Re: A Critique: Getting Started with GNUstep on Windows

On 17 Feb 2016, at 10:51, Thom Cherryhomes wrote: > > Why is it? I am trying to put a very visual and demonstrative face on what I > am experiencing, in the hope of disrupting the community out of its apparent > complacency. Because you claim that you want to

Re: A Critique: Getting Started with GNUstep on Windows

On 17 Feb 2016, at 06:37, Thom Cherryhomes wrote: > > Above all, the intent is to stimulate contemplation, self reflection, and > discussion of these issues. If this is your intention, then perhaps you could provide a transcript? A video is an awful medium for any

Re: Windows theme should be active by default on Windows...

imilar ideas. > > > On 14 February 2016 at 05:26, David Chisnall <thera...@sucs.org> wrote: > On 12 Feb 2016, at 23:04, Gregory Casamento <greg.casame...@gmail.com> wrote: > > > > If this works would it suggest that we don't really need msys anymore? > &g

Re: Windows theme should be active by default on Windows...

On 12 Feb 2016, at 23:04, Gregory Casamento wrote: > > If this works would it suggest that we don't really need msys anymore? I’m not sure. It looks as if MS is shipping their own POSIX layer (they’re using pthreads and similar in various places in their code).

Re: Windows theme should be active by default on Windows...

On 10 Feb 2016, at 22:21, Ivan Vučica wrote: > > I guess it's time to try giving Windows some love. If someone has easy access to a Windows 10 machine, it would be worth trying to build GNUstep with Visual Studio and the Objective-C runtime from WinObjC. The MS folks are a

Re: libdispatch-gnustep/libdispatchgs-etoilé vs libdispatch0

On 31 Jan 2016, at 23:52, Gregory Casamento wrote: > > Indeed. Niels is correct. I thought at one point libdispatch was working > without the need for a modified libBlocksRuntime, but after some testing it > appears I'm incorrect. It will sometimes work. It depends

Re: Debian 8 (Not really SOLVED)

On 29 Jan 2016, at 21:57, Svetlana A. Tkachenko wrote: > >> From the output that Svetlana has shown me, it also looks as if the >> Debian clang packages are no longer able to find any headers. > > I don't remember specifics. Do you have the output still? If you can

Re: Debian 8 (Not really SOLVED)

On 28 Jan 2016, at 19:52, Niels Grewe <niels.gr...@halbordnung.de> wrote: > >> >> Am 25.01.2016 um 16:29 schrieb David Chisnall <thera...@sucs.org>: >> >> On 25 Jan 2016, at 15:12, Tristan Bellogi <bell...@orange.fr> wrote: >>> >>>

Re: Debian 8 (Not really SOLVED)

On 25 Jan 2016, at 15:12, Tristan Bellogi wrote: > > Debian clang.deb install libobjc4, Apple has libobjc4, Gunstep recommends its > own libobjc2 ? Debian appends a version number to the GCC libobjc package. It supports the same features of Objective-C as GCC (i.e. nothing

Re: Only build llvm backend parts needed for GNUstep

On 14 Dec 2015, at 04:54, Patryk Laurent <plaur...@me.com> wrote: > > On Dec 13, 2015, at 08:42 PM, Patryk Laurent <plaur...@me.com> wrote: >> On Dec 13, 2015, at 12:25 PM, David Chisnall <thera...@sucs.org> wrote: >>> 20GB sounds very big. Debug bui

Re: Only build llvm backend parts needed for GNUstep

On 13 Dec 2015, at 19:06, Patryk Laurent wrote: > > Does anyone know, offhand, how to have clang/llvm only build what is required > for GNUstep? I recently built the latest clang/llvm from sources, and found > that the whole thing requires about 20GB! But I saw plenty of other

Re: Savannah vs. Gitlab

On 7 Dec 2015, at 22:07, Matt Rice wrote: > > and if they fail to live up to this obligation they have many > copyright assigners holding a contract the FSF would be in violation > of, They never sent me the $1 that the copyright assignment form required that they pay (you

Re: Savannah vs. Gitlab

On 7 Dec 2015, at 11:05, Luboš Doležel wrote: > > I concur. We're talking about a "number of people" who would prefer Savannah, > but the amount of people actually submitting code is very low. Just going to > Savannah's website makes me feel like this is a dead project. >

Re: Anyone using Atom with autocomplete-clang?

GNUstep Make can’t produce JSON compilation databases, which makes it quite difficult to integrate with this kind of tool. I’ve experimented with using CMake to build GNUstep tools, which makes it work a lot better. Building .app and .framework bundles is more complex though. David > On 4

Re: Savannah vs. Gitlab

On 4 Dec 2015, at 09:55, Ivan Vučica wrote: > > Primary hosting on Github? *shrug* there are no strong benefits, I think, and > sufficient number of people would be happier if we hosted elsewhere. I strongly disagree with this. The benefits for collaboration from being able

Re: Setting up GNUstep.sh on system profile loses LD_LIBRARY_PATH

On 3 Dec 2015, at 09:37, Luis Garcia Alanis wrote: > > On Linux, when a setuid (or setgid) program is ran, it does not receive > certain environment variables (including LD_LIBRARY_PATH, several other LD_* > variables, and HOSTALIASES), in order to reduce the possible attack

Re: Plans for ahead

On 3 Dec 2015, at 13:17, Tito Mari Francis Escaño wrote: > > For usability, I would like to raise my question: what's with Etoile? It’s not dead, though it is somewhat refocussed. CoreObject, for example, is still very actively developed, though I don’t think that

Re: [FOSDEM] Desktops DevRoom 2016: Call for Participation

On 1 Dec 2015, at 07:09, H. Nikolaus Schaller wrote: > > Wouldn't that be something to get a bigger audience to GNUstep? We’ve had a GNUstep devroom a few years, but it’s largely been ignored by people other than regular GNUstep contributors. It would probably be better to

Re: [FOSDEM] Desktops DevRoom 2016: Call for Participation

On 1 Dec 2015, at 12:14, Niels Grewe wrote: > > [0] Cf. https://github.com/SSheldon/rust-objc/pull/27#issuecomment-160052370 > if you need an example. I’m going to completely ignore the rest of your email, because having an ObjC<->Rust bridge that works with

Re: Plans for ahead

On 29 Nov 2015, at 19:40, James Carthew wrote: > > I'm going to weigh in here in the UI discussions. I come from a mainly user > background. A couple observations. Yes gnustep is a bit ugly, but that's not > really my biggest concern. I have been trying to run GNUStep as a

Re: Plans for ahead

On 28 Nov 2015, at 14:44, Gregory Casamento wrote: > > Which theme are you using? A modern theme is sorely needed. Rik. It’s not quite polished, but it looks better than anything else that I’ve tried. David -- Sent from my Difference Engine

Re: Build GNUStep

On 16 Nov 2015, at 14:25, Adam S wrote: > > So I have built GNUStep on both Raspbian (Raspberry Pi2) and Lubuntu (Banana > Pi) and get the same three errors each time. I feel a lot like I’m talking to myself here, but I’ll try one more time: If you install the

Re: NeXT GNU Homage Project Work

I’m not sure why you’d want to. FreeBSD produces images for RPi and RPi2, and has ports for all of GNUstep (not yet packages, but a source build won’t take too long on a RPi 2 and we should have packages building soon). Installing the gnustep meta-port will install over a hundred GNUstep

Re: NeXT GNU Homage Project Work

On 15 Nov 2015, at 16:38, Liam Proven wrote: > > NetBSD is quite a lot more primitive than Linux. E.g. Linux (and only > Linux) has graphical acceleration on RasPi This is not true. We have working 3D on FreeBSD with the RPi. The code for the 3D accelerator on the RPi 2

Re: ANN: PikoPixel pixel-art editor

On 12 Oct 2015, at 11:15, Josh Freeman wrote: > > - Requires a newer version of the GNUstep development environment: Oct. 11, > 2015 or later Unfortunately, this requirement now means that I can’t package it. And, given that the last version didn’t work, I

Re: ANN: PikoPixel pixel-art editor

Having now tried testing PikoPixel, I discover that it segfaults on startup. It turns out that this is because it relies on undefined behaviour, which may or may not work depending on linker behaviour. 1. The PPGNUstepGlue_TitleablePopUpButton category on NSObject has a +load method 2. This

Re: problem compiling base with clang / libobjc2

On 7 Oct 2015, at 18:15, Riccardo Mottola wrote: > > I then used the "deprecated" gnustep make file of libobjc which says "gnustep > environment detected" and with that one it appears to work! thus using the > "cmake" is actually problematic. This will install

Re: Problems building gnustep-base on Ubuntu Linux with clang/llvm-3.7

On 30 Sep 2015, at 17:09, David Lobron wrote: > > Ah, you are right. I modified the conftest.c compile command so that it has > -L/usr/lib as the first -L argument, and it now compiles fine (my libobjc > from libobjc2 lives in /usr/lib, whereas the gcc versions are in >

Re: Problems building gnustep-base on Ubuntu Linux with clang/llvm-3.7

On 29 Sep 2015, at 16:21, David Lobron wrote: > > configure:14837: clang -o conftest -m64 -march=opteron -mno-3dnow -ggdb -O2 > -Wall > -I/home/dlobron/build/clangport/akamai/common/GNUstep/Local/Library/Headers >

Re: Problems building gnustep-base on Ubuntu Linux with clang/llvm-3.7

On 28 Sep 2015, at 19:02, David Lobron wrote: > > Hey David, > >> /usr/local/{include,lib} is not in the default search path for gcc / clang >> on Ubuntu. You should install libobjc2 with the prefix set to /usr/ or >> explicitly add /usr/local to the search paths when

Re: Problems building gnustep-base on Ubuntu Linux with clang/llvm-3.7

On 28 Sep 2015, at 17:36, David Lobron wrote: > >> It sounds as if the configure script is not picking up the new runtime (and >> is picking up a libobjc from a gcc install). Where did you install it? > > I have it at: /usr/local/lib/libobjc.so.4.6. The >

Re: Problems building gnustep-base on Ubuntu Linux with clang/llvm-3.7

On 28 Sep 2015, at 17:18, David Lobron wrote: > > The above line appears under the "Cache variables" section in the config.log. > I do not have a cache file in my build tree, so I don't think I'm picking up > an obsolete value. Do you know how this variable would be set

Re: Problems building gnustep-base on Ubuntu Linux with clang/llvm-3.7

Hi David, > On 24 Sep 2015, at 20:14, David Lobron wrote: > > Hi All, > > I am trying to build gnustep-base-1.24.8 on Ubuntu Linux, and hitting a few > snags. I would be very grateful if someone could suggest where I'm going > wrong. > > I first installed the llvm

Re: Problem to compile gnustep-base on android

On 26 Sep 2015, at 23:37, Germán Arias wrote: > > What is the difference between -r and -pie? -r means do linking, but don’t expect the result to be a complete binary (partial link). -pie means generate a position-independent executable. Obviously, these are incompatible,

Re: Update libobjc2 build instructions for Ubuntu/Android/Darling?

On 25 Sep 2015, at 19:13, Josh Freeman wrote: > > The issue with building libobjc2 before GNUstep make (on Ubuntu at least) > is that libobjc2's install path depends on whether gnustep-config is present: > If it is, it installs to /usr/GNUstep/Local/Library/ -

Re: Problem to compile gnustep-base on android

On 26 Sep 2015, at 08:42, Germán Arias wrote: > > El sáb, 26-09-2015 a las 07:50 +0100, Richard Frith-Macdonald escribió: >>> On 26 Sep 2015, at 07:28, Germán Arias wrote: >>> >>> El vie, 25-09-2015 a las 21:27 -0600, Germán Arias escribió: First

Re: Update libobjc2 build instructions for Ubuntu/Android/Darling?

On 25 Sep 2015, at 05:55, Josh Freeman wrote: > > The wiki's Ubuntu script builds libobjc2 before GNUstep make, but I've had > issues building libobjc2 unless GNUstep make is built beforehand You should not have had these issues in the last five years. libobjc2

Re: ANN: PikoPixel pixel-art editor

On 21 Sep 2015, at 08:17, Josh Freeman wrote: > > 2. Screencasting popup (PPScreencastingController.m, > PPApplication_Screencasting.m) > > When enabled, a popup window shows up whenever a key is pressed or mouse > clicked. It's useful as a demonstration aid

Re: NSStepper Action Problem

The simplest fix may be to modify the stepper to use a double for the value. Given that steppers are commonly used with fixed increments that make sense in base 10, a better solution might be to modify it to store a base value and a number of applications of the increment, so that in your case

Re: ANN: PikoPixel pixel-art editor

On 20 Sep 2015, at 15:28, Fred Kiefer wrote: > > Strange enough this only works for the two instance method swizzling > methods, for the class methods I have to use just > > method_exchangeImplementations(method1, method2); Note that this is not actually a good thing to do

Re: ANN: PikoPixel pixel-art editor

Hi Josh, Thank you for taking the time to make this work with GNUstep and for the open source release. I’ve added it to the FreeBSD ports collection, it should appear as a binary package in the next couple of days. A few things that you could do to make life easier for people packaging it

Is anyone using -fobjc-gc?

Hi all, Apple has deprecated Objective-C GC on OS X a while ago and never supported it on iOS, and the memory model made it quite difficult to work with. There’s some talk of removing support for it entirely from clang. Does anyone depend on it working with GNUstep? If so, now would be a

Re: Is anyone using -fobjc-gc?

On 21 Aug 2015, at 10:55, Richard Frith-Macdonald richardfrithmacdon...@gmail.com wrote: While both David and I have put a lot of effort into *two* different sets of GC support (both Apple compatible and pre-Apple), and therefore I guess have an emotional investment in it, my objective

Re: Seg fault in GSIMapCleanMap in gnustep-base-1.24.8

On 3 Aug 2015, at 21:24, David Lobron dlob...@akamai.com wrote: I tried this, and I did not see an unbalanced release. Actually, I'm now thinking that Valgrind was pointing me in the wrong direction. The stack trace after my code seg faults suggests that GSArray might be to blame: #0

Re: Seg fault in GSIMapCleanMap in gnustep-base-1.24.8

On 30 Jul 2015, at 17:56, David Lobron dlob...@akamai.com wrote: I tried Valgrind, and it did indeed report an invalid read of size 4. It also reports a double free: ==24092== Address 0xab6cd98 is 16 bytes inside a block of size 68 free'd ==24092==at 0x48CE06C: free (in

Re: Seg fault in GSIMapCleanMap in gnustep-base-1.24.8

On 29 Jul 2015, at 16:47, David Lobron dlob...@akamai.com wrote: #0 0xf73fc056 in objc_msg_lookup () from /usr/lib32/libobjc.so.3 #1 0xf749e290 in GSIMapCleanMap (map=0x818efb4) at ../Headers/GNUstepBase/GSIMap.h:1188 #2 GSIMapCleanMap (map=0x818efb4) at GSDictionary.m:117 #3

Re: viewFortableColumn method invocation

On 16 Jul 2015, at 14:22, chandan bp cbp...@gmail.com wrote: Hi, I am implementing NSTableCellView class i.e for view based NSTableView. I added NSTableCellView header and implementation files. I added viewForTableColumn method in NSTableViewDelegate protocol and

[ANN] GNUstep Objective-C Runtime 1.8

Hi Everyone, I’ve just pushed out a new release of the GNUstep Objective-C runtime. There are no changes between this and the release candidate that’s currently in FreeBSD ports, so hopefully there will be no surprises. David GNUstep Objective-C Runtime 1.8 ===

Re: Debian and Ubuntu packages - 2015/06/03

On 7 Jul 2015, at 17:53, Ivan Vučica i...@vucica.net wrote: - It's still using libobjc2 from GNUstep's Subversion repository, and not the one in David's repository on Github FYI: This is now moved back to the GNUstep organisation on GitHub[1] and that’s where the FreeBSD port fetches it

Re: Implementing tableview(GORM and PROJECTCENTER)

On 7 Jul 2015, at 07:55, chandan bp cbp...@gmail.com wrote: I am able to build and run the application, however when I run the app using the project center, the table is empty. In the GUI after app launch I get 'Data source doesnt respond to numberOfRowsInTableView'. Can

Re: Status of GNUstep

On 6 Jul 2015, at 06:57, Dr. H. Nikolaus Schaller h...@goldelico.com wrote: This will list the “undefined” references that are to be satisfied by the shared libraries. Of course it is a mix of system libraries and GNUstep libraries, but you will see the format of Obj-C methods. No you

Re: Status of GNUstep

On 5 Jul 2015, at 11:06, James Carthew jcart...@gmail.com wrote: Hello, Does GNUstep have a status page similar to: http://nouveau.freedesktop.org/wiki/FeatureMatrix/ Which outlines the current API status of GNUstep in relation to NextStep/Cocoa APIs? I haven't been able to locate

Re: [ANN] GNUstep FreeBSD ports updated

/usr/local/GNUstep/Local). David On 2 Jul 2015, at 09:02, David Chisnall thera...@sucs.org wrote: Hi All, I’ve now updated the GNUstep ports in FreeBSD. Binary package building runs roughly every 30 hours, and I updated them in the middle of a run yesterday, so hopefully the binary

[ANN] GNUstep FreeBSD ports updated

Hi All, I’ve now updated the GNUstep ports in FreeBSD. Binary package building runs roughly every 30 hours, and I updated them in the middle of a run yesterday, so hopefully the binary packages should appear some time tomorrow. As part of this update, we’re shipping an unreleased libobjc2

Re: libdispatch problem

I suspect that this is a bug in GNUstep where a +initialize method invoked by the NSLog() is creating an autoreleased object that is then destroyed prematurely. You could try with NSZombieEnabled and see if that produces more useful debugging info. David On 24 Jun 2015, at 21:08, allynm

Re: cmake can't find library to compile libobjc2

On 8 Jun 2015, at 23:55, allynm mark.al...@verizon.net wrote: ../../Headers/ObjectiveC2/objc/runtime.h:89:15: error: typedef It looks as if the configure script has picked up the wrong Objective-C runtime (an old GCC one) and so is trying to build a compatibility layer that implements

Re: cmake can't find library to compile libobjc2

Please just turn off building the LLVM opts and do not build LLVM 3.5. This will be done by default in the GitHub version if it doesn’t find a version of LLVM that it’s tested to work with. David On 7 Jun 2015, at 21:08, allynm mark.al...@verizon.net wrote: Hello all, Trying to build

Re: Debian and Ubuntu packages - 2015/06/03

On 4 Jun 2015, at 08:51, Philippe Roussel p.o.rous...@free.fr wrote: On Thu, Jun 04, 2015 at 05:11:21PM +1000, Svetlana A. Tkachenko wrote: Oh, and my understanding is that Debian's gnustep-* packages are compiled using gcc. Am I wrong? If not, then how may this valuable set of packages

Re: Proposal: Switch back to savannah using GIT

On 25 May 2015, at 13:48, Ivan Vučica i...@vucica.net wrote: On Mon, May 25, 2015 at 1:40 PM, Gregory Casamento greg.casame...@gmail.com wrote: The issue with savannah at this point is that they only allow one repository per user. This is quite different from git hub which allows as many

Re: ANN: GNUstep GUI/Back 0.24.1

I’m currently travelling, but will try to update the FreeBSD ports when I get back. If anyone else wants to have a go, feel free to file a FreeBSD bug report with a patch. David On 17 May 2015, at 13:06, Fred Kiefer fredkie...@gmx.de wrote: This is version 0.24.1 of the GNUstep GUI

Re: art backend : NSView crashes

On 30 Apr 2015, at 23:11, Fred Kiefer fredkie...@gmx.de wrote: Looks like gcc 4.9 has gotten very strict about method calls on nil that return a structure. It has always been better to check that case, but now it has become a must. This has always been the case with gcc: it will call a

Re: Microsoft is working on iOS SDK for Windows 10 apps

On 30 Apr 2015, at 23:43, Fred Kiefer fredkie...@gmx.de wrote: On the other hand, I find it very frustrating to see how fast they implemented something that we have been working on for years. I am led to believe that Microsoft employs some full-time developers. David -- Sent from my IBM

Re: GDL2 compilation errors

On 30 Apr 2015, at 14:42, Edwin Ancaer eanc...@gmail.com wrote: EOGenericRecord.m:1526:19: error: incomplete definition of type 'struct objc_class' size = selfClass-instance_size; EOGenericRecord.m:1622:26: error: incomplete definition of type 'struct objc_class' size

Re: GDL2 compilation errors

On 30 Apr 2015, at 16:31, Edwin Ancaer eanc...@gmail.com wrote: David, indeed, it is working now. Thanks. I've got a new problem now: ../EOControl/EOQualifier.h:239:12: error: cannot define category for undefined class 'NSArray' @interface NSArray (EOQualifierExtras) There

Re: libobjc2's cmake installation method seems to be broken on FreeBSD 10.1

This is fixed in Git: https://github.com/davidchisnall/libobjc2 David On 19 Apr 2015, at 00:11, Lars Sonchocky-Helldorf lars.sonchocky-helld...@hamburg.de wrote: Hi, when trying to install libobjc2 from SVN trunk into the Local Domain of a FreeBSD 10.1 I encountered the following

Re: libobjc2's cmake installation method seems to be broken on FreeBSD 10.1

On 19 Apr 2015, at 15:12, Lars Sonchocky-Helldorf lars.sonchocky-helld...@hamburg.de wrote: Thanks for the hint. Does this mean that the master site for libobjc2 is no longer svn://svn.gna.org/svn/gnustep/trunk ? I’m developing on GitHub, as it makes collaboration easier (and makes it much

Re: weak redefinition

On 14 Apr 2015, at 15:39, Riccardo Mottola riccardo.mott...@libero.it wrote: do you have an authoritative reference to that? I wound weak and not __weak. A language spec or a doc from Apple would be excellent, so I can argument to the NetBSD guys. See:

Re: weak redefinition

On 14 Apr 2015, at 15:04, Riccardo Mottola riccardo.mott...@libero.it wrote: Hi, David Chisnall wrote: It appears that all NetBSD headers are incompatible with Objective-C. Probably worth filing a bug with them - this is even harder to work around than the glibc issue of using __block

Re: FreeBSD GNUstep ports proposal

dependencies of the FreeBSD's GNUstep ports. Therefore the designated recipient of this mail is actually the maintainer of the GNUstep ports for FreeBSD, thera...@freebsd.org who happens to be David Chisnall. I cc'ed this mail to the GNUstep list for the record. I am currently

Re: Using GCC

On 8 Apr 2015, at 08:41, Luboš Doležel lu...@dolezel.info wrote: I'd also refrain from calling it kernel GCD. It is true that some parts needed for GCD to function are in the kernel (pthreads and pthread workqueues), but that's it. It's a pure userspace library implemented on top of these

Re: GNUstep's lack of progress and Apple's holding back Swift language

On 3 Apr 2015, at 00:27, Gregory Casamento greg.casame...@gmail.com wrote: It seems that the reason so many people are happy about Swift is because it's NOT ObjC, for whatever sense that makes. From talking to our local Cocoa dev group and a few others, there seems to be a trend in

Re: Building libobjc2 on ubuntu

I recently committed some tweaks to the CMake that will allow it to continue working if the LLVM CMake files are not correctly installed: https://github.com/davidchisnall/libobjc2 I hope to push out a 1.8 release soon containing this and other fixes. David On 27 Mar 2015, at 04:53, Luis

Re: Nonfragile ABI libobjC flags for cmake

On 23 Mar 2015, at 14:30, Jacek Wisniowski oper...@baltobor.com wrote: Here is a short compiler dump: [ 48%] Building CXX object Test/CMakeFiles/BlockTest_arc.dir/BlockTest_arc.m.o clang: warning: argument unused during compilation: '-fobjc-runtime=gnustep-1.7' error: -fobjc-arc is not

Re: data type polymorphism

On 20 Mar 2015, at 19:39, Scott Christley schrist...@mac.com wrote: Hello, This is more a generic Objective-C question versus GNUstep but maybe some experts here have a suggestion. I have a bunch of code that looks l like this: if ([encode isEqual: [BioSwarmModel floatEncode]]) {

Re: GNUStep 20th anniversary - GNUstep summit

On 8 Mar 2015, at 01:14, Gregory Casamento greg.casame...@gmail.com wrote: another cometogether in Cambridge would be the the best option for me, but David seems have to lost interest in GNUstep :-( I hope he hasn't lost all interest. Once a GNUstepper always a GNUstepper. ;) Fred isn't

Re: Is there any way to use ARC in 32-bit?

On 30 Jan 2015, at 09:40, Richard Frith-Macdonald richardfrithmacdon...@gmail.com wrote: On 30 Jan 2015, at 07:34, David Chisnall thera...@sucs.org wrote: On 29 Jan 2015, at 09:10, Richard Frith-Macdonald richardfrithmacdon...@gmail.com wrote: Different worlds ... on FreeBSD that's

Re: Is there any way to use ARC in 32-bit?

On 29 Jan 2015, at 09:10, Richard Frith-Macdonald richardfrithmacdon...@gmail.com wrote: Different worlds ... on FreeBSD that's roughly 2:1 cmake to autotools, but I guess it looks different in non-bsd systems. No. This is the FreeBSD *ports* collection (i.e. third-party code, most of

Re: Is there any way to use ARC in 32-bit?

It sounds like you're using an OS where clang defaults to the GCC runtime (some Linux variant probably). You'll need to add -fobjc-runtime=gnustep-1.7. Note that this has nothing to do with 32-bit vs 64-bit, it is purely about the Objective-C runtime that the compiler is targeting. David On

Re: Is there any way to use ARC in 32-bit?

On 28 Jan 2015, at 16:55, Jens Alfke j...@couchbase.com wrote: Sorry if this sounds entitled, but shouldn't the GNUstep install process already configure the right compiler options? I'm a bit frazzled after spending a day messing with environment variables and paths to even get as far as

Re: Is there any way to use ARC in 32-bit?

On 28 Jan 2015, at 17:49, Richard Frith-Macdonald richardfrithmacdon...@gmail.com wrote: When tryiung to move people over to using clang and libobjc2, everyone seems to have problems with building libobjc2 because of its cmake dependency :-( People who can't manage: $ mkdir Build ; cd Build

Re: Problem Compiling Objective C code

Please don't send big images to mailing lists when a couple of lines of text would convey exactly the same information. It looks as if you compiled the runtime with a different way to the rest of your code. The code in the runtime seems to have the right ABI though. What happens if you add

Re: CoreFoundation

On 7 Jan 2015, at 10:05, Dr. H. Nikolaus Schaller h...@goldelico.com wrote: and with base+cf I could have a universal way to access and modify system preferences in plists from both ObjC and C. I image there might be a situation where I wanna use plists for compatibility with gnustep but

Re: CoreFoundation

On 7 Jan 2015, at 09:28, Lundberg, Johannes johan...@brilliantservice.co.jp wrote: I think plists is a great way to store preferences and with base+cf I could have a universal way to access and modify system preferences in plists from both ObjC and C. I image there might be a situation

Re: Font issues

On 6 Jan 2015, at 10:19, Lundberg, Johannes johan...@brilliantservice.co.jp wrote: The text rendering does not look correct to me. You can see in the font panel how ee, re and ra stick together with no space between. I believe that this is because the Cairo back end uses the 'toy' font API,

Re: signature error NSPopUpButton with clang

On 29 Dec 2014, at 19:09, Riccardo Mottola riccardo.mott...@libero.it wrote: a cast didn't help, however the correct type did. This seems to have other positive effects too, I'll check more before rejoicing too much though. I believe that clang has a warning about ambiguous selectors on

Re: How to enable shorthand access element in NSArray

Are you targeting the non-fragile ABI? This array accessor notation is a terrible idea, so I'd recommend that you avoid it, as it (along with the property notation) destroy one of the key strengths of Objective-C: the orthogonality of syntax. In particular, with the fragile ABI, you were

Re: signature error NSPopUpButton with clang

On 28 Dec 2014, at 19:40, Riccardo Mottola riccardo.mott...@libero.it wrote: Riccardo Mottola wrote: The type for the selector used in the call will be the static type of wherePopUp. What is the type declared as, what is the actual run-time type of the object? wherePopUp is IBOutlet id

Re: How to enable shorthand access element in NSArray

On 28 Dec 2014, at 20:59, Rael Bauer supp...@bauerapps.com wrote: I was not aware of whether I was or wasn't (I am very new to gnustep/objective-C). After some research it seems that I was not. (just using the default). If I add this line to the GNUmakefile: test_OBJCFLAGS +=

Re: How to enable shorthand access element in NSArray

On 28 Dec 2014, at 22:57, Lars Sonchocky-Helldorf lars.sonchocky-helld...@hamburg.de wrote: Well, I also don't like Windows but I know how to run it in a virtual machine (even VirtualBox is sufficient here) when I need to. Here in Germany you can get a legal copy of Windows 7 Professional

Re: gworkspace / clang / libobjc2 hang problem

On 25 Dec 2014, at 11:08, Riccardo Mottola riccardo.mott...@libero.it wrote: Hi, Riccardo Mottola wrote: The patch seems to work with one or more operations on both gcc computers. However on NetBSD with clang, on copy I get a hang even with at the first copy operation! Why? nothing

Re: signature error NSPopUpButton with clang

On 25 Dec 2014, at 19:26, Riccardo Mottola riccardo.mott...@libero.it wrote: if no specifier is given, this should be an integer constant, not an unisgned one! So why the error? Also, I tried casting to (NSInteger) and still get a mismatch warning. It is most probably harmless, but since it

Re: How use -indexOfObjectPassingTest:?

On 7 Dec 2014, at 06:36, Germán Arias germanan...@gmx.es wrote: error: expected expression before ‘^’ token This looks like a GCC error. If you want to use methods that take blocks as arguments, then you need to use Clang. David ___

Re: Determinate progress bar

On 22 Nov 2014, at 00:37, Ivan Vučica i...@vucica.net wrote: I'd guess that under OS X this is done with Core Animation, as many animations don't stop even if the UI thread is blocked. But -- we don't have that, so ... CoreAnimation is part of the story, multithreaded rendering is another.

Re: Hanging multi-threaded tools

Hi Andreas, It's difficult to diagnose the cause of a deadlock without backtraces from *all* participating threads... David On 3 Nov 2014, at 15:53, Andreas Höschler ahoe...@smartsoft.de wrote: Hi all, I still have problems with multi-threaded tools that run into the following trap on a

Re: What would be the most complete GNUStep system?

On 27 Oct 2014, at 21:04, Asiga Nael asigan...@yahoo.com wrote: Btw, nobody suggested Darwin. It's open-source, Well, kind of. The audio subsystem, for example, is proprietary, so if you want to have working sound then you're going to need to write a load of kernel code. I'm also not sure

Re: profiling gnustep apps and libraries

Last time I tried, replacing the NSMutableDictionary class with a very thin wrapper around their C++ standard library counterparts gave a fairly significant speedup. A good place to look for poor performance would be in the standard collection classes, as a win there can be a big win elsewhere

Re: Objective-C without an actual compiler

On 25 Oct 2014, at 01:10, Ingwie Phoenix ingwie2...@googlemail.com wrote: I just came across an email on Clang, where they talked about how ObjC is translated into C - then compiled. I just wondered if there is a general-purpose tool to „convert“ ObjC code into plain C (or C++)? This was

Re: Phoenix language

On 23 Oct 2014, at 07:37, Dr. H. Nikolaus Schaller h...@computer.org wrote: If you like to study an Obj-C Interpreter that is starting to do useful things: https://github.com/goldelico/mySTEP/tree/master/ObjC It is designed as a pipeline: preprocessor - lexer - parser - abstract

Re: Phoenix language

On 22 Oct 2014, at 18:14, Dr. H. Nikolaus Schaller h...@computer.org wrote: John Siracusa's review of Yosemite is extensive; for this thread, the embedded extensive overview of Swift is more interesting. I would highly recommend reading pages 21, 22, 23 as an overview of what makes Swift

Re: Build a complete gnustep app in a shared object

I'd probably suggest that you rethink your aversion to Objective-C++. With ARC, you can safely put Objective-C object pointers in C++ classes (they become non-POD) and you can use C++ classes as instance variables for Objective-C objects. Most importantly, you can now put Objective-C object

<    1   2   3   4   5   6   7   8   9   10   >