Re: Cross platform code
Gilboa Davara <[EMAIL PROTECTED]> writes: > CPP/GCC/CC all define _WIN32 by default under all the environment I ever > worked on. (which is pretty extensive.) Really? Right now I am in an environment where several different version of cpp and gcc don't define that. It seems to me that the range of environments where gcc defines _WIN32 is pretty narrow. To complement that, I have no idea what you mean by CPP or CC. > By design, compilers running under Windows (be that Watcom, Borland or > MCVC) must define _WINTVER, _WIN32 and M_IX86/M_ALPA/etc in-order to > include the Winxxx.h files. (which are required to access any OS > functions) That's fine, this is exactly one platform that we have covered in the passage above. > So yes, these are compiler defined macros that can be relayed upon. Yes, indeed. However, I know of no rule that says that, e.g., unix must be defined for any of the unices or for Linux, as the snippet of your earlier mail that follows below seems to imply. On my RH7.3 gcc does define unix, __unix, __unix__, linux, __linux, and __linux__. However, it is enough to use the -ansi option to make unix and linux disappear. There you go. None of the standard glibc or glibc-kernel headers (meaning linux-specific headers exposed to applications programmers) uses unix, __unix, or __unix__ (some applications do, but it's their business). This tells me that gcc only uses those internally, and not to access any library or system facilities. This means that there may be compilers that will not define these macros and still work just fine. Gilboa Davara <[EMAIL PROTECTED]> writes: > Better yet: > > #ifdef unix > > // UNIX/Linux/*BSD. > > #endif > > #ifdef _WIN32 > > // Window stuff > > #endif > > Both symbols are auto defined by the compiler. -- Oleg Goldshmidt | [EMAIL PROTECTED] = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
Just two more inputs to the bunch of very good answers already provided. 1. If you are using C++, there is a library called ACE that provides a platform independant wrapper for TCP communications, as well as some shortcuts. You might want to check it out. I have never worked with it, but I heard it may be a bitch to compile. 2. In any case, if you are going to be creating more than the two platforms, you may want to consider including an autoconf script in your project. This makes it ever so much easier to tell the various minor differences apart. Shachar Voguemaster wrote: Hi all, For the life of me, I can't seem to find anything that will help me write code that can compile under Linux AND Win32. Basically I want to write a small network application (details later) that will be cross platform. The problem is very basic: Linux and Win32 have different include files for some things and placing #include directives inside #ifdef doesn't do the trick (it nullifies the #ifdef possibly ?). I'm not sure how to go about this and I couldn't find anything useful elsewhere! Most projects just have a *nix source tree and a Win32 source tree (that I could see). I don't want to use autotools as this project is small, quiter small. Any tips ?? Thanks, Eli -- Shachar Shemesh Open Source integration consultant Home page & resume - http://www.shemesh.biz/ = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Israeli DMCA looming?
Stanislav Malyshev wrote on 2003-07-03: > BC>> Well, I see no reason why the person reviewing the software > BC>> patents shouldn't be an expert in programming (but not knowing > BC>> every existing program, of course). He should be required to be > BC>> an expert, else how can we entrust him to grant patents in this > BC>> field? > > OK. Then let's see - first of all, there is no such thing as "expert in > programming" - no more than "expert in medicine" - there are > specialisations. So either patent office should have a staff of an > university and more, or it should retain the services of external experts. > Which means patent application is going to cost _a real lot_ of money and > take a real lot of time to complete. Which is not exactly what the patents > are meant for. BTW, even with this scheme nobody really guarantees that > expert will do thorough work (after all, what means of control over him > does the office have?) or that he will not just be bribed by the patent > submitter. > Fair points. They apply to all kinds of patents, BTW. Anyway currently the US patent offices pass through some patents (especially but not only in software) that are ridiculous even to people with basic programming background. See http://lpf.ai.mit.edu/Patents/anatomy-trivial-patent.txt for an example. -- Beni Cherniavsky <[EMAIL PROTECTED]> "Reading the documentation I felt like a kid in a toy shop." -- Phil Thompson on Python's standard library = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Making linux behave as a real multiuser multitask os.
On Tuesday 01 July 2003 22:52, Meir Michanie wrote: > in the file /etc/security/limits.conf you can setup: > "cpu - max CPU time (MIN)" or "nproc - max number of processes" I do not > know if when a proccess reach this limit is killed or suspended. anyway cpu When the user maxes out his "max number of processes", subsequent fork()s would fail. As to "max CPU time", here's a simple experiment: $ ulimit -t 2 $ cat /dev/zero > /dev/null Killed > How can I limit the percentage of CPU for a regular user? A better approach would be: how do I make the system divide CPU time equally between users rather than between processes? > If the proccess that is limited jump to the execution to system level (like > writing to hdd) will still hang the PC. Shouldn't happen. Make sure your hard drive and CDROM use the optimal settings; especially 32-bit transfers and DMA. Use the 'hdparm' utility to tune your IDE device settings. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
I should add that you can always use your own macro (defined inside your makes files) in-case you don't want to count on the "compiler" doing the work for you. E.g. ifdef _WINNT_ #include #endif #ifdef _LINUX_ #include #include #include #include #endif #ifdef _FREEBSD_ // Same as Linux... .. #endif #ifdef _HPUX_ ..etc On Thu, 2003-07-03 at 21:08, Oleg Goldshmidt wrote: > Gilboa Davara <[EMAIL PROTECTED]> writes: > > > Better yet: > > > > #ifdef unix > > > > // UNIX/Linux/*BSD. > > > > #endif > > > > #ifdef _WIN32 > > > > // Window stuff > > > > #endif > > > > Both symbols are auto defined by the compiler. > > What is "the compiler"? How can you be sure that 6 months from now > someone in New Zealand will be using "the compiler"? > > There is nothing I know of that will guarantee existence of "unix" or > "_WIN32". Even worse, what about the cases where Solaris, AIX, HPUX, > Linux, DG/UX, and assorted BSDs all work differently, what good is > "unix" then? If a compiler defines something like that, it is most > likely for its own internal use, and to comply with the system > facilities, *not* to be exported into client code. > > To the OP: > > Create your own compilation environment, call your CPP macros whatever > you want, in your Makefile or configuration files define the proper > flags depending on the output of things like uname -s etc. In short, > do whatever is needed but do not rely on things that *a* compiler > defines automatically. > > Finally, if you suspect that something or somebody is #undefing or > #defining your macros without you knowing it, try to make the compiler > #as verbose as possible - maybe there will be a clue there. Often you > #will see what it silently #defines or #undefs on the command line then. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
CPP/GCC/CC all define _WIN32 by default under all the environment I ever worked on. (which is pretty extensive.) By design, compilers running under Windows (be that Watcom, Borland or MCVC) must define _WINTVER, _WIN32 and M_IX86/M_ALPA/etc in-order to include the Winxxx.h files. (which are required to access any OS functions) So yes, these are compiler defined macros that can be relayed upon. Gilboa On Thu, 2003-07-03 at 21:08, Oleg Goldshmidt wrote: > Gilboa Davara <[EMAIL PROTECTED]> writes: > > > Better yet: > > > > #ifdef unix > > > > // UNIX/Linux/*BSD. > > > > #endif > > > > #ifdef _WIN32 > > > > // Window stuff > > > > #endif > > > > Both symbols are auto defined by the compiler. > > What is "the compiler"? How can you be sure that 6 months from now > someone in New Zealand will be using "the compiler"? > > There is nothing I know of that will guarantee existence of "unix" or > "_WIN32". Even worse, what about the cases where Solaris, AIX, HPUX, > Linux, DG/UX, and assorted BSDs all work differently, what good is > "unix" then? If a compiler defines something like that, it is most > likely for its own internal use, and to comply with the system > facilities, *not* to be exported into client code. > > To the OP: > > Create your own compilation environment, call your CPP macros whatever > you want, in your Makefile or configuration files define the proper > flags depending on the output of things like uname -s etc. In short, > do whatever is needed but do not rely on things that *a* compiler > defines automatically. > > Finally, if you suspect that something or somebody is #undefing or > #defining your macros without you knowing it, try to make the compiler > #as verbose as possible - maybe there will be a clue there. Often you > #will see what it silently #defines or #undefs on the command line then. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Israeli DMCA looming?
BC>> Well, I see no reason why the person reviewing the software BC>> patents shouldn't be an expert in programming (but not knowing BC>> every existing program, of course). He should be required to be BC>> an expert, else how can we entrust him to grant patents in this BC>> field? OK. Then let's see - first of all, there is no such thing as "expert in programming" - no more than "expert in medicine" - there are specialisations. So either patent office should have a staff of an university and more, or it should retain the services of external experts. Which means patent application is going to cost _a real lot_ of money and take a real lot of time to complete. Which is not exactly what the patents are meant for. BTW, even with this scheme nobody really guarantees that expert will do thorough work (after all, what means of control over him does the office have?) or that he will not just be bribed by the patent submitter. -- [EMAIL PROTECTED] \/ There shall be counsels taken Stanislav Malyshev /\ Stronger than Morgul-spells phone +972-50-624945/\ JRRT LotR. whois:!SM8333 = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Israeli DMCA looming?
Stanislav Malyshev wrote on 2003-07-03: > EM>> Maybe conflicts between patents and technology advances were caused > EM>> because of non-professional staff of the patent office? > > Asking that, let's ask what kind of "professionality" is required from > them? Should they be an experts in software knowing to the least detail > every part of every program existing, for example? What would be > requirements to professional patent office worker? Note that patent office > is not exclusively software - they should also work on other things too. > Well, I see no reason why the person reviewing the software patents shouldn't be an expert in programming (but not knowing every existing program, of course). He should be required to be an expert, else how can we entrust him to grant patents in this field? > EM>> Maybe a better computerization will help the patent office > EM>> reject patents that were "prior art"ed by open technologies? > > It is pretty hard to know this - for this, detailed analysis of the > functions of patented and potentially prior-art (meaning, just > almost every other in the field) applications required. I fear for a > patent office which is not really overbudgeted it is not actually > feasible. > Better computerization will most probably not solve the issue. It would help the staff, sure. But the staff should be competent and should have the correct guiding rules. And these should either exclude SW patents completely or be very demanding about them. There is no reason to expect better equipped staff whose task is to allow trivial patents to filter - but there is no reason to under-equip them instead of redefining their task. -- Beni Cherniavsky <[EMAIL PROTECTED]> "Reading the documentation I felt like a kid in a toy shop." -- Phil Thompson on Python's standard library = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: [OT] Single Board Computers
Quoth Baruch Shpirer: > If anyone found in the end some company who deals with this I'd be happy > to hear about it > Iam looking for p3-450-650mhz probes distributed around the world in > atleast 3 different regions for some application Baruch - there IS a company building such buggers, in Israel. -- ---OFCNL This is MY list. This list belongs to ME! I will flame anyone I want. Official Flamer/Cabal NON-Leader [EMAIL PROTECTED] = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
Gilboa Davara <[EMAIL PROTECTED]> writes: > Better yet: > > #ifdef unix > > // UNIX/Linux/*BSD. > > #endif > > #ifdef _WIN32 > > // Window stuff > > #endif > > Both symbols are auto defined by the compiler. What is "the compiler"? How can you be sure that 6 months from now someone in New Zealand will be using "the compiler"? There is nothing I know of that will guarantee existence of "unix" or "_WIN32". Even worse, what about the cases where Solaris, AIX, HPUX, Linux, DG/UX, and assorted BSDs all work differently, what good is "unix" then? If a compiler defines something like that, it is most likely for its own internal use, and to comply with the system facilities, *not* to be exported into client code. To the OP: Create your own compilation environment, call your CPP macros whatever you want, in your Makefile or configuration files define the proper flags depending on the output of things like uname -s etc. In short, do whatever is needed but do not rely on things that *a* compiler defines automatically. Finally, if you suspect that something or somebody is #undefing or #defining your macros without you knowing it, try to make the compiler #as verbose as possible - maybe there will be a clue there. Often you #will see what it silently #defines or #undefs on the command line then. -- Oleg Goldshmidt | [EMAIL PROTECTED] who is currently trying to make some code he did not write work on Linux without screwing up 5 or 6 other operating systems... = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Israeli DMCA looming?
EM>> Regarding the first one, I divide between patents that STOP EM>> technological advances (like LZW) and patents that HELP EM>> technological advances (like RSA). I don't agree that EVERYTHING Hm. Only difference I see is that the owner of the RSA was less greedy than owner of the LZW, but that difference does not lie in the field of patents but of owners. EM>> Maybe conflicts between patents and technology advances were caused EM>> because of non-professional staff of the patent office? Asking that, let's ask what kind of "professionality" is required from them? Should they be an experts in software knowing to the least detail every part of every program existing, for example? What would be requirements to professional patent office worker? Note that patent office is not exclusively software - they should also work on other things too. EM>> Maybe a better computerization will help the patent office EM>> reject patents that were "prior art"ed by open technologies? It is pretty hard to know this - for this, detailed analysis of the functions of patented and potentially prior-art (meaning, just almost every other in the field) applications required. I fear for a patent office which is not really overbudgeted it is not actually feasible. -- [EMAIL PROTECTED] \/ There shall be counsels taken Stanislav Malyshev /\ Stronger than Morgul-spells phone +972-50-624945/\ JRRT LotR. whois:!SM8333 = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Israeli DMCA looming?
Gilad Ben-Yossef wrote: > All of this assumes that a better efficient patent office is a good > thing. I was trying to suggest tht IMHO, doe to how we see patent today > - as "property", as opposed to the original intent of the patent system, > making it more efficient is actually a bad thing. > > I guess my point was missed. It was, but this is not the issue. There are two claims that are required to be proved: 1. That everything in patents is bad. 2. That efficiency always causes a patent office to be "worse" even further. I don't agree with both claims. Regarding the first one, I divide between patents that STOP technological advances (like LZW) and patents that HELP technological advances (like RSA). I don't agree that EVERYTHING with patenting is bad. But even if the first claim was right, then you still have to prove the second claim. It's like the argument regarding the Palestinians, whether money helps them to bring more order and stop the terror, or to develop more advanced weapon. Maybe conflicts between patents and technology advances were caused because of non-professional staff of the patent office? Maybe better staff will prevent such conflicts? Maybe a better computerization will help the patent office reject patents that were "prior art"ed by open technologies? Many budgets are given in return to drops of activities or payments. For example, the special American support for the Israeli Government was given on condition that many expenses will be dropped, for example the progressive support for children (Kizbaot-yeladim), so this money not only that it didn't help the kizbaot-yeladim, it even hurt them; Maybe in this case too, this budget will cause the patent office to be "better" for us? We can't oppose something automaticall, just because of the name behind it. If you want to oppose this budget, you must check exactly where the money goes to. -- Eli Marmor [EMAIL PROTECTED] CTO, Founder Netmask (El-Mar) Internet Technologies Ltd. __ Tel.: +972-9-766-1020 8 Yad-Harutzim St. Fax.: +972-9-766-1314 P.O.B. 7004 Mobile: +972-50-23-7338 Kfar-Saba 44641, Israel = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
Quoting Omer Zak <[EMAIL PROTECTED]>: > > On Thu, 3 Jul 2003, Voguemaster wrote: > > > The problem is very basic: Linux and Win32 have different include files > > for some things and placing #include directives inside #ifdef doesn't > > do the trick (it nullifies the #ifdef possibly ?). > as a good solution i can advice u to look at apache new socket library. it is cross platfor, and it is a core engine of the apache 4 = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
On Thu, 03 Jul 2003 18:48:25 +0300, Nadav Har'El <[EMAIL PROTECTED]> wrote: First, thanks for all the replies, the ifdef thing works now, let me explain what happened. Consider the following code: #ifdef WIN32 #include #else // any other include for linux #endif #ifdef WIN32 int APIENTRY WinMain( ) { #else int main( int argc, char *argv[]) { #endif return 0; } Trying to compile this file in VC++ 6 gives some error about #else not expected. The weird thing is that moving the first portion (the one with the include directives) to a header file works like a charm. Can anyone explain this ??? At least the problem is no longer a problem :) Eli PS. A google search showed that the header file features.h might contain some definitions that can be used to determine the platform, at least on Linux. The way the ifdefs are set up at the moment take into consideration only Win32 and EVERYTHING else. I'd like to somehow detect Linux and glibc (yes i know about __GLIBC__). Well, thanks for the help On Thu, Jul 03, 2003, Voguemaster wrote about "Cross platform code": The problem is very basic: Linux and Win32 have different include files for some things and placing #include directives inside #ifdef doesn't do the trick (it nullifies the #ifdef possibly ?). You probably made some mistake - #include doesn't nullify #ifdef or anything of that sort :) (you might want to refer to any C book, or the "cpp" info-page, for more information) You can have something like #ifdef LINUX_SYSTEM #include #else #include #endif And when you compile on the Linux system, add a "-DLINUX_SYSTEM" in the command line. Alternatively, you can use predefined macros that are automatically defined on one system and not on the other. For example, last time I checked, the C preprocessor defines "linux" on linux systems. So you can replace the above example with #ifdef linux #include #else #include #endif The macro __linux__ is also defined in Linux, I believe. A similar macro (whose name I don't remember) is defined by default on Microsoft's C compiler on Windows. -- Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/ = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
Better yet: #ifdef unix // UNIX/Linux/*BSD. #endif #ifdef _WIN32 // Window stuff #endif Both symbols are auto defined by the compiler. Gilboa On Thu, 2003-07-03 at 18:48, Nadav Har'El wrote: > On Thu, Jul 03, 2003, Voguemaster wrote about "Cross platform code": > > The problem is very basic: Linux and Win32 have different include files > > for some things and placing #include directives inside #ifdef doesn't > > do the trick (it nullifies the #ifdef possibly ?). > > You probably made some mistake - #include doesn't nullify #ifdef or > anything of that sort :) (you might want to refer to any C book, or the > "cpp" info-page, for more information) > > You can have something like > > #ifdef LINUX_SYSTEM > #include > #else > #include > #endif > > And when you compile on the Linux system, add a "-DLINUX_SYSTEM" in the > command line. Alternatively, you can use predefined macros that are > automatically defined on one system and not on the other. For example, > last time I checked, the C preprocessor defines "linux" on linux systems. > So you can replace the above example with > > #ifdef linux > #include > #else > #include > #endif > > The macro __linux__ is also defined in Linux, I believe. A similar macro > (whose name I don't remember) is defined by default on Microsoft's C compiler > on Windows. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
On 2003-07-03 Voguemaster wrote: > The problem is very basic: Linux and Win32 have different include files > for some things and placing #include directives inside #ifdef doesn't > do the trick (it nullifies the #ifdef possibly ?). What exactly is not working? For me it always worked just fine like this: #ifdef WIN32 #include #else #include #endif = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
On Thu, Jul 03, 2003, Voguemaster wrote about "Cross platform code": > The problem is very basic: Linux and Win32 have different include files > for some things and placing #include directives inside #ifdef doesn't > do the trick (it nullifies the #ifdef possibly ?). You probably made some mistake - #include doesn't nullify #ifdef or anything of that sort :) (you might want to refer to any C book, or the "cpp" info-page, for more information) You can have something like #ifdef LINUX_SYSTEM #include #else #include #endif And when you compile on the Linux system, add a "-DLINUX_SYSTEM" in the command line. Alternatively, you can use predefined macros that are automatically defined on one system and not on the other. For example, last time I checked, the C preprocessor defines "linux" on linux systems. So you can replace the above example with #ifdef linux #include #else #include #endif The macro __linux__ is also defined in Linux, I believe. A similar macro (whose name I don't remember) is defined by default on Microsoft's C compiler on Windows. -- Nadav Har'El| Thursday, Jul 3 2003, 4 Tammuz 5763 [EMAIL PROTECTED] |- Phone: +972-53-245868, ICQ 13349191 |Guarantee: this email is 100% free of http://nadav.harel.org.il |magnetic monopoles, or your money back! = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: XDMCP problem
On Thu, Jul 03, 2003 at 05:17:17PM +0300, Sobolevsky Vladek wrote: > >Hi All, > >I have M$ Terminal Server , and I want users that connects to it >will be able connect to Linux machine through XDMCP session using >ReflectionX, >I have set it up , but I have a problem: only one user can connect >to Linux from Terminal Server >simultaneously . I am using gdm on RH 7.2 . I suspect that the problem >is , in opening >multiple sessions from the same ip to gdm . Here xdmcp section from my >gdm.conf : > >[xdmcp] ># Distributions: Ship with this off. It is never a safe thing to >leave ># out on the net. Alternatively you can set up /etc/hosts.allow and ># /etc/hosts.deny to only allow say local access. >Enable=true >HonorIndirect=true >MaxPending=4 >MaxPendingIndirect=4 >MaxSessions=16 >MaxWait=15 >MaxWaitIndirect=15 >DisplaysPerHost=10 >Port=177 ># Willing script, none is shipped and by default we'll send ># hostname system id >Willing=/etc/gdm/Xwilling > >Maybe someone can suggest different setup . I guess the problem is on the ReflectionX side, not gdm. Each of your users needs to have a different X display on the machine. I never used ReflectionX, so I don't know if it allows that. I did work with others, both ones that allow that (XFree86 under cygwin, Xoftware), and ones that didn't (eXcursion). If you want to check that the problem is not on the gdm side, try to run multiple X's from Linux (there you set the display number with :n in the command line, e.g. X :1 -query mygdmserver & X :2 -query mygdmserver & X :3 -query mygdmserver & In fact, if I am correct, you won't even be able to run 2 ReflectionX's on the same machine, even without getting to the gdm, because they will try to listen on the same port (6000), and only the first one will succeed. Can you describe exactly what you try to do and what happens? -- Didi > >Thanks > >Vladek = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
Voguemaster <[EMAIL PROTECTED]>: > For the life of me, I can't seem to find anything that will help me > write code that can compile under Linux AND Win32. Basically I want > to write a small network application (details later) that will be > cross platform. Python is your friend. http://www.python.org If you wish to use C then I guess cygwin (as previously offered). = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Israeli DMCA looming?
Eli Marmor wrote: Gilad Ben-Yossef wrote: Eli Marmor wrote: This doesn't have anything to do with HaMakor. Currently, the Israeli patents office is very primitive, and not professional. There is almost no computerization. The procedures take MANY years to complete. Usually, instead of scanning or reviewing patents, the staff prefer to adopt decisions of other offices regarding the specific patent. This is "bizayon", and makes a bad name for the state of Israel. Fortunately, it is going to change. Finally. It's a good improvement for everybody. I'm sorry Eli, that I don't share your opptimism. The US patent office, to takr a not so rendom example, has been very very active. I don't think I like the results much and being the pessimist that I am I'm having a hard time believing that the Israeli PO will do any better. Gilad, Either my English is bad (it is), or you didn't read my message. The original poster wondered what was the meaning of a quotation that he brought from the newspaper. No matter what are the plans of the Israel patents office, the specific quotation referred to a critical budget for computerization of the patents office, and for hiring more professional staff. These steps have been planned for about 10 years, and were postponed again and again, because of the financial limitations of Israel. All of this assumes that a better efficient patent office is a good thing. I was trying to suggest tht IMHO, doe to how we see patent today - as "property", as opposed to the original intent of the patent system, making it more efficient is actually a bad thing. I guess my point was missed. Gilad. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Multiple GLIBCs
Isaac Aaron wrote: Hi Gilad I tried the dual installation of glibc-2.2 and glibc-2.3. rpm -ivh fails the upgrade because the depend on their glibc-common packages (which don't want each other whatsoever). When I installed glibc with --nodeps I got a fine list of file conflicts. It seems like it's not intended to work. Are you talking about just the runtime library or the development headers and all? It seems like my only resort is to create a "compat-" RPM which only installs the stuff under /lib. You can try re-locating thr RPM to a different root. Gilad. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
RE: Multiple GLIBCs
Hi Gilad I tried the dual installation of glibc-2.2 and glibc-2.3. rpm -ivh fails the upgrade because the depend on their glibc-common packages (which don't want each other whatsoever). When I installed glibc with --nodeps I got a fine list of file conflicts. It seems like it's not intended to work. It seems like my only resort is to create a "compat-" RPM which only installs the stuff under /lib. Thanks, Isaac Aaron > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] Behalf Of Gilad Ben-Yossef > Sent: Thursday, July 03, 2003 11:28 AM > To: Isaac Aaron > Cc: [EMAIL PROTECTED] > Subject: Re: Multiple GLIBCs > > > Isaac Aaron wrote: > > > >>-Original Message- > >>From: Gilad Ben-Yossef [mailto:[EMAIL PROTECTED] > >>Sent: Thursday, July 03, 2003 10:56 AM > >>To: Isaac Aaron > >>Cc: [EMAIL PROTECTED] > >>Subject: Re: Multiple GLIBCs > >> > >> > >>Isaac Aaron wrote: > >> > >>>Hi List > >>> > >>>Running Red Hat 7.3 requires me to recompile everything I want > >> > >>from Red Hat 9 (or that is avaliable to Red Hat) before I install it. > >> > >>>I'm looking for a way to co-install GLIBC 2.3 with the existing > >> > >>GLIBC 2.1. > >> > >>>Any opinions? > >>> > >> > >>No problem at all. Linux (like other Un*x like systems) has a rather > >>elegant way to deal with libraries. When you install the RPM (I'm > >>assuming you using RPM) make sure to "install" with -i rather then > >>upgrade with -U. > >> > > > > > > Won't I need a --force there? > > Not because of the exitence of an older library. > > > How about if I just "steal" the libc* files (not links) and > drop them on the /usr/lib directory. > > Would that satisfy the RPM "Requires" directive? > > The application will be able to use them (once you run ldconfig at > least), but RPM will have no knowledge on it so it wont let you install > (unless you use --nodeps of course). > > > >>When the RPM will run ldconfig it will set things up so that > any program > >>that explicitly asks for the older version will get it, a program that > >>does not specify a specific version will get the latest, and if a > >>program seems to mis behave with the newer lib you can always use > >>LD_PRELOAD (man ld.so) to override. > > > > > > That seems to do the trick as long as the small stuff > "util-linux" "fileutils" are all linked in a way > > that won't mess up the whole installation. > > I don't think they will but don't trust me. Copy the binaries to some > bizzare location and use LD_PRELOAD to force load the library with the > stuff you need to test it BEFORE "installation". > > > Red hat used to have compat-glibc-* packages which contained > older glibc libraries. > AFAIR that was for REALLY old libc that had major binary interface > changes from the newer ones. > > > For some reason they don't do that anymore (to promote the > advanced server?). For 99% of the cases it's not needed. > > > > What would really be nice is if I had a way to prepare such > RPMS by myself (install cleanly, work with any version). > Of course you can! That's the whole point of using Free Software. Just > download the SRPMS use to create the RPMs you need and generate the RPMs > localy. man rpm for the gory details. > > Gilad > > > > = > To unsubscribe, send mail to [EMAIL PROTECTED] with > the word "unsubscribe" in the message body, e.g., run the command > echo unsubscribe | mail [EMAIL PROTECTED] To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Cross platform code
On Thu, 3 Jul 2003, Voguemaster wrote: > The problem is very basic: Linux and Win32 have different include files > for some things and placing #include directives inside #ifdef doesn't > do the trick (it nullifies the #ifdef possibly ?). Very strange. > I'm not sure how to go about this and I couldn't find anything useful > elsewhere! Most projects just have a *nix source tree and a Win32 > source tree (that I could see). If the #ifdef thing absolutely cannot be made to work, and cygwin (suggested by Beni Cherniavsky) is not your cup of tea, consider development of an abstraction layer - call the problematic functions using your names. Those names will resolve into either Unix or Win32 functions by means of macro definitions and/or inline procedures. --- Omer My opinions, as expressed in this E-mail message, are mine alone. They do not represent the official policy of any organization with which I may be affiliated in any way. WARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
The Internet Tool needs your help!
In the past few months the Kinneret development team worked on the Israeli internet connection tool that should gather all ISPs, all connection methods and all modems into one application, that will create everything necessary for newbies to connect to the internet. Today the tool advances to it's second phase. After the core tool has been written, we are calling for the community to submit new entries to the database in order to support more ISPs and more modems. The current status of the database, a guide on how to submit new entries (Hebrew PDF + example entries) and more details about the tool can be found on the tool's development page at http://z9u2k.homelinux.com/iwiz/ NOTE: The site is hosted on my local machine which as a SLOW upload limit of 12KBps... be patient... :) Regards, Itay 'z9u2K' Duvdevani, GNU/Linux Kinneret. Public GPG Key: ftp://ftp.berlios.de/pub/kinneret/z9u2k.asc = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
XDMCP problem
Title: XDMCP problem Hi All, I have M$ Terminal Server , and I want users that connects to it will be able connect to Linux machine through XDMCP session using ReflectionX, I have set it up , but I have a problem: only one user can connect to Linux from Terminal Server simultaneously . I am using gdm on RH 7.2 . I suspect that the problem is , in opening multiple sessions from the same ip to gdm . Here xdmcp section from my gdm.conf : [xdmcp] # Distributions: Ship with this off. It is never a safe thing to leave # out on the net. Alternatively you can set up /etc/hosts.allow and # /etc/hosts.deny to only allow say local access. Enable=true HonorIndirect=true MaxPending=4 MaxPendingIndirect=4 MaxSessions=16 MaxWait=15 MaxWaitIndirect=15 DisplaysPerHost=10 Port=177 # Willing script, none is shipped and by default we'll send # hostname system id Willing=/etc/gdm/Xwilling Maybe someone can suggest different setup . Thanks Vladek
Re: Cross platform code
Voguemaster wrote on 2003-07-03: > Hi all, > > For the life of me, I can't seem to find anything that will help me > write code that can compile under Linux AND Win32. Basically I want > to write a small network application (details later) that will be > cross platform. > The problem is very basic: Linux and Win32 have different include files > for some things and placing #include directives inside #ifdef doesn't > do the trick (it nullifies the #ifdef possibly ?). > > I'm not sure how to go about this and I couldn't find anything useful > elsewhere! Most projects just have a *nix source tree and a Win32 > source tree (that I could see). > > I don't want to use autotools as this project is small, quiter small. > Any tips ?? > If all your app needs are network functions and you have a command-line interface, consider cygwin. You code for unix and cygwin will compile your program with zero changes for windows. The users won't need a full-blown cygwin installations; for most simple cases just distributing a single cygwin DLL will your exe will be enough. -- Beni Cherniavsky <[EMAIL PROTECTED]> "Reading the documentation I felt like a kid in a toy shop." -- Phil Thompson on Python's standard library = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Anyone has filter files for Privoxy to block adult content?
On Thursday 03 July 2003 00:22, Amit Margalit wrote: AM>Hello all, AM> AM>I am looking for Privoxy config files that will block adult sites, etc. AM>I also need them to allow users to use webmail sites, etc, and I don't AM>care much about the privacy features of Privoxy. I saw someone compiling one for Proxomitron, which is a freeware(almost) web filter for Win..*cough*. These files can be converted quite easily to fit into Privoxy. I remember I followed the link from the main Proxomitron site to get there. Unfortunately the website is down, and I can't pull it out of google cache either :(( But how effective do you want it to be? There are thousands if not millions of adult websites, some of them have innocent names that probably will not be blocked by pattern-matching engine. When adding them into our filters, i found that it grows very fast causing slower responce times, thus becoming less effective. I really think that one should really concentrate on blocking the nasty code that leads to adult sites (popups etc..). -- Sincerely Yours, Vasiliev Michael = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Cross platform code
Hi all, For the life of me, I can't seem to find anything that will help me write code that can compile under Linux AND Win32. Basically I want to write a small network application (details later) that will be cross platform. The problem is very basic: Linux and Win32 have different include files for some things and placing #include directives inside #ifdef doesn't do the trick (it nullifies the #ifdef possibly ?). I'm not sure how to go about this and I couldn't find anything useful elsewhere! Most projects just have a *nix source tree and a Win32 source tree (that I could see). I don't want to use autotools as this project is small, quiter small. Any tips ?? Thanks, Eli -- Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/ = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Israeli DMCA looming?
Gilad Ben-Yossef wrote: > > Eli Marmor wrote: > > > > > This doesn't have anything to do with HaMakor. > > > > Currently, the Israeli patents office is very primitive, and not > > professional. There is almost no computerization. The procedures take > > MANY years to complete. Usually, instead of scanning or reviewing > > patents, the staff prefer to adopt decisions of other offices regarding > > the specific patent. > > > > This is "bizayon", and makes a bad name for the state of Israel. > > > > Fortunately, it is going to change. Finally. > > It's a good improvement for everybody. > > I'm sorry Eli, that I don't share your opptimism. The US patent office, > to takr a not so rendom example, has been very very active. > > I don't think I like the results much and being the pessimist that I am > I'm having a hard time believing that the Israeli PO will do any better. Gilad, Either my English is bad (it is), or you didn't read my message. The original poster wondered what was the meaning of a quotation that he brought from the newspaper. No matter what are the plans of the Israel patents office, the specific quotation referred to a critical budget for computerization of the patents office, and for hiring more professional staff. These steps have been planned for about 10 years, and were postponed again and again, because of the financial limitations of Israel. Have I ever written anything about optimism, pessimism, the plans of the patents office, etc.? I just answered the original poster, regarding the specific quotation. I didn't see the article in Globes, but I know the plan to give this specific budget. And this specific budget really doesn't have anything to do with HaMakor. -- Eli Marmor [EMAIL PROTECTED] CTO, Founder Netmask (El-Mar) Internet Technologies Ltd. __ Tel.: +972-9-766-1020 8 Yad-Harutzim St. Fax.: +972-9-766-1314 P.O.B. 7004 Mobile: +972-50-23-7338 Kfar-Saba 44641, Israel = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Patents or copyright ? [was Re: Israeli DMCA looming?]
Gilad Ben-Yossef wrote on 2003-07-03: > 2. The second thing is that our time and resources are limited. We are > ALL vulenteers. No one gets paid and that includes lawyer and > accountant. We do this with pleasure, but we also have to work and live. > Therefore the next thing is organise activities on your own. We will be > glad to help you in things like finanace, getting recipiets, printing > material, web storage etc. But you'll have to be the one to *organise* > the activity. > I don't currently have the time to organize any such activity but more importantly I don't have the expretise. I don't even have the slightest idea what kind of activity could help here. One thing I can think of that would be useful is putting information - at least links - about the legal IP situtuation in Israel on Hamakor's website. Carefully explained opnions of those issues would be even better than raw information, of course. But currently I feel there is not even enough raw information availiable - unless one looks very hard. Some monthes ago I tried to learn what's the situation on software patents in Israel, I searched the web for about a day I think and didn't find much. IANAL and it's problematic for me to even talk with somebody about these issues when I can't ground my claims since I don't even know the current laws and the threatening changes. Paradoxically (or not), I know much more about US laws than Israel's... -- Beni Cherniavsky <[EMAIL PROTECTED]> = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Patents or copyright ? [was Re: Israeli DMCA looming?]
Beni Cherniavsky wrote: Guy Baruch wrote: Patents, on the other hand, can (and often do) directly affect SW, and especially _new_ SW (as opposed with ripped copies of *.exe or *.wav ) There is one more kind of IP crippling that affects software, even more than patents: DMCA's anti-circumvention clauses. I could reluctantly live with the need to pay royalities for running certain algorithms but I don't want to go to prison for even documenting them! What are the chances of such things getting into Israel? Unless people do something about it the chances are very high. The WTO has inserted an obligation to introduce DMCA-like provisions to local state law into several international treates that governs the subject of copyright in international context (WEPA and related treaties). Israel is (or plans to be) a side to these treaties. Moreover, as those who were present in the last annual ISOC-IL convention at the panel tprevious to the Open SOurce one heared, the top Israeli lawyers who deal with copyright issues push very hard for this legalization to happen. They see it as a good thing and generally do not understand or are aware of the issues involved. The people of HaMakor have (including btw, in that ISOC-Il panel) participated in discussions about these issues and generally tried to raise awareness to the whole matter and other related issues (like digital signatures and the dangers of misusing them, e-voting issues etc.) Quote frankly, this isn't enough. In fact, it's almost nothing. There are two reasons for this, and this is where the part of doing something gets in: 1. HaMakor curretnly counts something like 30 registered members and friends. That is a really low number. We have a very hard time convincing people that we represent anyone when only 30 people have bothered registering. So the first thing to do in join. 2. The second thing is that our time and resources are limited. We are ALL vulenteers. No one gets paid and that includes lawyer and accountant. We do this with pleasure, but we also have to work and live. Therefore the next thing is organise activities on your own. We will be glad to help you in things like finanace, getting recipiets, printing material, web storage etc. But you'll have to be the one to *organise* the activity. I relaise that 1 and 2 are connected - people do not want to join a body that they don't yet precieve as done enough for them, but this is very much a chicken and egg problem. So, the bottom line - put your money and time where your mouth is and MAYBE we'll be able to do something about it. Everyone we'll have a good oppertunity to register at August Penguin, if sending that 100 NIS was what stopping you. And no, this isn't personally directed to the authors of this message. thanks, Gilad. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Multiple GLIBCs
Isaac Aaron wrote: Hi List Running Red Hat 7.3 requires me to recompile everything I want from Red Hat 9 (or that is avaliable to Red Hat) before I install it. I'm looking for a way to co-install GLIBC 2.3 with the existing GLIBC 2.1. Any opinions? No problem at all. Linux (like other Un*x like systems) has a rather elegant way to deal with libraries. When you install the RPM (I'm assuming you using RPM) make sure to "install" with -i rather then upgrade with -U. When the RPM will run ldconfig it will set things up so that any program that explicitly asks for the older version will get it, a program that does not specify a specific version will get the latest, and if a program seems to mis behave with the newer lib you can always use LD_PRELOAD (man ld.so) to override. In short - don't worry, be happy ;-) Cheers, Gilad. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Israeli DMCA looming?
Eli Marmor wrote: This doesn't have anything to do with HaMakor. Currently, the Israeli patents office is very primitive, and not professional. There is almost no computerization. The procedures take MANY years to complete. Usually, instead of scanning or reviewing patents, the staff prefer to adopt decisions of other offices regarding the specific patent. This is "bizayon", and makes a bad name for the state of Israel. Fortunately, it is going to change. Finally. It's a good improvement for everybody. I'm sorry Eli, that I don't share your opptimism. The US patent office, to takr a not so rendom example, has been very very active. I don't think I like the results much and being the pessimist that I am I'm having a hard time believing that the Israeli PO will do any better. Gilad = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Strange Locale ( was Re: gentoo and Xkb )
On Thu, Jul 03, 2003 at 12:07:14PM +0300, [EMAIL PROTECTED] wrote: > > http://syscalltrack.sf.net > > Thanks. Sounds like just what I was after, including a Debian > package. The debian package is badly out of date... I really recommend compiling from source. > For instance, from reading the docs of syscalltrack all I'll have > to do is to configure it with: > > rule { > syscall_name = execle > when = before > action { type = LOG } > } > > And get exactly what I want (hmmm, lots of shell/perl hacking on the output, > yummi yummi:) You can also set the log format, to make it easier for you to parse. -- Muli Ben-Yehuda http://www.mulix.org http://www.livejournal.com/~mulix/ pgp0.pgp Description: PGP signature
RE: Multiple GLIBCs
> -Original Message- > From: Gilad Ben-Yossef [mailto:[EMAIL PROTECTED] > Sent: Thursday, July 03, 2003 10:56 AM > To: Isaac Aaron > Cc: [EMAIL PROTECTED] > Subject: Re: Multiple GLIBCs > > > Isaac Aaron wrote: > > Hi List > > > > Running Red Hat 7.3 requires me to recompile everything I want > from Red Hat 9 (or that is avaliable to Red Hat) before I install it. > > > > I'm looking for a way to co-install GLIBC 2.3 with the existing > GLIBC 2.1. > > Any opinions? > > > > No problem at all. Linux (like other Un*x like systems) has a rather > elegant way to deal with libraries. When you install the RPM (I'm > assuming you using RPM) make sure to "install" with -i rather then > upgrade with -U. > Won't I need a --force there? How about if I just "steal" the libc* files (not links) and drop them on the /usr/lib directory. Would that satisfy the RPM "Requires" directive? > When the RPM will run ldconfig it will set things up so that any program > that explicitly asks for the older version will get it, a program that > does not specify a specific version will get the latest, and if a > program seems to mis behave with the newer lib you can always use > LD_PRELOAD (man ld.so) to override. That seems to do the trick as long as the small stuff "util-linux" "fileutils" are all linked in a way that won't mess up the whole installation. Red hat used to have compat-glibc-* packages which contained older glibc libraries. For some reason they don't do that anymore (to promote the advanced server?). What would really be nice is if I had a way to prepare such RPMS by myself (install cleanly, work with any version). > > In short - don't worry, be happy ;-) I WILL! I PROMISE! Thanks Isaac Aaron > > Cheers, > Gilad. > To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Strange Locale ( was Re: gentoo and Xkb )
On Thu, Jul 03, 2003 at 11:10:45AM +0300, Muli Ben-Yehuda wrote: > On Thu, Jul 03, 2003 at 10:56:29AM +0300, [EMAIL PROTECTED] wrote: > > What was that syscalltrace? > > http://syscalltrack.sf.net Thanks. Sounds like just what I was after, including a Debian package. > > > I'm looking for tools in that area and couldn't find such a program > > in google or otherwise (specifically - I'm thinking of something which > > will report to me all the paths passed to execve(2) anywere in the > > system, the closest I found so far is the LinuxTraceToolkit (LTT) but > > I suspect it's too heavy for my needs). > > Why is ltt too heavy? Because all I want is to have a list of file names (and maybe times), LTT gives much much more about many more things and looks farheavier to setup and run. For instance, from reading the docs of syscalltrack all I'll have to do is to configure it with: rule { syscall_name = execle when = before action { type = LOG } } And get exactly what I want (hmmm, lots of shell/perl hacking on the output, yummi yummi:) Thanks, --Amos = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Patents or copyright ? [was Re: Israeli DMCA looming?]
Guy Baruch wrote: Actually, although there is a reference to IP in general, from all the types of IP (patents, trademarks, copyrights, trade-secrets, are those I'm aware of), only the first gets mentioned specificly. AFAIK, Patents are largely irrelevant to music-copying (a copyright issue), or to DMCA (again, mainly a copyright-aimed law) Patents, on the other hand, can (and often do) directly affect SW, and especially _new_ SW (as opposed with ripped copies of *.exe or *.wav ) So, any change w.r.t. patent-policy should interest HAMAKOR greatly. In fact, it should, IMHO, be very high on HAMAKOR's agenda. It is high on our Agenda. Very high, in fact and we are trying to find out at the moment what was the remark refering to, excacly. However, we're only a bunch of vulenteers. If you want something to be done about it - why don't you come forward and organise what needs to be done and we will gladly provide any help and "offical" status to back up the effort? BTW, I very much assume that the reason patents were mentioned was because the speakers has no idea what the difference between the various types of the so called Intelectual property are. Gilad. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Patents or copyright ? [was Re: Israeli DMCA looming?]
Guy Baruch wrote: > Patents, on the other hand, can (and often do) directly affect SW, > and especially _new_ SW (as opposed with ripped copies of *.exe or > *.wav ) > There is one more kind of IP crippling that affects software, even more than patents: DMCA's anti-circumvention clauses. I could reluctantly live with the need to pay royalities for running certain algorithms but I don't want to go to prison for even documenting them! What are the chances of such things getting into Israel? -- Beni Cherniavsky <[EMAIL PROTECTED]> = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Patents or copyright ? [was Re: Israeli DMCA looming?]
Geoffrey S. Mendelson wrote on 2003-07-03: > One of the big problems is that Israel's newest copyright law makes it > illegal to SELL copyrighted software that you don't have the rights > to sell, e.g. "bootleg", but does not make it illegal to BUY it. > > While this protects the person who buys a bootleg copy in a box that looks > like a legal copy, it also protects the person who buys a copy at the > Tel Aviv bus station. > So what do you propose? It's not always easy to know whether your copy is bootleg when it does come in a box. It doesn't sound fair to blame the first person. But how would you distinguish him from the second person in a law? > On the other hand, it still is illegal to buy one copy and propigate it > throughout your network if it comes with a single user license. (or to > use a five user license 20 times). > What do you want to say here? Is good/bad? > One of the nice things about Windows XP and Office XP, is the first time > you use them they register with a server at Microsoft. When a single user > key starts showing up multiple times, a real person gets notified. > Wouldn't it also complain when you replace your hard disk and install and register it again? While we are talking about licenses, what's the legal status of shrink-wrap EULAs in Israel? Are there legal restrictions on what kinds of restrictions can EULAs contain? -- Beni Cherniavsky <[EMAIL PROTECTED]> "Reading the documentation I felt like a kid in a toy shop." -- Phil Thompson on Python's standard library = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Multiple GLIBCs
Isaac Aaron wrote: -Original Message- From: Gilad Ben-Yossef [mailto:[EMAIL PROTECTED] Sent: Thursday, July 03, 2003 10:56 AM To: Isaac Aaron Cc: [EMAIL PROTECTED] Subject: Re: Multiple GLIBCs Isaac Aaron wrote: Hi List Running Red Hat 7.3 requires me to recompile everything I want from Red Hat 9 (or that is avaliable to Red Hat) before I install it. I'm looking for a way to co-install GLIBC 2.3 with the existing GLIBC 2.1. Any opinions? No problem at all. Linux (like other Un*x like systems) has a rather elegant way to deal with libraries. When you install the RPM (I'm assuming you using RPM) make sure to "install" with -i rather then upgrade with -U. Won't I need a --force there? Not because of the exitence of an older library. How about if I just "steal" the libc* files (not links) and drop them on the /usr/lib directory. > Would that satisfy the RPM "Requires" directive? The application will be able to use them (once you run ldconfig at least), but RPM will have no knowledge on it so it wont let you install (unless you use --nodeps of course). When the RPM will run ldconfig it will set things up so that any program that explicitly asks for the older version will get it, a program that does not specify a specific version will get the latest, and if a program seems to mis behave with the newer lib you can always use LD_PRELOAD (man ld.so) to override. That seems to do the trick as long as the small stuff "util-linux" "fileutils" are all linked in a way > that won't mess up the whole installation. I don't think they will but don't trust me. Copy the binaries to some bizzare location and use LD_PRELOAD to force load the library with the stuff you need to test it BEFORE "installation". Red hat used to have compat-glibc-* packages which contained older glibc libraries. AFAIR that was for REALLY old libc that had major binary interface changes from the newer ones. For some reason they don't do that anymore (to promote the advanced server?). For 99% of the cases it's not needed. What would really be nice is if I had a way to prepare such RPMS by myself (install cleanly, work with any version). Of course you can! That's the whole point of using Free Software. Just download the SRPMS use to create the RPMs you need and generate the RPMs localy. man rpm for the gory details. Gilad = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Patents or copyright ? [was Re: Israeli DMCA looming?]
Guy Baruch wrote: > Patents, on the other hand, can (and often do) directly affect SW, and > especially > _new_ SW (as opposed with ripped copies of *.exe or *.wav ) One of the big problems is that Israel's newest copyright law makes it illegal to SELL copyrighted software that you don't have the rights to sell, e.g. "bootleg", but does not make it illegal to BUY it. While this protects the person who buys a bootleg copy in a box that looks like a legal copy, it also protects the person who buys a copy at the Tel Aviv bus station. On the other hand, it still is illegal to buy one copy and propigate it throughout your network if it comes with a single user license. (or to use a five user license 20 times). One of the nice things about Windows XP and Office XP, is the first time you use them they register with a server at Microsoft. When a single user key starts showing up multiple times, a real person gets notified. Geoff. -- Geoffrey S. Mendelson [EMAIL PROTECTED] 972-54-608-069 Do sysadmins count networked sheep? = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Mail RFC compliance
It's maybe off-topic for this thread, but I have a very similar situation at work. I can't work out the solution for the problem. Someone out there is sending spam with arbitrary names at the Reply-To field. The problem is, that the mail domain names there are of our main mail server. Typical name would be [EMAIL PROTECTED] The messages are returned from the spam destinations with 'user unknown' status, and naturally return to my mail server. The headers show only the spam destination and the mail proxy of the technion. As a result, in the lucky hours the server recieves ten extra mails per minute (with all the javascript, etc, as it's usually done at spam messages), and never less then 2 or three mails per minute. In addition, I recieve about 5 abuse complaints a day from all over the world. I can't figure out the solution for the problem. I configured the mail server to drop the 'unknown-user' mails before the body download occures, but agree, that it's still not the solution. :) Hope someone can help me. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Strange Locale ( was Re: gentoo and Xkb )
On Thu, Jul 03, 2003 at 10:56:29AM +0300, [EMAIL PROTECTED] wrote: > On Thu, Jul 03, 2003 at 09:42:51AM +0300, Oron Peled wrote: > > (Ok, I know we can trap acceess to it via syscalltrace but) > > What was that syscalltrace? http://syscalltrack.sf.net > I'm looking for tools in that area and couldn't find such a program > in google or otherwise (specifically - I'm thinking of something which > will report to me all the paths passed to execve(2) anywere in the > system, the closest I found so far is the LinuxTraceToolkit (LTT) but > I suspect it's too heavy for my needs). Why is ltt too heavy? -- Muli Ben-Yehuda http://www.mulix.org http://www.livejournal.com/~mulix/ pgp0.pgp Description: PGP signature
Re: Strange Locale ( was Re: gentoo and Xkb )
Can you explain me or even better point me to some documents that can explain how can I generate a new locale ? I tried to search over the internet for a document that will explain me what exactly is i18n / locale and how is it implanted on the system. does any one know of such document ? > > Generate one. But use the name: "he_IL.UTF-8". ("utf8" vs. "UTF-8" > shouldn't make a difference, but then again, someone might actually rely > on the charset part in the future) > -- Canaan Surfing Ltd. Internet Service Providers Ben-Nes Michael - Manager Tel: 972-4-6991122 Fax: 972-4-6990098 http://sites.canaan.co.il -- = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Strange Locale ( was Re: gentoo and Xkb )
On Thu, Jul 03, 2003 at 09:42:51AM +0300, Oron Peled wrote: > (Ok, I know we can trap acceess to it via syscalltrace but) What was that syscalltrace? I'm looking for tools in that area and couldn't find such a program in google or otherwise (specifically - I'm thinking of something which will report to me all the paths passed to execve(2) anywere in the system, the closest I found so far is the LinuxTraceToolkit (LTT) but I suspect it's too heavy for my needs). --Amos = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
RE: Multiple GLIBCs
>> I'm looking for a way to co-install GLIBC 2.3 with the >> existing GLIBC 2.1. >> Any opinions? Since you can control all GCC's flags from environment variables you should add to your profile a menu to choose what compilation environment you like (and set it to the prompt or title of the terminal). This e-mail message has been sent by Elbit Systems Ltd. and is for the use of the intended recipients only. The message may contain privileged or confidential information . If you are not the intended recipient you are hereby notified that any use, distribution or copying of this communication is strictly prohibited, and you are requested to delete the e-mail and any attachments and notify the sender immediately. = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
Re: Strange Locale ( was Re: gentoo and Xkb )
On Thu, Jul 03, 2003 at 09:42:51AM +0300, Oron Peled wrote: > On Thu, 03 Jul 2003 00:54:18 +0300 > Tzafrir Cohen <[EMAIL PROTECTED]> wrote: > > > But then I tried to locate the locale files, I found none, ascept > > /usr/lib/locale/locale-archive > > First time I bumped into this file. Thought it might be some gzipped > tar or something, but guess what? > > $ file /usr/lib/locale/locale-archive > /usr/lib/locale/locale-archive: PDP-11 separate I&D executable not stripped > > So, obviously file(1) doesn't reconize the format of this "archive". > Anybody knows it's function/format? > > (Ok, I know we can trap acceess to it via syscalltrace but) No need to do that. It is indeed where the locales are kept. Or so it seems from strace of any simple command (try 'strace date'). Anybody wants to dive into glibc's source? Searching, I see that rawhide's rpm packages now contain /usr/sbin/build-locale-archive . I could not find this binary on my debian-unstable. -- Tzafrir Cohen +---+ http://www.technion.ac.il/~tzafrir/ |vim is a mutt's best friend| mailto:[EMAIL PROTECTED] +---+ = To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]