Post by Evan LeibovitchI'm looking at stats related to web browser use such as
netmarketshare<http://marketshare.hitslink.com/report.aspx?qprid=0>.
They show browsers that don't run on Linux such as Safari and IE counting
for two-thirds of all browser hits. And that's not counting non-Linux
installations of Firefox and Opera.
this primarily counts the markets I already mentioned we aren't doing well in
on the desktop: North America and much/most of Europe.
it doesn't capture, however, the markets we shine in: educational deployments,
the developing world and many government offices.
web site hits are easy to measure and so are intriguing and a nice data point.
but they are not reflective of global PC sales and usage. they are reflective
of the computers used to visit those sites.
Post by Evan LeibovitchKonqueror, the only Linux-only browser
in the group, scored a whopping 0.03%.
Konqueror has completely "lost out" to Firefox even on Linux, so that's not an
interesting number. (reasons for that left as an exercise to the interested ;)
Post by Evan LeibovitchPost by Aaron J. SeigoPost by Evan LeibovitchDoes this mean conceding the Linux desktop to being a niche rather than
mainstream player? I don't think you need to look far to see that this
reality hasn't changed significantly in a decade.
that's the problem, isn't it: we aren't looking far. if we look to
Brazil, Spain, Turkey, China and elsewhere, a lot has changed in the
last decade .. and universally in favour of F/OSS on the desktop.
Indeed. Linux has gone from off the charts to on the charts, and the wins
are extremely impressive. But private sector adoption is still glacially
slow.
again, go to some of these countries and be amazed to see adds for Linux from
local/regional companies on clocks in the main airports, for instance. private
sector adoption is glacially slow ... here. it's not nearly as slow elsewhere,
even in some developed nations like Germany where >20% of desktops use
OpenOffice and a bit more than 10% of office and governmental PCs in the last
survey i saw run Linux (our lawyers in Mainze happened to be running Linux; we
picked them for other reasons and in our first meeting in their offices
noticed what they were using on all their desktops; they didn't seem to think
it was anything special ;)
there is, btw, a common thread between the countries where successes are being
seen: government support from the top. in Germany there is a huge emphasis on
electronic security with the government sponsoring work that results in public
advisories on the matter which much of the population takes quite seriously.
in Brazil there are huge investments in public education of public officials,
such as a run of 10,000 booklets authored (very professionally) by prominent
members of the F/OSS community there that was paid for and disseminated by the
government to all public managers above a certain level in the national
bureaucracy.
in Spain and Portugal, the regional governments are working tightly with
regional companies to find affordable and culturally relevant answers. (we're
about see another 400,000 seat (not user, seat!) deployment in Portugal, btw;
laptops to students of all ages in the public school system running Linux +
KDE)
in Turkey we have a "national operating system" project that's turning out
Pardus, which is used pretty widely even outside of Turkey.
the moment we take our eyes off of North America we see a growing landslide of
wins. the question is: why should we bother doing that?
because we have a lot to learn from those wins.
because those wins will translate into new sources of F/OSS participation and
powerhouses.
because those wins show it can be done to those here.
because those wins will be critical to ensuring F/OSS' rise to dominance in
places like mobile, where so much work in that space happens in places such as
Brazil (look up iNdt on the Internet)
Post by Evan LeibovitchContrast this with the cloud where FOSS platfoms dominate, even in
the private sector.
oh, absolutely, we do quite well as the infrastructure for the cloud. not that
the cloud itself is made up of much F/OSS. though efforts such as Open
Collaboration Services are trying to address those issues as well.
Post by Evan LeibovitchThe success of FOSS servers and desktop apps indicate that the fear of open
source per-se is gone. The slowness to adopt comes from other factors.
absolutely.
Post by Evan LeibovitchI guess my move to cynic happened when the OLPC project had to jump through
hoops to enable Windows to run on the XO1, because of customer pressure.
none of the actual customers really cared; this was much more about pressure
at the international governmental level as focused around the US efforts of
Microsoft.
Post by Evan LeibovitchThat IMO was a fairly depressing indicator on the (in)ability of Linux to
be accepted as a mainstream system, even in an environment that should be
friendliest to it.
the OLPC project had many other issues inherent to the project that ensure it
would not get far. speaking of "other factors" this is certainly part of our
problem: we love to find ways to creatively shoot ourselves in the foot.
Post by Evan LeibovitchPost by Aaron J. Seigoout of a few hundred thousand seats serving over 50 million students in
Brazil, which saw the participation of a Canadian company no less
(UserFul based in Calgary), we see such statements as "reality hasn't
changed significantly in a decade".
I stand behind what I said. It's a growing niche, but the transition from
niche to mainstream is still a prospect that's far away if even possible.
when we own 100% of the public school systems in multiple countries (and in
Brazil, huge chunks of the University system as well), those are hardly
"niches". the problem is that we tend to focus on North America and Western
Europe and discount what the rest of the world is up to. it results in us
focusing on failure and not learning very much from the successes we are
having.
Post by Evan LeibovitchMy premise is that Linux will end up dominating, but mainly because the
desktop will become less and less relevant.
something to revisit in 5 years, i suppose. i've been hearing "the PC is dead"
mantra for 15 years now :)
Post by Evan LeibovitchFOSS is winning, but the most
significant wins are happening on apps and mobile.
again, sorry to disagree, but F/OSS is the system that the kids in
Extremadura, the Canary Islands, Brazil, Georgia and elsewhere are using.
those are significant wins in terms of reachs (10s of millions of people who
will be entering the adult population in this decade).
and yes, we have great wins on mobile and some app categories. which is great.
Post by Evan LeibovitchBy the time Linux wins
on the desktop, it won't matter because the function of the OS will be
little more than launching a browser.
remember Larry Ellison's thin client revolution in the mid-90s? :)
what will happen at best is that rich client computing will move into the
browser and use it as a delivery mechanism, but the software itself will
become no less complex and resemble more and more what we are already doing
(and in many cases probably -be- what we are already doing now). the cloud is
being pushed so hard because it is seen as an economic opportunity and the
next way to develop captive audience with the rise of F/OSS "destroying" that
idea on the client side.
Post by Evan LeibovitchPost by Aaron J. Seigoyou also mentioned netbooks, where Microsoft trotted out numbers stating that
in North America. in many places in the world, Linux is doing more than great
on netbooks. in Norway, Linux has kept pace with Windows on netbooks and
globally we see over 30% of the total shipments being Linux based.
Source?
first hit on google: http://www.desktoplinux.com/news/NS5114054156.html
Post by Evan LeibovitchIn North America it's no contest.
agreed. (yet :)
Post by Evan Leibovitchsee with my own eyes. I'm typing this message on an Ubuntu Netbook Remix,
and it too months from the time I bought my Asus to when the wireless was
properly supported. Suspend works most of the time now but not always. And
making the external display work well is like stepping up to a roulette
wheel.
that part i mentioned earlier about shooting ourselves in the foot? yeah, we
like to destroy our chances in these markets with the uncoordinated
introduction of things like xrandr 1.2, pulseaudio and more. in places where
F/OSS on the desktop -is- succeeding, the to-retail companies move much more
slowly and conservatively in their products they ship. Red Flag, racking up 5
million+ retail units / year in big box stores in S. America and China (with
~80% F/OSS retention rate on those boxes in S. America), is only now moving to
KDE 4, for instance. very smart. we have a lot to learn from those successes,
but we'll only learn if we pay attention and stop these "at least we have
other markets" cop outs. it isn't sexy work to look hard at what we are doing
that isn't working. it's more fun to watch where things are already working
right here in our back yard. but there is a VERY compelling model/story around
'device spectrum thinking' that no other technology out there can deliver, but
it requires we do the non-sexy work of paying attention to our achiles heel
here in our home markets.
Post by Evan LeibovitchThat's not Microsoft hype, that's someone with 15 years of admining Linux
systems fighting with wireless drivers on a supposedly-well-supported
EeePC. If I was not technically minded and was told I'd have to read the
WiFi Howto<https://help.ubuntu.com/community/WifiDocs/WiFiHowTo>(and
follow its many instructions) just to make my laptop work right --
something that comes naturally under Windows -- I'd return it too.
yes, this is sad. unfortunately much of the use market here has gotten behind
Linux variants that are not suitable for the public. again, shooting ourselves
in the foot.
Post by Evan LeibovitchPost by Aaron J. Seigothis is precisely what the remaining two
proprietary vendors want to happen: pay no mind to the global market, keep the
North American and, hopefully, much of the European market in their
pocket because "it isn't being done, so it can't be done".
Apple and Microsoft/Dell/HP/etc are having enough fun going at each others'
throats. Playing this as a conspiracy between them against Linux is
ludicrous.
it is absolutely part of it, though we have our own hand on our throat as
well. it's not a conspiracy theory, either, i get to deal with it first hand
in my dealings around the world.
Post by Evan LeibovitchBut it doesn't matter. More people own mobile phones than PCs worldwide,
and I'm quite happy with FOSS being as dominant there as it is in server
space.
while this is true, the "desktop" side does matter to me. it does matter, for
instance, that Apple still has the only compelling application store for
mobile. that's a "desktop" issue (though it doesn't look it at first blush,
it's in the "desktop" space that we have all the expertise needed for that;
which is why a "desktop" company made that successful app store in the first
place and not a mobile company.) it also matters that 80+ million "PCs" are
shipped every quarter. it matters that those PCs and those mobile devices are
supposed to work together. it matters that in addition to phones, there are
many other device categories just as compelling which even more so have to
work with PCs well.
the F/OSS desktop is about a lot more than giving Linux icons and panels on
the screen. it's all woven together.
btw, we haven't "won" mobile quite yet: RIM still owns the lion's share of the
market F/OSS is most suited for (smart phones and above), we are uncoordinated
and even nascent in non-phone markets and Apple is growing very fast in these
areas as well with their hyper-lock-in offerings. Android is full of
anachronistic practices (http://www.kroah.com/log/linux/android-kernel-
problems.html) and so needs continued "support" to go in the "right"
directions.
if it weren't for Symbian, with over 50% of phones world wide, joining in the
F/OSS parade things would be even more gritty for F/OSS in mobile. what we do
have now is momentum, however. with Symbian GPL'd and, more importantly, Qt
becoming a universal toolkit that bridges Symbian, MeeGo and much more; Linux
is a checkbox entry now in the mobile space that has as much or more respect
as an option than any other platform in mobile. no other sector has the
momentum we have there. and it's almost entirely due to people believing in
and supporting F/OSS in that area that we are here now with a rosy future in
mobile (even though it isn't a "done deal" quite yet :). pragmatism, effort
and support have been the keys.
this is what we need to do across the board. including large form factor rich
client. (and much of the mobile Linux stack came directly from, and is still
part of and maintained as part of, the F/OSS desktop stack).
Post by Evan LeibovitchIn time it will compete effectively with Apple (and in devices like
the Droid and Nexus One already does), and poses real threats to
Microsoft, and Palm.
well, Palm isn't a hard target and Microsoft is busy killing itself in the
mobile space. Apple and sort-of-open mobile plays based on Linux (which
Android is, btw) are the real issues.
Post by Evan LeibovitchNokia will be interesting to watch to see whether
they have enough confidence in MeeGo as a replacement for Symbian. (Mind
you, Symbian is open source so even its use is a win)
i don't think MeeGo will replace Symbian, per se; it will displace it. Symbian
will continue to have it's place. thankfully Nokia has GPL'd it and brought Qt
to it. but yes, it's all good news there. :) Intel pushing on MeeGo will also
help.
Post by Evan LeibovitchPost by Aaron J. SeigoPost by Evan LeibovitchChromeOS and Android show
promise as conventional OS replacements, but the biggest encroachment
of Linux is just beginning, and it fits in your pocket.
until a proprietary vendor in North America tells us it isn't really happening
and we buy that line too?
Sigh. It's this level of conspiracy theory that hurts the advocacy movement.
i'm afraid you missed my point, which was: if you aren't interested in the FUD
that exists around mobile, then why do the same around the desktop? there was
rampant enthusiasm with no basis some years ago around the desktop, and it
hurt us. now some (many of the same people involved with the rampant
enthusiasm, btw) are abandoning the desktop because "it's un-winnable, but
doesn't matter" just when we are actually getting real, signficant wins and in
a time where it does, really still matter.
Post by Evan LeibovitchDon't look for competing vendors, look for objective stats.
i agree; where we differ, it seems, is that i'm interested in the global
market and sales figures rather than trying to measure indirectly via web
browser stats.
Post by Evan LeibovitchListen to friends and neighbours whose fear of Linux has little to do
anymore with Microsoft-spread FUD.
much of my day-to-day work is about exactly this. and what concerns me is that
when we take lines such as the ones you've been making here ("desktop doesn't
matter, and we can't make inroads there anyways") it entrenches these
problems. we need to take the desktop a lot more seriously (not the hype from
the early 2000s, which i was similarly against), realize it's interplay with
the device spectrum and start supporting Linux variants that actually play
responsibly in terms of granting a reliable user experience (and demanding
such behaviour as the status quo instead of accepting more "oh look, another
breakage shipped via automatic updates")
Post by Evan Leibovitchwhat you say here is true, though: mobile is huge for F/OSS right now and
Post by Aaron J. Seigoit's
only going to get bigger. at the same time, we're doing quite well across the
device spectrum, and that includes those laptops and desktops in the middle.
Agreed. But "quite well" is, in the big picture, still a niche. It will be
intetresting to see what happens as the cloud expands and we move further
into OS-as-thin-client, as the OS choice will diminish as the function of
the desktop OS diminishes.
i've been waiting for OS-as-thin-client for 15 years. (the first 5 or so years
i was involved in the industry, it was "all" rich client on local systems with
virtually no thin-client hype.) it's a pipe dream. technologies displace, not
replace, and thin-client computing will not displace rich client nearly as
significantly as people say it will. we're just in that part of the thin-
client/rich-client cycle right now.
Post by Evan LeibovitchPost by Aaron J. Seigolet's not soft peddle our hardest fought wins, those we have on the desktop /
laptop market, which still represents a ~80mm units/quarter (yes,
quarter) market globally.
Most of that market is replacement for old computers, where inertia is
still a major (probably the buggest) impediment.
if your point is that the desktop hardware market is saturated in the
developing world, then yes. but that's a bit like saying "new cars are
generally bought as replacements for old cars" and using that as a reason
Toyota/ GM / etc ought to forget cars and focus on bicycles. it's a very
flawed way of looking at a market.
Post by Evan LeibovitchPost by Aaron J. Seigoand as such, as we make bold strides on mobile, we have no need to
downplay our progress across the board just because the poorly written
rhetoric is getting louder by the day. the fact that it is getting
louder ought to tell us
all something.
To say that the rhetoric is getting louder really makes me concerned, for
it forgets just how bad things were a decade ago. Linux is competing now
on its practical merits. And it has to fight aversion to change rather
than any real love for Windows. Inertia is now an infinitely bigger
impediment than Microsoft propaganda.
If Microsoft hype was as effective as you think it is, the Mac would be a
niche too.
depends on how you define niche, but globally the Mac definitely is a niche if
we consider desktop Linux a niche. we sell ~as much retail Linux, globally, as
Apple sells Macs.
Post by Evan LeibovitchStop blaming vendors for the slowness in Linux adoption. The
other factors are much more important now, and ignoring them to focus on MS
is a mistake.
i'm not focussing on MS, i'm focussing on people within our own community
making boldly incorrect statements about the reality of the market today and
tomorrow, about writers in the media who are led into writing articles that
pile on the misinformation and about people who are giving up because numbers
spewed out by various vendors[1] that are skewed are taken as dogma. i'm
focusing on people who want to see the desktop surrendered because it's "too
hard", "too inconvenient" or "not important". there is a much bigger picture
here where we are moving forward significantly on all fronts and require all
fronts to move forward significantly to continue to see F/OSS succeed in a
sustainable fashion.
[1] yes, including MS, such as the netbook numbers that said Linux was 3-7% of
the market. those numbers came directly from the pocket book of Microsoft.
that isn't a "conspiracy theory", this is what i know from dealing with that
situation first hand, personally.
--
Aaron J. Seigo
humru othro a kohnu se
GPG Fingerprint: 8B8B 2209 0C6F 7C47 B1EA EE75 D6B7 2EB1 A7F1 DB43
KDE core developer sponsored by Qt Development Frameworks