecoin-qt guys for help setting this up?
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Mon Mar 31, 2014 9:10 pm    Post subject: Reply with quote

Add the following to /etc/portage/package.accept_keywords to accept the lack of keywords in that ebuild (as by default, you only have stable; to get live, you need to accept the lack of keywords):

Code:
=net-p2p/feathercoin-qt-9999 **
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Wed Apr 09, 2014 11:01 am    Post subject: Reply with quote

Sorry for the delay, I was setting the video card which caused more trouble than expected. :roll:

Anyways.. did not work, after adding that line to packages.accept_keywords, I tried emerging all of the following:

feathercoin,
net-p2p/feathercoin-qt-9999
net-p2p/feathercoin-qt
net-p2p/feathercoinqt

the 9999 one is not a valid atom, and of all the others none were found. The guy there didn't add it to portage I suppose? I guess I'll have to compile it myself...
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Wed Apr 09, 2014 11:20 am    Post subject: Reply with quote

Indeed isn't in Portage; as for the invalid atom, use a = in front.
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Wed Apr 09, 2014 11:25 am    Post subject: Reply with quote

just like in the package.accept line you suggested? or before =9999?
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Wed Apr 09, 2014 11:57 am    Post subject: Reply with quote

Same way as in the line, =net-p2p/feathercoin-qt-9999
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Fri Apr 11, 2014 11:48 am    Post subject: Reply with quote

Didn't work. Onwards to self-compilation, here is what happened:
Code:

In file included from db.cpp:7:0:
db.h:14:20: fatal error: db_cxx.h: No such file or directory
compilation terminated.
make: *** [obj/db.o] Error 1

I did notice something weird when getting the depencies, berkeley DB was version: 0.53 on portage. while depencies lists Berkeley DB 4.8.30.NC

I tried the following to get that package:
Code:
# emerge -a BerkeleyDB

These are the packages that would be merged, in order:

Calculating dependencies... done!
[ebuild   R    ] dev-perl/BerkeleyDB-0.530.0

Would you like to merge these packages? [Yes/No]


readme file listing dependencies is here:
https://github.com/FeatherCoin/Feathercoin/blob/master-0.8/doc/build-unix.md

Should I try and compile that or am I getting the wrong package? (maybe I need a dev package or something?)
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Fri Apr 11, 2014 11:50 am    Post subject: Reply with quote

You'll want sys-libs/db as that provides /usr/include/db*/db_cxx.h
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Fri Apr 11, 2014 12:59 pm    Post subject: Reply with quote

same problem :cry:
Code:
src # make -f makefile.unix
/bin/sh ../share/genbuild.sh obj/build.h
which: no git in (/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/opt/bin:/usr/x86_64-pc-linux-gnu/gcc-bin/4.7.3)
g++ -c -O2 -pthread -Wall -Wextra -Wformat -Wformat-security -Wno-unused-parameter -g -DBOOST_SPIRIT_THREADSAFE -D_FILE_OFFSET_BITS=64 -I/root/FTC/qt/Feathercoin-master-0.8/src -I/root/FTC/qt/Feathercoin-master-0.8/src/obj -DUSE_UPNP=0 -DUSE_IPV6=1 -I/root/FTC/qt/Feathercoin-master-0.8/src/leveldb/include -I/root/FTC/qt/Feathercoin-master-0.8/src/leveldb/helpers -DHAVE_BUILD_INFO -fno-stack-protector -fstack-protector-all -Wstack-protector -D_FORTIFY_SOURCE=2  -MMD -MF obj/db.d -o obj/db.o db.cpp
<command-line>:0:0: warning: "_FORTIFY_SOURCE" redefined [enabled by default]
db.cpp:1:0: note: this is the location of the previous definition
In file included from db.cpp:7:0:
db.h:14:20: fatal error: db_cxx.h: No such file or directory
compilation terminated.
make: *** [obj/db.o] Error 1


emerge brought a version which coincides with the dependency though, so that's reassuring. : )
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Fri Apr 11, 2014 3:50 pm    Post subject: Reply with quote

You need to edit the Makefile such that the directory that contains that file is included in the include path. Paths can be added to the include path with the -I parameter.
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Sat Apr 12, 2014 12:10 pm    Post subject: Reply with quote

tried editing the file, but didn't see any obvious modification (there is a relevant variable name, more on this below), so I went for the -I option.

This did not work. Reading the help it says:
Quote:
-I DIRECTORY, --include-dir=DIRECTORY
Search DIRECTORY for included makefiles.

Which is somewhat ambiguous. "included makefiles" would mean it is only looking for makefiles in that directory, no? Rather than include directory to be used by makefiles?

At least confirmed the missing file is really there, which is reassuring as well : ) So it would indeed be a matter of editing the makefile. (thx ^_^)

but... why it can find all other libs no problem? (maybe it doesn't? maybe db just happens to be the first it tries to use)
why doesn't emerge set the path variable for this lib as well, which I assume is the variable the makefile tries to use? I mean, I see nowhere in the makefile it setting any of those variables, so it must be getting them somewhere?

The relevant variables being: $(BDB_INCLUDE_PATH) and/or $(BDB_LIB_PATH)

the makefile uses:
Code:
DEFS += $(addprefix -I,$(CURDIR) $(CURDIR)/obj $(BOOST_INCLUDE_PATH) $(BDB_INCLUDE_PATH) $(OPENSSL_INCLUDE_PATH))
LIBS = $(addprefix -L,$(BOOST_LIB_PATH) $(BDB_LIB_PATH) $(OPENSSL_LIB_PATH))


I could hardcode it in, if everything else fails, but what if it is systemic and other libs are also not found? There must be a better way to set those 2 variables properly =)

edit: hmm, that -I prefix it adds seem to suggest it is passing those variables as a -I parameter to another makefile? so maybe that's why -I does not work for this one, it ignores it.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Apr 12, 2014 3:31 pm    Post subject: Reply with quote

MarcoMarin wrote:
Quote:
-I DIRECTORY, --include-dir=DIRECTORY
Search DIRECTORY for included makefiles.

Which is somewhat ambiguous.


That in `man make`; but CFLAGS can be found in `man gcc`, this is what it it lists for -I:

Code:
       -I dir
           Add the directory dir to the list of directories to be searched for header files.  Directories named by -I are searched before the standard system include directories.  If the
           directory dir is a standard system include directory, the option is ignored to ensure that the default search order for system directories and the special treatment of system
           headers are not defeated .  If dir begins with "=", then the "=" will be replaced by the sysroot prefix; see --sysroot and -isysroot.


MarcoMarin wrote:
but... why it can find all other libs no problem? (maybe it doesn't? maybe db just happens to be the first it tries to use)


They might be in the right path, therefore not needing special adjustment to find it; or, they might be yet to be used in a later step in the compilation.

MarcoMarin wrote:
why doesn't emerge set the path variable for this lib as well, which I assume is the variable the makefile tries to use? I mean, I see nowhere in the makefile it setting any of those variables, so it must be getting them somewhere?


It receives them through various variables, these differ sometimes from Makefile to Makefile; usually, these are filling in using pkg-config calls in ./configure; for example, for a random package`pkg-config --cflags freetype2` yields "-I/usr/include/freetype2 -I/usr/include/libpng16" and `pkg-config --libs freetype2` yields "-lfreetype". As a result of that, I know what I need to pass to the compiler (as CFLAGS) and the linker (as LFLAGS) in order to respectively find the headers as well as find the library of this random package.

MarcoMarin wrote:
The relevant variables being: $(BDB_INCLUDE_PATH) and/or $(BDB_LIB_PATH)

the makefile uses:
Code:
DEFS += $(addprefix -I,$(CURDIR) $(CURDIR)/obj $(BOOST_INCLUDE_PATH) $(BDB_INCLUDE_PATH) $(OPENSSL_INCLUDE_PATH))
LIBS = $(addprefix -L,$(BOOST_LIB_PATH) $(BDB_LIB_PATH) $(OPENSSL_LIB_PATH))


I could hardcode it in, if everything else fails, but what if it is systemic and other libs are also not found? There must be a better way to set those 2 variables properly =)


In this case I think that BDB_INCLUDE_PATH might be incorrectly or unset; you can define that by passing it after make, like `make BDB_INCLUDE_PATH=/usr/include/db*/` (replace * by the proper version). Though BDB stands for Berkeley DB and not necessarily DB; and thus, I'm not sure if that's entirely correct. You could just add the path at the end of the DEFS line; for example, like this (again, replace *):

Code:
DEFS += $(addprefix -I,$(CURDIR) $(CURDIR)/obj $(BOOST_INCLUDE_PATH) $(BDB_INCLUDE_PATH) $(OPENSSL_INCLUDE_PATH) /usr/include/db*/)
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Sat Apr 12, 2014 4:55 pm    Post subject: Reply with quote

Cool, I'll try setting it for make before I hardcode it (though technically not much different, right? :))

But before I do it, look what I found in the readme file above...

Quote:
if you have to build Berkeley DB yourself:

../dist/configure --enable-cxx
make


is it just a coincidence it cant find (though it is there) exactly a cxx.h? (with the db prefix for the entire lib, thus db_cxx.h)

maybe I should set USE=cxx or enable-cxx for that package.use and re-emerge?

otherwise I'll proceed to manually setting the variable.

PS.: about the uncertainty of which db it is, I think the versioning confirms this is the one, it would be too much a coincidence if it was not.. right? :?
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Apr 12, 2014 5:04 pm    Post subject: Reply with quote

Yes, that might we something worth trying; either adjust the ebuild to add --enable-cxx to ./configure or call emerge with EXTRA_ECONF="--enable-cxx" in front of it, but I think it is already enabled by default so alternatively --disable-cxx could be considered.
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Sat Apr 12, 2014 7:53 pm    Post subject: Reply with quote

Confirmed through -v that cxx was enabled (red font, no minuses). decided not to test disabling it, went for directing make and it worked \o/ At least it proceeded way longer! Thanks.

Next problem is with the boost lib it seems, but it doesn't seem to be a location problem this time? perhaps...

Check it out:
Code:
obj/db.o: In function `wait_for<long int, boost::ratio<1l, 1000000000l> >':
/usr/include/boost/thread/pthread/condition_variable_fwd.hpp:177: undefined reference to `boost::chrono::system_clock::now()'


though the beginning says "db.o".. but I suppose this is the object being built... in fact there are 4 objects with the exact same problem (db.o, init.o, main.o, walletdb.o)

and for each of those 4, not only on line 177 but also on 178 and 181, but this is less important as it is the same file having problems with that same function (now())

I guess it could after all be the same -I problem, if this means the headers for boost isn't being found and thus this reference.

echo $BOOST_INCLUDE_PATH is also empty, just as with BDB..

So I'm typing in:
Code:
# make BDB_INCLUDE_PATH=/usr/include/db4.8/ BOOST_INCLUDE_PATH=/usr/include/boost/ -f makefile.unix


and..... nope :( same problem...

edit: Maybe it's $(BOOST_LIB_PATH) now?
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Apr 12, 2014 8:46 pm    Post subject: Reply with quote

That edit is right, BOOST_LIB_PATH.

As you see, it starts with "db.o" which means it is already an object; what it is trying to do here is "link" the multiple objects together, like putting together the pieces of a puzzle. But, there appears to be a puzzle piece missing; that are the objects of Boost, which are in the Boost libraries. They are usually at `ls -1 /usr/lib*/libboost_*`; in this example, the undefined reference boost::chrono is to /usr/lib/libboost_chrono.so, this (with help of `man gcc`, which reveals -L for specifying the library path and -l for specifying the library itself) translates to -L/usr/lib/ -llibboost_chrono.

Now, it might be you only need -L/usr/lib/ and thus only need to add /usr/lib/ to the variable; but I'm not entirely sure.

If you need more, then you'll soon run from one error to the other; so, to avoid doing that, we can expand to just include all of the libraries at once:

Quote:
( echo -L/usr/lib/ ; cd /usr/lib/ ; for lib in libboost_* ; do echo -l${lib/.so*/} ; done ) | tr '\n' ' '


In my case, this outputs the following list:

Quote:
-L/usr/lib/ -llibboost_atomic-mt -llibboost_atomic -llibboost_atomic -llibboost_chrono-mt -llibboost_chrono -llibboost_chrono -llibboost_date_time-mt -llibboost_date_time -llibboost_date_time -llibboost_exception-mt.a -llibboost_exception.a -llibboost_filesystem-mt -llibboost_filesystem -llibboost_filesystem -llibboost_graph-mt -llibboost_graph -llibboost_graph -llibboost_iostreams-mt -llibboost_iostreams -llibboost_iostreams -llibboost_locale-mt -llibboost_locale -llibboost_locale -llibboost_log-mt -llibboost_log -llibboost_log -llibboost_log_setup-mt -llibboost_log_setup -llibboost_log_setup -llibboost_math_c99-mt -llibboost_math_c99 -llibboost_math_c99 -llibboost_math_c99f-mt -llibboost_math_c99f -llibboost_math_c99f -llibboost_math_c99l-mt -llibboost_math_c99l -llibboost_math_c99l -llibboost_math_tr1-mt -llibboost_math_tr1 -llibboost_math_tr1 -llibboost_math_tr1f-mt -llibboost_math_tr1f -llibboost_math_tr1f -llibboost_math_tr1l-mt -llibboost_math_tr1l -llibboost_math_tr1l -llibboost_prg_exec_monitor-mt -llibboost_prg_exec_monitor -llibboost_prg_exec_monitor -llibboost_program_options-mt -llibboost_program_options -llibboost_program_options -llibboost_random-mt -llibboost_random -llibboost_random -llibboost_regex-mt -llibboost_regex -llibboost_regex -llibboost_serialization-mt -llibboost_serialization -llibboost_serialization -llibboost_signals-mt -llibboost_signals -llibboost_signals -llibboost_system-mt -llibboost_system -llibboost_system -llibboost_test_exec_monitor-mt.a -llibboost_test_exec_monitor.a -llibboost_thread-mt -llibboost_thread -llibboost_thread -llibboost_timer-mt -llibboost_timer -llibboost_timer -llibboost_unit_test_framework-mt -llibboost_unit_test_framework -llibboost_unit_test_framework -llibboost_wave-mt -llibboost_wave -llibboost_wave -llibboost_wserialization-mt -llibboost_wserialization -llibboost_wserialization


As a final note, you might need to replace /usr/lib/ by /usr/lib32/ or /usr/lib64/ as appropriate; depending on the architecture of the application, don't worry as only picking the right one will work.

Since these already have parameters; if you plan to add the whole version, do it at the end of the line after anything else (thus after the last ")" character).
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Sat Apr 12, 2014 9:08 pm    Post subject: Reply with quote

Cool, will try that later/tomorrow morning.

Didn't get this last bit.. "after the last )", you mean after done )? and by "the whole version", you mean together with the previous solutions for the db lib, or you mean the whole line inside the makefile where it places all those variables? (so this should be concatenated at the end?)

thnx : )
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Apr 12, 2014 9:13 pm    Post subject: Reply with quote

No, in the LIBS line; in the place of the "HERE" at the end of "LIBS = $(addprefix -L,$(BOOST_LIB_PATH) $(BDB_LIB_PATH) $(OPENSSL_LIB_PATH)) HERE".
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Sun Apr 13, 2014 12:45 pm    Post subject: Reply with quote

/bin/sh: -c: line 0: syntax error near unexpected token `('

after original line inserted, no 32 nor 64 needed, gave a similar output to yours, so lib/ alone is right. But also tried the output using 32 and 64 (substituting into the 3 libs/), 64 gave a string of empty -l, 32 was even less verbose, giving only 1 -l.

once inside make, the original line gave the error above. Are you sure it doesn't need to call bash from inside the script so it actually execute the line instead of pasting it verbatum, because that is what it seems to be doing when calling g++ (I dunno how to do that though, maybe sh(command) ? )

moreover, those -l in front of the libs, shouldn't it be -L now?

thx.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sun Apr 13, 2014 1:42 pm    Post subject: Reply with quote

MarcoMarin wrote:
/bin/sh: -c: line 0: syntax error near unexpected token `('


Yeah, it is written with Bash instead of POSIX in mind; to quickly take both through tr '\n' ' '

MarcoMarin wrote:
after original line inserted, no 32 nor 64 needed, gave a similar output to yours, so lib/ alone is right. But also tried the output using 32 and 64 (substituting into the 3 libs/), 64 gave a string of empty -l, 32 was even less verbose, giving only 1 -l.


Yeah, lib/ alone sounds good then.

MarcoMarin wrote:
once inside make, the original line gave the error above. Are you sure it doesn't need to call bash from inside the script so it actually execute the line instead of pasting it verbatum, because that is what it seems to be doing when calling g++ (I dunno how to do that though, maybe sh(command) ? )


Either copy the results, or indeed you could add $(bash -c '...') (but you'll need to escape the ' from tr then.

MarcoMarin wrote:
moreover, those -l in front of the libs, shouldn't it be -L now?


The -l are the libraries themselves, the -L are the directories the libraries can be found in; in other words, -L specifies where and -l specifies what.
Back to top
View user's profile Send private message
MarcoMarin
Apprentice
Apprentice


Joined: 12 Mar 2014
Posts: 151

PostPosted: Sun Apr 13, 2014 1:54 pm    Post subject: Reply with quote

TomWij wrote:
MarcoMarin wrote:
/bin/sh: -c: line 0: syntax error near unexpected token `('


Yeah, it is written with Bash instead of POSIX in mind; to quickly take both through tr '\n' ' '


Did not get this or how it relates to the error message?

Quote:

Either copy the results, or indeed you could add $(bash -c '...') (but you'll need to escape the ' from tr then.


ok, help me understand this, please.. '\n' is the parameter to tr, so both should be escaped.. \'\n\' .. right? what are those last 2 '' for? or is ' escaped by repeating it, like ''?
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Desktop Environments All times are GMT
Goto page Previous  1, 2, 3, 4, 5, 6  Next
Page 5 of 6

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum