[Mono-list] Daily build errors
Title: Daily build errors Hi! When I want to install the daily build of 2004-10-26 (and -24), I get following error while make install: ---snip--- /bin/sh ../mkinstalldirs /usr/etc/mono /usr/bin/install -c -m 644 config /usr/etc/mono/config /usr/bin/install -c -m 644 browscap.ini /usr/etc/mono/browscap.ini make[3]: Verlassen des Verzeichnisses Verzeichnis »/root/mono-26-Oct-04/mono/data« make[2]: Verlassen des Verzeichnisses Verzeichnis »/root/mono-26-Oct-04/mono/data« make[1]: Verlassen des Verzeichnisses Verzeichnis »/root/mono-26-Oct-04/mono/data« Making install in runtime make[1]: Wechsel in das Verzeichnis Verzeichnis »/root/mono-26-Oct-04/mono/runtime« Making install in . make[2]: Wechsel in das Verzeichnis Verzeichnis »/root/mono-26-Oct-04/mono/runtime« test -f ../../mcs/monoresgen/monoresgen.exe.mdb make[2]: *** [monoresgen.exe.mdb] Fehler 1 make[2]: Verlassen des Verzeichnisses Verzeichnis »/root/mono-26-Oct-04/mono/runtime« make[1]: *** [install-recursive] Fehler 1 make[1]: Verlassen des Verzeichnisses Verzeichnis »/root/mono-26-Oct-04/mono/runtime« make: *** [install-recursive] Fehler 1 [EMAIL PROTECTED] mono]# ---snip--- The file monoresgen.exe.mdb hasn't been there... Thanks! Jochen
RE: [Mono-list] Ask Microsoft: Mono support
Not only can't it come as a surprise, we should all nod our heads in agreement... There's no question that MSFT .Net on Windows is more complete and more solid than Mono right now. It's ok, we shouldn't expect it to be different yet. I would argue that even Linux, with it's much longer history and incredibly greater amount of resource and effort, is just over the last year approaching the stability and completeness of Windows in certain markets. Somebody said in IRC that based on code count, mono so far is estimated to have cost under $1m in Novell-paid time to write. While our leverage is much higher in open source, this is still a far cry from the kind of money Microsoft has thrown behind .Net -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dan Maltes Sent: Wednesday, October 27, 2004 10:13 AM To: DAN MALTES; 'Carl Olsen'; 'Jonathan Pryor'; 'Jonathan Stowe' Cc: 'RoBiK'; 'Mono-List' Subject: RE: [Mono-list] Ask Microsoft: Mono support I will add to this that Amanda's use of the word attempt is obviously meant to communicate that Mono is not as complete or solid as the MS .NET framework. Of course an MS rep will say something like this, of course they want folks to use .NET on Windows. That's a given folks. No need to be offended, or galled, or whatever else. MS was the first to do .NET and they will continue to make the point(right, wrong or otherwise) that it's the most feature rich and stable. Come now, this can't come as a surprise. -Dan -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dan Maltes Sent: Wednesday, October 27, 2004 9:47 AM To: 'Carl Olsen'; 'Jonathan Pryor'; 'Jonathan Stowe' Cc: 'RoBiK'; 'Mono-List' Subject: RE: [Mono-list] Ask Microsoft: Mono support I agree, Amanda Morgan actually says in the second paragraph: Frankly, Mono is just one example of the level of excitement within the developer community around .NET. That sure doesn't sound like a FUD of Mono to me. It sounds to me like Amanda is simply saying we're not involved with Mono, but they are creating something just like the .NET framework because it's an exciting platform. As far as FUD goes, I think people will pull what they want out of what other people say because they have a bias and it serves their agenda. Sociology 101 really. -Dan -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Carl Olsen Sent: Wednesday, October 27, 2004 8:15 AM To: 'Jonathan Pryor'; 'Jonathan Stowe' Cc: 'RoBiK'; 'Mono-List' Subject: RE: [Mono-list] Ask Microsoft: Mono support I didn't find Microsoft's position to be negative at all. I agree that ease of use and standardization is needed on Linux platforms. Mono provides exactly that. I've been programming mostly with Microsoft Visual C# for the past year or two and I find it fairly easy to learn to program with C# on Mono. If I had to use another language, I'd be completely lost. I'm only guessing, but Mono could lead to a flood of applications being developed for Linux platforms which would only increase the attraction to Linux. I like Microsoft products because so many people are using them that there's quite a body of knowledge floating around amongst general users, but I'm not opposed to competition. I find Linux to be difficult to work with, but I'm sure it's only because I'm unfamiliar with it. Competition is a good thing for consumers. I want to thank Microsoft for making C# open source and I want to thank Novell and all the people who have worked on Mono for giving us more options instead of less. -Original Message- From: [EMAIL PROTECTED] [mailto:mono-list- [EMAIL PROTECTED] On Behalf Of Jonathan Pryor Sent: Wednesday, October 27, 2004 6:00 AM To: Jonathan Stowe Cc: RoBiK; Mono-List Subject: Re: [Mono-list] Ask Microsoft: Mono support On Wed, 2004-10-27 at 06:48, Jonathan Stowe wrote: snip/ Yes this all true but (and possibly only in my reading) the attempt by Novell phrase appears to be trying to give it an entirely negative connotation. I suppose that does have a slightly negative connotation. More galling to me is that Novell isn't the only entity involved; lots of individuals are also involved, as is Mainsoft (and likely other companies). - Jon ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list smime.p7s Description: S/MIME cryptographic signature
Re: [Mono-list] Ask Microsoft: Mono support
I think that you are missing the most important words in your final assertion, I have rephrased it into what I think is the more realistic phrasing. Microsoft is happy that the Mono project gets more people to learn and use .Net. We can point to it if anybody complains that .Net isn't portable. But it ever poses a threat to us we reserve the right to ATTEMPT to pull the rug out from under it, and with the bumbling of our legal counsel, there is a good chance that all we would do is smear a bit of mud on our own faces, and all of this assumes that these actions don't place us in further danger of severe Justice Department and EU sanctions.. smime.p7s Description: S/MIME cryptographic signature
Re: [Mono-list] Daily build errors
Hi, When I want to install the daily build of 2004-10-26 (and -24), I get following error while make install: The file monoresgen.exe.mdb hasn't been there... Four questions. First - why are you doing this as root? Root is there for a purpose and it isn't just so you don't have to keep typing su to change to it. On a machine, it can be considered a major deity. It has the power of life, death and existance on a box. Mess something up as root and you can (and probably *will*) do some serious damage! Second, is this a compile from cvs and if it is, is it via and update and if it is from an update, what happens if you ditch that version of mono/mcs and do a fresh checkout? Third, what distro are you using? I know with Fedora Core, things aren't always happy in that /usr/local/lib/pkgconfig isn't always checked - this could be something similar. Fourth, when you ran autogen.sh, did you specify anything with -- prefix=? (ie. --prefix=/usr). Again, problems can occur if you're in as root or have specified somewhere insane for your libraries and apps to live. TTFN Paul -- Trust me, I know what I'm doing - Det. Sledgehammer signature.asc Description: This is a digitally signed message part
[Mono-list] Mono, ARM and .NETCF
Hi! I'm new to this list - hello everybody :) I'm .net programmer for 1.5 year, now I'm testing mono and I like it very much. I'm also developing some software for Pocket PC using MS .NET CF. And here are my questions - I couldn't find answers in archives of this list: I would like to write one application and run it both on MS Pocket PC and some linux distro for handhelds (ie. on Sharp Zaurus or on iPAQ with Familiar). It would be realy nice :). But how does support of ARM look like in mono? AFAIK there is now JIT compiler for this platform... I saw this page http://amy.udd.htu.se/~malte/mono/, but mono version described there is 0.26 and it is using mint to run. How does things changed since 0.26? Another thing - as I suppose, mono on ARM (linux) is the full version of mono platform, not like the .NETCF, which is only a subset of full .NET? And most important thing: how to run mono (especially GTK#) on Pocket PC 2002/2003? Is there any way to do this? Or maybe I can write and compile code using mono (and SWF) and run it on .NETCF (the same binary on linux/mono)? Thanks in advance for any answers:) Best regards, Lukasz Piatkowski ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
[Mono-list] DirectoryServices class
I was just wondering, since the class status is down right now for 2.0 head.., what the current state of the System.DirectoryServices namespace is... I've had quite a bit of success working with some test programs and OpenLdap, but the only thing I cannot do is write to the directory store. DirectoryEntry.CommitChanges() runs and exits ok, but doesn't seem to really do anything. Then next time I run it, or read the directory in GQ the entry is still exactly the same. Right now I'm using Mono-1.0 with Linux Mandrake 10.1 (actually Cooker, I'm going to redo my Laptop with 10.1 PowerPack now that its out) I don't see any listing for the Commit method in the class status mentioned earlyer, and I'm actually using a slightly modified example from the MSDN site. I just modified to support a specific entry and added authentication to have write access (was getting a Novell.Ldap blah blah error when it wasn't authenticating properly) ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
[Mono-list] DirectoryServices class
I was just wondering, since the class status is down right now for 2.0 head.., what the current state of the System.DirectoryServices namespace is... I've had quite a bit of success working with some test programs and OpenLdap, but the only thing I cannot do is write to the directory store. DirectoryEntry.CommitChanges() runs and exits ok, but doesn't seem to really do anything. Then next time I run it, or read the directory in GQ the entry is still exactly the same. Right now I'm using Mono-1.0 with Linux Mandrake 10.1 (actually Cooker, I'm going to redo my Laptop with 10.1 PowerPack now that its out) I don't see any listing for the Commit method in the class status mentioned earlyer, and I'm actually using a slightly modified example from the MSDN site. I just modified to support a specific entry and added authentication to have write access (was getting a Novell.Ldap blah blah error when it wasn't authenticating properly) ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
[Mono-list] Re: Ask Microsoft: Mono support
-Original Message- From: Paolo Molaro [EMAIL PROTECTED] Java would be close, but it's limited in many ways, so it wouldn't be good enough. Can you be more specific about why Java is close [but not] good enough for a good development platform? I'm not trying to be argumentative, just to understand your position. Another way to put this is, say .Net didn't exist, why would I develop a product today with mono instead of Java, strictly on its own merits? IMHO, if they decide to attack Mono, as detailed in the FAQ, if they have patents that can stand in court, we'll remove the features or work around the patents. This may break compatibility, but we have our own code and we'll adapt it. I read that in the FAQ. Why not do it now and make mono completely unencumbered? You say that compatibility isn't the primary concern. An IP-free implementation would eliminate any legal doubt that is slowing down commercial adoption. More adoption means more users and developers and a better environment, right? I think you underestimate the paranoia of lawyers, especially at big companies. Although you and I can recognize FUD for what it is, lawyers take more convincing. ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
Re: [Mono-list] Re: Ask Microsoft: Mono support
Ed Burnette wrote: Can you be more specific about why Java is close [but not] good enough for a good development platform? I'm not trying to be argumentative, just to understand your position. Another way to put this is, say .Net didn't exist, why would I develop a product today with mono instead of Java, strictly on its own merits? Almost every aspect of Mono/.NET is Java-like but better. P/Invoke vs JNI Reflection (Attributes vs. null) Remoting Value types are actually types The class library is (seems?) more complete (e.g. System.Web). Extremely easy web integration (mod_mono included, vs. Tomcat, etc.) Mono is open source I mean, come on! :) But, look, it's not up to the Mono developers to justify to the world why everyone should use Mono over something else. If you like Mono, you use it. If not, you use something else, and everyone is still happy and life goes on. -- - Joshua Tauberer http://taubz.for.net ** Nothing Unreal Exists ** ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
Re: AW: [Mono-list] Daily build errors
Hi, First - why are you doing this as root? Root is there for a purpose and it isn't just so you don't have to keep typing su to change to it. On a machine, it can be considered a major deity. It has the power of life, death and existance on a box. Mess something up as root and you can (and probably *will*) do some serious damage! As in the real world: if you have rights you can missuse them. If you don't have any rights, you can't do your job. You've missed the point. You do everything as a user and until you need to (for instance, installing into a directory owned by root, such as /usr or /var), you stay like that. As a user, you have all the rights you need. Second, is this a compile from cvs and if it is, is it via and update and if it is from an update, what happens if you ditch that version of mono/mcs and do a fresh checkout? This is a compile from the daily build downloaded from the mono website. Nothing CVS or similar. Try the cvs route - it's much easier to bugtrack! Third, what distro are you using? I know with Fedora Core, things aren't always happy in that /usr/local/lib/pkgconfig isn't always checked - this could be something similar. Sorry, I had really forgotten this important information. Here it is: RedHat 9 RH9 was fine with pkgconfig in /usr/local/lib Fourth, when you ran autogen.sh, did you specify anything with -- prefix=? (ie. --prefix=/usr). Again, problems can occur if you're in as root or have specified somewhere insane for your libraries and apps to live. I used following lines: ./autogen.sh --prefix=/usr make make install here it crashes because of the missing file As I say, use the cvs route. Full instructions are at http://www.all-the-johnsons.co.uk/mono/mono-compiling.html TTFN Paul -- Trust me, I know what I'm doing - Det. Sledgehammer signature.asc Description: This is a digitally signed message part
Re: [Mono-list] Re: Ask Microsoft: Mono support
On Thu, 28 Oct 2004 10:45:40 -0400, Ed Burnette [EMAIL PROTECTED] wrote: -Original Message- From: Paolo Molaro [EMAIL PROTECTED] IMHO, if they decide to attack Mono, as detailed in the FAQ, if they have patents that can stand in court, we'll remove the features or work around the patents. This may break compatibility, but we have our own code and we'll adapt it. I read that in the FAQ. Why not do it now and make mono completely unencumbered? You say that compatibility isn't the primary concern. An IP-free implementation would eliminate any legal doubt that is slowing down commercial adoption. More adoption means more users and developers and a better environment, right? Thats not really how it works. If there were four active patents total, and they were all by Microsoft about things within the .Net Framework - then sure, it would make sense to avoid it. But patents are a horrible reality of the software engineering environment today. You really cannot tie your own shoes without violating someone's patent anymore - and at that point, it really becomes a business and legal problem. In addition. the penalties for patent infringement skyrocket if you knowingly violate a patent. It is far better to just not look, play ignorant to the whole issue, and be reactionary. Or to summarize, in response to the 'why not switch now' question, my follow-on question would be 'to what?' What is there in any software engineering vertical which is not potentially patent-encombered in the US? _Every_ open-source project has an equal potential for patent infringement. Mono just gets a much larger amount of FUD over it because villifying Microsoft is so easy and fashionable. I think you underestimate the paranoia of lawyers, especially at big companies. Although you and I can recognize FUD for what it is, lawyers take more convincing. When there are enforced patents on ideas like 'plugins' and 'having one program help another', lawyers really need to change their outlook on what is the best outlook on patents. -David Waite ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
AW: AW: [Mono-list] Daily build errors
Hi Paul, I needed to update my mono runtime because I cannot compile via CVS. In folder mono, I started with: ./autogen.sh --prefix=/usr make make install [..] /bin/sh ../../mkinstalldirs /usr/lib/mono/1.0 /usr/bin/install -c -m 644 mscorlib.dll /usr/lib/mono/1.0/mscorlib.dll /usr/bin/install -c -m 644 ./mscorlib.dll.mdb /usr/lib/mono/1.0/mscorlib.dll.mdb MONO_PATH=. ../../mono/mini/mono --config ../../data/config ./../gacutil.exe /i ./Accessibility.dll /f /package 1.0 /gacdir /usr/lib /root /usr/lib Corlib not in sync with this runtime: expected corlib version 28, found 26. Download a newer corlib or a newer runtime at http://www.go-mono.com/daily. make[3]: *** [install-data-local] Fehler 1 make[3]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/runtime/net_1_1« make[2]: *** [install-am] Fehler 2 make[2]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/runtime/net_1_1« make[1]: *** [install-recursive] Fehler 1 make[1]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/runtime« make: *** [install-recursive] Fehler 1 For testing purposes, I also started that: [EMAIL PROTECTED] mono]# make bootstrap rm -f ./../mcs/mcs/mcs.exe ./../mcs/class/lib/default/mscorlib.dll ./../mcs/class/lib/default/System.dll ./../mcs/class/lib/default/System.Xml.dll ./../mcs/class/lib/default/Mono.CompilerServices.SymbolWriter.dll for dir in jay mcs class/corlib class/System; do \ (cd ./../mcs/$dir make PROFILE=default NO_SIGN_ASSEMBLY=yes USE_BOOT_COMPILE=yes BOOTSTRAP_MCS=mcs -d:BOOTSTRAP_WITH_OLDLIB) || exit 1 ; \ done make[1]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mcs/jay« make all-local make[2]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mcs/jay« make[2]: Für das Target »all-local« gibt es nichts zu tun. make[2]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mcs/jay« make[1]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mcs/jay« make[1]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mcs/mcs« make all-local make[2]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mcs/mcs« mcs -d:BOOTSTRAP_WITH_OLDLIB -d:NET_1_1 -d:ONLY_1_1 -g /target:exe /out:mcs.exe cs-parser.cs @mcs.exe.sources Corlib not in sync with this runtime: expected corlib version 28, found 22. Download a newer corlib or a newer runtime at http://www.go-mono.com/daily. make[2]: *** [mcs.exe] Fehler 1 make[2]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mcs/mcs« make[1]: *** [all.real] Fehler 2 make[1]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mcs/mcs« make: *** [mcs-do-basic-build] Fehler 1 [EMAIL PROTECTED] mono]# I'm wondering why it finds sometimes runtime version 22, and the next time 26... In any case, this is the cause why I tried to build from the daily builds completely new. I also have tried to remove anything which has been installed with mono (/usr and /usr/local). This helped last time when there was mono installed in /usr and in another version in /usr/local. After this, I removed all mono-packages with red carpet and then reinstalled them again to get a working start point. But it hasn't helped this time :( [EMAIL PROTECTED] mono-cvs-auth]# cat cleanupmonoinstallation.sh #!/bin/sh cd /usr rm -f -R -d bin/*.exe bin/*.dll bin/mbas* bin/mcs* bin/mono* bin/libmono* bin/libmint* lib/mono include/mono etc/mono share/mono cd /usr/local rm -f -R -d bin/*.exe bin/*.dll bin/mbas* bin/mcs* bin/mono* bin/libmono* bin/libmint* lib/mono include/mono etc/mono share/mono [EMAIL PROTECTED] mono-cvs-auth]# -Ursprüngliche Nachricht- Von: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Im Auftrag von Paul Gesendet: Donnerstag, 28. Oktober 2004 17:47 An: Mono List Betreff: Re: AW: [Mono-list] Daily build errors Hi, First - why are you doing this as root? Root is there for a purpose and it isn't just so you don't have to keep typing su to change to it. On a machine, it can be considered a major deity. It has the power of life, death and existance on a box. Mess something up as root and you can (and probably *will*) do some serious damage! As in the real world: if you have rights you can missuse them. If you don't have any rights, you can't do your job. You've missed the point. You do everything as a user and until you need to (for instance, installing into a directory owned by root, such as /usr or /var), you stay like that. As a user, you have all the rights you need. Second, is this a compile from cvs and if it is, is it via and update and if it is from an update, what happens if you ditch that version of mono/mcs and do a fresh checkout? This is a compile from the daily build downloaded from the mono website. Nothing CVS or similar. Try the cvs route - it's much easier to bugtrack! Third, what distro are you using? I know
WG: AW: [Mono-list] Daily build errors
Hi Paul, again, Now I removed my complete old content of my CVS folder and checked out again. export CVS_RSH=ssh export [EMAIL PROTECTED]:/cvs/public rm -f -R -d mcs mono xsp monkeyguide monocov cvs -z3 co mcs mono xsp monkeyguide monocov cd mono ./autogen.sh --prefix=/usr make This is the new and well-known error message: -2.0 -I/usr/lib/glib-2.0/include -DGC_LINUX_THREADS -D_GNU_SOURCE -D_REENTRANT -D_FILE_OFFSET_BITS=64 -DUSE_COMPILER_TLS -g -O2 -fno-strict-aliasing -g -Wall -Wunused -Wmissing-prototypes -Wmissing-declarations -Wstrict-prototypes -Wmissing-prototypes -Wnested-externs -Wpointer-arith -Wno-cast-qual -Wcast-align -Wwrite-strings -c mono-cov.c -MT mono-cov.lo -MD -MP -MF .deps/mono-cov.TPlo -o mono-cov.o /dev/null 21 mv -f .libs/mono-cov.lo mono-cov.lo /bin/sh ../../libtool --mode=link gcc -g -O2 -fno-strict-aliasing -g -Wall -Wunused -Wmissing-prototypes -Wmissing-declarations -Wstrict-prototypes -Wmissing-prototypes -Wnested-externs -Wpointer-arith -Wno-cast-qual -Wcast-align -Wwrite-strings -o libmono-profiler-cov.la -rpath /usr/lib mono-cov.lo ../../mono/mini/libmono.la -lnsl -lpthread -lm -lrt rm -fr .libs/libmono-profiler-cov.la .libs/libmono-profiler-cov.* .libs/libmono-profiler-cov.* gcc -shared mono-cov.lo -Wl,--rpath -Wl,/mono-dev/mono-cvs-auth/mono/mono/mini/.libs -L/usr/lib ../../mono/mini/.libs/libmono.so -lnsl -lpthread -lm -lrt -Wl,-soname -Wl,libmono-profiler-cov.so.0 -o .libs/libmono-profiler-cov.so.0.0.0 (cd .libs rm -f libmono-profiler-cov.so.0 ln -s libmono-profiler-cov.so.0.0.0 libmono-profiler-cov.so.0) (cd .libs rm -f libmono-profiler-cov.so ln -s libmono-profiler-cov.so.0.0.0 libmono-profiler-cov.so) ar cru .libs/libmono-profiler-cov.a mono-cov.o ranlib .libs/libmono-profiler-cov.a creating libmono-profiler-cov.la (cd .libs rm -f libmono-profiler-cov.la ln -s ../libmono-profiler-cov.la libmono-profiler-cov.la) make[3]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/mono/profiler« make[3]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mono/mono« make[3]: Für das Target »all-am« gibt es nichts zu tun. make[3]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/mono« make[2]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/mono« Making all in docs make[2]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mono/docs« make[2]: Für das Target »all« gibt es nichts zu tun. make[2]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/docs« Making all in data make[2]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mono/data« Making all in net_1_1 make[3]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mono/data/net_1_1« make[3]: Für das Target »all« gibt es nichts zu tun. make[3]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/data/net_1_1« Making all in net_2_0 make[3]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mono/data/net_2_0« make[3]: Für das Target »all« gibt es nichts zu tun. make[3]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/data/net_2_0« make[3]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mono/data« make[3]: Für das Target »all-am« gibt es nichts zu tun. make[3]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/data« make[2]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/data« Making all in runtime make[2]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mono/runtime« Making all in . make[3]: Wechsel in das Verzeichnis Verzeichnis »/mono-dev/mono-cvs-auth/mono/runtime« test -f ../../mcs/monoresgen/monoresgen.exe make[3]: *** [monoresgen.exe] Fehler 1 make[3]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/runtime« make[2]: *** [all-recursive] Fehler 1 make[2]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono/runtime« make[1]: *** [all-recursive] Fehler 1 make[1]: Verlassen des Verzeichnisses Verzeichnis »/mono-dev/mono-cvs-auth/mono« make: *** [all] Fehler 2 [EMAIL PROTECTED] mono]# -Ursprüngliche Nachricht- Von: Jochen Wezel (CompuMaster GmbH) Gesendet: Donnerstag, 28. Oktober 2004 18:48 An: 'Paul'; 'Mono List' Betreff: AW: AW: [Mono-list] Daily build errors Hi Paul, I needed to update my mono runtime because I cannot compile via CVS. In folder mono, I started with: ./autogen.sh --prefix=/usr make make install [..] /bin/sh ../../mkinstalldirs /usr/lib/mono/1.0 /usr/bin/install -c -m 644 mscorlib.dll /usr/lib/mono/1.0/mscorlib.dll /usr/bin/install -c -m 644 ./mscorlib.dll.mdb /usr/lib/mono/1.0/mscorlib.dll.mdb MONO_PATH=. ../../mono/mini/mono --config ../../data/config ./../gacutil.exe /i ./Accessibility.dll /f /package 1.0 /gacdir /usr/lib /root /usr/lib Corlib not in sync with this runtime:
[Mono-list] Custom Marshalling
I've implemented some custom marshalling for an external call so that the parameters being sent gets massaged as required by the external routine. In addition, it is supposed to massage data after the call. However, while it will do so on the way out, it's not being driven the other way. I found that if I specified ref on the parameters being passed then the MarshalManagedToNative method will not be driven, if I specify nothing or out then neither the ToNative or ToManaged methods are being called. [DllImport(libname, EntryPoint = scumbag)] private unsafe static extern void scumbag (ref XXXCB cb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaByteBuffer))] XXXBuffer fb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaByteBuffer))] XXXBuffer rb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaByteBuffer))] XXXBuffer sb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaByteBuffer))] XXXBuffer vb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaIntBuffer))] XXXIB ib); Neale ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
Re: AW: AW: [Mono-list] Daily build errors
Hi, I needed to update my mono runtime because I cannot compile via CVS. What is causing the problems with cvs? In folder mono, I started with: Corlib not in sync with this runtime: expected corlib version 28, found 26. Download a newer corlib or a newer runtime at http://www.go-mono.com/daily. Grab a copy of monocharge-latest (wget http://www.go- mono.com/daily/monocharge-latest.tar.gz), install and you should find that the above error just vanishes. TTFN Paul -- Trust me, I know what I'm doing - Det. Sledgehammer signature.asc Description: This is a digitally signed message part
Re: AW: AW: [Mono-list] Daily build errors
Hi, add this to your build commands: make uninstall make clean make bootstrap-clean From rom memory, that won't get around the corlib being out of sync problem though. TTFN Paul -- Trust me, I know what I'm doing - Det. Sledgehammer signature.asc Description: This is a digitally signed message part
Re: WG: AW: [Mono-list] Daily build errors
Hi, export CVS_RSH=ssh export [EMAIL PROTECTED]:/cvs/public Try rm -rf mcs mono xsp (I don't know what the others are - I'd imagine though one of them is monodoc) export CVSROOT=:pserver:[EMAIL PROTECTED]:/mono cvs -z3 co mcs mono xsp monodoc gtk-sharp cd mono ./autogen.sh --prefix=/usr make get-monolite-latest make bootstrap-monolite su make install See what that gives you. TTFN Paul -- Trust me, I know what I'm doing - Det. Sledgehammer signature.asc Description: This is a digitally signed message part
Re: AW: AW: [Mono-list] Daily build errors
Hi, I wonder if it really actually works now, after you said it might not. how do we confirm? :) Dunno is the simplest answer. Actually, it's simple. Find a few bods willing to try both hypothesis who are not using the same distro. I can only speak from a RH/FC front and on that side, it was a pain in the backside over the install of monocharge. It gets even messier if monocharge wasn't installed in /usr, but in /usr/local! TTFN Paul -- Trust me, I know what I'm doing - Det. Sledgehammer signature.asc Description: This is a digitally signed message part
Re: AW: AW: [Mono-list] Daily build errors
I think there should be no room left for guessing. it is clear that one must uninstall previous installation. cleaning up in build environment was always a must if you want to get a clean build. by actually works I wanted to know if what I suggested worked for Jochen. On 10/28/04 12:19 PM, Paul [EMAIL PROTECTED] wrote: Hi, I wonder if it really actually works now, after you said it might not. how do we confirm? :) Dunno is the simplest answer. Actually, it's simple. Find a few bods willing to try both hypothesis who are not using the same distro. I can only speak from a RH/FC front and on that side, it was a pain in the backside over the install of monocharge. It gets even messier if monocharge wasn't installed in /usr, but in /usr/local! TTFN Paul ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
Re: AW: AW: [Mono-list] Daily build errors
Hi, I think there should be no room left for guessing. it is clear that one must uninstall previous installation. cleaning up in build environment was always a must if you want to get a clean build. I agree. What would be useful would be a digest (as I get from the rawhide autobuilder on the Fedora test list) with a change log and any new files. That way the missing file problem for cvs users would vanish. For this problem (corlib out of date), I think the instructions need to be made clearer. Perhaps I'll add it to the mono-compiling page of mine. From rom what I can see, the problem is three fold 1. If you have the rpms installed 2. If you are installing to /usr/local 3. If you have something in /usr/local, haven't deinstalled and install into /usr 1. is simple enough. Combine it with 2, it becomes a pain. 3 is a case of getting a list of binaries and libs (etc) created and manually removing them. The corlib problem comes (from my experience) when combining 1 and 2. mcs gets confused and screams! by actually works I wanted to know if what I suggested worked for Jochen. It will be interesting to see. TTFN Paul -- Trust me, I know what I'm doing - Det. Sledgehammer signature.asc Description: This is a digitally signed message part
Re: [Mono-list] Custom Marshalling
I'm a little confused by your description. I am not trying to be difficult, but you use the pronoun it in many places where the antecedent is unclear. Also, the phrase on its way out is not clear. Perhaps if you demonstrated the problem using a P/Invoke method with fewer parameters and explained any problems with in, out, and ref parameters separately, I could better understand. Marcus On Thursday 28 October 2004 1:12 pm, [EMAIL PROTECTED] wrote: I've implemented some custom marshalling for an external call so that the parameters being sent gets massaged as required by the external routine. In addition, it is supposed to massage data after the call. However, while it will do so on the way out, it's not being driven the other way. I found that if I specified ref on the parameters being passed then the MarshalManagedToNative method will not be driven, if I specify nothing or out then neither the ToNative or ToManaged methods are being called. [DllImport(libname, EntryPoint = scumbag)] private unsafe static extern void scumbag (ref XXXCB cb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaByteBuffer))] XXXBuffer fb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaByteBuffer))] XXXBuffer rb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaByteBuffer))] XXXBuffer sb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaByteBuffer))] XXXBuffer vb, [MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (AdaIntBuffer))] XXXIB ib); ___ Mono-list maillist - [EMAIL PROTECTED] http://lists.ximian.com/mailman/listinfo/mono-list
[Mono-list] Compile from CVS on Windows
I tried to run the win32 script that gets everything from CVS and does a build. Everything seems to be ok in the CVS part but when compiling it gets stuck on the following screen make[3]: Leaving directory `/cygdrive/c/sandbox/monosrc/mono/mono/interpreter' Making all in mini make[3]: Entering directory `/cygdrive/c/sandbox/monosrc/mono/mono/mini' /bin/bash ../../libtool --mode=link --tag=CC gcc -mno-cygwin -g -O -I /cygdrive/ c/sandbox/monosrc/install/include -I /cygdrive/c/sandbox/monosrc/install/icu/inc lude -fno-strict-aliasing -g -Wall -Wunused -Wmissing-prototypes -Wmissing-decla rations -Wstrict-prototypes -Wmissing-prototypes -Wnested-externs -Wpointer-ari th -Wno-cast-qual -Wcast-align -Wwrite-strings -L/cygdrive/c/sandbox/monosrc/in stall/lib -L/cygdrive/c/sandbox/monosrc/install/icu/lib -o genmdesc.exe genmdes c-genmdesc.o genmdesc-helpers.o ../../mono/utils/libmonoutils.la ../../mono/meta data/opcodes.lo -lm -L/cygdrive/c/sandbox/monosrc/install/lib -lgthread-2.0 -lgl ib-2.0 -lintl -liconv -Wl,--export-dynamic -L/cygdrive/c/sandbox/monosrc/insta ll/lib -lgmodule-2.0 -lglib-2.0 -lintl -liconv -lws2_32 -lpsapi -lole32 gcc -mno-cygwin -g -O -I /cygdrive/c/sandbox/monosrc/install/include -I /cygdriv e/c/sandbox/monosrc/install/icu/include -fno-strict-aliasing -g -Wall -Wunused - Wmissing-prototypes -Wmissing-declarations -Wstrict-prototypes -Wmissing-prototy pes -Wnested-externs -Wpointer-arith -Wno-cast-qual -Wcast-align -Wwrite-strings -o genmdesc.exe genmdesc-genmdesc.o genmdesc-helpers.o ../../mono/metadata/.lib s/opcodes.o -Wl,--export-dynamic -L/cygdrive/c/sandbox/monosrc/install/lib -L/c ygdrive/c/sandbox/monosrc/install/icu/lib ../../mono/utils/.libs/libmonoutils.a /usr/lib/libgthread-2.0.dll.a -L/usr/lib /usr/lib/libgmodule-2.0.dll.a /usr/lib/ libglib-2.0.dll.a /usr/lib/libintl.dll.a /usr/lib/libiconv.dll.a -lws2_32 -lpsap i -lole32 ./genmdesc ./cpu-pentium.md cpu-pentium.h pentium_desc
RE: [Mono-list] Custom Marshalling
I'll attempt to think before I type this time :-)
I have a a C routine I wish to call that takes the following parameters:
typedef struct XXXCB {
short a;
char b[2];
};
void scumbag(XXXCB *cb, char *v);
for our purposes v is an arbitrarily long string of bytes in the range 0x00-0x7f.
Depending on the contents of b variable v can be input to or output from the
function scumbag.
In C# I'd code XXXCB as:
public struct XXXCB {
public short a;
[MarshalAS(UnmanagedType.ByValArray, SizeConst=2)}
public byte[] b;
}
Now, because the variable v can be a different length of each call (up to 32K is
size) I can't use MarshalAs(UnmanagedType.ByValArray) as mcs demands that SizeConst=
be coded (using SizeParamIndex= doesn't work, in fact I don't think it's supported).
So what I've done is created a class XXXByteBuffer to Marshal the data using
ICustomMarshaler.
Therefore, I've coded the call prototype as:
[DllImport(libname, EntryPoint = scumbag)]
private static extern void scumbag (ref XXXCB cb,
[MarshalAs(UnmanagedType.CustomMarshaler, MarshalTypeRef = typeof (XXXByteBuffer))]
v);
I created the necessary methods (like ManagedToNative and NativeToManaged) in
XXXByteBuffer. When I call scumbag using parameters conforming to the above prototype
the ManagedToNative method is being driven prior to the call but NativeToManaged is
not being called on the return. If I change the prototype to specify ref v or out
v then neither method is being driven.
Neale
-Original Message-
I'm a little confused by your description. I am not trying to be difficult,
but you use the pronoun it in many places where the antecedent is unclear.
Also, the phrase on its way out is not clear.
Perhaps if you demonstrated the problem using a P/Invoke method with fewer
parameters and explained any problems with in, out, and ref parameters
separately, I could better understand.
___
Mono-list maillist - [EMAIL PROTECTED]
http://lists.ximian.com/mailman/listinfo/mono-list
RE: [Mono-list] Custom Marshalling
You're making this far too difficult on yourself. See below.
On Thu, 2004-10-28 at 19:57, [EMAIL PROTECTED] wrote:
I'll attempt to think before I type this time :-)
I have a a C routine I wish to call that takes the following parameters:
typedef struct XXXCB {
short a;
char b[2];
};
void scumbag(XXXCB *cb, char *v);
snip/
In C# I'd code XXXCB as:
public struct XXXCB {
public short a;
[MarshalAS(UnmanagedType.ByValArray, SizeConst=2)}
public byte[] b;
}
Good declaration.
Now, because the variable v can be a different length of each call
(up to 32K is size) I can't use MarshalAs(UnmanagedType.ByValArray) as
mcs demands that SizeConst= be coded (using SizeParamIndex= doesn't
work, in fact I don't think it's supported).
There's a bigger problem: ByValArray doesn't mean what you think it
does. At least, not entirely.
ByValArray means that it's a contiguous blob of memory, which is why you
used it in your XXXCB structure declaration -- you needed to specify the
size of b. Perfectly natural.
However, it means the *same thing* no matter where it's applied. So
applied to a method argument, it means push a blob of data SizeConst
bytes onto the stack). This is NOT what C does; C always pushes a
*pointer* to an array onto the stack. Compare these two (C-like)
declarations:
// ByValArray equivalent
void scumbag (XXXCB *cb, struct {char data[SOME_LENGTH];} v);
// LPArray equivalent:
void scumbag (XXXCB *cb, char *v);
The first one probably isn't valid C, and certainly isn't what you want
when invoking existing C code (but may be useful when dealing with other
languages).
Obviously, you don't want ByValArray for method arguments. You want
LPArray. Fortunately, this is the default, and marshals the number of
elements inside the array:
[DllImport (libname)]
private static extern void scumbag (ref XXXCB cb, byte[] v);
Usage is straight-forward:
XXXCB cb = new XXXCB();
scumbag (ref cb, Encoding.ASCII.GetBytes (this is my v arg1));
scumbag (ref cb, Encoding.ASCII.GetBytes (this is another v));
Consequently, you don't need the custom marshaler.
- Jon
___
Mono-list maillist - [EMAIL PROTECTED]
http://lists.ximian.com/mailman/listinfo/mono-list
