BYTE interested in Linux?

BYTE interested in Linux?

Post by Doug Ridgw » Wed, 13 Nov 1996 04:00:00

I was wandering by Byte's web site, and did a search on Linux.
There was a fair number of hits, but the November 96 / Web Project item
was interesting ... I quote a bit here:

``Plugging in the Linux Box

A few months back, analysis of the keywords used to
search The BYTE Site revealed that Linux ranked fifth [An earlier
item said third, after Microsoft and Java. Doesn't really matter...].
Clearly, a lot of visitors knew something I didn't, and
I resolved to find out what.

Today, a P150-based Dell running the Caldera distribution
of Linux is an increasingly important pillar of The BYTE Site. All our
conferences run there; INND supports newsreaders, and
Apache provides an alternate Web-based view. I transfer files between
Windows NT servers and Unix servers using Samba, a
nifty SMB utility that makes the Linux server look
like an NT peer file server.

And when I found that the key component of Metasearch
(a Perl 5 module called didn't like my NT Perl setup, I didn't
waste time figuring out why. (A lot of Perl tools,
though in principle are portable, in practice work better with
Unix.) It was easier to
build and run Metasearch over on the Linux box, so I did. ''

A few points:

  Linux is being used in conjunction with other systems. Interoperability
  is key. Whoever plays well with the most others wins.

  Computer journalists often do take long term looks at things. Depth
  counts for a lot when you use something heavily for months.

  And finally -- it's amusing that the article described doing exactly
  what I was doing when I found it: searching Byte on ``Linux''.




1. Q: How do I read linux dirs byte-by-byte?

I've recently written a simple program to make hexadecimal dumps of
files. (I know there are probably others out there, but I was rusty and
needed the exercise.)

I uploaded it from my 386 running linux to a academic DEC running OSF,
and my program functions identically in both locations, except for
reading directories. On the DEC, the directories are displayed
byte-by-byte as expected, just as an ordinary file. I see lots of
miscellaneous characters representing the i-node numbers and so forth,
and the file names. On my 386, I seem to be hitting EOF immediately.

I've read the man pages for opendir(), readdir(), etc, but those seem to
return the directory information as directory entries, not as a raw
character stream. Can somebody tell me what I'm doing wrong or where to
find the right functions?



   "Paths don't work. The traveler who walks the path is what works."
                  -- Brigit Brat, God's Girlfriend --

2. fonts & "scalabale alias"

3. Sherlock linux plug-ins / sherlock for linux?

4. groff

5. RAW mode & byte-by-byte I/O

6. Cloning F50 to F80?

7. 4 bytes from 6 byte mac address

8. Intel EtherExpress PRO 10 NIC, please help

9. Byte 4 byte copy of DAT tape? how?

10. FreeBSD File System 512 byte-blocks --> 1024 byte-blocks

11. Accessing memory addresses byte-by-byte

12. Serial IO Byte by Byte

13. List of plug-ins for linux/netscape please!