The Grumpy Editor's guide to 2005 [LWN.net]

An inevitable part of the new-year ritual is the posting of predictions for the coming twelve months. Your editor, having access to a moderately high soapbox, feels it would be morally wrong to fail to make use of that soapbox to make an absolute fool out of himself. To that end, here are a few ideas for what we might see in the coming year. As always, these predictions are offered in the hope that they will be useful, but they carry NO WARRANTY regarding any correspondence with reality as experienced in your timezone or as to whether they make sense at all.

Development

This will be the year for free desktop infrastructure. Yes, there will be a long series of high-profile application releases, with OpenOffice.org 2.0 being, arguably, at the top of the list. But 2005 will be the year when projects like HAL and D-BUS stabilize and see wide deployment, and when the reinvigorated X.Org development team starts making some truly big strides. The kernel's support for contemporary video cards will be rationalized and improved. Perhaps there will even be a place for Mono. The convergence of all this new, low-level support code, combined with increased cooperation between desktop projects for low-level support, will build the base for the next generation of amazing free desktop applications.

Free databases will see some high-profile deployments. The adoption of free database management systems is still in an early stage. Things will progress in 2005, to the point that some proprietary database vendors will see the need to start competing directly against the free alternatives. Perhaps 2005 is when we'll see some real free database FUD.

There will be no 2.7 kernel in 2005, despite the requests for such a release from some quarters. The 2.6 process will continue to merge changes at a staggering rate, and nothing will come along which is so disruptive that it forces the creation of a new development series. The steady series of complaints about the quality of the 2.6 mainline releases will force some changes to the process - we may see more frequent releases or true "release candidates" for wider testing. But the simple fact is that the kernel developers - and the distributors who have the job of delivering stable kernels to their customers - are happy with things as they are, and will not be in a hurry to go back to the older way of doing things.

Commercial

Red Hat will find something to do with its cash pile. The company currently has about $1 billion (almost half of its market capitalization) in the bank - much of that cash is the result of a debt sale one year ago. As Red Hat's management tries to push the company's stock price back up, it will have to find something more productive to do with that money. It would not be surprising to see an acquisition or two happen in the near future.

The market for not-quite-enterprise distributions will grow. There are no end of companies looking to gain the benefits of switching to Linux, but who do not want to pay the hefty "enterprise Linux" price tag. Many of these companies will realize that high-quality Linux can be had for less, and will look to companies with credible support offerings. Companies like Progeny, Ubuntu, and Specifix may be well placed to thrive in this market. The UserLinux distributed support network model looks an awful lot like the early Red Hat "support partner" program, and risks ending up the same way.

Embedded Linux will gain a higher profile, especially as a base for a new round of "personal media player" gadgets. Expect some fireworks as some of these devices - and their built-in DRM schemes - prove to be more hackable than the entertainment industry would like.

Very few companies will buy Linux indemnification policies, making life difficult for insurance vendors like OSRM.

Distributions

Debian will get a new stable release out, one way or another. Much of the user base for stable Debian releases will, however, have moved on to offshoot distributions like Ubuntu. There will be a new round of soul-searching within the Debian Project over the value of its stable distribution and what that distribution should be.

Community involvement in Fedora will increase, mostly through outside maintenance of some non-core packages. Red Hat will maintain a firm grip on important decisions, however. Don't expect to see an open Fedora developers' conference in 2005.

Legal and political

Thanks to serious activism and the entry of several countries into the EU, software patents will not be enacted in Europe in 2005. One thing your editor has seen many times, however, is that the commercial forces behind this kind of legislation do not ever give up. While their current push looks to be headed for failure, the issue will remain, and the fight will go on.

A new round of copyright legislation will hit the U.S. Congress. The entertainment industry will attempt to strengthen its control and find some sort of legislative solution to file sharing over increasingly decentralized networks. Fair use activists will try again for copyright and DMCA reform. Neither side is likely to get far. The entertainment industry may get caught engaging in increasingly dirty denial of service attacks on peer-to-peer networks and their users.

This one should be fairly obvious: 2005 will see the end of SCO. The company's remaining cases will fall apart in court, and its cash will run out. In retrospect, it will become clear that the SCO lawsuit has actually been a good thing for free software: it has proved how clean our code is now, made developers more aware of the potential for such lawsuits in the future, and has made many large companies take a clear position in the defense of free software. The next company that tries to extract payments from the free software business world will find a climate which is far less hospitable to that sort of litigation; for this reason, your editor believes there will not be a new major intellectual property suit related to Linux in the coming year.

In conclusion...

More people will notice that Linux users don't have spyware and adware problems, which will be getting steadily worse on other platforms. This issue, alone, will cause more people to look at free software. Many will get their feet wet with Firefox and stop there, but others will take the full plunge. As proprietary systems are turned into zombies which spam and spy on their alleged owners, pure exasperation will push a new round of Linux adoption.

Your editor expects many things to continue as they have been. An increasing number of developers will work to create ever more powerful applications. More and more people will awaken to the value of free software, and they will look seriously at using it. Some people will even figure out ways to make money from it. And, inevitably, Linux will continue to be fun - even for a grumpy editor.

(Log in to post comments)

The End of SCO

Posted Jan 6, 2005 7:38 UTC (Thu) by LinuxLobbyist (guest, #6541) [Link]

Oh, I'm sure that there's some Canopy company out there willing to be it's first victim^Wcustomer. Unless, of course Mr. Mustard puts a stop to that kind of nonsense. ;-)

The End of SCO

Posted Jan 6, 2005 12:57 UTC (Thu) by Duncan (guest, #6647) [Link]

While some might bite, it'll be pretty limited to a subset of current
customers, I'll predict. As the legal side folds, there will simply be no
reason for uptake by anyone else, and all sorts of reasons /not/ to.

Consider for a moment the situation SCO is in. They've demonstrated that
they won't hesitate to cannibalize their own licensees, so for new
customers to touch what has become the third rail of Unix everybody knows
will be essentially begging SCO to sue them. At present, the question of
whether Novel actually turned over the Unix copyrights and etc is still up
in the air, but all the legal opinion I read says there's at least a 50/50
chance that Novel still owns it. That doesn't leave SCO in a good
position there relative to others (say SUN), altho to be fair, there's no
question that SCO /does/ have the right to use the Unix IP for its own
work.

The effect of the USL/BSD settlement becoming public shouldn't be ignored
either. Basically, what Novell bought and /might/ /have/ transferred to
SCO was certain legal rights as pertain to UNIX licensees. The terms of
the settlement make it /far/ more difficult for them to claim violation
from those who /have/ no such licenses, making anything related to a Unix
license even /more/ difficult to sell, now, particularly to those not
already caught in the net of current licenses. Any new customer buying a
Unix license at this point /has/ to have rocks in their collective head.
Anybody with an ounce of sense not already caught in the net, had best be
staying /far/ /far/ away from it. That and there's already precedent for
SCO foisting off licenses on customers doing unrelated business, such that
said customers claim they weren't even aware of purchasing them. Anyone
doing business with SCO therefore has to be worried about unwittingly
being caught in the UNIX licensing net and therefore becoming subject to
suit based on violations thereof, even when they /thought/ they were
buying something entirely unrelated. This is /not/ a good position to be
caught in, for either party.

SCO's Unix has always been an x86 play -- they don't have their own
hardware as does SUN -- and in all honesty, there's nothing left of the
x86 Unix market save for an ever smaller legacy installation set.
Everyone's turning to Linux or one of the other open source *ix players
(the BSDs, now SUN Solaris as well). Those who in times past might have
gone for SCO's offering will probably go for Solaris instead. Otherwise,
they'll either go for Sun or IBM "big iron" on the high end, or go Novell
or Red Hat for Enterprise Linux on x86. Of course, there are other
offerings available below that, from a number of other suppliers, but
while they might have been SCO customers in the late 80s and 90s, they
really aren't in the SCO market at all now.

That pretty much leaves... an ever shrinking current customer base as the
market for Legend or anything else SCO may come up with. As LWN has
predicted, SCO money will probably run out before it could really take
off. What that means is that currently, this product serves as a way for
SCO to continue to claim it hasn't put all its eggs in the litigation
basket, while in the future, it's likely to be the one asset with some
arguable value left, to pass on to whoever gets the SCO remnants when it
dies. Thus, what is Legend from SCO today will most likely be wrapped
into some other company's product in 2006. Who'd step up to by the
remnants as SCO folds? Maybe it'll pass to some other Canopy group. Or,
maybe SUN will buy it. Or... maybe some other independent lawyer group
might see some IP or other still there and worth litigating over, and will
buy it to expand their litigation only IP holdings.

My prediction, FWLIW...

Duncan

The Grumpy Editor's guide to 2005

Posted Jan 6, 2005 14:08 UTC (Thu) by Duncan (guest, #6647) [Link]

OK, I might step on some toes here, but of the areas I know enough about
to make any educated guesses of my own, the two hardest to call here might
be the no 2.7 kernel series (I agree out thru say 3Q, but think something
might develop by 4Q), and that Debian will actually get a stable release
out (it has been delayed so often, will this /actually/ be the year it
happens?). Still, assuming Debian does get one out, the rest of that
prediction isn't even really a prediction -- the delay has been so long
that the relevance of Debian stable has been in question for some time.
Perhaps they will end up doing basically what the kernel did, dumping the
idea of totally stable, while leaving ultra-stable support, for those
nervous about the stability and testing of a more fast-moving product, to
the distributions based on the product. As with the kernel, there are
enough distributions based on it to make that at least an option.

Back to the kernel, I think it's possible there'll be something big enough
to warrant a 2.7 (under the new definition) by 4Q2005. Again, as those
following kernel developments certainly already know by now, this 2.7
won't be a traditional development kernel on the order of 2.5, but would
be a branch to develop a specific, very kernel code invasive, feature.
Changes other than that feature would be synced between the two kernels
very frequently, such that said developing feature was the only difference
and backporting to the stable kernel wouldn't be an issue. Upon
maturation of the feature into stability, it would be merged back into
mainline which then might become 2.8 or 3.0. Alternatively, if the
feature proved a bad decision, 2.7 would be scrapped and we'd continue
with 2.6. While this certainly isn't a traditional development kernel
2.7, a lot can happen in 9 months, and it's just possible something like
this might. Thus, predicting that it won't is IMO a pretty bold
prediction, even if the likelihood is only perhaps 10-25%, which is where
I'd put it, because it could seem so obvious in hindsight.

Many of the predictions seem fairly safe. That xorg will break into its
own after a year of coming to terms with the state of things when it
originated, given that there has already been a release with /some/ new
code, seems pretty safe, even if /very/ significant. Desktop
infrastructure such as HAL and D-BUS finally becoming semi-standard seems
pretty safe as well, given they seem on the cusp of it at this point and
that's what everyone is already planning for. Still, again, despite it
being obvious, its actual occurrence is quite significant.

That Red Hat will find something significant to do with its cash pile also
seems obvious, tho everybody's kind of in suspense as to /what/ that might
be. Anybody care to predict that what was marketed as simply maximizing
their ability to take advantage of any opportunities that present
themselves might remain that way thru the entire year -- that they /won't/
find such a natural opportunity of any significance? Now /that/ would be
a bold prediction, even tho no one has any idea what they'll do with that
money at present.

The personal media player prediction seems one of the more fascinating in
what it might bring in terms of SCO level legal "entertainment". Of
significance this year is that SCO and such direct legal threats (even
from MS) seem less of a menace, while the greatest danger seems to be
indirect, patent and other IP legislation changes. Of course, at the same
time, these present the greatest opportunity for positive change. As SCO
was the area of potentially greatest danger last year, changing IP laws
must be the greatest danger yet opportunity this year.

Of interest to me because I predicted it and switched to Linux and
libreware as a result of that prediction, the spyware/adware problem is
significant in its mention. Back in the lead-up to what became eXPrivacy,
I found I could not personally accept the precedents set by the
anti-privacy/anti-piracy forced registration. To me, that set a precedent
in disrespect for the user that I feared to think of the consequences of.
There were all sorts of reasons not to do it (including the fact that to
my mind, there was no clearer demonstration that MS /knew/ it was a
monopoly, since it had time and again demonstrated it was to smart to
shoot itself in the competitive foot like that when there was viable
competition around), and no valid reason to actually /do/ it, save to
demonstrate that it /could/ do it, by shear power of its monopoly, and
that users (to its way of thinking) had no viable recourse at all.

Several years later, and that disrespect for the user being master of his
own system has lead to everyone and their brother trying to get a piece of
the malware action. I jumped off the MS platform because they pushed me,
by pushing terms I could not and would not accept. Now, I look at all
those that didn't jump, and see what they are subject to, and am glad
I /did/ jump. MS opened the gates and gave its official blessing, and
it's only going to get worse for those remaining on the platform.
"MSWormOS" indeed, never could I have dreamed that nick would be so
apropos when I coined it.

What with that for the home crowd, and licensing 6.0 for the corporate
crowd, I couldn't believe my eyes how bumbling MS was being at the time.
They had the market monopoly, and must have truly believed that made them
invincible. Now their actions are coming home to rest, and now with
Firefox, for the first time in desktop history AFAIK, MS is losing market
share in real, measurable, numbers. This prediction may prove I'm an
optimist, but I predict this is the first of a long line of similar
numbers, that computer historians will point to the MS actions with XP and
Licensing 6.0 as the turning point, with spyware/adware the ultimate
result of the policies of eXPrivacy, and the Firefox/IE share numbers that
came out at the end of 2004 the first real numbers indication of a
monopoly in decline. Linux was the right alternative at the right time,
they will say, as MS' numbers continue to plunge past Linux parity... tho
obviously that won't be /this/ year, but I predict we /will/ see a few
more numbers indicating this shift, /this/ year. That's MY bold
prediction.

One more thing. The thing that took everybody by surprise last year was
the xfree/xorg development. In hindsight, the signs were there, but
either nobody saw them, or nobody was bold enough to make the prediction.
Anybody care to make a bold prediction of an "upset development" for this
year? I'm afraid I don't see anything /that/ big /I/ can predict.

Duncan

Kernel forking

Posted Jan 6, 2005 16:45 UTC (Thu) by farnz (guest, #17727) [Link]

I disagree on when we'll see a 2.7; if a feature is invasive enough to need 2.7, it'll probably be so invasive that syncing between 2.6 and 2.7 is very difficult. Anything else is likely to be developed in a -mm or -ac style branch, then merged back.

MS market share

Posted Jan 11, 2005 22:07 UTC (Tue) by roelofs (guest, #2599) [Link]

now with Firefox, for the first time in desktop history AFAIK, MS is losing market share in real, measurable, numbers.

Perhaps, but keep in mind that it's a free-as-in-beer market, so it's not really in the same class as if they were losing OS or office-suite market shared (i.e., their cash cows).

Indeed, I'd argue that their ongoing losses in the web-server space are far more interesting. (Of course, that's not on the desktop.)

Greg

MS market share

Posted Jan 12, 2005 11:59 UTC (Wed) by Duncan (guest, #6647) [Link]

Well, you make a good point, but keep in mind who it was that /made/ it a
"free as in beer" market, the investment they've sunk, and the integration
they've achieved. If there's one thing we know about MS, it's that it did
NOT do all this "out of the goodness of their heart"! What do /they/
think about the browser?

1) It might not be MS Office, but keep in mind how much elements of MS
Office depend on IE and that integration.

2) Wean users of that integration, and you've already gone a good way to
prying them out of MS' grasp, OOo may be next, then it becomes
increasingly easy to jump off the platform altogether, particularly for
those many corporates where the browser/internet and Office suite are
really all that functionally matter on 90% of their desktops. At that
point, the increased stability and security of Linux (plus the multiple
vendor sources = no lock-in thing) begin to loom ever larger in their
sights, because the functionality of the apps is no longer locked onto one
platform, and they /can/ look elsewhere. Employees take OOo and FF home
with them, first to use on their own MSWormOS desktops, then eventually
they too consider the switch to Linux not that big a deal, after they see
that it too has a nice point and click interface, unlike the command line
they were scared of years before when they first heard the term "Linux".

BTW, it seems some mainline sites are quoting Firefox in the 20s percent
now and still rising, with IE 70s and dropping. When I first posted, FF
was in the teens and the IE drop was still single digits, just barely out
of the 5%-ish margin of error, with its share still in the 80s. The FF
share has doubled since then, nearly all of it coming from IE. It's not
coming from the others (like the Konqueror/KHTML rendering I use, I'm not
just a FF zealot by any means) as they've hardly budged.

Duncan

The "SCO" sources

Posted Jan 6, 2005 17:34 UTC (Thu) by vonbrand (subscriber, #4458) [Link]

By now, the Unix source code is of purely archeologic interest. What matters new developments inside that code (developments that by no stretch belong to SCOX), like journalling filesystems, SMP, NUMA support, drivers for sundry exotic (and not so exotic) iron.

Nope, it isn't worth much. SCOX has done no real development on that, whatever was useful in there has long been taken over (or surpassed) by *BSD and Linux. It is said that IBM's AIX has almost no original Unix code left in it, and it sounds credible. If/when Sun opens up Solaris, it would be interesting to let the code comparators (developed for the SCOX nonsense) loose on it vs the (open-sourced) ancient Unix versions...

The "SCO" sources, "archeologic interest only"

Posted Jan 12, 2005 11:34 UTC (Wed) by Duncan (guest, #6647) [Link]

I'd tend to agree. Like it or not, wherever you are in relation to FLOSS
and Linux, in the *ix universe, Linux seems to be fast replacing any form
of Unix as /the/ reference implementation. SCOG, SUN, IBM/AIX, the BSDs,
they /all/ make a pretty big deal of their Linux compatibility in some way
or another. Why? Because the simple fact of the matter is, Linux now has
momentum behind it like nothing else, proprietary or libre/free/open, and
pointedly including not only *ix but MS (which has pretty much stagnated,
by virtue of the fact that where it is strong, there's nowhere else left
to grow, and where it isn't, it just doesn't seem to have had as good a
luck lately, probably because everybody knows how they play and is wary
enough to keep them out of monopoly contention in any market they don't
already super-saturate) as well.

It's scary in many ways for the uninitiated, because Linux looks about to
steam roll everything just as MS did. However, the big and defining
difference, and really, the only reason Linux has been able to build the
momentum it has, is because of that copy-left license -- the distribution
model and license means it's very much the /user/ in charge, and enough
big companies have realized that fact makes Linux the one and only
really /safe/ investment because it cannot be co-opted, that the Linux
movement is quickly not just "snowballing" but "avalanche-ing" (that is,
becoming many snowballs, each of which is snowballing, compounding the
effect).

Duncan

How about a Ubuntu feature?

Posted Jan 12, 2005 11:11 UTC (Wed) by Duncan (guest, #6647) [Link]

Ubuntu flown under the LWN radar? Are you /kidding/? For its age, it's
doing /amazingly/ well PR-wise, in the entirely Linux community including
on LWN. Go try the LWN search. Excluding comments and security alert
and vuln references (but including security updates as there seems no
quick way to exclude them), Ubuntu is /already/ mentioned in 100 articles,
including this one. I dare say that's a better record for its age than
probably any of the other distributions, including the big five
(SuSE/Novell, RH, Mandrake, Debian, Slack, in the order they popped into
my head), which now should really be the big seven, including Gentoo and
Ubuntu.

Glancing at the list, LWN is now announcing the regular Ubuntu traffic,
has coverage of two December community meetings, is mentioned in the Sept.
and Oct. 2004 timelines, is mentioned multiple times in relation to
Rosetta, there's heavy coverage of the Matari (I think that last is an i
with some sort of accent, from other coverage, LWN's coverage seems to
have it as a missing-char square, on my setup) Conferences, lots of
coverage of Hoary and Warty including a Warty Starter's Guide, coverage of
the Ubuntu images controversy both in Ubuntu Traffic and a mention from an
article covering the Debian "Hot Babe" controversy, more community
meetings (Oct/Nov)... and I quit scanning at that point.

By comparison with Debian, which returns the search-limit 500 hits with
the last one on June 2, meaning just over seven months to get 500 hits,
Ubuntu, based on Debian, gets 100 hits in ~3.5 months, so will have
probably over 200 hits in a similar 7 month period. Yet, as I already
mentioned, Ubuntu is based on Debian so any articles mentioning that will
be a hit for Debian as well. Again, how much bigger is Debian than Ubuntu
and what's based on what? 2/5 the hits is certainly beyond any reasonable
expectations, I'd say.

I'm not complaining that it's too much. Good for Ubuntu! However, saying
it has flown under the radar IMO more correctly reveals simply that the
LWN coverage has apparently flown under /your/ radar, as LWN coverage has
been far higher than many would argue Ubuntu deserves at this point. In
fact, it could be easily argued that if coverage continues at this torrid
pace, there's serious potential for a backlash, as people simply get tired
of seeing the name.

Hmm... That Warty Starter's guide looks like it might be what you are
after. The LWN article is a very short pointer to the off-site article,
but here it is. I guess you can follow the link from there, and make your
comment on the LWN pointer page when you are done, for anyone else that
might find it and be interested. http://lwn.net/Articles/115349/

Duncan