- WLUG - wlug.mailman3.com

kernel - Help!

by avedis＠rcn.com

I'm in windows right now, which is where I automatically boot to if I don't use the floppy. I only get a choice of boots with floppy. Tried recompiling the kernel with sound, added the lbn32(?) line to /etc/lilo and the cylinder error went away, but somehow I lost support for ipchains and lost the suse firewall. Also lost support for vfat. I went back into make xconfig and started messing around with ppp support, vfat and dos support. Got my vfat back ok. Also had to go in and re-edit /etc/rc.config and /etc/rc.config.d/firewall- somethingorother.config. I think I need to review my selections in those files. At first, I could dialup and as soon as it would connect, it would disconnect, and pppd would give error message 4, no support for ppp in kernel. I went back in and selected netfilter with ipchains support, and ppp with async support. Now it will connect and stay connected, but I can't send or receive anything (mail or web). Oh, and I removed the sound support from the kernel, but I'm still getting error messages about trying to load the sound modules. Let me know if I did something obvious. Meanwhile, I'll review what I did. Thanks, Greg

Dec. 22, 2002

Re: [Wlug] "LN" question

by Bill Mills-Curran

On Sat, 21 Dec 2002, Michael LaPointe wrote: > My original site had all of the images in an images directory. To > avoid copying all of the images to each new sub domain/directory I > created an images directory for each sub directory and ran ln *.* > /path/to/sub/images from within the real images directory. > > This worked great but is it only a one shot deal? I.E. if I add > another image to the main directory do I need to run ln again? Michael, What you have created in each directory is an images directory populated with links. As you supposed, each time you add something to the "real" images directory, you have to add links to it everywhere. Here's a more efficient approach. 1) get rid of all these directories of links and 2) in each site directory, make single link to the "real" images directory (not to the images themselves). For example, cd <site_directory> ln -s <path_to_images_directory> images That's it. You won't have to worry about maintaining links to the individual files. Hope this helps. Bill

Dec. 21, 2002

[Off-Topic] Free Software Consortium in search of Consultants and Agents in your area.

by Consorcio Software Libre

We would like to invite you to be a founding members of the FSC either as a Consultant and/or an Agent. http://www.fsconsortium.com On the main page you will find links to subscribe as consultant and/or agent. Free Software Consortium is a multinational consortium established to link the companies and the independent consultants that use and promote free software. One of the bigger obstacles to the development of the free software services market is the fear of many companies to establish contracts with independent consultants who do not have support of a well known Company, Corporation or Consortium. Questioning quality, support, experience and capability to successfully complete a project are the companies manager's main concerns when considering the option of migrating their systems to free software. Like you, we know the excellent quality that Free Software offers. We have complete confidence that Free Software is going to be the platform of the future. That is why it's so important to work and consolidate together to solve the problem mentioned above. Free Software Consortium intends to fill this void joining the best free software companies and consultants at the national and international level. By joining to Free Software Consortium you or your company will have the following benefits: 1. Possibility of being hired by other Consortium members. 2. Possibility of hiring other Consortium members. 3. Use of the Consortium image and name with your clients. 4. Use of the Consortium's resources like Publicity Brochures, Business Cards, Banners for presentations and events, financial presentations, publicity articles for the press and magazines, support via mailing list and forums with the Consortium's members, marketing and communication assistance, CD's and manuals from the Consortium, site and mail hosting (yourarea.fsc.cc) and other future services. 5. Representation of the FS Consortium in your area to share administrative, legal and marketing resources between the Consultant members in your area. Free Software Consortium will establish a contract with an Agent company which will have exclusive representation in it's area, in charge of organizing the consultants in the area. This Agent company will provide the administrative services, marketing and sales with the local customers.

Dec. 21, 2002

"LN" question

by Michael LaPointe

I just recently upgraded my hosting plan to include sub domains, shell access and a few other goodies. My original site had all of the images in an images directory. To avoid copying all of the images to each new sub domain/directory I created an images directory for each sub directory and ran ln *.* /path/to/sub/images from within the real images directory. This worked great but is it only a one shot deal? I.E. if I add another image to the main directory do I need to run ln again? Server is FreeBSD 4.7 with Apache/1.3.27 I didnt try hard linking the directories since my books say that you need to be root to do so but I think that is what I really needed. Thank you, ~ Michael LaPointe

Dec. 21, 2002

Re: [Wlug] kernel questions

by Charles R. Anderson

On Sat, Dec 21, 2002 at 11:52:21AM -0500, Gregory Avedissian wrote: avedis> 1. I get errors on this, because I'm above cylinder 1023 on my dual-boot avedis> drive. I boot linux from floppy with lilo. Error message suggests using avedis> lba32 option, but it doesn't say where or how to use it. How do I get avedis> around this? /etc/lilo.conf, put lba32 on a line by itself (like right below the boot = line). Remove any "linear" line you see at the same time. Then rerun "/sbin/lilo". Then you may be able to boot using lilo on the hard drive instead of using a floppy (if your BIOS is new enough). avedis> 2. Instructions say next to install the kernel. Is that done by typing avedis> make install? In addition to the make install that comes after make avedis> modules? (see below) make install does work with recent kernels, yes. It should put the kernel into /boot for you. You might or might not still need to create an initial ramdisk with the "mkinitrd" command... You also will probably have to add an entry to /etc/lilo.conf and rerun /sbin/lilo. -- Charles R. Anderson <cra(a)wpi.edu> / http://angus.ind.wpi.edu/~cra/ PGP Key ID: 49BB5886 Fingerprint: EBA3 A106 7C93 FA07 8E15 3AC2 C367 A0F9 49BB 5886

Dec. 21, 2002

RE: [Wlug] kernel questions

by Skip Gaede

Hi Greg, > -----Original Message----- > From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org]On > Behalf Of Gregory Avedissian > Sent: Saturday, December 21, 2002 11:52 AM > To: wlug(a)mail.wlug.org > Subject: [Wlug] kernel questions > > > I'm trying to get oss sound modules compiled into kernel. > make xconfig was easy enough to get through. > instructions say to then make dep and make bzImage or bzlilo. > 1. I get errors on this, because I'm above cylinder 1023 on > my dual-boot > drive. I boot linux from floppy with lilo. Error message > suggests using > lba32 option, but it doesn't say where or how to use it. How do I get > around this? The best option would be to shrink your primary partition by a few MB, leaving room to create a small primary partition for the Linux /boot partiition. You can easily do this with Partition Magic. ---------------------- Primary | ext2 | /boot ---------------------- Primary | FAT or NTFS or ??? ---------------------- Extended| swap ---------------------- Extended| ext2 or ext3 | / ---------------------- other partitions as desired... > > 2. Instructions say next to install the kernel. Is that done > by typing > make install? In addition to the make install that comes after make > modules? (see below) No, you need to copy the arch/i386/boot/bzImage --> /boot/<some unique name> and then add an entry to /etc/lilo.conf or /boot/grub/menu.lst. If you are using lilo, you then need to install the new boot sector by typing lilo. > > 3. I haven't been able to create a lilo boot disk for awhile. > I'm using > Yast2 to configure boot mode, and it keeps telling me that my > floppy is > write protected and it can't access the boot sector. I've tried this > with new floppies that have been formated with fdformat /dev/fd0. You shouldn't need to format it first, and the disk should not be mounted. Refer to help on www.tldp.org on how to make a boot disk. > > Thanks for any insight. > > Here are the instructions I'm looking at: > > 1. To make the kernel sound modules, cd to your /usr/src/linux > > directory (typically) and type make config, make menuconfig, > > or make xconfig (to start the command line, dialog, or x-based > > configuration tool). > > > > 2. Select the Sound option and a dialog will be displayed. > > > > 3. Select M (module) for "Sound card support". > > > > 4. Select your sound driver(s) as a module. For ProAudio, Sound > > Blaster, etc., select M (module) for OSS sound modules. > > [thanks to Marvin Stodolsky <stodolsk(a)erols.com>]A > > > > 5. Make the kernel (e.g., make dep ; make bzImage), and install > > the kernel. > > > > 6. Make the modules and install them (make modules; make > modules_install). > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Dec. 21, 2002

kernel questions

by Gregory Avedissian

I'm trying to get oss sound modules compiled into kernel. make xconfig was easy enough to get through. instructions say to then make dep and make bzImage or bzlilo. 1. I get errors on this, because I'm above cylinder 1023 on my dual-boot drive. I boot linux from floppy with lilo. Error message suggests using lba32 option, but it doesn't say where or how to use it. How do I get around this? 2. Instructions say next to install the kernel. Is that done by typing make install? In addition to the make install that comes after make modules? (see below) 3. I haven't been able to create a lilo boot disk for awhile. I'm using Yast2 to configure boot mode, and it keeps telling me that my floppy is write protected and it can't access the boot sector. I've tried this with new floppies that have been formated with fdformat /dev/fd0. Thanks for any insight. Here are the instructions I'm looking at: > 1. To make the kernel sound modules, cd to your /usr/src/linux > directory (typically) and type make config, make menuconfig, > or make xconfig (to start the command line, dialog, or x-based > configuration tool). > > 2. Select the Sound option and a dialog will be displayed. > > 3. Select M (module) for "Sound card support". > > 4. Select your sound driver(s) as a module. For ProAudio, Sound > Blaster, etc., select M (module) for OSS sound modules. > [thanks to Marvin Stodolsky <stodolsk(a)erols.com>]A > > 5. Make the kernel (e.g., make dep ; make bzImage), and install > the kernel. > > 6. Make the modules and install them (make modules; make modules_install).

Dec. 21, 2002

Re: [Wlug] The Package Chase

by Andy Stewart

On Friday 20 December 2002 9:25 pm, Gregory Avedissian wrote: > OK, I've been going along merrily with my SuSE 7.2 for about a year, and > I just now ran into a problem that a lot of you folks must know all too > well. Install a new program and it calls for a package you don't have. > You've all seen the lists, I'm sure. > > I'm trying to install ecawave and ALSA 0.9, and I've had to install some > gtk and gnome stuff, and I went to look at upgrading the whole gnome > package. There's a list of more than 50 things to compile, and they must > be done in the correct order. Do people actually do that? Is there any > advantage of doing that over upgrading the whole distribution? HI Greg, It might be easier to find a more recent SuSE CD and upgrade just the packages about which you care. You could probably download them, but that may be slow and cumbersome given the number of packages you describe. If its just one or two packages, I'll "do it the hard way" and download the package and upgrade it. On occasion, it turns into a rat's nest of dependencies, but typically not. Later, Andy -- Andy Stewart, Founder Worcester Linux Users' Group Worcester, MA USA http://www.wlug.org

Dec. 21, 2002

The Package Chase

by Gregory Avedissian

OK, I've been going along merrily with my SuSE 7.2 for about a year, and I just now ran into a problem that a lot of you folks must know all too well. Install a new program and it calls for a package you don't have. You've all seen the lists, I'm sure. I'm trying to install ecawave and ALSA 0.9, and I've had to install some gtk and gnome stuff, and I went to look at upgrading the whole gnome package. There's a list of more than 50 things to compile, and they must be done in the correct order. Do people actually do that? Is there any advantage of doing that over upgrading the whole distribution?

Dec. 21, 2002

Re: [Wlug] bioswriter

by Wes Allen

This is what I was afraid of. Oh well, maybe one of the freeBIOS' will come along and I can try flashing this thing out of existence (if my chip is supported, that is...). Wes On Thursday 19 December 2002 10:13 pm, Skip Gaede wrote: > Wes, > > The location in BIOS for the PNP setting is not cast in concrete. > Unless you know which bit does the job, it is pretty hard to > figure it out. > > When you do figure it out, each block of 16 bytes (I think) includes > a checksum. so in addition to changing the bit, you'll also need to > change the checksum. > > --Skip > > > -----Original Message----- > > From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org]On > > Behalf Of Wes Allen > > Sent: Thursday, December 19, 2002 5:19 PM > > To: wlug(a)mail.wlug.org > > Subject: [Wlug] bioswriter > > > > > > Has anyone ever used bioswriter or a similar utility to get > > your bios settings > > from CMOS? I'm trying to read it with bioswriter with little > > success as I'm > > not sure what to do with the image file. > > > > I'm wondering there is a way to hack this bios to let me turn > > "plug and play > > os" off, as I'm fairly certain it's on by default and the > > given interface is > > pretty limited. > > > > Or am I breaking the law to edit my bios settings.... on the > > computer I OWN. > > > > Wes > > -- > > ------ > > This message may be signed using GPG, for my public key please > > send me a message with "Key Request" in the subject line. > > > > > > _______________________________________________ > > Wlug mailing list > > Wlug(a)mail.wlug.org > > http://mail.wlug.org/mailman/listinfo/wlug > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug -- ------ This message may be signed using GPG, for my public key please send me a message with "Key Request" in the subject line.

Dec. 20, 2002

RE: [Wlug] bioswriter

by Skip Gaede

Wes, The location in BIOS for the PNP setting is not cast in concrete. Unless you know which bit does the job, it is pretty hard to figure it out. When you do figure it out, each block of 16 bytes (I think) includes a checksum. so in addition to changing the bit, you'll also need to change the checksum. --Skip > -----Original Message----- > From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org]On > Behalf Of Wes Allen > Sent: Thursday, December 19, 2002 5:19 PM > To: wlug(a)mail.wlug.org > Subject: [Wlug] bioswriter > > > Has anyone ever used bioswriter or a similar utility to get > your bios settings > from CMOS? I'm trying to read it with bioswriter with little > success as I'm > not sure what to do with the image file. > > I'm wondering there is a way to hack this bios to let me turn > "plug and play > os" off, as I'm fairly certain it's on by default and the > given interface is > pretty limited. > > Or am I breaking the law to edit my bios settings.... on the > computer I OWN. > > Wes > -- > ------ > This message may be signed using GPG, for my public key please > send me a message with "Key Request" in the subject line. > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Dec. 20, 2002

recording in linux

by Gregory Avedissian

Someone asked about recording under linux recently, and I recall that several of us said we'd never been able to do it. I just did it. In alsamixer, space bar sets the channel you want to use for capture. (line in is a reasonable choice) Had to manually edit /etc/asound.conf to turn on the input gain and set the level to 10,10. It looks like this: element("Input Gain Switch",0,101,Switch2(on)) ; Voice 0 : Min 0 Max 15 ; Voice 1 : Min 0 Max 15 element("Input Gain Volume",0,200,Volume1(10,10))

Dec. 20, 2002

bioswriter

by Wes Allen

Has anyone ever used bioswriter or a similar utility to get your bios settings from CMOS? I'm trying to read it with bioswriter with little success as I'm not sure what to do with the image file. I'm wondering there is a way to hack this bios to let me turn "plug and play os" off, as I'm fairly certain it's on by default and the given interface is pretty limited. Or am I breaking the law to edit my bios settings.... on the computer I OWN. Wes -- ------ This message may be signed using GPG, for my public key please send me a message with "Key Request" in the subject line.

Dec. 19, 2002

RE: [Wlug] Cross Compile Question

by Scott Venier

Depends on the language... if it's C, yes. if it's C++, no way. Scott On Thu, 19 Dec 2002, Marc Hughes wrote: > If I compile some source files with gcc2.x and some with gcc3.x, can I > link the object files together? > > -Marc > > -----Original Message----- > From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org] On > Behalf Of Skip Gaede > Sent: Thursday, December 19, 2002 12:23 PM > To: wlug(a)mail.wlug.org > Subject: RE: [Wlug] Cross Compile Question > > > I recently saw that the GCC 3.x compiler isn't cross compile friendly. > Try using GCC 2.95.3. > > --Skip > > -----Original Message----- > > From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org]On > > Behalf Of Marc Hughes > > Sent: Thursday, December 19, 2002 11:45 AM > > To: wlug(a)wlug.org > > Subject: [Wlug] Cross Compile Question > > > > > > I'm trying to make a cross-compiler on an i386 linux box to a > > ppc darwin > > (OSX) box. I downloaded binutils, configured them with > > ./configure --target=powerpc-apple --prefix=/usr/local > > Build them, and installed them. > > > > Then I went and got gcc (with g++), configured it with: ./configure > > --target=powerpc-apple-darwin --prefix=/usr/local > > > > I also made a symlink from /usr/local/powerpc-apple -> > > /usr/local/powerpc-apple-darwin since it seemed to fix a problem in > > the gcc compile. > > > > When I try to build gcc, it goes for a minute or so, and then > > I get this > > towards the end: > > > > libbackend.a(varasm.o): In function `assemble_alias': > > /data/hughesm/tmp/gcc-3.1/gcc/varasm.c:5184: undefined reference to > > `ASM_OUTPUT_DEF' > > collect2: ld returned 1 exit status > > make[1]: *** [cc1] Error 1 > > make[1]: Leaving directory `/data/hughesm/tmp/gcc-3.1/gcc' > > make: *** [all-gcc] Error 2 > > > > I'm not really asking for the exact cause of the error... I'm just > > trying to figure out if I'm even going about this the correct way. > > The documentation on cross compiling is a bit sparse. It seems odd to > > > me that the targets don't match (but there is no powerpc-apple-darwin > > in binutils, and no powerpc-apple in gcc) > > > > All that being said, I'm doing this so I can add my linux boxes to my > > OSX build farm. Check out distcc.samba.org for a wicked cool and easy > > > way to distribute compiles. > > > > -Marc > > > > > > > > _______________________________________________ > > Wlug mailing list > > Wlug(a)mail.wlug.org > > http://mail.wlug.org/mailman/listinfo/wlug > > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug > > > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Dec. 19, 2002

RE: [Wlug] Cross Compile Question

by Marc Hughes

If I compile some source files with gcc2.x and some with gcc3.x, can I link the object files together? -Marc -----Original Message----- From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org] On Behalf Of Skip Gaede Sent: Thursday, December 19, 2002 12:23 PM To: wlug(a)mail.wlug.org Subject: RE: [Wlug] Cross Compile Question I recently saw that the GCC 3.x compiler isn't cross compile friendly. Try using GCC 2.95.3. --Skip > -----Original Message----- > From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org]On > Behalf Of Marc Hughes > Sent: Thursday, December 19, 2002 11:45 AM > To: wlug(a)wlug.org > Subject: [Wlug] Cross Compile Question > > > I'm trying to make a cross-compiler on an i386 linux box to a > ppc darwin > (OSX) box. I downloaded binutils, configured them with > ./configure --target=powerpc-apple --prefix=/usr/local > Build them, and installed them. > > Then I went and got gcc (with g++), configured it with: ./configure > --target=powerpc-apple-darwin --prefix=/usr/local > > I also made a symlink from /usr/local/powerpc-apple -> > /usr/local/powerpc-apple-darwin since it seemed to fix a problem in > the gcc compile. > > When I try to build gcc, it goes for a minute or so, and then > I get this > towards the end: > > libbackend.a(varasm.o): In function `assemble_alias': > /data/hughesm/tmp/gcc-3.1/gcc/varasm.c:5184: undefined reference to > `ASM_OUTPUT_DEF' > collect2: ld returned 1 exit status > make[1]: *** [cc1] Error 1 > make[1]: Leaving directory `/data/hughesm/tmp/gcc-3.1/gcc' > make: *** [all-gcc] Error 2 > > I'm not really asking for the exact cause of the error... I'm just > trying to figure out if I'm even going about this the correct way. > The documentation on cross compiling is a bit sparse. It seems odd to > me that the targets don't match (but there is no powerpc-apple-darwin > in binutils, and no powerpc-apple in gcc) > > All that being said, I'm doing this so I can add my linux boxes to my > OSX build farm. Check out distcc.samba.org for a wicked cool and easy > way to distribute compiles. > > -Marc > > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug > _______________________________________________ Wlug mailing list Wlug(a)mail.wlug.org http://mail.wlug.org/mailman/listinfo/wlug

Dec. 19, 2002

RE: [Wlug] Cross Compile Question

by Skip Gaede

I recently saw that the GCC 3.x compiler isn't cross compile friendly. Try using GCC 2.95.3. --Skip > -----Original Message----- > From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org]On > Behalf Of Marc Hughes > Sent: Thursday, December 19, 2002 11:45 AM > To: wlug(a)wlug.org > Subject: [Wlug] Cross Compile Question > > > I'm trying to make a cross-compiler on an i386 linux box to a > ppc darwin > (OSX) box. I downloaded binutils, configured them with > ./configure --target=powerpc-apple --prefix=/usr/local > Build them, and installed them. > > Then I went and got gcc (with g++), configured it with: > ./configure --target=powerpc-apple-darwin --prefix=/usr/local > > I also made a symlink from /usr/local/powerpc-apple -> > /usr/local/powerpc-apple-darwin since it seemed to fix a > problem in the > gcc compile. > > When I try to build gcc, it goes for a minute or so, and then > I get this > towards the end: > > libbackend.a(varasm.o): In function `assemble_alias': > /data/hughesm/tmp/gcc-3.1/gcc/varasm.c:5184: undefined reference to > `ASM_OUTPUT_DEF' > collect2: ld returned 1 exit status > make[1]: *** [cc1] Error 1 > make[1]: Leaving directory `/data/hughesm/tmp/gcc-3.1/gcc' > make: *** [all-gcc] Error 2 > > I'm not really asking for the exact cause of the error... I'm just > trying to figure out if I'm even going about this the correct > way. The > documentation on cross compiling is a bit sparse. It seems odd to me > that the targets don't match (but there is no powerpc-apple-darwin in > binutils, and no powerpc-apple in gcc) > > All that being said, I'm doing this so I can add my linux boxes to my > OSX build farm. Check out distcc.samba.org for a wicked cool and easy > way to distribute compiles. > > -Marc > > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Dec. 19, 2002

Cross Compile Question

by Marc Hughes

I'm trying to make a cross-compiler on an i386 linux box to a ppc darwin (OSX) box. I downloaded binutils, configured them with ./configure --target=powerpc-apple --prefix=/usr/local Build them, and installed them. Then I went and got gcc (with g++), configured it with: ./configure --target=powerpc-apple-darwin --prefix=/usr/local I also made a symlink from /usr/local/powerpc-apple -> /usr/local/powerpc-apple-darwin since it seemed to fix a problem in the gcc compile. When I try to build gcc, it goes for a minute or so, and then I get this towards the end: libbackend.a(varasm.o): In function `assemble_alias': /data/hughesm/tmp/gcc-3.1/gcc/varasm.c:5184: undefined reference to `ASM_OUTPUT_DEF' collect2: ld returned 1 exit status make[1]: *** [cc1] Error 1 make[1]: Leaving directory `/data/hughesm/tmp/gcc-3.1/gcc' make: *** [all-gcc] Error 2 I'm not really asking for the exact cause of the error... I'm just trying to figure out if I'm even going about this the correct way. The documentation on cross compiling is a bit sparse. It seems odd to me that the targets don't match (but there is no powerpc-apple-darwin in binutils, and no powerpc-apple in gcc) All that being said, I'm doing this so I can add my linux boxes to my OSX build farm. Check out distcc.samba.org for a wicked cool and easy way to distribute compiles. -Marc

Dec. 19, 2002

Re: [Wlug] MORE Apache [PHP, MySQL] on Debian --Update

by Josh Huber

"Malcolm H. Beaulieu" <pctech(a)ece.wpi.edu> writes: > I add 'use strict;' and 'use CGI;' at the top, nothing changes and I > still receive the same error. BUT, when they are removed, the > script then works fine. WITHOUT changing anything else. Any ideas? Well, does it from the command line fine with both those options? You mentioned that you installed everything from source...I assume you you're not referring to perl as well? Perhaps trying to run the script as the user apache is running as? Do you have the CGI module installed for perl? (isn't it included with the default installation? I forget...) -- Josh Huber

Dec. 19, 2002

Re: [Wlug] MORE Apache [PHP, MySQL] on Debian --Update

by Malcolm H. Beaulieu

So this is unusual ... I add 'use strict;' and 'use CGI;' at the top, nothing changes and I still receive the same error. BUT, when they are removed, the script then works fine. WITHOUT changing anything else. Any ideas? On Wed, 18 Dec 2002, Malcolm H. Beaulieu wrote: > > I recently compiled Apache (with PHP, SSL, Perl) on a Debian box (I > downloaded all the sources BTW) and it has been running fine, for the most > part. The recent discussions reminded me that I had been having problems > with the perl part of the build. Whenever a script is executed, I receive > a "500 Internal Server Error" with the following in the logs: > > [DATE] [error] (2)No such file or directory: exec of /usr/local/apache/cgi-bin/test-cgi failed > [DATE] [error] [client IP] Premature end of script headers: /usr/local/apache/cgi-bin/test-cgi > > The script is actually a stripped down version of my original one (pasted > at the end of this email). The permissions are a+rx on the file, the path > to perl and the file are correct, the directory is the one defined under > ScriptAlias in the httpd.conf and the script executes fine from a prompt, > ie 'perl test-cgi'. mod_perl didn't complain about anything and compiled > correctly. Does anyone have any ideas about what the problem could be? > Thanx in advance for your help! > > -Malcolm > > > > =================== BEGIN FILE > > #!/usr/bin/perl > > print "Content-type: text/html \n\n"; > print "<HTML>\n"; > print "<HEAD>\n"; > print "</HEAD>\n"; > print "<BODY BGCOLOR=WHITE>\n\n"; > print "<BR><BR>\n\n"; > print "<CENTER><H3>Form Error</H3></CENTER><BR>\n"; > print "<H2>\nErrors were encountered in the following fields:\n"; > print "</BODY>\n"; > print "</HTML>\n"; > > > =================== END FILE > > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Dec. 18, 2002

MORE Apache [PHP, MySQL] on Debian

by Malcolm H. Beaulieu

I recently compiled Apache (with PHP, SSL, Perl) on a Debian box (I downloaded all the sources BTW) and it has been running fine, for the most part. The recent discussions reminded me that I had been having problems with the perl part of the build. Whenever a script is executed, I receive a "500 Internal Server Error" with the following in the logs: [DATE] [error] (2)No such file or directory: exec of /usr/local/apache/cgi-bin/test-cgi failed [DATE] [error] [client IP] Premature end of script headers: /usr/local/apache/cgi-bin/test-cgi The script is actually a stripped down version of my original one (pasted at the end of this email). The permissions are a+rx on the file, the path to perl and the file are correct, the directory is the one defined under ScriptAlias in the httpd.conf and the script executes fine from a prompt, ie 'perl test-cgi'. mod_perl didn't complain about anything and compiled correctly. Does anyone have any ideas about what the problem could be? Thanx in advance for your help! -Malcolm =================== BEGIN FILE #!/usr/bin/perl print "Content-type: text/html \n\n"; print "<HTML>\n"; print "<HEAD>\n"; print "</HEAD>\n"; print "<BODY BGCOLOR=WHITE>\n\n"; print "<BR><BR>\n\n"; print "<CENTER><H3>Form Error</H3></CENTER><BR>\n"; print "<H2>\nErrors were encountered in the following fields:\n"; print "</BODY>\n"; print "</HTML>\n"; =================== END FILE

Dec. 18, 2002

Re: [nylug-talk] Re: Help installing Apache on Debian

by bruce ingalls

> On Wed, Dec 18, 2002 at 10:53:42AM -0800, jose sanchez wrote: > > I did a search "installing php with oracle support on > > debian" and didn't get much on it. > > You need to have a whole bunch of oracle's libraries installed on the > system you're building on. Its safest to copy the entire installation > from the other systems (except for the datafiles) and set ORACLE_BASE > and ORACLE_HOME in a manner similar to how they're set on your > production system, then re-run the build. What kind of database driver are you using? You might also need to set LD_LIBRARY_PATH Oracle also has some configuration files, but generally, they don't need tweaking. However, they could be useful for turning on logging.

Dec. 18, 2002

Re: [nylug-talk] Re: Help installing Apache on Debian

by Ruben I Safir

I love Oracle. Fast, secure, easy. Do your job and get out of the way. Ruben On 2002.12.18 14:15 Peter C. Norton wrote: > On Wed, Dec 18, 2002 at 10:53:42AM -0800, jose sanchez wrote: > > I did a search "installing php with oracle support on > > debian" and didn't get much on it. > > You need to have a whole bunch of oracle's libraries installed on the > system you're building on. Its safest to copy the entire installation > from the other systems (except for the datafiles) and set ORACLE_BASE > and ORACLE_HOME in a manner similar to how they're set on your > production system, then re-run the build. > > -Peter > > -- > The 5 year plan: > In five years we'll make up another plan. > Or just re-use this one. > _______________________________________________ > The nylug-talk mailing list is at nylug-talk(a)nylug.org > To subscribe or unsubscribe: http://herzl.nylug.org/mailman/listinfo/nylug-talk > -- __________________________ Brooklyn Linux Solutions __________________________ DRM is THEFT - We are the STAKEHOLDERS http://fairuse.nylxs.com http://www.mrbrklyn.com - Consulting http://www.inns.net <-- Happy Clients http://www.nylxs.com - Leadership Development in Free Software http://www2.mrbrklyn.com/resources - Unpublished Archive or stories and articles from around the net http://www2.mrbrklyn.com/downtown.html - See the New Downtown Brooklyn.... 1-718-382-0585

Dec. 18, 2002

Re: [nylug-talk] Re: Help installing Apache on Debian

by Peter C. Norton

On Wed, Dec 18, 2002 at 10:53:42AM -0800, jose sanchez wrote: > I did a search "installing php with oracle support on > debian" and didn't get much on it. You need to have a whole bunch of oracle's libraries installed on the system you're building on. Its safest to copy the entire installation from the other systems (except for the datafiles) and set ORACLE_BASE and ORACLE_HOME in a manner similar to how they're set on your production system, then re-run the build. -Peter -- The 5 year plan: In five years we'll make up another plan. Or just re-use this one.

Dec. 18, 2002

Re: Help installing Apache on Debian

by jose sanchez

> apt-get install libc6-dev After installing libc6-dev the configuration now works. Now I have another problem. I'm trying to configure PHP with Oracle 8i support: ./configure --with-apache=../apache_1.3.27 --enable-track-vars --with-oracle --with-mysql --enable-url-includes and get this error: checking for Oracle-ORACLE support... yes checking Oracle Install-Dir... checking Oracle version... configure: error: Oracle needed libraries not found Do I need Oracle server installed on the same box Apache and php are or could I just install a third party library to support Oracle? I have Oracle 8i installed on a different box running RH 7.2 and I would like to connect to this box through a php webpage from the Debian box. I did a search "installing php with oracle support on debian" and didn't get much on it. Can it be done? Thanks everyone for your time and input. __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com

Dec. 18, 2002

Re: [Wlug] Apache, PHP, MySQL on Debian

by jose sanchez

Hi Josh: I don't know if you already received my last post on "installing apache on Debian" but, I am having some problems getting it configured. I would like to know if you can please help me. I already did a search on google.com but didn't get much help. This is what I'm getting when I type ./configure -v --prefix=/usr/apache + Module setenvif: yes cd ..; gcc -DLINUX=22 -DUSE_HSREGEX -DUSE_EXPAT -I./lib/expat-lite -DNO_DL_NEEDED `./apaci` -o helpers/dummy helpers/dummy.c -lm /usr/bin/ld: cannot open crt1.o: No such file or directory collect2: ld returned 1 exit status make: *** [dummy] Error 1 Creating Makefile in src + configured for Linux platform + setting C compiler to gcc + setting C pre-processor to NOT-AVAILABLE + checking for system header files + adding selected modules + using builtin Expat + checking sizeof various data types + doing sanity check on compiler and options ** A test compilation with your Makefile configuration ** failed. The above error output from the compilation ** test will give you an idea what is failing. Note that ** Apache requires an ANSI C Compiler, such as gcc. ======== Error Output for sanity check ======== cd ..; gcc -DLINUX=22 -DUSE_HSREGEX -DUSE_EXPAT -I./lib/expat-lite -DNO_DL_NEEDED `./apaci` -o helpers/dummy helpers/dummy.c -lm /usr/bin/ld: cannot open crt1.o: No such file or directory collect2: ld returned 1 exit status make: *** [dummy] Error 1 ============= End of Error Report ============= Aborting! Thanks in advance. ===== "An ounce of gold cannot buy an ounce of time." - Anonymous www.whmicro.com __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com

Dec. 18, 2002