From: ibuprofin@painkiller.example.tld.invalid   
      
   On Tue, 19 Feb 2013, in the Usenet newsgroup alt.os.linux.mandriva, in article   
   , TJ wrote:   
      
   >Moe Trin wrote:   
      
   >> TJ wrote:   
      
   >>> I've heard that some users have had mixed results with the open   
   >>> source firmware and drivers and Mageia 2.   
      
   >> I think it's more the old saw "if it works, don't fix it".   ;-)   
      
   >Well, you know how some Linux folks are. They're dead set against   
   >using anything proprietary, and the b43 driver I use needs proprietary   
   >firmware to work.   
      
   Some of that works both ways. In some cases, you get binary-only stuff   
   that is built by the hardware manufacturer for use in one specific   
   distribution/release.  Had this problem with one of the video card   
   manufacturers who created/provided a binary that worked with a   
   2.2.14-5.0 kernel ONLY.   What's the matter?  Doesn't EVERYONE run the   
   out-of-box kernel in Red Hat Linux 6.2 ("zoot" from March 2000)?  Just   
   because there were 8 kernel updates over the (RH) supported life of   
   that distribution is irrelevant.   
      
   In another case, a (high priced) box assembler was selling boxes with   
   pre-installed and configured Linux (I think it was also a Red Hat   
   release - it certainly used rpm packages), and it came with a USR PCI   
   modem, back when such modems were rare.  It included a package that   
   did some simple magic to get the modem recognized (PCI modems don't use   
   the old ISA serial port I/O and IRQs).  Investigation of the package   
   showed it was a boot script that looked at the contents of /proc/pci   
   searching for the "Serial controller:" device.   
      
      
   Bus  0, device  15, function  0:   
       Serial controller: Unknown vendor Unknown device (rev 1).   
         Vendor id=12b9. Device id=1008.   
         Medium devsel.  IRQ 10.   
         I/O at 0x1890 [0x1891].   
      
      
   (That vendor and device ID says it's a USR 5610 family.)  The script   
   detected the IRQ and "I/O at" strings, and plugged that data into a   
   'setserial' command:   
      
   [fermi ~]$ whatis setserial   
   setserial         (8)  - get/set Linux serial port information   
   [fermi ~]$   
      
   The need for this hack went away when the Linux kernel got to a version   
   5.0 serial driver, which knew about PCI port weirdness.  The box shop   
   finally released the script, but it wasn't needed any more.   
      
   The main point is that if you are able to get some hardware running   
   using a "stock" kernel (without a proprietary driver) in distribution   
   A, you should be able to get in running in any OTHER distribution   
   running a similar vintage kernel.  What it might be is a secret script   
   (as with the PCI modem above) or similar, but that's it.   
      
           Old guy   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)