[linux-elitists] Are we Dead Yet? (or "For every sprinkle I find, I shall kill you!)

Greg Folkert greg@gregfolkert.net
Fri Jan 21 07:46:05 PST 2005


On Fri, 2005-01-21 at 08:57 +0000, Phil Mayers wrote:
> On Thu, Jan 20, 2005 at 09:34:21AM -0500, Greg Folkert wrote:
> >
> >I am just wondering how other list members felt about 2004.
> 
> [Hmm - message rejected by filter match, I was under the impression 
> Thunderbird was Free. Oh well]
> 
> Pretty good, though some of these are a bit nebulous and some even 
> contentious to "power users" (Gnome? :o). And obviously it's from my 
> personal PoV and thus does not mention e.g. the good progress the *BSDs 
> have been making

I have not been tracking the *BSDs as much the past coupla years. Takes
more time than I have to just keep up with the FOSS+Linux and SCOG
stuff. It is a shame really, I know some of the perennial issues with
some of the *BSDs have been resolved recently; SMP, Thread speed
improvements, Streams, ...etc. *BSDs mostly had superior implementations
of these elements than other OSes.

> o) 2.6 kernel with some very useful stuff (udev+hotplug, sysfs, better 
> power management, better IPv6, IPSec, SElinux) and the first round of 
> 2.6-based distros (see below).

Again, hit the good buttons.

> o) Ongoing maturity of desktops (Gnome, KDE, XFCE as the primary trio) 
> and the applications (OpenOffice became useful, Firefox and Thunderbird 
> to 1.0, Evolution and the Exchange Connector being FOSS)

Of which some think TOO MUCH progress towards "Sane Defaults" has been
made. Restrictive to what you can change easily. Ehh. If it works, it
works. I like them, mainly Gnome for me. GTK just feels cleaner to me.

> In particular, the freedesktop.org portfolio including thing like HAL 
> and interoperable session management

HAL+udev+hotplug+discover+usbstorage+gphoto == *THE* Shizzit 

From rom more than one standpoint; HAL needs much refining yet to make it
easy enough for everyone to use it. But for now, it makes it much less
of a headache for me and many others to deal with camera, USB Fobs,
Firewire devices.

> o) Early progress in some promising directions e.g. desktop search - 
> Beagle/Dashboard, it's Mono dependency aside; and Mono itself, which is 
> less of a conflicting viewpoint than it might seem.

I haven't seen any real issues with Mono and Dashboard. Mono, hopefully
will preempt Microsoft's setting of the bar, as they use a crooked bar,
FOSS people tend to set a straight and simple bar without strange twists
and turns.

> o) Strong progress in some server environments such as Apache, 
> mod_LANGUAGE, general Perl/Python/PHP progress, Zope, Twisted, etc.

Twisted == Python based
Zope == Python based
Mason == Perl based
Everything == Perl based
TomCat == Java based
JBoss == Java based
<insert ones I forgot> == <something> based

Funstuffs. I have am support at least two of those right now. Typically
the are (from and machine admin perspective) Setup the Service and
Forget about it. I am truly amazed at how flexible these kinds of things
really are. Better than many "Established" competing Commercial Products
I can think of. 

> o) People started to realise that ActiveX and so forth were a really, 
> really bad idea, and thus that locking into an IE solution was worse 
> than worthless, it was actively harmful

You hit someone over the head enough times with a dead Mackerel they
usually notice.

> o) Related, the rise of very credible XMLHttpRequest-based webapps like 
> gmail, google autocomplete, etc. - and frameworks to implement such like 
> mod_pubsub, Twisted's nevow+LivePage (check out the chatola sample 
> application - it's a *sample* and in some ways more promising than 
> thick-client IM systems!)

These scare me. But scaring me, typically makes me really try to
understand the implications, benefits and problems that are there. Good
case of that: Microsoft's Passport.

> o) Again related, the web standards wrangling is more or less over for 
> the current generation. XHTML+DOM+CSS is finally a credible platform to 
> implement serious applications on, and if the SVG folks can correct 
> their mis-steps, we're gold. Care is needed to not be surprised by XAML 
> or whatever, but for the moment I'm pleased.

It really has been credible in certain circles, for a while. General
Public (well IT Public that is) consumption is starting to come forth
and show the effect of hard work and effort. There is no magic bullet in
these kinds of things... but a Teflon Coated, Armor Piercing, Hollow
Point bullet sure does a lot of damage in that area.

> And most amazing of all
> 
> o) In stark contrast to a couple of years ago, time-to-crack for an 
> average Unix system went up to 3 months (I'd seen instances of 20 
> minutes ~2000/2001), while that for XP dropped into the single-digit 
> minutes range. Partly of course that's based on economies of scale for 
> the new, organised cracking botnet "salesmen", but bear in mind: For 
> that period both Windows and Unix systems have been fixing bugs, but one 
> has got better, the other worse

This is NO accident. Believe it or not, SCOG has something to do with
this. SCOG yelling FIRE in the theater about 2^256 times has cause quite
a few people to step-up and hit those marginal areas with much gusto.
And more companies are actually stepping up and literally giving more
support for code auditing and coding for many, many projects.

> Less obvious happenings include the progress on Xen and virtualisation 
> in general; recent results showing 2.6 scalability to 64 cores for some 
> workloads; various vendors shipping not just FOSS-based solutions, but 
> solutions that aren't crippleware e.g. you *can* access and modify the 
> underlying systems, and warranty aside are often encouraged to do so.

Schway stuff that is for sure, tells us we are coming along better than
our commercial counter part for the x86 hardware.

I hope other share what they feel was bright about 2004.
-- 
greg, greg@gregfolkert.net

The technology that is
Stronger, better, faster: Linux
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://allium.zgp.org/pipermail/linux-elitists/attachments/20050121/0abeb803/attachment.pgp 


More information about the linux-elitists mailing list