|
|
|
|
|
|
|
|
We would like to thank you very much for publishing (with a wonderful cover) our article ``Wireless Networking in Africa'' in the December 1998 issue. We have received many compliments and requests for more information. I wish to mention that Dr. Emmanuel Ekuwem's name was omitted in the acknowledgments and request you add his name to that section:
Our appreciation also goes to Dr. A. Nobile, one of the coordinators of the Programme, and Dr. E. Ekuwem, Mr. A. Maggi and C.E. Onime who are actively participating in this Project.This is very important to us as he has done a lot for the project and merits the acknowledgment. Thanks very much.
--
Enrique Canessa
canessae@ictp.trieste.it
Regarding Mr. Havlik's letter in the December 1998 issue complaining about the presence of advertising, it is easy for readers who don't like the ads to skip them. I personally value the ads. I want to know about the commercial products available for the Linux world, and I want to know about the firms that deliver these products. I vote for the ads.
--
Chuck Jackson
chuck@jacksons.net
I was just re-reading Mike Thomas' article ``Virtual Domains and qmail'' in the June 1998 issue of Linux Journal and noticed that the qmail_db_lookup script contains the test:
elseif ($db_local_address eq $recipient) {
which could fail if the values of the two variables do not have the same
case. This could happen, since e-mail addresses are case insensitive.
Assuming all the entries in the database table are lower case, I would do earlier on:
$recipient = lc @_[scalar(@_)-1];or something equivalent. Of course, defensive programming would require that we not make the assumption and change the test to:
elseif ("\L$db_local_address" eq $recipient) {
so one can sleep easy at night.
--
Gyepi Sam
gyepi@praxis-sw.com
I am writing to salute the support team at Xi Graphics. Never in my computing career have I worked with a more prompt or helpful support staff. Their response time is incredible; I submitted a Support Request to their web site and got a response in five minutes! Via e-mail, Bryan (the Xi Graphics support engineer) and I worked on a weird refresh rate problem I was having with Laptop Accelerated X. It took a few rounds of e-mail back and forth, but we got the problem straightened out and Bryan was patient and helpful through it all.
Three cheers to Xi Graphics support!
--
John Duksta
jduksta@bbn.com
I am disappointed that the first substantive mention of Linux for the Macintosh (``Linux for Macintosh 68K Port'' by Alan Cox, January 99) in the several months I have subscribed is an aggressively hostile article about several generations of old Macintosh hardware. Sure, people who know anything about Macintoshes will also know the current hardware and Linux picture for Macintoshes, but a very significant fraction, likely a large majority of the readers, will read the article as being less about the task of porting Linux (which it is about) and more about how inferior the Macintosh is.
There is a vigorous Macintosh Linux community, both an official Apple version (http://www.mklinux.apple.com/) and the even more popular LinuxPPC (http://www.linuxppc.org/). Current Apple hardware with Linux is as good as any. The author's second sentence, ``... Apple does not want other operating systems on its machines'' is refuted by the aforementioned official Apple Linux.
I would say more, but I am probably going to get relegated to the legion of Macintosh zealots anyway. Believe it or not, I don't want Apple or Wintel to take over the world.
--
Jeffrey L. Wragg
wraggj@cofc.edu
Alan did mention the many users of Linux on the Macintosh, and as
you point out, his
article was meant only as a report on how to port the kernel, not how to use
Linux on the Macintosh. We have had articles in the past on using Linux on
the Macintosh (see issues 19, 31, 37 and 45).
--Editor
I would like to join the chorus of voices of frustration with the ``Red Hat'' phenomenon. If you follow other UNIX journals, you will see a growing annoyance amongst UNIX users with Red Hat Linux. There does not seem to be any reasonable explanation anywhere for the strange Red Hat implementations. To quote Mike Borowiec on page 9 of the December 1998 Performance Computing Magazine: ``I've since wiped the stain of [Red Hat] Linux from my machine ...''
Mike's problems are the same as everyone else's in the UNIX community. Red Hat is unconventional in layout, difficult to install, extremely difficult to reconfigure and deficient in basic tools. The worst problem is that Red Hat requires extensive editing of C source code and rebuilding of the kernel. This is not a Linux problem--it is a Red Hat problem (other Linux distributions follow conventional UNIX methods). What is Red Hat's objective in making and promoting such a difficult implementation? Where are we heading if Red Hat is creating such a negative reaction against Linux among professionals? None of it makes even the slightest bit of sense to a UNIX user.
The question you need to answer for us (since we cannot figure it out on our own) is: why is Red Hat doing this to the Linux community? And why are so many industry and media types (including your magazine) getting behind Red Hat? Assuming that what Red Hat is doing makes sense to somebody, we readers really ought to have some kind of explanation so we can understand and evaluate what's going on.
--
Reilly Burke
reilly@aerotraining.com
Sorry, you will have to ask Red Hat about their policies.
I am not in their confidence. However, Red Hat does seem to be the most
popular distribution available, so they must be doing something right.
--Editor
Thanks! Jawed Karimi's September 1998 LJ article (``Porting MS-DOS Graphics Applications'') on basic graphics got me going in the right direction with SVGALIB. (I am a Linux newcomer and had never heard of it before then.) Sometimes, short and sweet articles are the best.
--
John E. McGovney
compgraf@bellatlantic.net
I'd like to mention that I disagree with author John Blair regarding his article ``Samba's Encrypted Password Support'' in the December 1998 issue.
In the last paragraph, he states ``Finally, to allow users to update their encrypted password, set the permissions on smbpasswd to be setuid root ..''.
This should not be necessary, since smbpasswd acts like an SMB client, asking over an SMB session to change a user's password. This is the reason you should include in smb.conf the ``allow hosts = 127'' line, or leave it blank.
I have also used pam_smb, which can be found on the main Samba FTP site, so that other services can authenticate with Samba. On my system, my users all have their passwords in /etc/shadow ``starred'' (*).
About the magazine, it's great!
--
Celso Kopp Webber
webber@sj.univali.rct-sc.br
The ``free'' controller shipped with a lot of scanners is the Symbios 53C400A. It is not, as the writer of the reply (``Best of Technical Support'', January 1999) suggested, 53c8xx compatible but is instead a 5380 variant.
It is supported by current Linux 2.1.x series kernels, although its lack of an IRQ and 8-bit ISAness make it a good candidate for the bin, not a multiuser OS.
There is an equivalent patch for 2.0.x in the SANE mail archive.
--
Alan Cox
alan@terrorserver.swansea.linux.org.uk
I think the supplement is a very good idea-I very much enjoyed this special issue.
I would like the addition of a Table of Contents as in the regular issues. This would be helpful because I (and perhaps others) photocopy the Tables of Contents and put them in a three-ring binder, where they are easier to scan than the issues themselves.
Thanks for an excellent publication.
--
Ray Liere
lierer@cs.orst.edu
One of the weird rules of the Post Office is, if it has a table
of contents, it cannot be called a supplement. Go figure.
--Editor
I just wanted to say ``Thank you'' to all at LJ who did such a great job presenting my article in February. The magazine keeps getting better and better.
By the way, my bio states that I maintain the ``official'' Linux Csound, but that is no longer true. As of late 1998 I have worked mainly with the development version available from Nicola Bernardini's site at AIMI. The official version is currently maintained at Bath by John Fitch.
Best of the New Year to all of you!
--
Dave Phillips
dlphilp@bright.net
For years I've maintained a dual boot machine at home with both Windows 95 and Linux because my spouse didn't feel she could learn the intricacies of Linux. So, it was a great pleasure to see KDE chosen in your 1998 Reader's Choice poll. It was with KDE and StarOffice that I finally won her over to the good side.
The ease of a standard desktop which was more intuitive than Windows 9x has her singing its praises. I thought I would never see the day. It should be pointed out that a big turning point for her was the day she accidentally started three different StarOffice sessions and was amazed that the system didn't crash. It did take awhile for those monsters to load, but once they finished, she was back in business. Windows 9x would have died at this point.
Now all I need is to find a calendar program which comes close to Calendar Creator Plus and Windows 9x will just be a bad memory in our house. Maybe korganizer will fill the bill when it's completed.
--
Randy Kyrk
rkyrk@anet-chi.com