[Wed Jun 29 10:42:57 CDT 2005]
Howard Fosdick writes a piece in NewsForge about the Rexx
scripting language. It is not so much the comments about Rexx that I find interesting as his thoughts on programming
in the object-oriented paradigm:
Python is easy to learn. But new users and occasional programmers don't think
in object-oriented terms. One must be taught to think this way. My friend, a
high school computer science teacher, tells me that new users state their
programming problem and then ponder the steps to resolve it. They don't
identify the objects of a problem space and then consider the methods they
need. At the end of the semester, a few advanced students love
object-oriented programming and are off and coding Java. The vast majority
would have been better served by having learned to script procedural solutions
for common programming problems.
I have great respect for the enthusiastic Python developers who tell me that
anyone can program in the language, and that ease of use is one of Python's big
advantages. But I believe they are wrong in thinking that even casual users
must love (or be made to love) the OO paradigm.
Yes, object-oriented is powerful, but it just does not come naturally.
{
link to this story}
[Wed Jun 29 08:57:02 CDT 2005]
Hannibal
reports in ArsTechnica about the latest Wired NextFest that
took place in Chicago over the weekend, and how disappointing the "future"
looks. Some of the paragraphs are quite funny.
In the future, the airport security checkpoint will look and function exactly
the same way as does now, except that the scanning technology that powers it
will be different. For instance, at the GE-manufactured checkpoint that I saw,
the machine supposedly sniffs you for bomb residue.
Interestingly enough, there was a long line of people waiting to go through
that checkpoint and be checked for bomb residue, which is something that just
baffled me. I mean, don't people dread going through the checkpoint at airport
security? Why voluntarily stand in line in order to pass through an airport
security scanner if you don't have to? It's not like the machine did anything
other than flash a little green light saying you were free of bomb residue.
Truly, the long line of people who just couldn't wait to go through that
security checkpoint was probably the most bizarre thing that I saw at the
entire NextFest. I wonder if it was a kind of programmed reaction like, "oh
look, a security checkpoint. I'd better get in that line and go through it.
Everybody else is." If that's the way we've all been conditioned, then I fear
for the future of the Republic.
(...)
Speaking of the elderly, the senior citizens of the future won't roll
around in wheelchairs —not even cool robotic wheelchairs like those
invented by Dean Kamen. Instead, they'll have robotic exoskeletons that
will make them much stronger and faster than the non-elderly. So in
addition to being the largest voting block in future elections, they'll
also have superhuman strength and speed. If I were a politician, I'd make
sure that the elderly of the future get great healthcare coverage, and I
wouldn't even think about doing anything to reduce their social security
benefits. You do not want to incur the wrath of our robotically
enhanced, geriatric overlords (or their Phillip K. Dick android companions).
In the future, most robots will look pretty much like the robots of the future
have looked since at least the 1970's. About the only difference is that
any antennae attached to a 1970's future robot were spiral shaped and had a
tiny ball on the tip. The current thinking is that future robots will have
straight antennae with no ball, and maybe a plastic coating instead of just
bare wire.
In other words, Hannibal has seen the future, and it ain't pretty.
I am
not sure what this means, but it is true that in the last two years or so
the future has become more and more prosaic. The excitement of the dot-com
bubble years have definitely worn out.
{
link to this story}
[Mon Jun 27 09:48:36 CDT 2005]
ITWeb tells us how the Linux desktop is being
slowed down, among other things, by a lack of Linux skills among those who
work in the IT. I suppose it makes sense, but what I found most
interesting of all was the words of Yossi Hasson, sales and marketing
director at Synaq, a Linux support and
services company:
"Because Windows is so simple, in a sense it is un-educating. Linux
training empowers people as it allows interested students to go beyond just
learning how to use the desktop applications".
This has always been my main argument in favor of using Linux in our schools.
What is our objective, teaching the kids how to use this or that application
or teaching them how computers (or the Internet) work? If the latter, there
is no doubt in my mind that using Linux would be the best choice so our kids
can learn not only how to run Microsoft Office or Internet Explorer, but also
how a web server works and the concept of network ports.
{
link to this story}
[Mon Jun 27 08:52:18 CDT 2005]
CNet News published an interview with Scott McNealy discussing the future
of Sun that includes a few interesting
comments:
What are out assets? We're one of three processor architectures
—Intel-AMD being one, Sun and Fujitsu's Sparc being another and Power
being another— that are going to survive. We have one of the two,
maybe three operating systems that are going to survive: Windows, Solaris
and maybe Red Hat. I'll be happy to compare Solaris vs. Red Hat.
(...)
I'll start with the vision. We believe we're moving out of the Ice Age, the
Iron Age, the Industrial Age, the Information Age, to the participation age.
You get on the Net and you do stuff. You IM (instant message), you blog, you
take pictures, you publish, you podcast, you transact, you distance learn, you
telemedicine. You are participatinig on the Internet, not just viewing stuff.
We build the infrastructure that goes in the data center that facilitates the
participation age. We build that big friggin' Webtone switch. It has
security, directory, identity, privacy, storage, compute, the whole Web
services stack. We build that infrastructure piece.
The vision is clear, and one would have to agree that our current social
trends definitely point to something like what McNealy describes here. Yet,
it is far from clear that Sun will be the only (or even the main) company
providing the infrastructure for that participation age. For one
thing, it is clear to everyone that we are moving towards it, and as a
consequence every other vendor is already positioning itself to be there
too. I simply cannot see why anyone would think of Sun as the only vendor
capable of providing that infrastructure. Likewise, it is far from clear
that those three architectures McNealy mentions will be the survivors in the
end.
He also had some words to say about Java,
perhaps their most popular technology today:
We are now at 2.5 billion Java devices on the planet —(including) 700
million cell phones, 700 million PCs. We had 17 million and 20 million
downloads in the last couple months of the J2SE environment. That is a
stunning number. The new Blu-Ray spec is going to put a Java virtual
machine in every new next-generation DVD player, and all your DVDs are
going to have Java bytecode on (them) that gets executed.
By the way, one of the readers of
CNet
News posted
a comment to the interview that does
indeed clarify some of the reasons why Sun (just like other high-end UNIX
vendors, just like
SGI) is in trouble:
Sun's hardware has always been {censored}ing expensive just because of the
Sun brand name. Sun continued that even after companies were getting rid
of Sun workstations in droves. Let's see here. A 3+ GHz (enter brand name)
PC/workstation loaded with a huge hard drive and gobs of memory for less than
$2,000 or a 1.2 GHz Sun Blade with 1/2 the memory and 1/2 the hard
drive for $5,000. Can you say "no brainer"?
Even now, a 650 MHz Sun Blade 100 **STARTS** at $1,350. I just within the
past two months built a crushing Athlon 64 3200, 1 GB RAM, GeForce
6600GT, and 360GB of total drive space for less than $1,000. If I built the
same kind of system that was 100% compatible with Solaris x86, I'd have a
homebrew system that CRUSHES current Sun Workstations in that price range.
Oh, wait —there ARE NO Sun Workstations in that price range.
In other words, no different than
SGI's
problem with its own line of visual workstations.
Let us face it: the
Intel- (and AMD-) based PCs running Linux or Windows have killed the old
UNIX workstations as we knew it.
{
link to this story}
[Thu Jun 23 08:22:26 CDT 2005]
Tom Yager
writes a piece for InfoWorld explaining why it is not so obvious that
Linux will end up displacing UNIX from the enterprise market and competing
against Windows on its own.
This business about Linux effortlessly coasting to a silver medal is crap.
It's based on the presumption that Linux is running for that position
unopposed; if Linux avoids getting creamed by Windows, victory is assured.
(...)
With Sun's recent delivery of Open Solaris, Unix, which Linux was created
to counter, has surpassed Linux. Solaris 10 and Mac OS X Server 10.4
(Tiger, whose open source counterpart is Darwin 8, based on BSD Unix) were
born as mission-critical, scalable, secure, stable, and portable OSes.
Linux grew into those roles, but while Linux was firming its foundation, Sun,
Apple, and, before Apple, NeXT were building on OSes that were already
bulletproof. Now that Solaris and Darwin are open source, they share
Linux's endearing traits and have the benefit of being developed and
supported in-house by the fourth- and fifth-largest makers of computer
systems (ranked by sales: IBM, HP, Dell, Sun, and Apple).
(...)
If I seem to be blending the open and commercial editions of Solaris and
OS X/Darwin too freely, I admit I see them as I see Linux. From IT's
point of view, OS freeware is a try-and-buy edition of the supported
commercial software IT trusts. Open Solaris is a lure for Solaris 10, just
as Darwin is a lure for OS X. Freeware Linux is the gateway to Red Hat
and Novell/Suse. Freeware always attracts developers. Linux didn't start
life that way, but it's in that position now.
(...)
... but now Linux's appeal is diluted by the opening up and broad
availability of enterprise Unix from Sun and Apple. So the formula
has changed: Closed Unix will lose market share, but Linux will not take
over those numbers and dollars by default. It'll have to fight for them,
and Sun and Apple are well-positioned to make IT consider open Unix
alongside Linux as it phases out its proprietary solutions.
The point is well taken. However,
we ought to remember in the first
place that without Linux those changes would have never taken place.
It is precisely Linux that built enough pressure to lead
Sun and
Apple to "open up".
Second,
I beg to differ with Yager's representation of free software
(he calls it "freeware") as a developer's toy or something that IT simply
uses for testing purposes. To start with, if that were the case Apple
itself would have not based its own OS on a "developer's toy". The reality
is that, according to his own words, this "freeware" is mission-critical,
scalable, secure, stable and portable. Now, those are some big arguments
for any IT department to consider Linux. Yes, they do care about support,
which is why the likes of Red Hat are still in business. However, there
are also many small and midsize business out there (not to talk about
governments) running "freeware" Linux. In other words, it seems to me
that Yager is too fixated on just one single sector (the high-end enterprise
market), and even there he does not take into account that there are already
companies such as
IBM, HP and
SGI selling and supporting Linux with
big iron.
{
link to this story}
[Mon Jun 20 21:11:10 CDT 2005]
Now, who said there was no such a thing as good geeky humor? How about this
Mail Reader Comparison?:
- Outlook: Always looks normal, then suddenly e-mails your 6GB PST file to
Russian spammers or Microsoft and deletes it.
- Thunderbird: Your choice of 42 fully themable crash messages in 30
languages
- Gnus: Crashes, after first consuming 2GB of system RAM
- Lotus Notus: Makes you wander through virtual filing cabinets for 2 hours
to find your e-mail, *THEN* crashes.
- Pine: Sends a satisfaction survey to the University of Washington every
time it crashes.
- Elm: Impossible to distinguish a crash from regular usage.
- Eudora: Crashing the same way since 1985.
- Mac Mail.App: Can handle up to 5 simultaneous animated segfaults at
once.
- Evolution: Crashes, but never really dies.... or does it???
- Hotmail: Crashes are "to serve you better".
- GMail: You must receive an invitation before we prevent you from deleting
your mail... and no, that is not Mr. Ashcroft at the keyboard.
- KMail: Pops up a new dialog box every 5 minutes to inform you that it's
crashing again, "just in case you didn't know".
- Mutt: Crashes can be removed by just the right .muttrc
- MH: Crashes, then presents the core file as new mail.
{
link to this story}
[Mon Jun 20 20:48:52 CDT 2005]
A quick note about a few pieces of software I just came across of while
browsing around this evening. First of all, SchilliX, an OpenSolaris-based live CD and distribution that
should make it quite easy for anyone to give Solaris a try. It should prove especially handy to software
developers who need to port applications. Second, Apachetop, a curses-based real time utility
to display information from a running copy of Apache. I still have not given
it a try because it only appears to be included in Debian's Sarge repositories, and I am still running Woody. Still, it
sounds quite interesting and it should prove very useful for web admins. Third,
if you are a fan of Apple's iCal, you will appreciate the Monket Calendar, an AJAX-enabled online calendar fully compatible with Firefox. Finally, while perusing the entries on
Planet Debian, I noticed a funny entry about cowsay:
# apt-get update && apt-get install cowsay
# cowsay "Cowsay is a pretty funny program. I mean,
it generates talking cows -- how could it get any cooler?"
________________________________________
/ Cowsay is a pretty funny program. I \
| mean, it generates talking cows -- how |
\ could it get any cooler? /
----------------------------------------
\ ^__^
\ (oo)\_______
(__)\ )\/\
||----w |
|| ||
# cowsay -f bunny "Did you know there's lots of other animals
you can draw with cowsay? Have a look at /usr/share/cowsay/cows
and use the -f option. There are even web frontends and a Wikipedia
article for cowsay..."
________________________________________
/ Did you know there's lots of other \
| animals you can draw with cowsay? Have |
| a look at /usr/share/cowsay/cows and |
| use the -f option. There are even web |
| frontends and a Wikipedia article for |
\ cowsay... /
----------------------------------------
\
\ \
\ /\
( )
.( o ).
Yeah, a total waste of time. It makes you wonder about the guy who wrote
the program, huh? In any case, if you still liked it and would like to burn
some of your own time, feel free to check out the
cowsay web front-end.
{
link to this story}
[Mon Jun 20 11:57:44 CDT 2005]
Sun has launched a Share
campaing recently where they show a picture of some of their key
employees together with a paean to innovation and how Sun has been sharing ideas with the community for years now. We may
think whatever we want about the prospects of the company, but one must admit
that they truly have innovated quite a bit in the last decade or so. One of
the pictures shows Bill Joy, Jon Bosak (XML) and James Gosling, among others.
At the same time, I recently read in Software Development Times that
Sun is
developing a new technical language, called Fortress, to replace Fortran.
Like I said, no matter what you think of these guys, at the very least one
has to admit they do work on some exciting stuff.
{link to this story}
[Mon Jun 20 11:23:20 CDT 2005]
Information Week has published a story about an interesting
alternative to offshoring: rural outsourcing. A company named Startup
Rural Outsourcing is pitching the idea of outsourcing jobs to workers in
rural America, who are still cheaper than those in the big cities. I ignore
how much this is helping to keep jobs in the USA, but the idea is intriguing.
{link to this story}
[Mon Jun 20 08:31:13 CDT 2005]
Now, here is a great story of entrepeneurial success in the open source
world (something we are not used to, since all we hear over and over
again is that there is no way anybody can make money out of it). Freelance software
developer Matthew Allum was "scratching an itch" when he created his Matchbox
Window Manager that has ended up being used by Nokia for its handheld
devices.
Allum became enamored with the idea of running Linux on a Compaq Ipaq in 2000
when he saw screenshots published by Compaq that showed the Ipaq happily
running Linux. He bought one and installed Debian, but found that a lot of the
Linux-based window managers didn't work with the small 240x320 display.
Frustrated, he "bought a book on xlib," sat down, and in 2001 wrote Matchbox,
a 50KB highly flexible window manager that depends only on xlib, which makes
it lightweight enough to run on small devices without using too many
resources.
Matchbox "stacks" open windows one on top of another and allows access to each
through the use of a drop-down menu on the title bar. Users can't move or
resize windows, which sounds restrictive, but actually works well because of
the extremely limited screen space on small devices.
Head to
the Matchbox Window
Manager official website for more information and some screenshots. It
looks surprisingly user-friendly in spite of its lightweight size and
simplicity.
{
link to this story}
[Mon Jun 20 08:23:32 CDT 2005
Somebody wrote a story in
KernelTrap about splitting swap to improve its performance that looks
quite interesting. Unless the swap algorithms in the Linux kernel are
somehow inefficient, this should not be the case. Yet, the reality appears
to be that it does indeed improve swap performance:
In any case, in 2.4, if I break up my 512 MB swap partition into 4 different
128 MB partitions, and then give them all the same priority, it changes my
swap-in wait period quite a bit. This is with all the swap partitions on the
same drive, so the limiting factors are not IDE transfer speeds here. With the
smaller partitions my max wait goes down from 60 seconds to 30 seconds (on the
old laptop), and the average wait of 30 seconds is down to about 10. That makes
it a lot more usable. It is much nicer. And on the old laptop there is no
possibility of adding RAM, which would help more if it could be done.
{
link to this story}
[Fri Jun 17 11:27:30 CDT 2005]
Apple's announcement that it would be releasing its MacOS X for the Intel
architecture (careful there, this does not mean you will be able to
run their OS on your plain vanilla IA-32 box any time soon, since it will
only work on specific "MacIntel" machines) got a lot of people excited. So
much so that David Kirkpatrick, who writes for Fortune, asked Michael
Dell about it and the Chairman of Dell
said that he would be willing to license MacOS X if Steve Jobs is OK with
it.
P.S.: by the way, check out Fortune's How Big Can Apple Get? article too now that you are at it.
{link to this story}
[Fri Jun 17 10:13:54 CDT 2005]
I read that Eric Raymond recently came up with a fierce defense of the BSD license while
giving a speech in Brazil, comparing it to the GPL license and stressing its business-friendly features.
His argument can be summarized as follows:
Freedom and choice are pretty cool. But we should talk about many other
things. GPL is based on the belief that open source software is weak and needs
to be protected. With it, we continue injuring ourselves, cutting ourselves
from the economic benefits of BSD license.
Of course, the first question we should ask ourselves is whether the open
source community would be as strong as Raymonds this it is now without the
help of the GPL license. However, Raymond's point is well worth some
thoughts.
Should the open source community adopt the BSD license for most of its projects,
many businesses out there (perhaps even including Microsoft) might have a vested interest in promoting it, therefore
investing in it and contributing to its overall stability and maturity. The
price we pay is that those same companies can take our software and make it
proprietary without any problems, which would most likely affect the chances
of free (as in beer) software out there. Yet, does it really matter that
much in a world where the likes of Red Hat
and SUSE are making it more and more
difficult to run their products for free anyways?. Sure, there are still
free distributions to choose from, but the BSD license would not preclude that
from happening either.
{
link to this story}
[Wed Jun 15 20:46:07 CDT 2005]
I just came across an interesting project. Koders.com aspires to be like Google to source code searches. Simply select
a programming language, a license type and enter a term to search for. Their
engine will return relevant files from open source projects that were added
to their database, displaying the source file in a nicely formatted page with
syntax coloring, line numbers and, most importantly, cross-referencing it
with other files from the same project. They also plan to release free
plugins for both Visual Studio
and Eclipse pretty soon. So, how
do they plan to make money? They will release a paid Enterprise Edition of
the software that allow companies to include their own code repositories in
the searches. Who knows? It may even work.
{link to this story}
[Wed Jun 15 08:47:17 CDT 2005]
NewsForge published yesterday a short interview with Linus Torvalds where they
asked him to compare Linux and BSD, and while he did not come up with any
groundbreaking news there is usually something to learn from his pragmatic
attitude. It also gives us a glimpse into what is it that has made Linux
so successful in the last few years:
Linux has a much wider audience, in many ways. That ranges from supporting much
wider hardware (both in the driver sense and in the architecture sense) to
actual uses. The BSDs tend to be focused in specific areas, while I have
always personally felt that any particular focus on any particular use is a bad
thing.
Which one is "better"? To me, Linux is much better, since to me, the
important thing for an OS is how well it performs under different patterns, be
they embedded, server, or desktop, or just some totally crazy person in a
basement trying something new....
To me, it's largely a mentality issue. I said "good enough," and that's really
telling. The BSD people (and keep in mind that I'm obviously generalizing) are
often perfectionists. They hone something specific for a long time, and then
they frown on anything that doesn't meet their standards of perfection. The
OpenBSD single-minded focus on security is a good example.
In contrast, one of my favorite mantras is "perfect is the enemy of good,"
and the idea is that "good enough" is actually a lot more flexible than some
idealized perfection. The world simply isn't black-and-white, and I recognize a
lot of grayness. I often find black-and-white people a bit stupid, truth be
told.
To me, this "grayness" (others may prefer to call it "flexibility" or
"malleability") is the key to Linux's success. There are, of course,
many other elements that also had an influence, but this single factor can
be pointed out as the key component of its sucess, I believe. It is, after
all, an OS that can be used as a server, desktop, for embedded devices,
appliances, live CDs, diskless routers and statiosn... you name it.
Today, NewsForge published the follow-up, interviewing OpenBSD's The de Raadt
and NetBSD's Christos Zoulas. In this case, it is Zoulas who came up
with the more insightful comments:
My biggest gripes about the NetBSD kernel are:
- Multiprocessing issues: although NetBSD supports multiprocessing and
has thread support, threads from a single process cannot use more than
one CPU. Only one process can be in the kernel at a time.
- No journaled filesystem or support for very large filesystems.
- General device driver availability.
Linux's code is much newer and it keeps constantly being re-factored. This
has the nice side effect of keeping the code simple and readable (at the base
system layers such as VM and FS), but stability is suffering. While 2.4.x was a
monotonic climb to stability, the road of 2.6.x has been very bumpy. My
biggest gripes about the Linux kernel are:
- OOM killer (memory leaks?)
- Filesystem stability: Linux has far too many filesystems and each
distribution is promoting a different one (for political reasons mostly,
not because of technical merit). Most of them support large filesystems
and are journaled. Unfortunately some of them are not safe to use, but
there are no true stress tests available to the general user population
to help them decide which one to use.
- General device driver stability.
In general, it looks as if the actual Linux and BSD developers have no
major issues among themselves, no matter how difficult it is to believe
when listening to the fanatics from both camps out there in the wild. By
the way, it seems to me that Zoulas is a little bit confused about the
nature of the so called "OOM killer", which kicks in whenever a Linux system
is running out of memory and needs to make a decision about what to do with
the processes that are hogging memory. It does not imply that the kernel
itself is ridden with memory leaks, as Zoulas appears to believe.
{
link to this story}
[Fri Jun 10 14:00:38 CDT 2005]
This week's big news was without any trace of a doubt Apple's announcement that it will release MacOS X for Intel chips
by the end of 2007. However, there is a catch: MacOS X will only run on Intel-based Mac systems. In other
words, you will not be able to install it on a regular plain vanilla Intel
box. According to Steve Jobs, they had been running alpha versions of the
OS inside Apple for a long time now,
and they had also made it part of the release process for quite some time
now to make sure that each and every software product they release ran just
fine on these systems. Now, while this may look like a revolutionary
announcement, the reality is that once we consider it at some length it may
not be such a big issue after all: what is the advantage of using MacOS X
on Intel instead of Apple's own PowerPC architecture? It used to be the
cost, but by now it is perfectly possible to buy a Mac Mini at a very decent price. Will I get better
performance? I doubt it. More applications? Nope. So, other than
increasing the chances of Apple surviving a potential death of their own
line of hardware, I just cannot see the advantage for the end users here.
Now, if they released MacOS X for regular Intel machines, that would be a
different story. By the way, the Motley Fools have an interesting take on the whole
thing. According to them, Apple's true objective is none other than
Dell. As I said, I do not doubt it is a
move that makes business sense. I am just not so sure there is much to be
excited about as an end-user.
{link to this story}
[Fri Jun 10 12:46:16 CDT 2005]
Russell Dyer writes a
quick overview of the first Red Hat Summit that also includes a few
interesting comments regarding the commercialization of Linux in the last
few years:
Linux purists have long been aware of this developing pattern. They regularly
guffaw at Red Hat, GNOME and other such commercializations of Linux and GNU
software. They stick with Slackware for their Linux distribution, Enlightenment
for window management and Emacs for text editing and even word processing.
Maybe I'm a little slow, but I'm starting to see their point of view and the
validity of it. If the big software companies are to take over the
revolution —as implied in Szulik's keynote comments— what will be
the results? Will they be what Linus Torvalds set out to achieve 14 years ago?
We seem to be long past that point. More importantly, will the many thousands
of volunteers that donated their time over the last decade or so have done so
in the end to make big corporations richer? Also, if we concede to the
overpowering marketing strategies and business savvy of the technology giants,
what will become of us? Are we simply to become their employees? Are our
opinions in the future to be written on cards to be dropped in company
suggestion boxes and thereby ignored? Or, maybe we will merely grumble for a
few decades until another Linus Torvalds comes forward and starts a new
revolution? I don't know what the answers are, and I don't really know what
should or can be done--or if anything needs to be done. I do think, however,
that we need to pay attention to what's happening to our revolution, and these
are the kinds of questions that should be discussed at a "summit" on Linux--and
the answers shouldn't be told to us by corporate executives.
It is an issue I myself have raised here a few times. It is not so much a
problem with
Red Hat or their attitude
at all. They are doing what any publicly owned company ought to do. As a
matter of fact, they are behaving much better than most other companies too.
Still, as Dyer asks, what will become of us if this trend continues?
It
seems clear that Linux and the open source movement are at risk of losing
their idealistic edge, and I am not so sure that is a good thing in the long
term.
{
link to this story}
[Fri Jun 10 11:35:04 CDT 2005]
Debian finally released Sarge (it was
about time!), but it has immediately hit a couple of big roadblocks: first
of all, a configuration mistake prevented it from using security updates by
default and then today we read reports of the upgrade breaking perhaps for
as many as 30% of the users. The picture is not pretty, that is for sure,
especially for a distribution that speaks so proudly about considering
stability as its first priority. All I can say is that, at least for the
time being, I'm sticking to my good old Woody, just in case.
{link to this story}
[Fri Jun 3 14:38:47 CDT 2005]
If you run Fedora in a network where
security is not an issue and need to configure rshRed Hat or even Fedora itself, but
version 3 broke it. What to do now? Well, I ran into that problem myself
when I needed to open up rsh so that NetWorker could do its thing. Here is the
solution, directly from one of the Red Hat mailing lists:
Ah, you need to add "promiscuous" to the set of options which you're
passing to the pam_rhosts_auth.so module in /etc/pam.d/rsh. The module
deviates from the historic behavior by NOT treating "+" as a wildcard
unless this option is given.
Note that if you run "rsh" without any arguments, it invokes "rlogin",
so you may want to modify the PAM configuration for rlogind similarly.
{link to this story}
[Fri Jun 3 11:17:02 CDT 2005]
Red Hat
released yesterday the directory server code they bought from AOL (i.e.,
the old Netscape code, part of their
Netscape Enterprise Server product, which as far as I know is the original
implementation of directory services). More information can be found on the
Fedora Directory
Server website. Apparently, it runs on Fedora Core 3, Solaris 8 and 9,
and also HP-UX 11i. Mike Ferris, Red Hat product marketing manager for
identity, security and systems management, made some interesting comments
regarding this release and Red Hat's
twofold strategy in releasing this product to the open source community:
One is certainly to build some community around the technology, to continue
its evolution and the creation of additional plug-ins and types of uses for
the directory itself. But secondly, and probably more importantly, to
encourage the adoption of LDAP as a standard for identity management
overall.
It definitely is in Red Hat's best interest to promote an open standard
such as LDAP in the enterprise,
especially now that Microsoft and
Sun appear to be working together on
interoperability issues and Novell has
its own directory services. Once again, it is not only Red Hat that
benefits from this move, but also the open source community at large. We
simply cannot let two or three vendors take over such a key component of
the enterprise network as the directory services. Interestingly enough,
Red Hat also appears to be preparing the release of a Fedora Global
Filesystem based on their
Global File System (GFS) product that they bought from Sistina. This
would be one more solid brick on the open source building.
{link to this story}