From gambas at ...1... Sun Jul 8 16:42:59 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 16:42:59 +0200 Subject: [Gambas-devel] About gb.gsl components Message-ID: <4FF99C73.5040909@...1...> Hi Randall, I have no news from you for a long time. I hope everything is ok for you! Since revision #4908, the interpreter allows any native class to handle the standard arithmetic operators - at least some of them at the moment: +, -, *, /, =, <>, and Abs(). This is done through the "_operators" interface. Moreover, the "_convert" interface now allows to convert any native object to any other datatype, and is called by Str() and Print to convert an object to a localized string representation. I used that in the Complex class, so that you can do: Dim A As New Complex(1, 4) Dim B As New Complex(3, -1) Print A;; B;; A * 2;; A * B;; A / B --> 1+4i 3-i 2+8i 7+11i -0,1+1,3i I have ',' inside numbers because this the french decimal separator. Tell me if you want to continue working on gb.gsl. That new feature could be implemented in polynomials, matrix... with automatic conversions between them maybe! Regards, -- Beno?t Minisini From emil.lenngren at ...176... Sun Jul 8 18:01:36 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Sun, 8 Jul 2012 18:01:36 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FF99C73.5040909@...1...> References: <4FF99C73.5040909@...1...> Message-ID: Hi. Some null checks seems to be required. For example: Dim A As Complex = NULL Print Abs(A) -> Segmentation fault Another question: If I have a class I want to implement _operators in, do I have to implement all 13 or can I set some of them to NULL? /Emil 2012/7/8 Beno?t Minisini > Hi Randall, > > I have no news from you for a long time. I hope everything is ok for you! > > Since revision #4908, the interpreter allows any native class to handle > the standard arithmetic operators - at least some of them at the moment: > +, -, *, /, =, <>, and Abs(). This is done through the "_operators" > interface. > > Moreover, the "_convert" interface now allows to convert any native > object to any other datatype, and is called by Str() and Print to > convert an object to a localized string representation. > > I used that in the Complex class, so that you can do: > > Dim A As New Complex(1, 4) > Dim B As New Complex(3, -1) > > Print A;; B;; A * 2;; A * B;; A / B > > --> 1+4i 3-i 2+8i 7+11i -0,1+1,3i > > I have ',' inside numbers because this the french decimal separator. > > Tell me if you want to continue working on gb.gsl. That new feature > could be implemented in polynomials, matrix... with automatic > conversions between them maybe! > > Regards, > > -- > Beno?t Minisini > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Sun Jul 8 18:15:15 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 18:15:15 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> Message-ID: <4FF9B213.5050303@...1...> Le 08/07/2012 18:01, Emil Lenngren a ?crit : > Hi. Some null checks seems to be required. For example: > > Dim A As Complex = NULL > Print Abs(A) > > -> Segmentation fault > > Another question: If I have a class I want to implement _operators in, > do I have to implement all 13 or can I set some of them to NULL? > > /Emil > I think some operators will not be mandatory, like min + max + inequality operators. Or maybe I won't allow inequality operators and min + max. I don't know yet. At the moment, except inequality things, all must be implemented. -- Beno?t Minisini From gambas at ...1... Sun Jul 8 18:23:56 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 18:23:56 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> Message-ID: <4FF9B41C.2040206@...1...> Le 08/07/2012 18:01, Emil Lenngren a ?crit : > Hi. Some null checks seems to be required. For example: > > Dim A As Complex = NULL > Print Abs(A) > > -> Segmentation fault > OK, should be fixed in revision #4912. -- Beno?t Minisini From emil.lenngren at ...176... Sun Jul 8 18:34:21 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Sun, 8 Jul 2012 18:34:21 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FF9B41C.2040206@...1...> References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> Message-ID: More segfaults: Dim C As Complex = Null Print 1 * C And Dim A As New Complex(1, 4) Dim C As Complex = A Dim I As Integer For I = 1 To 2 Print Abs(C) ' or for example Print C * C C = Null Next /Emil 2012/7/8 Beno?t Minisini > Le 08/07/2012 18:01, Emil Lenngren a ?crit : > > Hi. Some null checks seems to be required. For example: > > > > Dim A As Complex = NULL > > Print Abs(A) > > > > -> Segmentation fault > > > > OK, should be fixed in revision #4912. > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Sun Jul 8 18:56:00 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 18:56:00 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> Message-ID: <4FF9BBA0.4020600@...1...> Le 08/07/2012 18:34, Emil Lenngren a ?crit : > More segfaults: > > Dim C As Complex = Null > Print 1 * C > > And > > Dim A As New Complex(1, 4) > Dim C As Complex = A > Dim I As Integer > > For I = 1 To 2 > Print Abs(C) ' or for example Print C * C > C = Null > Next > > /Emil > > 2012/7/8 Beno?t Minisini > > > Le 08/07/2012 18:01, Emil Lenngren a ?crit : > > Hi. Some null checks seems to be required. For example: > > > > Dim A As Complex = NULL > > Print Abs(A) > > > > -> Segmentation fault > > > > OK, should be fixed in revision #4912. > > -- > Beno?t Minisini > > Yep. Should really be better in revision #4913! -- Beno?t Minisini From emil.lenngren at ...176... Sun Jul 8 22:25:04 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Sun, 8 Jul 2012 22:25:04 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FF9BBA0.4020600@...1...> References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> Message-ID: What is the point having "has_convert" and "has_operators" and so on? Isn't it easier to simply check if the "operators" and "convert" pointers are NULL or not? That would also save processor cycles on x86 and x86_64 :) /Emil 2012/7/8 Beno?t Minisini > Le 08/07/2012 18:34, Emil Lenngren a ?crit : > > More segfaults: > > > > Dim C As Complex = Null > > Print 1 * C > > > > And > > > > Dim A As New Complex(1, 4) > > Dim C As Complex = A > > Dim I As Integer > > > > For I = 1 To 2 > > Print Abs(C) ' or for example Print C * C > > C = Null > > Next > > > > /Emil > > > > 2012/7/8 Beno?t Minisini > > > > > > Le 08/07/2012 18:01, Emil Lenngren a ?crit : > > > Hi. Some null checks seems to be required. For example: > > > > > > Dim A As Complex = NULL > > > Print Abs(A) > > > > > > -> Segmentation fault > > > > > > > OK, should be fixed in revision #4912. > > > > -- > > Beno?t Minisini > > > > > > Yep. Should really be better in revision #4913! > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Sun Jul 8 22:32:10 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 22:32:10 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> Message-ID: <4FF9EE4A.6090706@...1...> Le 08/07/2012 22:25, Emil Lenngren a ?crit : > What is the point having "has_convert" and "has_operators" and so on? > Isn't it easier to simply check if the "operators" and "convert" > pointers are NULL or not? That would also save processor cycles on x86 > and x86_64 :) > > /Emil The last time I checked with valgrind, x86 and X86_64 are faster to test a bit than to check an entire pointer. -- Beno?t Minisini From emil.lenngren at ...176... Sun Jul 8 22:46:46 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Sun, 8 Jul 2012 22:46:46 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FF9EE4A.6090706@...1...> References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> <4FF9EE4A.6090706@...1...> Message-ID: Hmm. I just checked it up. Null pointer comparison ends up in a "cmp" instruction, and bit test ends up in a "test" instruction. Both take the same number of clock cycles.. So in theory it should make no difference at all. 2012/7/8 Beno?t Minisini > Le 08/07/2012 22:25, Emil Lenngren a ?crit : > > What is the point having "has_convert" and "has_operators" and so on? > > Isn't it easier to simply check if the "operators" and "convert" > > pointers are NULL or not? That would also save processor cycles on x86 > > and x86_64 :) > > > > /Emil > > The last time I checked with valgrind, x86 and X86_64 are faster to test > a bit than to check an entire pointer. > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Sun Jul 8 22:55:32 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 22:55:32 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> <4FF9EE4A.6090706@...1...> Message-ID: <4FF9F3C4.6040609@...1...> Le 08/07/2012 22:46, Emil Lenngren a ?crit : > Hmm. I just checked it up. Null pointer comparison ends up in a "cmp" > instruction, and bit test ends up in a "test" instruction. Both take the > same number of clock cycles.. > > So in theory it should make no difference at all. > Maybe the cache changes something then? -- Beno?t Minisini From emil.lenngren at ...176... Sun Jul 8 23:07:10 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Sun, 8 Jul 2012 23:07:10 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FF9F3C4.6040609@...1...> References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> <4FF9EE4A.6090706@...1...> <4FF9F3C4.6040609@...1...> Message-ID: Maybe.. I think it is very strange that interpreted functions with debugging flags sometimes are faster than without... This code: Dim i As Integer = 1 While i <= 80000000 Inc i Wend Takes 3.48 seconds with debugging flags, and 5.66 seconds without :/ I think it should be the opposite. /Emil 2012/7/8 Beno?t Minisini > Le 08/07/2012 22:46, Emil Lenngren a ?crit : > > Hmm. I just checked it up. Null pointer comparison ends up in a "cmp" > > instruction, and bit test ends up in a "test" instruction. Both take the > > same number of clock cycles.. > > > > So in theory it should make no difference at all. > > > > Maybe the cache changes something then? > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Sun Jul 8 23:12:02 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 23:12:02 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> <4FF9EE4A.6090706@...1...> <4FF9F3C4.6040609@...1...> Message-ID: <4FF9F7A2.8000205@...1...> Le 08/07/2012 23:07, Emil Lenngren a ?crit : > Maybe.. > > I think it is very strange that interpreted functions with debugging > flags sometimes are faster than without... > > This code: > Dim i As Integer = 1 > While i <= 80000000 > Inc i > Wend > > Takes 3.48 seconds with debugging flags, and 5.66 seconds without :/ > I think it should be the opposite. > > /Emil > So do I! -- Beno?t Minisini From gambas at ...1... Sun Jul 8 23:17:23 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 23:17:23 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FF9F7A2.8000205@...1...> References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> <4FF9EE4A.6090706@...1...> <4FF9F3C4.6040609@...1...> <4FF9F7A2.8000205@...1...> Message-ID: <4FF9F8E3.5040006@...1...> Le 08/07/2012 23:12, Beno?t Minisini a ?crit : > Le 08/07/2012 23:07, Emil Lenngren a ?crit : >> Maybe.. >> >> I think it is very strange that interpreted functions with debugging >> flags sometimes are faster than without... >> >> This code: >> Dim i As Integer = 1 >> While i <= 80000000 >> Inc i >> Wend >> >> Takes 3.48 seconds with debugging flags, and 5.66 seconds without :/ >> I think it should be the opposite. >> >> /Emil >> > > So do I! > Well, I cannot reproduce what you are talking about. Can you be more precise about what you do exactly? -- Beno?t Minisini From emil.lenngren at ...176... Sun Jul 8 23:25:17 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Sun, 8 Jul 2012 23:25:17 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FF9F8E3.5040006@...1...> References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> <4FF9EE4A.6090706@...1...> <4FF9F3C4.6040609@...1...> <4FF9F7A2.8000205@...1...> <4FF9F8E3.5040006@...1...> Message-ID: Ok. First I start a new Gambas Console project. MMain.module contains this: ' Gambas module file Public Sub Main() Dim i As Integer = 1 While i <= 80000000 Inc i Wend End Then I open a terminal and go to the project directory: emil at ...700...:~/Programmering/Gambas/DebugTest$ gbc3 -a -g OK emil at ...700...:~/Programmering/Gambas/DebugTest$ time gbx3 real 0m3.396s user 0m3.391s sys 0m0.002s emil at ...700...:~/Programmering/Gambas/DebugTest$ gbc3 -a OK emil at ...700...:~/Programmering/Gambas/DebugTest$ time gbx3 real 0m5.599s user 0m5.592s sys 0m0.002s The debug version is faster than the non-debug version... /Emil 2012/7/8 Beno?t Minisini > Le 08/07/2012 23:12, Beno?t Minisini a ?crit : > > Le 08/07/2012 23:07, Emil Lenngren a ?crit : > >> Maybe.. > >> > >> I think it is very strange that interpreted functions with debugging > >> flags sometimes are faster than without... > >> > >> This code: > >> Dim i As Integer = 1 > >> While i <= 80000000 > >> Inc i > >> Wend > >> > >> Takes 3.48 seconds with debugging flags, and 5.66 seconds without :/ > >> I think it should be the opposite. > >> > >> /Emil > >> > > > > So do I! > > > > Well, I cannot reproduce what you are talking about. Can you be more > precise about what you do exactly? > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Sun Jul 8 23:29:28 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 08 Jul 2012 23:29:28 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FF9B41C.2040206@...1...> <4FF9BBA0.4020600@...1...> <4FF9EE4A.6090706@...1...> <4FF9F3C4.6040609@...1...> <4FF9F7A2.8000205@...1...> <4FF9F8E3.5040006@...1...> Message-ID: <4FF9FBB8.8080501@...1...> Le 08/07/2012 23:25, Emil Lenngren a ?crit : > Ok. First I start a new Gambas Console project. MMain.module contains this: > ' Gambas module file > > Public Sub Main() > Dim i As Integer = 1 > While i <= 80000000 > Inc i > Wend > End > > Then I open a terminal and go to the project directory: > > emil at ...700...:~/Programmering/Gambas/DebugTest$ gbc3 -a -g > OK > emil at ...700...:~/Programmering/Gambas/DebugTest$ time gbx3 > > real 0m3.396s > user 0m3.391s > sys 0m0.002s > emil at ...700...:~/Programmering/Gambas/DebugTest$ gbc3 -a > OK > emil at ...700...:~/Programmering/Gambas/DebugTest$ time gbx3 > > real 0m5.599s > user 0m5.592s > sys 0m0.002s > > The debug version is faster than the non-debug version... > > /Emil > > I have the contrary (4.18s & 3.62s)... Very strange! You should use valgrind and kcachegrind to find where things are longer in the second case. -- Beno?t Minisini From gambas at ...1... Mon Jul 9 04:49:23 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Mon, 09 Jul 2012 04:49:23 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> Message-ID: <4FFA46B3.40506@...1...> Le 09/07/2012 04:05, Randall Morgan a ?crit : > Yes, the GSL is huge! Lots of functions and you can find some extensions > to it on the net for specialty uses. I need to spend some time really > getting in to the Gambas code. I am not familure with it at all. The gsl > project was larger than I expected it to be but at the time I had time. > Then I got side track by work. But that is widing down now. So give me a > month and I should be back working on it. In the meantime any help, or > additions you want to make will only aid me in completing the project > and understanding how you would like it to integrate with Gambas. > > I will look at your code a little later this evening. > > Thanks > As for the polynomial, I think it would be simpler to make 'Polynomial' a child of 'Float[]', and ComplexPolynomial a child of 'Complex[]'. Writing a _convert interface will allow to transparently convert between them. NewtonPolynomial are a special case. Another point: at the moment, as Complex is an object, a NULL Complex is NULL. Maybe I will be able to find a way to make NULL act like zero for Complex. It will be easier to use then! -- Beno?t Minisini From gambas at ...1... Mon Jul 9 19:36:08 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Mon, 09 Jul 2012 19:36:08 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FFA46B3.40506@...1...> References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> Message-ID: <4FFB1688.80000@...1...> Le 09/07/2012 04:49, Beno?t Minisini a ?crit : > > As for the polynomial, I think it would be simpler to make 'Polynomial' > a child of 'Float[]', and ComplexPolynomial a child of 'Complex[]'. > Hi Randall, I'm currently trying to reimplement Polynomial by inheriting Float[]... Surprisingly, there is no polynomial arithmetic in gsl. If you are aware of any polynomial arithmetic routines implemented in C (+,-,*,/,modulo), I'm interested! Regards, -- Beno?t Minisini From gambas at ...1... Tue Jul 10 00:23:31 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Tue, 10 Jul 2012 00:23:31 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FFB1688.80000@...1...> References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> Message-ID: <4FFB59E3.8050506@...1...> Le 09/07/2012 19:36, Beno?t Minisini a ?crit : > Le 09/07/2012 04:49, Beno?t Minisini a ?crit : >> >> As for the polynomial, I think it would be simpler to make 'Polynomial' >> a child of 'Float[]', and ComplexPolynomial a child of 'Complex[]'. >> > > Hi Randall, > > I'm currently trying to reimplement Polynomial by inheriting Float[]... > > Surprisingly, there is no polynomial arithmetic in gsl. If you are aware > of any polynomial arithmetic routines implemented in C (+,-,*,/,modulo), > I'm interested! > > Regards, > Finally inheriting Float[] is not interesting at all for Polynomial. I just have to put everything to the trash now, and do differently. A bit like you did before no doubt! -- Beno?t Minisini From emil.lenngren at ...176... Tue Jul 10 00:50:38 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Tue, 10 Jul 2012 00:50:38 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FFB59E3.8050506@...1...> References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> Message-ID: Some thoughts: Wouldn't it be better to be more generalized with all those operators? I mean now they have to be native functions in a special interface. Inheriting and virtual calls and other OO stuff doesn't seem to work with them. Why not simply see if an object has the method "_add" when you try to add an object with something? Then you are not restricted to native classes, but can use the "standard" principles and calling conventions and inheritance rules. Of course this will be slower than directly calling the correct C function through an interface but it won't be slower than a normal gambas call. /Emil 2012/7/10 Beno?t Minisini > Le 09/07/2012 19:36, Beno?t Minisini a ?crit : > > Le 09/07/2012 04:49, Beno?t Minisini a ?crit : > >> > >> As for the polynomial, I think it would be simpler to make 'Polynomial' > >> a child of 'Float[]', and ComplexPolynomial a child of 'Complex[]'. > >> > > > > Hi Randall, > > > > I'm currently trying to reimplement Polynomial by inheriting Float[]... > > > > Surprisingly, there is no polynomial arithmetic in gsl. If you are aware > > of any polynomial arithmetic routines implemented in C (+,-,*,/,modulo), > > I'm interested! > > > > Regards, > > > > Finally inheriting Float[] is not interesting at all for Polynomial. I > just have to put everything to the trash now, and do differently. A bit > like you did before no doubt! > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Tue Jul 10 01:13:19 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Tue, 10 Jul 2012 01:13:19 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> Message-ID: <4FFB658F.3000204@...1...> Le 10/07/2012 00:50, Emil Lenngren a ?crit : > Some thoughts: > > Wouldn't it be better to be more generalized with all those operators? > I mean now they have to be native functions in a special interface. > Inheriting and virtual calls and other OO stuff doesn't seem to work > with them. > > Why not simply see if an object has the method "_add" when you try to > add an object with something? Then you are not restricted to native > classes, but can use the "standard" principles and calling conventions > and inheritance rules. Of course this will be slower than directly > calling the correct C function through an interface but it won't be > slower than a normal gambas call. > > /Emil > Mainly because of speed. The '_operator' interface is a bit complex, but it is faster to call it than using the special methods. And I don't want to allow overriding operators in any class like in C++. It leads to nightmare-code. -- Beno?t Minisini From emil.lenngren at ...176... Tue Jul 10 01:20:46 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Tue, 10 Jul 2012 01:20:46 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FFB658F.3000204@...1...> References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> <4FFB658F.3000204@...1...> Message-ID: Ok, that's fine. Tell me when you think the specifications are finished so I can implement it in the jit. /Emil 2012/7/10 Beno?t Minisini > Le 10/07/2012 00:50, Emil Lenngren a ?crit : > > Some thoughts: > > > > Wouldn't it be better to be more generalized with all those operators? > > I mean now they have to be native functions in a special interface. > > Inheriting and virtual calls and other OO stuff doesn't seem to work > > with them. > > > > Why not simply see if an object has the method "_add" when you try to > > add an object with something? Then you are not restricted to native > > classes, but can use the "standard" principles and calling conventions > > and inheritance rules. Of course this will be slower than directly > > calling the correct C function through an interface but it won't be > > slower than a normal gambas call. > > > > /Emil > > > > Mainly because of speed. > > The '_operator' interface is a bit complex, but it is faster to call it > than using the special methods. > > And I don't want to allow overriding operators in any class like in C++. > It leads to nightmare-code. > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Tue Jul 10 17:26:32 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Tue, 10 Jul 2012 17:26:32 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> <4FFB658F.3000204@...1...> Message-ID: <4FFC49A8.3070204@...1...> Le 10/07/2012 01:20, Emil Lenngren a ?crit : > Ok, that's fine. Tell me when you think the specifications are finished > so I can implement it in the jit. > > /Emil > OK, I think that 'operators on object' is more or less finished. It can handle '+', '-', '*', '/', 'Abs', '=' and '<>' (that are mandatory) and '^' (that is optional). I will have to add the '%' (modulo) operator soon. Regards, -- Beno?t Minisini From rmorgan62 at ...176... Tue Jul 10 17:44:07 2012 From: rmorgan62 at ...176... (Randall Morgan) Date: Tue, 10 Jul 2012 08:44:07 -0700 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FFC49A8.3070204@...1...> References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> <4FFB658F.3000204@...1...> <4FFC49A8.3070204@...1...> Message-ID: That's great Benoit. I'll need to start reading your code today. On Tue, Jul 10, 2012 at 8:26 AM, Beno?t Minisini < gambas at ...1...> wrote: > Le 10/07/2012 01:20, Emil Lenngren a ?crit : > > Ok, that's fine. Tell me when you think the specifications are finished > > so I can implement it in the jit. > > > > /Emil > > > > OK, I think that 'operators on object' is more or less finished. > > It can handle '+', '-', '*', '/', 'Abs', '=' and '<>' (that are > mandatory) and '^' (that is optional). > > I will have to add the '%' (modulo) operator soon. > > Regards, > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -- If you ask me if it can be done. The answer is YES, it can always be done. The correct questions however are... What will it cost, and how long will it take? -------------- next part -------------- An HTML attachment was scrubbed... URL: From emil.lenngren at ...176... Wed Jul 11 22:06:30 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Wed, 11 Jul 2012 22:06:30 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> <4FFB658F.3000204@...1...> <4FFC49A8.3070204@...1...> Message-ID: Next bug: Public Sub Main() Dim a As Object = New Complex(2, 1) Dim i As Integer For i = 1 To 2 Print a + 2.0 a = Me Next End -> Segmentation fault. And some questions: I see that if a class inherits Complex (at least a non-native class), the operators are also inherited = good. When I implement this in the JIT, can I assume that child classes do not override operators, i.e. can I codegen a direct call to the operator (if the type is a "pure object")? Or do I have to use virtual dispatching? /Emil 2012/7/10 Randall Morgan > That's great Benoit. I'll need to start reading your code today. > > > > On Tue, Jul 10, 2012 at 8:26 AM, Beno?t Minisini < > gambas at ...1...> wrote: > >> Le 10/07/2012 01:20, Emil Lenngren a ?crit : >> > Ok, that's fine. Tell me when you think the specifications are finished >> > so I can implement it in the jit. >> > >> > /Emil >> > >> >> OK, I think that 'operators on object' is more or less finished. >> >> It can handle '+', '-', '*', '/', 'Abs', '=' and '<>' (that are >> mandatory) and '^' (that is optional). >> >> I will have to add the '%' (modulo) operator soon. >> >> Regards, >> >> -- >> Beno?t Minisini >> >> >> >> >> ------------------------------------------------------------------------------ >> Live Security Virtual Conference >> Exclusive live event will cover all the ways today's security and >> threat landscape has changed and how IT managers can respond. Discussions >> will include endpoint security, mobile security and the latest in malware >> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >> _______________________________________________ >> Gambas-devel mailing list >> Gambas-devel at lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/gambas-devel >> > > > > -- > If you ask me if it can be done. The answer is YES, it can always be done. > The correct questions however are... What will it cost, and how long will > it take? > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From tobias at ...692... Wed Jul 11 22:15:04 2012 From: tobias at ...692... (Tobias Boege) Date: Wed, 11 Jul 2012 22:15:04 +0200 Subject: [Gambas-devel] gb.ncurses continuation Message-ID: <20120711201504.GA1894@...693...> Hi all, just in case you wondered... My school year is finally about to end in 2 weeks. If I appear to be still alive after some driving lessons in the holidays I will happily attend to gb.ncurses again, there are a lot more things to do. I think, I will integrate the CDK instead of writing controls from scratch then. Regards, Tobias From gambas at ...1... Wed Jul 11 23:50:57 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Wed, 11 Jul 2012 23:50:57 +0200 Subject: [Gambas-devel] gb.ncurses continuation In-Reply-To: <20120711201504.GA1894@...693...> References: <20120711201504.GA1894@...693...> Message-ID: <4FFDF541.5000407@...1...> Le 11/07/2012 22:15, Tobias Boege a ?crit : > Hi all, > > just in case you wondered... > My school year is finally about to end in 2 weeks. If I appear to be still alive after some > driving lessons in the holidays I will happily attend to gb.ncurses again, there are a lot more > things to do. I think, I will integrate the CDK instead of writing controls from scratch then. > > Regards, > Tobias > Cool. :-) -- Beno?t Minisini From gambas at ...1... Thu Jul 12 00:11:01 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Thu, 12 Jul 2012 00:11:01 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> <4FFB658F.3000204@...1...> <4FFC49A8.3070204@...1...> Message-ID: <4FFDF9F5.1020000@...1...> Le 11/07/2012 22:06, Emil Lenngren a ?crit : > Next bug: > Public Sub Main() > Dim a As Object = New Complex(2, 1) > Dim i As Integer > For i = 1 To 2 > Print a + 2.0 > a = Me > Next > End > > -> Segmentation fault. OK, fixed in revision #4944. > > And some questions: > I see that if a class inherits Complex (at least a non-native class), > the operators are also inherited = good. > When I implement this in the JIT, can I assume that child classes do not > override operators, i.e. can I codegen a direct call to the operator (if > the type is a "pure object")? Or do I have to use virtual dispatching? > > /Emil > At first sight, I would say that you have to use virtual dispatching, as nothing prevents a class to reimplement the _operators interface. But in other way, I don't find any real use of that. If A inherits B, then being a B implies being an A, so adding Bs should be handled by the A addition operator. But we never use inheritance for numbers: otherwise an integer would inherit a floating point that woul inherit a Complex that would inherit a Matrix, and everything would be a Matrix internally! So, even if I would not prevent reimplementing the _operators interface, I will never do it, and so you can assume that you can do a direct call. If there is a problem in the future, we'll see... -- Beno?t Minisini From emil.lenngren at ...176... Thu Jul 12 00:19:29 2012 From: emil.lenngren at ...176... (Emil Lenngren) Date: Thu, 12 Jul 2012 00:19:29 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: <4FFDF9F5.1020000@...1...> References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> <4FFB658F.3000204@...1...> <4FFC49A8.3070204@...1...> <4FFDF9F5.1020000@...1...> Message-ID: OK, good. I will be on holiday from tomorrow until the first week in August, so I will start work when I come home. /Emil 2012/7/12 Beno?t Minisini > Le 11/07/2012 22:06, Emil Lenngren a ?crit : > > Next bug: > > Public Sub Main() > > Dim a As Object = New Complex(2, 1) > > Dim i As Integer > > For i = 1 To 2 > > Print a + 2.0 > > a = Me > > Next > > End > > > > -> Segmentation fault. > > OK, fixed in revision #4944. > > > > > And some questions: > > I see that if a class inherits Complex (at least a non-native class), > > the operators are also inherited = good. > > When I implement this in the JIT, can I assume that child classes do not > > override operators, i.e. can I codegen a direct call to the operator (if > > the type is a "pure object")? Or do I have to use virtual dispatching? > > > > /Emil > > > > At first sight, I would say that you have to use virtual dispatching, as > nothing prevents a class to reimplement the _operators interface. > > But in other way, I don't find any real use of that. > > If A inherits B, then being a B implies being an A, so adding Bs should > be handled by the A addition operator. > > But we never use inheritance for numbers: otherwise an integer would > inherit a floating point that woul inherit a Complex that would inherit > a Matrix, and everything would be a Matrix internally! > > So, even if I would not prevent reimplementing the _operators interface, > I will never do it, and so you can assume that you can do a direct call. > > If there is a problem in the future, we'll see... > > -- > Beno?t Minisini > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Gambas-devel mailing list > Gambas-devel at lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/gambas-devel > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Thu Jul 12 00:33:46 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Thu, 12 Jul 2012 00:33:46 +0200 Subject: [Gambas-devel] About gb.gsl components In-Reply-To: References: <4FF99C73.5040909@...1...> <4FFA3A08.9050805@...1...> <4FFA46B3.40506@...1...> <4FFB1688.80000@...1...> <4FFB59E3.8050506@...1...> <4FFB658F.3000204@...1...> <4FFC49A8.3070204@...1...> <4FFDF9F5.1020000@...1...> Message-ID: <4FFDFF4A.10803@...1...> Le 12/07/2012 00:19, Emil Lenngren a ?crit : > OK, good. > > I will be on holiday from tomorrow until the first week in August, so I > will start work when I come home. > > /Emil > So good holidays! -- Beno?t Minisini From lordheavym at ...176... Sat Jul 14 23:28:30 2012 From: lordheavym at ...176... (Laurent Carlier) Date: Sat, 14 Jul 2012 23:28:30 +0200 Subject: [Gambas-devel] Wrong build order for gb.xml.rpc component Message-ID: <2125838.CenZ0qhiHH@...676...> When packaging gb.xml.rpc, component was always failling to build since gb.xml.rpc component needs gb.xml component that is always built after it, because gb.xml.rpc is a pure gambas component, and it depends on gb.xml to be properly built. The solution is to move gb.xml component also in a subdir, so build order can be changed in the SUBDIR variable in the Makefile.am file src \_xml \_gb.xml \_html \_rpc \_gb.xml.rpc \_xslt Makefile.am in src dir: SUBDIR = xml @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc Sample of installation log of gb.xml.rpc component: ---8<--- make[2]: Leaving directory `/build/src/gambas3-3.2.0/gb.xml/src/xslt' Making install in rpc make[2]: Entering directory `/build/src/gambas3-3.2.0/gb.xml/src/rpc' make[3]: Entering directory `/build/src/gambas3-3.2.0/gb.xml/src/rpc' make[3]: Nothing to be done for `install-exec-am'. /bin/mkdir -p '/build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3' /usr/bin/install -c -m 644 gb.xml.rpc.component '/build/pkg/gambas3-gb-xml- rpc/usr/lib/gambas3' /bin/mkdir -p '/build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3' /usr/bin/install -c -m 644 gb.xml.rpc.component '/build/pkg/gambas3-gb-xml- rpc/usr/lib/gambas3' make install-data-hook make[4]: Entering directory `/build/src/gambas3-3.2.0/gb.xml/src/rpc' Compiling the gb.xml.rpc project... gbi3: warning: component gb.xml.rpc not found gb.xml.rpc gbc: error: Component not found: gb.xml Installing the gb.xml.rpc control icons if needed... Creating the information files for gb.xml.rpc component... gbi3: warning: component gb.xml.rpc not found gb.xml.rpc make[4]: Leaving directory `/build/src/gambas3-3.2.0/gb.xml/src/rpc' make[3]: Leaving directory `/build/src/gambas3-3.2.0/gb.xml/src/rpc' make[2]: Leaving directory `/build/src/gambas3-3.2.0/gb.xml/src/rpc' make[2]: Entering directory `/build/src/gambas3-3.2.0/gb.xml/src' make[3]: Entering directory `/build/src/gambas3-3.2.0/gb.xml/src' make[3]: Nothing to be done for `install-exec-am'. /bin/mkdir -p '/build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3' /usr/bin/install -c -m 644 gb.xml.component '/build/pkg/gambas3-gb-xml- rpc/usr/lib/gambas3' /bin/mkdir -p '/build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3' /usr/bin/install -c -m 644 gb.xml.component '/build/pkg/gambas3-gb-xml- rpc/usr/lib/gambas3' /bin/mkdir -p '/build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3' /bin/sh ../libtool --mode=install /usr/bin/install -c gb.xml.la '/build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3' libtool: install: /usr/bin/install -c .libs/gb.xml.so.0.0.0 /build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3/gb.xml.so.0.0.0 libtool: install: (cd /build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3 && { ln -s -f gb.xml.so.0.0.0 gb.xml.so.0 || { rm -f gb.xml.so.0 && ln -s gb.xml.so.0.0.0 gb.xml.so.0; }; }) libtool: install: (cd /build/pkg/gambas3-gb-xml-rpc/usr/lib/gambas3 && { ln -s -f gb.xml.so.0.0.0 gb.xml.so || { rm -f gb.xml.so && ln -s gb.xml.so.0.0.0 gb.xml.so; }; }) libtool: install: /usr/bin/install -c .libs/gb.xml.lai /build/pkg/gambas3-gb- xml-rpc/usr/lib/gambas3/gb.xml.la libtool: install: warning: remember to run `libtool --finish /usr/lib/gambas3' make install-data-hook make[4]: Entering directory `/build/src/gambas3-3.2.0/gb.xml/src' Compiling the gb.xml project... gb.xml OK Installing the gb.xml control icons if needed... Creating the information files for gb.xml component... gb.xml make[4]: Leaving directory `/build/src/gambas3-3.2.0/gb.xml/src' make[3]: Leaving directory `/build/src/gambas3-3.2.0/gb.xml/src' make[2]: Leaving directory `/build/src/gambas3-3.2.0/gb.xml/src' make[1]: Leaving directory `/build/src/gambas3-3.2.0/gb.xml/src' ---8<--- Regards, -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 490 bytes Desc: This is a digitally signed message part. URL: From gambas at ...1... Sun Jul 15 00:01:01 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 15 Jul 2012 00:01:01 +0200 Subject: [Gambas-devel] Wrong build order for gb.xml.rpc component In-Reply-To: <2125838.CenZ0qhiHH@...676...> References: <2125838.CenZ0qhiHH@...676...> Message-ID: <5001EC1D.9060305@...1...> Le 14/07/2012 23:28, Laurent Carlier a ?crit : > When packaging gb.xml.rpc, component was always failling to build since > gb.xml.rpc component needs gb.xml component that is always built after > it, because gb.xml.rpc is a pure gambas component, and it depends on > gb.xml to be properly built. > > The solution is to move gb.xml component also in a subdir, so build > order can be changed in the SUBDIR variable in the Makefile.am file > > src > > \_xml > > \_gb.xml > > \_html > > \_rpc > > \_gb.xml.rpc > > \_xslt > > Makefile.am in src dir: > > SUBDIR = xml @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc > > Sample of installation log of gb.xml.rpc component: > Normally you should be able to define when the top directory is built by adding "." to the SUBDIR variable. SUBDIR = . @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc should make the top directory built before all others. Regards, -- Beno?t Minisini From gambas at ...1... Sun Jul 15 00:07:34 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 15 Jul 2012 00:07:34 +0200 Subject: [Gambas-devel] Wrong build order for gb.xml.rpc component In-Reply-To: <5001EC1D.9060305@...1...> References: <2125838.CenZ0qhiHH@...676...> <5001EC1D.9060305@...1...> Message-ID: <5001EDA6.2040603@...1...> Le 15/07/2012 00:01, Beno?t Minisini a ?crit : > Le 14/07/2012 23:28, Laurent Carlier a ?crit : >> When packaging gb.xml.rpc, component was always failling to build since >> gb.xml.rpc component needs gb.xml component that is always built after >> it, because gb.xml.rpc is a pure gambas component, and it depends on >> gb.xml to be properly built. >> >> The solution is to move gb.xml component also in a subdir, so build >> order can be changed in the SUBDIR variable in the Makefile.am file >> >> src >> >> \_xml >> >> \_gb.xml >> >> \_html >> >> \_rpc >> >> \_gb.xml.rpc >> >> \_xslt >> >> Makefile.am in src dir: >> >> SUBDIR = xml @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc >> >> Sample of installation log of gb.xml.rpc component: >> > > Normally you should be able to define when the top directory is built by > adding "." to the SUBDIR variable. > > SUBDIR = . @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc > > should make the top directory built before all others. > > Regards, > Sorry, it's already the case. I have no built problem there anyway, so I don't understand why you have one... I join the output of my "make install": gb.xml is actually built and installed before gb.xml.rpc. Regards, -- Beno?t Minisini -------------- next part -------------- benoit at ...701...:~/gambas/3.0/trunk/gb.xml$ make install DESTDIR=~/root Making install in src make[1]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src ? Making install in . make[2]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src ? make[3]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src ? make[3]: Rien ? faire pour ? install-exec-am ?. test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /usr/bin/install -c -m 644 gb.xml.component '/home/benoit/root/usr/local/lib/gambas3' test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /usr/bin/install -c -m 644 gb.xml.component '/home/benoit/root/usr/local/lib/gambas3' test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /bin/bash ../libtool --mode=install /usr/bin/install -c gb.xml.la '/home/benoit/root/usr/local/lib/gambas3' libtool: install: /usr/bin/install -c .libs/gb.xml.so.0.0.0 /home/benoit/root/usr/local/lib/gambas3/gb.xml.so.0.0.0 libtool: install: (cd /home/benoit/root/usr/local/lib/gambas3 && { ln -s -f gb.xml.so.0.0.0 gb.xml.so.0 || { rm -f gb.xml.so.0 && ln -s gb.xml.so.0.0.0 gb.xml.so.0; }; }) libtool: install: (cd /home/benoit/root/usr/local/lib/gambas3 && { ln -s -f gb.xml.so.0.0.0 gb.xml.so || { rm -f gb.xml.so && ln -s gb.xml.so.0.0.0 gb.xml.so; }; }) libtool: install: /usr/bin/install -c .libs/gb.xml.lai /home/benoit/root/usr/local/lib/gambas3/gb.xml.la libtool: install: warning: remember to run `libtool --finish /usr/local/lib/gambas3' make install-data-hook make[4]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src ? Compiling the gb.xml project... gb.xml OK Installing the gb.xml control icons if needed... Creating the information files for gb.xml component... gb.xml make[4]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src ? make[3]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src ? make[2]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src ? Making install in html make[2]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/html ? make[3]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/html ? make[3]: Rien ? faire pour ? install-exec-am ?. test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /usr/bin/install -c -m 644 gb.xml.html.component '/home/benoit/root/usr/local/lib/gambas3' test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /usr/bin/install -c -m 644 gb.xml.html.component '/home/benoit/root/usr/local/lib/gambas3' test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /bin/bash ../../libtool --mode=install /usr/bin/install -c gb.xml.html.la '/home/benoit/root/usr/local/lib/gambas3' libtool: install: /usr/bin/install -c .libs/gb.xml.html.so.0.0.0 /home/benoit/root/usr/local/lib/gambas3/gb.xml.html.so.0.0.0 libtool: install: (cd /home/benoit/root/usr/local/lib/gambas3 && { ln -s -f gb.xml.html.so.0.0.0 gb.xml.html.so.0 || { rm -f gb.xml.html.so.0 && ln -s gb.xml.html.so.0.0.0 gb.xml.html.so.0; }; }) libtool: install: (cd /home/benoit/root/usr/local/lib/gambas3 && { ln -s -f gb.xml.html.so.0.0.0 gb.xml.html.so || { rm -f gb.xml.html.so && ln -s gb.xml.html.so.0.0.0 gb.xml.html.so; }; }) libtool: install: /usr/bin/install -c .libs/gb.xml.html.lai /home/benoit/root/usr/local/lib/gambas3/gb.xml.html.la libtool: install: warning: remember to run `libtool --finish /usr/local/lib/gambas3' make install-data-hook make[4]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/html ? Installing the gb.xml.html control icons if needed... Creating the information files for gb.xml.html component... gb.xml.html make[4]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/html ? make[3]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/html ? make[2]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/html ? Making install in xslt make[2]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/xslt ? make[3]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/xslt ? make[3]: Rien ? faire pour ? install-exec-am ?. test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /usr/bin/install -c -m 644 gb.xml.xslt.component '/home/benoit/root/usr/local/lib/gambas3' test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /usr/bin/install -c -m 644 gb.xml.xslt.component '/home/benoit/root/usr/local/lib/gambas3' test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /bin/bash ../../libtool --mode=install /usr/bin/install -c gb.xml.xslt.la '/home/benoit/root/usr/local/lib/gambas3' libtool: install: /usr/bin/install -c .libs/gb.xml.xslt.so.0.0.0 /home/benoit/root/usr/local/lib/gambas3/gb.xml.xslt.so.0.0.0 libtool: install: (cd /home/benoit/root/usr/local/lib/gambas3 && { ln -s -f gb.xml.xslt.so.0.0.0 gb.xml.xslt.so.0 || { rm -f gb.xml.xslt.so.0 && ln -s gb.xml.xslt.so.0.0.0 gb.xml.xslt.so.0; }; }) libtool: install: (cd /home/benoit/root/usr/local/lib/gambas3 && { ln -s -f gb.xml.xslt.so.0.0.0 gb.xml.xslt.so || { rm -f gb.xml.xslt.so && ln -s gb.xml.xslt.so.0.0.0 gb.xml.xslt.so; }; }) libtool: install: /usr/bin/install -c .libs/gb.xml.xslt.lai /home/benoit/root/usr/local/lib/gambas3/gb.xml.xslt.la libtool: install: warning: remember to run `libtool --finish /usr/local/lib/gambas3' make install-data-hook make[4]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/xslt ? Installing the gb.xml.xslt control icons if needed... Creating the information files for gb.xml.xslt component... gb.xml.xslt make[4]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/xslt ? make[3]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/xslt ? make[2]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/xslt ? Making install in rpc make[2]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/rpc ? make[3]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/rpc ? make[3]: Rien ? faire pour ? install-exec-am ?. test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /usr/bin/install -c -m 644 gb.xml.rpc.component '/home/benoit/root/usr/local/lib/gambas3' test -z "/usr/local/lib/gambas3" || /bin/mkdir -p "/home/benoit/root/usr/local/lib/gambas3" /usr/bin/install -c -m 644 gb.xml.rpc.component '/home/benoit/root/usr/local/lib/gambas3' make install-data-hook make[4]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/rpc ? Compiling the gb.xml.rpc project... gb.xml.rpc OK Installing the gb.xml.rpc control icons if needed... Creating the information files for gb.xml.rpc component... gb.xml.rpc make[4]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/rpc ? make[3]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/rpc ? make[2]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src/rpc ? make[1]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml/src ? make[1]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml ? make[2]: entrant dans le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml ? make[2]: Rien ? faire pour ? install-exec-am ?. make[2]: Rien ? faire pour ? install-data-am ?. make[2]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml ? make[1]: quittant le r?pertoire ? /home/benoit/gambas/3.0/trunk/gb.xml ? benoit at ...701...:~/gambas/3.0/trunk/gb.xml$ From lordheavym at ...176... Sun Jul 15 00:13:08 2012 From: lordheavym at ...176... (Laurent Carlier) Date: Sun, 15 Jul 2012 00:13:08 +0200 Subject: [Gambas-devel] Wrong build order for gb.xml.rpc component In-Reply-To: <5001EDA6.2040603@...1...> References: <2125838.CenZ0qhiHH@...676...> <5001EC1D.9060305@...1...> <5001EDA6.2040603@...1...> Message-ID: <2907788.fFgM6elziS@...676...> Le dimanche 15 juillet 2012 00:07:34 Beno?t Minisini a ?crit : > Le 15/07/2012 00:01, Beno?t Minisini a ?crit : > > Le 14/07/2012 23:28, Laurent Carlier a ?crit : > >> When packaging gb.xml.rpc, component was always failling to build since > >> gb.xml.rpc component needs gb.xml component that is always built after > >> it, because gb.xml.rpc is a pure gambas component, and it depends on > >> gb.xml to be properly built. > >> > >> The solution is to move gb.xml component also in a subdir, so build > >> order can be changed in the SUBDIR variable in the Makefile.am file > >> > >> src > >> > >> \_xml > >> > >> \_gb.xml > >> > >> \_html > >> > >> \_rpc > >> > >> \_gb.xml.rpc > >> > >> \_xslt > >> > >> Makefile.am in src dir: > >> > >> SUBDIR = xml @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc > > > >> Sample of installation log of gb.xml.rpc component: > > Normally you should be able to define when the top directory is built by > > adding "." to the SUBDIR variable. > > > > SUBDIR = . @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc > > > > should make the top directory built before all others. > > > > Regards, > > Sorry, it's already the case. > > I have no built problem there anyway, so I don't understand why you have > one... > > I join the output of my "make install": gb.xml is actually built and > installed before gb.xml.rpc. > > Regards, It install fine on your computer because you have already a gb.xml component installed, on a system without a gb.xml component installed, it will fail. ++ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 490 bytes Desc: This is a digitally signed message part. URL: From lordheavym at ...176... Sun Jul 15 00:18:13 2012 From: lordheavym at ...176... (Laurent Carlier) Date: Sun, 15 Jul 2012 00:18:13 +0200 Subject: [Gambas-devel] Wrong build order for gb.xml.rpc component In-Reply-To: <5001EDA6.2040603@...1...> References: <2125838.CenZ0qhiHH@...676...> <5001EC1D.9060305@...1...> <5001EDA6.2040603@...1...> Message-ID: <2610008.NFWlemsfQP@...676...> Le dimanche 15 juillet 2012 00:07:34 Beno?t Minisini a ?crit : > I join the output of my "make install": gb.xml is actually built and > installed before gb.xml.rpc. > > Regards, Funny, here it is built and installed after..... ++ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 490 bytes Desc: This is a digitally signed message part. URL: From lordheavym at ...176... Sun Jul 15 00:31:30 2012 From: lordheavym at ...176... (Laurent Carlier) Date: Sun, 15 Jul 2012 00:31:30 +0200 Subject: [Gambas-devel] Wrong build order for gb.xml.rpc component In-Reply-To: <5001EDA6.2040603@...1...> References: <2125838.CenZ0qhiHH@...676...> <5001EC1D.9060305@...1...> <5001EDA6.2040603@...1...> Message-ID: <2984930.5WhiNF3ekh@...676...> Le dimanche 15 juillet 2012 00:07:34 Beno?t Minisini a ?crit : > Le 15/07/2012 00:01, Beno?t Minisini a ?crit : > > Le 14/07/2012 23:28, Laurent Carlier a ?crit : > >> When packaging gb.xml.rpc, component was always failling to build since > >> gb.xml.rpc component needs gb.xml component that is always built after > >> it, because gb.xml.rpc is a pure gambas component, and it depends on > >> gb.xml to be properly built. > >> > >> The solution is to move gb.xml component also in a subdir, so build > >> order can be changed in the SUBDIR variable in the Makefile.am file > >> > >> src > >> > >> \_xml > >> > >> \_gb.xml > >> > >> \_html > >> > >> \_rpc > >> > >> \_gb.xml.rpc > >> > >> \_xslt > >> > >> Makefile.am in src dir: > >> > >> SUBDIR = xml @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc > > > >> Sample of installation log of gb.xml.rpc component: > > Normally you should be able to define when the top directory is built by > > adding "." to the SUBDIR variable. > > > > SUBDIR = . @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc > > > > should make the top directory built before all others. > > > > Regards, > > Sorry, it's already the case. > > I have no built problem there anyway, so I don't understand why you have > one... > > I join the output of my "make install": gb.xml is actually built and > installed before gb.xml.rpc. > > Regards, Ok, i've checked again the Makefile.am provided with the 3.2.0 release and the dot is missing: ---8<--- COMPONENT = gb.xml include $(top_srcdir)/component.am SUBDIRS = @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc gblib_LTLIBRARIES = gb.xml.la gb_xml_la_LIBADD = @XML_LIB@ gb_xml_la_LDFLAGS = -module @LD_FLAGS@ @XML_LDFLAGS@ gb_xml_la_CPPFLAGS = @XML_INC@ ---8<--- But it's present in the Makefile.am file in the trunk. -> Bug found and already fixed! ++ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 490 bytes Desc: This is a digitally signed message part. URL: From gambas at ...1... Sun Jul 15 00:45:42 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 15 Jul 2012 00:45:42 +0200 Subject: [Gambas-devel] Wrong build order for gb.xml.rpc component In-Reply-To: <2984930.5WhiNF3ekh@...676...> References: <2125838.CenZ0qhiHH@...676...> <5001EC1D.9060305@...1...> <5001EDA6.2040603@...1...> <2984930.5WhiNF3ekh@...676...> Message-ID: <5001F696.2000200@...1...> Le 15/07/2012 00:31, Laurent Carlier a ?crit : > > Ok, i've checked again the Makefile.am provided with the 3.2.0 release > and the dot is missing: > > ---8<--- > > COMPONENT = gb.xml > > include $(top_srcdir)/component.am > > SUBDIRS = @XMLHTML_DIR@ @XMLXSLT_DIR@ rpc > > gblib_LTLIBRARIES = gb.xml.la > > gb_xml_la_LIBADD = @XML_LIB@ > > gb_xml_la_LDFLAGS = -module @LD_FLAGS@ @XML_LDFLAGS@ > > gb_xml_la_CPPFLAGS = @XML_INC@ > > ---8<--- > > But it's present in the Makefile.am file in the trunk. > > -> Bug found and already fixed! > > ++ > I must remember to make a 3.2.1 with that fix... -- Beno?t Minisini From silvan.calarco at ...531... Fri Jul 20 11:11:45 2012 From: silvan.calarco at ...531... (Silvan Calarco) Date: Fri, 20 Jul 2012 11:11:45 +0200 Subject: [Gambas-devel] Gambas 3.2.0 (arm): gbc3: Program received signal SIGSEGV, Segmentation fault. Message-ID: <3479651.dMp41DqWio@...704...> Hi all. I have a problem building gambas 3.2.0 on arm, gcc is 4.7.1, because gbc3 crashes here: static void control_add_pos(short **tab_pos, short pos) { if (!(*tab_pos)) ARRAY_create(tab_pos); > *((short *)ARRAY_add(tab_pos)) = pos; } Full backtrace follows. Do you have any hints on how this could be fixed? Thanks. Regards, Silvan --- [sdk at ...702... gb.settings]$ gdb --args /var/tmp/gambas3-root/usr/bin/gbc3 -ag - r /var/tmp/gambas3-root/usr GNU gdb (GDB) 7.4.1-1mamba (openmamba) Copyright (C) 2012 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "arm-openmamba-linux-gnueabi". For bug reporting instructions, please see: ... Reading symbols from /var/tmp/gambas3-root/usr/bin/gbc3...done. (gdb) r Starting program: /var/tmp/gambas3-root/usr/bin/gbc3 -ag -r /var/tmp/gambas3- root/usr Program received signal SIGSEGV, Segmentation fault. 0x00020f04 in control_add_pos (tab_pos=0x393a6, tab_pos at ...703...=0x3ecec, pos=) at gbc_trans_ctrl.c:86 86 *((short *)ARRAY_add(tab_pos)) = pos; (gdb) bt full #0 0x00020f04 in control_add_pos (tab_pos=0x393a6, tab_pos at ...703...=0x3ecec, pos=) at gbc_trans_ctrl.c:86 array = 0x385bfff5 ptr = 0x0 old_count = #1 0x00020f60 in control_add_current_pos () at gbc_trans_ctrl.c:92 No locals. #2 0x00021020 in trans_else () at gbc_trans_ctrl.c:437 nobreak = #3 0x00021278 in trans_if () at gbc_trans_ctrl.c:494 mode = msg = 0x2c6c9 "OR IF" #4 0x000217bc in trans_else_if () at gbc_trans_ctrl.c:542 No locals. #5 TRANS_else () at gbc_trans_ctrl.c:568 No locals. #6 0xbe9e014c in ?? () No symbol table info available. #7 0xbe9e014c in ?? () No symbol table info available. Backtrace stopped: previous frame identical to this frame (corrupt stack?) (gdb) -- mambaSoft di Calarco Silvan Web: http://www.mambasoft.it mambaSoft Store @ http://www.mambastore.it openmamba GNU/Linux development @ http://www.openmamba.org From gambas at ...1... Sun Jul 22 20:53:39 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Sun, 22 Jul 2012 20:53:39 +0200 Subject: [Gambas-devel] Gambas 3.2.0 (arm): gbc3: Program received signal SIGSEGV, Segmentation fault. In-Reply-To: <3479651.dMp41DqWio@...704...> References: <3479651.dMp41DqWio@...704...> Message-ID: <500C4C33.9040200@...1...> Le 20/07/2012 11:11, Silvan Calarco a ?crit : > Hi all. > I have a problem building gambas 3.2.0 on arm, gcc is 4.7.1, because gbc3 > crashes here: > > static void control_add_pos(short **tab_pos, short pos) > { > if (!(*tab_pos)) > ARRAY_create(tab_pos); > >> *((short *)ARRAY_add(tab_pos)) = pos; > } > > Full backtrace follows. Do you have any hints on how this could be fixed? > Thanks. > > Regards, > Silvan > Please can you compile Gambas with no optimization and debugging information enabled? Otherwise the backtrace is almost useless. To do that, instead of doing just "make", do that: $ make "CFLAGS=-O0 -g -ggdb" "CXXFLAGS=-O0 -g -ggdb" Regards, -- Beno?t Minisini From silvan.calarco at ...531... Mon Jul 23 01:14:21 2012 From: silvan.calarco at ...531... (Silvan Calarco) Date: Mon, 23 Jul 2012 01:14:21 +0200 Subject: [Gambas-devel] Gambas 3.2.0 (arm): gbc3: Program received signal SIGSEGV, Segmentation fault. In-Reply-To: <500C4C33.9040200@...1...> References: <3479651.dMp41DqWio@...704...> <500C4C33.9040200@...1...> Message-ID: <2426564.GLkAgQczQF@...704...> In data domenica 22 luglio 2012 20:53:39, Beno?t Minisini ha scritto: > Le 20/07/2012 11:11, Silvan Calarco a ?crit : > > Hi all. > > I have a problem building gambas 3.2.0 on arm, gcc is 4.7.1, because gbc3 > > > > crashes here: > > static void control_add_pos(short **tab_pos, short pos) > > { > > > > if (!(*tab_pos)) > > > > ARRAY_create(tab_pos); > >> > >> *((short *)ARRAY_add(tab_pos)) = pos; > > > > } > > > > Full backtrace follows. Do you have any hints on how this could be fixed? > > Thanks. > > > > Regards, > > Silvan > > Please can you compile Gambas with no optimization and debugging > information enabled? Otherwise the backtrace is almost useless. > > To do that, instead of doing just "make", do that: > > $ make "CFLAGS=-O0 -g -ggdb" "CXXFLAGS=-O0 -g -ggdb" > > Regards, Here it is: [sdk at ...702... gb.qt4]$ gdb --args /var/tmp/gambas-root//usr/bin/gbc3 -ag -r /var/tmp/gambas-root//usr GNU gdb (GDB) 7.4.1-1mamba (openmamba) Copyright (C) 2012 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "arm-openmamba-linux-gnueabi". For bug reporting instructions, please see: ... Reading symbols from /var/tmp/gambas-root/usr/bin/gbc3...done. (gdb) r Starting program: /var/tmp/gambas-root/usr/bin/gbc3 -ag -r /var/tmp/gambas- root//usr Program received signal SIGSEGV, Segmentation fault. 0x0003b224 in control_add_pos (tab_pos=0x5cfe2, pos=7) at gbc_trans_ctrl.c:87 87 *((short *)ARRAY_add(tab_pos)) = pos; (gdb) bt full #0 0x0003b224 in control_add_pos (tab_pos=0x5cfe2, pos=7) at gbc_trans_ctrl.c:87 array = 0x3a13fff6 ptr = 0xffffffff old_count = 161 #1 0x0003b2e0 in control_add_current_pos () at gbc_trans_ctrl.c:93 No locals. #2 0x0003c8e4 in trans_else () at gbc_trans_ctrl.c:438 nobreak = 0 '\000' #3 0x0003cab8 in trans_if () at gbc_trans_ctrl.c:495 mode = 153 msg = 0x4fc88 "AND IF" #4 0x0003cc6c in TRANS_if () at gbc_trans_ctrl.c:551 No locals. #5 0x0003631c in translate_body () at gbc_trans_code.c:332 look = 0x6d1b4 is_proc = 0 '\000' test_newline = 1 '\001' line = 28 just_got_select = 0 '\000' #6 0x00036c9c in TRANS_code () at gbc_trans_code.c:536 i = 2 #7 0x0004b454 in compile_file (file=0x6343c "/usr/src/RPM/BUILD/gambas3-3.2.0/gb.qt4/src/gb.qt4/.src/Action.class") at gbc.c:336 i = 17175534 time_src = 0 time_form = 405744 time_pot = 398268 time_output = 309272 source = 0xbecd3184 "\340\271\004" #8 0x0004b9e0 in main (argc=4, argv=0xbecd3494) at gbc.c:472 __err = 0xbecd3198 __err_context = {prev = 0x0, code = 0, env = {{__jmpbuf = {0, 0, 0, 0, 0, 0, 1073905584, -1093848252, -1093848696, 309528, 0, 0, 0, 1073741824, 1073743676, 1073905616, 1073838196, -1093848560, 1073742584, 1073762232, 0, 0, 0, 0, -1093847920, 1073907760, 1073843988, 1073905584, 1073741824, 0, 1073905384, 0, 0, 0, 0, 0, 0, 0, 1073905448, 1073905440, 1073905392, 1073905408, 1073905416, 0, 0, 0, 1073905424, 1073905432, 0, 0, 1073905384, 0, 0, 1073905472, 1073905480, 1073905488, 1073905456, 0, 0, 1073905464, 1073905512, 0, 0, 0}, __mask_was_saved = 0, __saved_mask = {__val = {0, 0, 0, 0, 0, 0, 0, 1074993496, 3, 0, 1073875008, 4131212846, 1073783400, 0, 1074993496, 1073905000, 129100401, 4294966712, 1075026804, 3201118984, 1073865324, 1073907820, 1074993988, 0, 0, 0, 0, 1, 2051, 1073875064, 1073874240, 37704}}}}} i = 0 Regards, Silvan -- mambaSoft di Calarco Silvan Web: http://www.mambasoft.it mambaSoft Store @ http://www.mambastore.it openmamba GNU/Linux development @ http://www.openmamba.org From gambas at ...1... Mon Jul 23 02:05:53 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Mon, 23 Jul 2012 02:05:53 +0200 Subject: [Gambas-devel] Gambas 3.2.0 (arm): gbc3: Program received signal SIGSEGV, Segmentation fault. In-Reply-To: <2426564.GLkAgQczQF@...704...> References: <3479651.dMp41DqWio@...704...> <500C4C33.9040200@...1...> <2426564.GLkAgQczQF@...704...> Message-ID: <500C9561.9030902@...1...> Le 23/07/2012 01:14, Silvan Calarco a ?crit : > > Here it is: > > [sdk at ...702... gb.qt4]$ gdb --args /var/tmp/gambas-root//usr/bin/gbc3 -ag -r > /var/tmp/gambas-root//usr > GNU gdb (GDB) 7.4.1-1mamba (openmamba) > Copyright (C) 2012 Free Software Foundation, Inc. > License GPLv3+: GNU GPL version 3 or later > This is free software: you are free to change and redistribute it. > There is NO WARRANTY, to the extent permitted by law. Type "show copying" > and "show warranty" for details. > This GDB was configured as "arm-openmamba-linux-gnueabi". > For bug reporting instructions, please see: > ... > Reading symbols from /var/tmp/gambas-root/usr/bin/gbc3...done. > (gdb) r > Starting program: /var/tmp/gambas-root/usr/bin/gbc3 -ag -r /var/tmp/gambas- > root//usr > > Program received signal SIGSEGV, Segmentation fault. > 0x0003b224 in control_add_pos (tab_pos=0x5cfe2, pos=7) at gbc_trans_ctrl.c:87 > 87 *((short *)ARRAY_add(tab_pos)) = pos; > (gdb) bt full > #0 0x0003b224 in control_add_pos (tab_pos=0x5cfe2, pos=7) at > gbc_trans_ctrl.c:87 > array = 0x3a13fff6 > ptr = 0xffffffff > old_count = 161 > #1 0x0003b2e0 in control_add_current_pos () at gbc_trans_ctrl.c:93 > No locals. > #2 0x0003c8e4 in trans_else () at gbc_trans_ctrl.c:438 > nobreak = 0 '\000' > #3 0x0003cab8 in trans_if () at gbc_trans_ctrl.c:495 > mode = 153 > msg = 0x4fc88 "AND IF" > #4 0x0003cc6c in TRANS_if () at gbc_trans_ctrl.c:551 > No locals. > #5 0x0003631c in translate_body () at gbc_trans_code.c:332 > look = 0x6d1b4 > is_proc = 0 '\000' > test_newline = 1 '\001' > line = 28 > just_got_select = 0 '\000' > #6 0x00036c9c in TRANS_code () at gbc_trans_code.c:536 > i = 2 > #7 0x0004b454 in compile_file (file=0x6343c > "/usr/src/RPM/BUILD/gambas3-3.2.0/gb.qt4/src/gb.qt4/.src/Action.class") > at gbc.c:336 > i = 17175534 > time_src = 0 > time_form = 405744 > time_pot = 398268 > time_output = 309272 > source = 0xbecd3184 "\340\271\004" > #8 0x0004b9e0 in main (argc=4, argv=0xbecd3494) at gbc.c:472 > __err = 0xbecd3198 > __err_context = {prev = 0x0, code = 0, env = {{__jmpbuf = {0, 0, 0, 0, > 0, 0, 1073905584, -1093848252, > -1093848696, 309528, 0, 0, 0, 1073741824, 1073743676, > 1073905616, 1073838196, -1093848560, > 1073742584, 1073762232, 0, 0, 0, 0, -1093847920, 1073907760, > 1073843988, 1073905584, 1073741824, 0, > 1073905384, 0, 0, 0, 0, 0, 0, 0, 1073905448, 1073905440, > 1073905392, 1073905408, 1073905416, 0, 0, > 0, 1073905424, 1073905432, 0, 0, 1073905384, 0, 0, 1073905472, > 1073905480, 1073905488, 1073905456, > 0, 0, 1073905464, 1073905512, 0, 0, 0}, __mask_was_saved = 0, > __saved_mask = {__val = {0, 0, 0, 0, > 0, 0, 0, 1074993496, 3, 0, 1073875008, 4131212846, > 1073783400, 0, 1074993496, 1073905000, > 129100401, 4294966712, 1075026804, 3201118984, 1073865324, > 1073907820, 1074993988, 0, 0, 0, 0, 1, > 2051, 1073875064, 1073874240, 37704}}}}} > i = 0 > > Regards, > Silvan > Mmffff... Is it possible to use valgrind on ARM? Which ARM architecture do you use exactly? -- Beno?t Minisini From pata.karlsson at ...176... Mon Jul 23 11:46:41 2012 From: pata.karlsson at ...176... (Patrik Karlsson) Date: Mon, 23 Jul 2012 11:46:41 +0200 Subject: [Gambas-devel] Components name and installation Message-ID: Hi, I finally succeeded in writing my first 'dummy' component, it does not do much, but it looks to be a valid component. :) In the Packaging section of http://gambasdoc.org/help/dev/gambas?v3#t20 I am told to use the name convention gambas3-vendor-name, so I'm using gambas3-trixon-testa. On the same page it says: "Even if the packager wizard allows to insert the package vendor string into the package name, please avoid it." When I create an installation package, the name is gambas3-trixon-gambas3-trixon-testa-3.2.1 and the checkbox for prefix is selected and disabled. My global Preference/Packaging/Prefix... is set to "No". Once compiled and installed the component is named gambas3-trixon-testa. Is "gambas3-trixon-gambas3-trixon-testa" a bug in gambas or in my brain? One more thing regarding installing, http://gambasdoc.org/help/dev/gambas?v3#t19 "You can install the component in your home directory by checking the corresponding option in the "Make executable" dialog." I cant get this 'per user' install to work, it does not show up in gambas among the other components. I just put my gambas3-trixon-testa.gambas in ~/. What is the "corresponding option"? The same component works just fine when installed globally. /patrik ps I just returned to Gambas after almost four years I think and it's really nice to see that it still is alive and kicking, and have matured, a lot. Some facts... [System] OperatingSystem=Linux Kernel=3.2.0-26-generic Architecture=x86_64 Memory=8102592 kB DistributionVendor=Ubuntu DistributionRelease="Ubuntu 12.04 LTS" Desktop=Gnome [Gambas 2] Version=2.23.1 Path=/usr/bin/gbx2 [Gambas 3] Version=3.2.1 Path=/usr/bin/gbx3 [Libraries] Qt4=libQtCore.so.4.8.1 GTK+=libgtk-x11-2.0.so.0.2400.10 -------------- next part -------------- An HTML attachment was scrubbed... URL: From gambas at ...1... Fri Jul 27 16:42:24 2012 From: gambas at ...1... (=?ISO-8859-1?Q?Beno=EEt_Minisini?=) Date: Fri, 27 Jul 2012 16:42:24 +0200 Subject: [Gambas-devel] Components name and installation In-Reply-To: References: Message-ID: <5012A8D0.8020003@...1...> Le 23/07/2012 11:46, Patrik Karlsson a ?crit : > Hi, > I finally succeeded in writing my first 'dummy' component, it does not > do much, but it looks to be a valid component. :) > > In the Packaging section of > http://gambasdoc.org/help/dev/gambas?v3#t20 I am told to use the name > convention gambas3-vendor-name, so I'm using gambas3-trixon-testa. > > On the same page it says: "Even if the packager wizard allows to insert > the package vendor string into the package name, please avoid it." > > When I create an installation package, the name > is gambas3-trixon-gambas3-trixon-testa-3.2.1 and the checkbox for prefix > is selected and disabled. My global Preference/Packaging/Prefix... is > set to "No". > > Once compiled and installed the component is named gambas3-trixon-testa. > Is "gambas3-trixon-gambas3-trixon-testa" a bug in gambas or in my brain? > > One more thing regarding installing, > http://gambasdoc.org/help/dev/gambas?v3#t19 > > "You can install the component > in your home directory by > checking the corresponding option in the "Make executable" dialog." > > I cant get this 'per user' install to work, it does not show up in > gambas among the other components. I just put > my gambas3-trixon-testa.gambas in ~/. > What is the "corresponding option"? > The same component works just fine when installed globally. > > /patrik > > ps I just returned to Gambas after almost four years I think and it's > really nice to see that it still is alive and kicking, and have matured, > a lot. > > Some facts... > > [System] > > OperatingSystem=Linux > > Kernel=3.2.0-26-generic > > Architecture=x86_64 > > Memory=8102592 kB > > DistributionVendor=Ubuntu > > DistributionRelease="Ubuntu 12.04 LTS" > > Desktop=Gnome > > > [Gambas 2] > > Version=2.23.1 > > Path=/usr/bin/gbx2 > > > [Gambas 3] > > Version=3.2.1 > > Path=/usr/bin/gbx3 > > > [Libraries] > > Qt4=libQtCore.so.4.8.1 > > GTK+=libgtk-x11-2.0.so.0.2400.10 > > Can you send me your project? -- Beno?t Minisini From pata.karlsson at ...176... Sun Jul 29 20:29:42 2012 From: pata.karlsson at ...176... (Patrik Karlsson) Date: Sun, 29 Jul 2012 20:29:42 +0200 Subject: [Gambas-devel] Components name and installation In-Reply-To: <5012A8D0.8020003@...1...> References: <5012A8D0.8020003@...1...> Message-ID: 2012/7/27 Beno?t Minisini > Le 23/07/2012 11:46, Patrik Karlsson a ?crit : > > Hi, > > I finally succeeded in writing my first 'dummy' component, it does not > > do much, but it looks to be a valid component. :) > > > > In the Packaging section of > > http://gambasdoc.org/help/dev/gambas?v3#t20 I am told to use the name > > convention gambas3-vendor-name, so I'm using gambas3-trixon-testa. > > > > On the same page it says: "Even if the packager wizard allows to insert > > the package vendor string into the package name, please avoid it." > > > > When I create an installation package, the name > > is gambas3-trixon-gambas3-trixon-testa-3.2.1 and the checkbox for prefix > > is selected and disabled. My global Preference/Packaging/Prefix... is > > set to "No". > > > > Once compiled and installed the component is named gambas3-trixon-testa. > > Is "gambas3-trixon-gambas3-trixon-testa" a bug in gambas or in my brain? > > > > One more thing regarding installing, > > http://gambasdoc.org/help/dev/gambas?v3#t19 > > > > "You can install the component > > in your home directory by > > checking the corresponding option in the "Make executable" dialog." > > > > I cant get this 'per user' install to work, it does not show up in > > gambas among the other components. I just put > > my gambas3-trixon-testa.gambas in ~/. > > What is the "corresponding option"? > > The same component works just fine when installed globally. > > > > /patrik > > > > ps I just returned to Gambas after almost four years I think and it's > > really nice to see that it still is alive and kicking, and have matured, > > a lot. > > > > Some facts... > > > > [System] > > > > OperatingSystem=Linux > > > > Kernel=3.2.0-26-generic > > > > Architecture=x86_64 > > > > Memory=8102592 kB > > > > DistributionVendor=Ubuntu > > > > DistributionRelease="Ubuntu 12.04 LTS" > > > > Desktop=Gnome > > > > > > [Gambas 2] > > > > Version=2.23.1 > > > > Path=/usr/bin/gbx2 > > > > > > [Gambas 3] > > > > Version=3.2.1 > > > > Path=/usr/bin/gbx3 > > > > > > [Libraries] > > > > Qt4=libQtCore.so.4.8.1 > > > > GTK+=libgtk-x11-2.0.so.0.2400.10 > > > > > > Can you send me your project? > > -- > Beno?t Minisini > > Ok, I am attaching the source for a project generated by the component template and a screen shot of the installation package dialog. One more thing that might or might not be related. The menu option library/component properties is disabled but functional via short cuts for library and components, ctrl shift p. /patrik -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: gambas321_install_package_dialog.png Type: image/png Size: 60997 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: gambas3-trixon-testc-3.2.1.tar.gz Type: application/x-gzip Size: 7302 bytes Desc: not available URL: