Episode: 332
Title: HPR0332: Libre Planet 2009
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr0332/hpr0332.mp3
Transcribed: 2025-10-07 16:36:27

---

Hmmm...
So, this has been traditionally our FSF members meeting, and by the way I am Peter Brown,
the Executive Director of the Free Supply Foundation.
So this is normally our annual members meeting, I know that most of you are here, our members.
This year we have opened up to a bigger audience and we also have a more eclectic group of
speakers as well, and I just wanted to cover off a few items for you before we get started.
We have a Floss Manual book sprint going on, so if you feel like sharing some of your knowledge
about working on the command line to help empower newbies to GNU Linux, please spend some
time and head over to room 109 and 112 and do some authoring, and your name will be in a book
that we will sell from the GNU press. So Libre Planet, what is Libre Planet? Well, Libre Planet
is an idea that we have been percolating for a while, and it is based around this concept that
we are getting to a point where we have what we need to encourage our neighbours to adopt
FreeSoftware. We have been working hard to make sure we have had an operating system,
we have been working hard to make sure there is hardware that works with FreeSoftware,
we are very close to a lot of these realities becoming true. What we really need now more
than the girls is to use our influence and our connectivity to encourage other people to adopt
and start using FreeSoftware. I think a lot of people recognise that we are in an economic decline
at the moment, and some people say, well, there is a great opportunity to talk about FreeSoftware
based on economics, but I think more than anything else what we want to try and do is with Libre Planet
is encourage FreeSoftware activists like you lot to push forward FreeSoftware based on the values
that we care about. It is being in control of the technology we use, and that is really part of
the reason why the book that we are doing in the book sprint is about the command line,
it is about this idea of, you know, we have been encouraging people to click on the things,
well, in fact there is a great opportunity to say to people, well, yeah, that is great,
but imagine if you could do more, and that is really what Libre Planet is all together about.
We want to encourage people to be more in control and we want people to value the values of FreeSoftware.
So that is what Libre Planet is to start off today. We have got three main areas to talk about,
Freedom for Network Services, Libre Planet Activism, and High-Private Projects,
FreeSoftware Projects. So we have got some really great speakers, but first I want to introduce you
to FSF's campaigns manager, Matt Lee. Tomorrow we are setting aside for some of the core
and open space event. The idea is that we want people here to help lead and play a major part
in tomorrow's efforts to talk about those three topic areas. Each of us has some experience
or expertise to share, and we want to come out of the open space event with some progress and
real work. So we have described this as a bit of a working conference, so Matt is going to explain
to you some of that stuff, but first of all I just want to say thank you also very much for coming
along, and I also want to thank all the FSF staff for their hard work in making this event happen
and look so good. Thank you. Good morning. Welcome. People got to mention actually that Rob Myers,
the GNU Chief Webmaster is also here, and he's going to be helping me, well he's actually
indeed helping him, lead tomorrow's event. So Hannah the Rob. Hi, so tomorrow we've got three tracks,
we've got Track 4 FreeSoftware Activism, Track 4, High-Private Projects, and Track 4,
what's everything? Free Network Services. Absolutely. So at lunchtime we're going to put three big white
boards out in the hall, lots of sticky notes and pens, and if you've got an idea, something that you'd
like to work on tomorrow, let's have a meeting, have a hack sprint, just have an open discussion,
or give a lecture on it, write down your name, what you'd like to do, and put it on the relevant board,
and then tomorrow morning. Tomorrow morning. We're going to gather those up and schedule things out
for during the day. You can also nominate somebody else to talk, which could be kind of fun.
If there's something you want to hear about that's not on that board, but you can't talk about it,
do you really want to hear about it? Put it down anyway, some people already add things like licensing
to the list of things to talk about, so if Brett's here somewhere, he may not be talking about that,
but also the thing to do is at the end of the day tomorrow we're going to have these five minute
superfast lightning talk sessions. If you have something to announce, or something you just started
doing, or something you wanted to say, it's a good idea, is it a bad idea? You can announce it
of those slots as well. There are actually enough slots here for every two people here to give a talk,
so get talking. I think now we're going to hand over to John Sullivan, who is our MC for the day,
who's going to introduce the speakers. So, again, lunchtime and the corridor, big white boards,
enjoy. John. Thank you all for coming. I'm going to keep this short so we can get to the
main attraction here. First speaker this morning is Jeremy Allison, who is going to talk about the
elephant that's apparently in the room here, Microsoft, that's topic that he's very familiar with,
given his groundbreaking work on Samba, and thank you for coming.
So, this is more I didn't have the slides up to start with. I didn't want to distract people.
So, thank you very much for the introduction. So, when I came up with this title for this talk,
I know we're all free software people, but sooner or later you do have to talk about the elephant
in the room, which is Microsoft. Because they are the world's largest software company,
they do have a great effect, their actions and activities, have a great effect on what we do.
And so, I went googling for the right picture to use for the title slide of this talk.
And I came across this, and for some reason it struck a real chord. And I actually believed
that's Richard, under there. And I must have been asked to sit there and thought, can I get away
with this? Can I get away with this? And then I thought about my audience, thought, yes, I can
probably get away with this. So, this is almost certainly a talk, I shall not be repeating in a public
space. So, or if I do, I might heavily edit some of the pictures. And also, I actually worked
for my day job at Google. So, this is in no way associated with them. And I wanted to make sure
that people understood this was the perfectly delusional ravings of someone who has been driven
slowly but surely insane from working on the interoperability with Microsoft software.
So, why should we actually care about Microsoft at all?
Things are very different now from the time in the mid to late 80s and early 90s.
When essentially, you really did have to use Microsoft software in many ways,
unless you escaped into Unixland, which wasn't really affordable for most people on personal computers,
you really had to deal with Microsoft. But that's no longer the case. So, here's an interesting question.
Oh, and by the way, don't let me stand up here and pontificate for an hour or so. Please ask
questions, make this as interactive as possible. If you think I'm full of the stuff that I showed
on the previous slide, please stick up your hand, let me know, argue. I can drill down into just
about any topic that I have on the slides, except for some which I will defer to further speakers.
But I am not a lawyer, even though I will be talking about some legal things. So, I may have to
defer to some lawyers if they're here or otherwise. I can drill down into many of the legal
activities that have been going on, because I've been involved with quite a few of them.
I can't always tell you exactly what went on behind the scenes, but I can try and get it close.
So, as I said, GNU Linux is a perfectly wonderful operating system, and we simply don't need to run
windows anymore. So, why don't we just ignore them and let them do their own thing? So, by the way,
how many people here actually still do run windows for anything? Oh, that's more than I would
thought actually. My only sins are running the occasional piece of proprietary graphics driver code,
because I'm a fan of unbelievably violent first-person tutormops. And so, I won't run them on
windows anymore. I do run them on wine, but unfortunately, I have to commit the sin of loading
Nvidia drivers on my box. So, hey, so, kill me. But, essentially, we really don't have to do
with Microsoft anymore. Our system works well enough, and hopefully, as we will find out later
at this conference, is good enough so that anyone can use this without having to interact at all
with Microsoft software. In fact, the dirty secret of the Samba team, and in fact, the cause of some
of our regressions, which we're trying to fix, is that none of the Samba team members uses Microsoft
anymore for anything. So, we test against virtual machines in a test environment, but we don't
need that software anymore. So, we've unfortunately moved out of the realm of scratching our own
itch, to we're actually scratching other people's itchers, which is never quite as much fun.
I know. You know, and you have to wear rubber gloves and stuff. Because we simply don't care
about windows anymore. We don't need it. We don't run it. So, we could just leave them alone.
And I always regard it as seeing people complaining about Microsoft is like watching people waiting
in a sewer. It's kind of like, you know, you're walking on the sidewalk and you're watching them
down there sort of digging the way through the filth and complaining about the viruses and whatever.
And you're like, well, why don't you just get out of there? You don't have to be down there.
So James bottomly, who's with these scuzzy maintainer in the Linux kernel,
actually came up with a wonderful analogy, which I'm not ashamed to completely steal here.
He said, he said, it's interesting. If you look at the three most popular operating systems,
windows is a prison, right? You know, you can see it. Everyone knows it. It's grim,
the food is bad, the doors are made of metal. You know, there's no plastic on the toilet seats.
You know, it's grim. MacOS is kind of, it's still a prison, and this is what people don't get.
Including you Mac users up there, I can see you. I don't think I can't. I noticed on the way in.
Yeah, yeah, I could see that from the login screen. So MacOS, it's still a prison.
It's very nice. There are carpets on the floors. You know, there's plasma TV instead of windows or whatever,
but you can't look at and see what you want. You're still watching the Apple channel whenever you
look through those plasma TVs. So Linux, GNU Linux is really the only free system out there,
and that's the one, obviously, we should be encouraging everyone to adopt for the obvious benefit.
So yes, we could just leave MacOS after loan. So why don't we? Why am I sort of here giving
this offensive talk with a picture of an elephant at the beginning? Well, the problem.
They don't really want to leave us alone. And I have hopes that this will change in the future,
and I hope that this talk will end on a positive note. But right now, MacOS off scurribusiness model
is actually dependent on maintaining the monopoly on the desktop and extending it out into other
places. One must never forget Microsoft's mission statement. They actually do tend to forget it
these days. But the Microsoft mission statement is a computer on every home and on every desk,
and that's the bit that they quote quite freely, and they're very happy to say, look, that's
our mission statement, but they often leave off the last part of the sentence, running Microsoft
software. And that's their full mission statement, and that's the one that they believe wholeheartedly
internally. So Apple is, as I say, is tolerated and even encouraged, although there may be a little
upset about how much they're encouraging it these days. But they really aren't allowed to become
that much of a threat. Even now, Apple, I believe, is less than 10% of desktops. And if you remember
the massive jumbo-tron picture of Bill Gates appearing at the Apple conference and being booed
when he was giving some money in the days of the scully days of Apple, they're kind of used by
Microsoft as an antitrust to get out of jail free card for the various antitrust authorities.
This has not been working so well of late. But the thing that we have to remember, and that they know,
very, very deeply, is that free software, GPL at licensed free software, not all free software,
but copy-lapt free software, is absolutely intolerable to them. They can't cope with it,
it doesn't fit in their business model, it's completely outside their ecosystem. And you can see
this, not from the friendly engineers that they send out to talk to conferences or whatever,
but if you actually look at some of the antitrust documentation from the trial, where you actually
get to look at what Microsoft's real feelings are internally. And the one that surprised me until
I really tracked it down to the actual source, which is there on the slide, was GNU Linux installations
inside what Microsoft regard as their customers are called infestations, which is something that you
would normally regard, cockroaches, not competitors. Having said that, I have actually often compared
free software to the cockroaches living in the walls, so maybe it's not so bad. But emails like
this and comments at very, very high executive levels are a sign that really they're not interested
in peaceful coexistence, and they ain't going to leave us alone. And in fact, I've had many experiences
over the years to show that they really, really aren't intending to leave us alone, so I'll talk a
little bit more about that. But I mean, I'm sure you've all seen Sam Ranji doing Sam Ranji for
a Microsoft, giving his wonderful talks, but don't they love us? And I actually know Sam quite well,
and he's a nice guy, he makes wonderful positive statements about open source and even free software
and occasion. I believe that he's genuine in his tasks to change Microsoft to be a free software
friendly company, and I encourage him in whatever way I can. I'm probably going to give him a hard
time next time I see him, but it's, they have a definite distinction internally. They're very
encouraging a free software under the MIT or BSD licenses, which they can appropriate into
proprietary software. They are not so keen, as I've mentioned, on copy left and other forms
of GPL software, which they can't. I recently joke that the quickest, if you have a church
faith or something that you're running and you want to get sponsorship, the quickest way is to add
open source into the title, and a Microsoft representative will turn up with a blank check
to offer you. We should probably call this the open source Libra Planet or something, and then
Microsoft would be here sponsoring us and pay for some excellent coffee in free lunches and dinners
and whatever. It really is the easiest. You'll see this in many of these open source conferences.
I actually have Microsoft representation, and that's how they get a lot of their sponsorship.
But you actually have to give something up when you take that shilling I'm afraid.
Microsoft is very often compared to the Borg, and I think this is incorrect, actually.
And I like to say, we're the Borg, not Microsoft, because I actually were a lot better,
a more efficient at assimilating technology into our collectives than they are into theirs.
For instance, for the longest time, most of their engineers couldn't even look at free software at all.
I mean, it was a far, farable offence to do that. But I would argue that we're much, much
friendlier than the Borg. We turn up and we invite people to join the collective. It's not
so much of either all thing with a cube, with a large energy weapon pointed at you. It's more
sort of bearing gifts. So they really aren't like the Borg. So what actually is Microsoft?
And what is it like internally? Now, I've actually worked up at Microsoft back in the early 90s.
And Microsoft is not what you think. You see Microsoft as this one monolithic entity,
but it really isn't. Microsoft is actually the best described as a series of continually
warring tribes. And they hate each other. I mean, they really hate each other. They hate each other
far more than they hate free software and we people externally. It's like a family. You
really hate the people in the family much more than, you know, you hate the outsiders. But
whenever you're under a threat, they all pull together as a team. But what happens internally is
the various groups hate and despise each other. I mean, the thing that I came into contact with
was how much the Windows new technology team despised the engineers who were working on Windows 95.
I mean, they really regarded those people as, you know, dishwashers with computers,
kind of. They really regarded them as stupid and couldn't code, etc.
And all this and the Excel group hates the word group and they all hate the access group,
you know, and the exchange group hate despises the file server team and actually codes exchange
so it won't even run off a Microsoft file server. And, you know, so you have all these
complete hostilities. But they use that internally to generate competition to make each group work
harder. And what this means is that they're actual external actions and much harder to predict
than you would expect because they don't always do things in rational self-interest or at least
not what you would consider rational self-interest. It's in the rational self-interest of whichever
group that is currently got the ear of the VP who's doing whatever it is that they'd be able to
be doing. But it's not necessarily in the rational self-interest of the company as a whole.
But as I said, the company as a whole isn't really one whole company. Now Gase has gone,
there is a lesser-can barmer at running things and boy, do they hate him.
They really want him out. But, you know, being the co-founder he's hard to dislodge.
So, if you look back and see when Microsoft first became aware of free software,
it was, I would probably date it to around 1998. Now, the Samba team had contact with Microsoft
and we'd work with them as far back as say in 1994. And they were actually incredibly
friendly at that time. They paid Trich, Andrew Trigel, who's the co-founder of Samba,
to the original author, to go business class from Australia to Seattle for this first SIFS conference
they ran. You know, they used to funders, they gave a documentation. Funders in travel,
modern actual pay for code. But what you have to remember is at that time, they were competing
very heavily against the entrenched Nobel Network. So, it was in their interest to open their
protocols as much as possible to be friendly, you know, to help interact with the free software
community. Because they saw us as interesting curiosity as sort of rather like creatures in the zoo
that, you know, you throw a few peanuts too for amusement. Which we were really, I mean, let's
be honest. You know, we were doing this because we wanted to and having fun with it, we had no
great dreams of world domination. You know, we were just riding code that we wanted to use because
we needed it and we wanted to use it. Things had changed by about 1998. Network was defeated,
completely in retreat. Windows NT had become dominant. And at that point, we weren't so funny
anymore. The Smiths song, that joke isn't funny anymore. And we were less encouraged at that point.
And they, I think, I think the change began from a little, a fair call, the Minecraft benchmark,
if anyone remembers this. And this was just after Herb Lewis, a Samber team member and I were both
working for SGI at the time. And we'd managed to coax Irix, which was a very nice unit at the time
to run Samber as a file server, something like twice as fast as Windows NT could serve
Windows clients. And we got a lot of press over that. And apparently, Mr. Gates walked into the
file store in teams slammed down the paper and said, fix this. And so they began to be concerned
about they actually had some more competition. And if you've read some of the early Microsoft
documents, you know, Microsoft's feeling that their legitimate share of any market is 100 percent
no competitor can be allowed to exist. So that's a very interesting read. And I read that going
back for this talk. And one of the things that struck me going through that, it's been a long time
since I read it, was how all free software at that point, this was very unsophisticated thinking,
all free software, including BSD, MIT, everything was lumped in under the enemy. And what do we do
about it? Interestingly enough, the author of that paper now works for a free software company,
as it seems to be the fate of many Microsoft engineers, including some of the people who wrote
Windows NT because I work with them. But all free software was considered a threat and they were
trying to decide what to do about it. So they came up with an interesting strategy, which they
have pursued. And I'll still pursuing to this day, although I'm going to examine what effects
exactly it's had. And so this, let me read this out for you, decommoditized protocols and
applications, open source projects have been able to gain a foot hold in many server applications,
Apache, Samber, et cetera, because of the wide utility of highly commoditized simple protocols
by extending these protocols and developing new protocols, we can deny open source software
projects entry into the market. And so this was the birth of the strategies behind Windows 2000
and what they did there. And so let me talk, this is the area that I'm most familiar with.
So when they introduced Windows 2000, what they did was they took MIT Kerberus and corrupted it
essentially. They added what's called a privilege authentication certificate pack field into the
ticket granting ticket, which they refused to document. And then essentially they tried
as hard as they could to prevent any other Kerberus server from serving Kerberus legitimate Kerberus
tickets to Windows clients. But they did this in an incredibly sophisticated way because what you
have to remember is that Microsoft is incredibly smart, they're full of very, very smart people,
they're not dumb. And you have to realize what kind of a threat you're dealing with. They
actually allowed MIT Kerberus servers to serve tickets to Windows clients, but without the privilege
authentication certificate, there was no way each of those clients was essentially an individual
workstation. So what that meant is you would have to manage the accounts on each individual
Windows machine. So every individual Windows machine would have to have full account details
of anyone who wanted to log in, which of course completely defeats the purpose of having Kerberus
in the first place. It shares the authentication, but you can't manage the users in one place.
So what they did essentially, if you're familiar with the Linux world of PAM and NSS, was they split
out NSS from PAM, and they made the PAM part work with MIT Kerberus servers, but they deliberately
obfuscated the NSS part. And another way they did this was to co-mingle commodity protocols.
So active directory is a mixture of LDAP, Kerberus, DCRPC, DNS, all of these things, all of which
were not developed at Microsoft, all of which are commodity protocols fully documented,
but they mingled them together in such a way that clients were unable to use those services
independently from non-microsoft implementations. And the way they did this was, so for instance,
clients would query services in a particular way, one after the other, some of which were expected
to fail, if they didn't fail with the correct error messages, the clients would assume that they
were not talking to a Microsoft server, and then refused to continue. So it was extremely
sophisticated. And this is one of the reasons that Samba 4 is so late and so hard to do because
that, the technical decisions they made there are still here to this day and still have to be
obeyed. Their clients have not changed. One of the other things that they did was extended
the ability of mail protocols. And so rather than to that, being honest here, some of the protocols
that they needed to do, calendaring were not available at the time that they developed this software,
but rather than working with standards groups to develop them openly, sort of CalDAV and WebDAV
or whatever, they completely invented their own protocol running on top of DCRPC, remote
procedure course, that I believe to this day they have not documented, although they may have
documented that one. This is not my particular technical area, so I would have to go look at
their documentation. The other thing that they did, and you will notice this, when you have
single sign-on to Microsoft IIS, service servers and not to Apache, at least until we fix this,
they extended standard HTML by adding a proprietary authentication protocol HTML.
This all was incredibly successful, it really was for them.
They continued to do this, and for a while I believe the open internet was in danger,
because the monopoly position of the Windows client and the fact that everybody codes for it,
really at one point was, I think it was in the balance that the internet would be
essentially become a closed Microsoft bulletin board, essentially. Internet Explorer only
HTML rendering, all of the browsers denied, active plugins for authentication and government
national security, I mean look at Korea, for instance, you still can't get onto many Korean
government websites or use much of the internet infrastructure in Korea without the correct
active plugins, which of course freezes out every other competitor in the market.
I believe that someone just lost a lawsuit in Korea over that very fact, so Korea is captured,
completely, there is no way for them to get out. The more modern attacks are silver light
to attack, which is currently proprietary, well hopefully we are going to hear more about that
the flash video format, and the standardization that I paying people to make a ubiquitous
of the Windows Media formats in place of open ones, and this came out in the European Union,
Antitrust Trail with Microsoft, Microsoft would go to broadcasters, and they would essentially they
would pay them to replace real media streaming with Windows Media streaming in order to capture that
market. So how it's worked in practice? This is, I'm going to sort of talk about three separate
cases, case number one, which is the work group server market, and as I mentioned they've
combing all the protocols, and what they're trying to do here is leverage their client monopoly
into a server monopoly, and they're kind of winning in many cases. If you look at the authentication
market now, there is in a single organization that doesn't have active directory in there somewhere
or other, and it used to be, I think, an available lead directory had a 30% share active directory,
had 40%, and then Sons and IS and other Unix had the rest. Now it's probably 18 and 90% active
directory. They've won this market, they really have, and we managed, Samo managed to catch up
reasonably quickly to do Windows NT4 domain controllers, which were just based on the DCOPC
protocols, which is the Samo3 timeline, but the complexity of what they did with active directory
means that we have yet to catch up there, and so they really, this is probably their most successful
attack. But then let's look what actually the results of that was, and the results of it were
a couple of lawsuits. Well, the US lawsuit, we're actually about browsers, so I'll mention that
later, but the really fact, the one that they lost badly was the EU case, and that was actually
brought originally by Sun, and everybody else piled in, real networks, piled in with the
complaint, Naveau piled in, everybody jumped on that bandwagon, and it was actually interesting
enough, it was Sun that invited the Samo team to participate. I took a bunch of Sun lawyers out
for a very nice Indian meal in Kapitino, which they weren't used to, and they actually said,
will you help us with this? And I still remember that my comment to the lawyer, and I said,
I said, okay, we'll help you, but we'd seen what had happened in the Department of Justice
Lawsuits, we said, but you have to be in this for the long haul. If you take the money and walk
away, then you will have achieved nothing. This will have been pointless. And sort of about four or
five years later, when Sun took two billion dollars to walk away, I emailed that lawyer and said,
do you remember? And I never got a reply. Funny that. And then that people forget, but that
lawsuit nearly collapsed, because every single participant in that lawsuit was bought off,
everyone. The only one that really remained, even the computer industry association was bought
off, and the head of that got, I think, 20 million in pay off, for accepting to settle on behalf
of his members. Samba and the European Free Software Europe were the only people who were there,
and as I joked at the time, it's because we couldn't be bought off. It's not that we didn't want
to be. It's that we didn't. Hey, let's be honest, two billion dollars, right? I'm thinking about
my own island there. But we don't own the code. We can't be bought off. We were representing the
people who wrote the code. We have nothing to sell to them. And so we actually won that case,
and we won the appeal of that case. And the appeal, although that was not
the judge who wrote the opinion on the appeal retired later, and I believe that the appeal was
much, much closer than people think. I believe it was like a seven to six decision. It was a very,
very close decision with the judge writing the opinion being the deciding vote. So they nearly won
that one. But years later, it was a bad legal loss. So the long term results of that mixed.
This is probably their most successful strategy that they've achieved. So they still, as I've said,
they still dominate the workgroup server market. But in doing so and concentrating so much on
authentication, it's like playing whack-a-mole, or you know, you hit the cockroaches one place,
and they come out from behind the cooker. They took their eyes off the ball in the appliance,
the server appliance area. And that's where a lot of the money is. So this will add their
eyes to companies like EMC, NetApp, who sell an awful lot of simple file service, not authentication.
They don't compete with Microsoft in the dangerous area, which is authentication. And in fact,
I always joke that any file server appliance that costs less than $5,000 is guaranteed to be running
Samber inside it. In fact, it's quite funny. It's sort of like any file server appliance that costs
less than $5,000 or more than $250,000 is running Samber. Because IBM have this unbelievable
product, they call scale out file system, which is built on a clustered Samber that sort of scales,
they can saturate 10 gigabit ethernet with it. But they, I mean, they're charging millions of
dollars for this thing. So it's like, we dominate both ends, the middle is, not so much.
But very few people buy generic Windows file servers anymore. They buy Windows authentication
servers, they buy Windows clients, but they buy generic file serving appliances. So, yeah, it was
like, as I said, whack them all. You hit it somewhere else, we pop up somewhere, we hit it one
place, we pop up somewhere else. The other thing that really cost them, well, either got hit by a
huge fine, a fine that was large enough that it tipped, it caused currency traders to speculate
on the value between the EU and the dollar on the day that they got announced, because they knew
that Microsoft would have to scramble to find a bunch of euros. And the other thing that is still
hurting is that they were forced to document the proprietary protocols, completely and fully
document, which they are doing in good faith. And the thing that I think probably cost them,
well, cost them longer in the longer than the fine, is the ongoing effort. They have to fund
an entire engineering team to do this. And much of that work, and I know, because I've been watching
them do it on the mailing lists, is forensic work, because the people who worked at Microsoft
who wrote that code have gone and bought their own private islands by now, because they wrote the
stock up in the 90s, nobody's left who knows how that code works, and so you have to have people
who go in and do code forensics to find out what error messages are created for certain things.
And the other thing that really hurt them is prior to that, they really had
a very, very high reputation as a company. And this, I think, was the first real public awareness
that, hey, these guys, everyone, just scumbags. Perhaps not as much as the bankers or AIG, but
well, actually, it's Goldman Sachs who are the ones who are getting paid behind the list, but
they really took a massive hit on their reputation for this. And the other thing, because as I said,
they're incredibly smart, they realize that trade secrets are not an adequate defense.
Trade secrets fail. They get worked out, they can be reverse engineered, although that's not what
we did, and I hate that word, because it's a very bad word to use in the US. We like the term
network analysis, but it means that it's not, well, because we don't just assemble like, oh,
we just look at the network, you see. But what it means is that this is not an adequate defense
against their real competition, which is free software. So case number two, and there was a wonderful
quote in the press recently from Linus, two of those about the OOXML thing where he said,
anyone who was involved in that had to be insane. That was a horrible mess. Well, I can probably say
I was involved in that. And yes, I am insane. So the genesis behind this was that
Sun's open office products, and also backed by IBM and their symphony product, they standardized
a document file format called ODF, which is actually quite sane, because I had to read both these
standards as part of this effort. And this is actually quite a sane standard. It's not great,
nothing is not perfect, nothing is, but you could actually hand write ODF files, you know,
in EMAX, for instance, you could sit down and write an ODF file that would be read perfectly
adequately by any of these work processes. I don't think you could do that with Microsoft's OOXML.
So they pushed to get, because ODF was standardized through ISO, Microsoft have a captured
organization called EMA, that essentially is a puppet organization that dances. And they have
a fast track into ISO. It's a little more complex than this, but I'm giving you the raw.
I'm filtering out all the polite stuff, and I'm telling you what really happens.
Microsoft says, EMA pushed this through, EMA pushed this through. And they tried to push through
a 6,000 page specification in less than a year through ISO. Now, as anyone here, actually,
don't you work with ISO? Yeah, so you know, essentially that's like trying to, I don't know,
it's like trying to eat an elephant in one sitting. It's very fun to do. Standards bodies don't
move that fast. So what it resolved in was a massive corruption of the standards process.
All the major competitors rallied against them. And there was an incredible public
opposition in the standards community and the people surrounding it. Having said that,
they still won. Now, think about that for a second. The entire free world, as it were,
was rallied against them. With every resource, the resource of Google, IBM, Oracle,
you name it. They still won. That's how powerful they were.
So I know a lot of incredibly dirty stories about this process, some of which I believe
border on illegality, but nobody can prove anything this is always the way. But I will tell you
one example of how this was done. So an African nation who shall not be named decided to become
a voting member on office file formats, which is their right. And the reason they did so was,
and of course they voted yes on OOXML. And the reason they did so was because a bunch of Microsoft
partners funded a free training seminar in XML file formats focused directly on OOXML. And
they ran this free training seminar for a month or two. And then they went to the government and
they said, we want to standardize office file formats. And we think you should have a standards
organized, standards, your standards, what you should do this. And look, here's a committee
of people who know a lot about this. Look, because we just trained them. Isn't that convenient?
So essentially, and the reason they can get away with this, of course, is because nobody else
was putting, and you can't blame the government for doing that. Nobody else put money into their
computing infrastructure. Nobody else offered to give them free computers and a computer lab for
the universities. You know, so why wouldn't they take this? It's only sensible for them to do that.
But this is, that's not corruption, but it's incredibly unethical, I think. And so they did win,
they did get this through by corrupting the ISO, who as one of the no OOXML people created this
wonderful ISO logo as it should now be myself. But it kind of backfired in that governments began to
actually take notice of office file formats and the fact that they were creating billions of
documents that they couldn't necessarily read without paying a tax. And I believe, I believe
Massachusetts had something to do with that, which I'm sure you have your own little dirty stories
about. So the long-term result of that, I would say it's poor for Microsoft. That was, even though
it was technically a win, I think that was a loss, really. They still own a monopoly in office
productivity suites. Everyone still buys office. They still make a massive fortune from it. However,
it stood up the EU again. Never a good idea that EUs get beginning to get pissed with them.
So one of the results of the EU starting to investigate office file formats in Microsoft
Office is Microsoft suddenly said, oh, we're going to support ODF inside Microsoft Office.
Now it remains to be seen how complete or good that implementation is going to be.
Early results that I've heard are that it's extraordinarily poor. What is the price? Yeah, question.
Yes, so that's interesting in comment. I believe what they said was not
that it couldn't be implemented, but that ODF wasn't to quote rich enough to document the full
horror of the Microsoft Office file. Yeah, good question. Now, isn't the ODF implementation
of ODF in Microsoft Office? Isn't that a one-in-one? Okay, so the comment was, isn't the ODF
implementation in Microsoft Office of plug-in that's sun-made? Actually, no, not the one that Microsoft
is coming out with. So there is an ODF plug-in into office that's sun-made. It can't do everything,
and the reason for that is, and I know there's going to be a terrible surprise to you,
there are hidden APIs within office that nobody else knows how to use. But Microsoft is actually
coding this up from scratch. And so one of the things that I think we need to examine very closely
when this comes out, and to give them a lot of bad press over, is that their current ODF implementation,
for instance, doesn't handle spreadsheet formulas at all, or at least, I believe it puts the
spreadsheet formulas in a completely different XML namespace so that they can't be read and used
by open office. So what they will have is an ODF plug-in that actually is less interoperable
than the clever age plug-in that they funded is free software, and the sun plug-in that sun did,
you know, to prove it could be done. So we need to point that out. The other reason I think it
was a family for them is that they had complete binary locking.doc.xls.ppt were their formats,
and they were essentially a memory dump of Microsoft Office, and nobody could implement them.
They could change them at will. As part of this episode, they were forced to completely and fully
document those formats. That was, I think, a significant loss for them. Now, admittedly,
much of that was already worked out by open office, but as part of claiming that OXML was open,
people said, well, it doesn't cover the legacy binary formats, so they actually opened the legacy
binary formats. They would not have done that had they not been pushed. And the other thing that's
happened is many more governments, and I believe Texas State Government is a ladies one of, now,
it's not going to pass because it's Texas, but it is raised by a Democratic in Texas, so it's
got no chance. But there are actually bills appearing like in Massachusetts suggesting open file
formats, and OOXML doesn't really fit that. Right now, not even Microsoft Office implements the OOXML.
So, case number three, and I think this is the one that they've probably fed the poorest on,
was corrupting the open internet. And as I said, for a time, this was a real threat.
It looked like they might win this. They were refusing to follow HTML standards,
i.e. only websites, which there are still many of today. Windows Media Format was becoming dominant.
.NET was a fairly thinly disguised effort to destroy Java, and he's even creeping into the
GNU Linux system, which kind of worries me a little. It didn't worry me until the Tom Tom lawsuit,
but the other thing that they're trying to do is they're attempting to dominate. Now, if somebody
can come up with words of power, right? Microsoft has this concept of what they call rich internet
applications, and I really despise and hate that word, but I can't think of any of the really
good way to describe the sort of complex applications that you, if anyone could come up with a better
word that we can use than rich, because that really is adopting the language of the competition.
Anyone who listens to Fox News knows what a bad idea that is, because he who controls the language,
sorry, bloated. Yes, bloated doesn't, bloated is a nice idea, but unfortunately they do have some
positive connotations, and bloated doesn't really carry those positive connotations. So we have to
find something that has both positive and negative connotations. Yeah, who controls the language,
just control the conversation. So why I think they failed on this one? It results in ongoing
battle, but I think they're going to fail. Firefox really broke open the I only HDML problem,
and of course they rattled the EU's cage again, and Ajax applications now. Yeah, come in.
So the comment was that the proprietary browser Opera was also doing a lot to break that open.
Yeah, but the problem is Opera has like 2% market share. You know, it has, if you look at all the
statistics, it barely scratches the surface. So yes, I mean Opera will behind the EU case,
definitely, and Firefox probably wouldn't have instigated that. So I'm grateful to Opera for that,
but as an actual competitor in terms of breaking open the market, Opera was near relevance.
Really, Firefox, I think, is, because I mean Firefox in some European countries is up to 20,
30% now, probably not in the US, but it really is beginning to break open the I only web.
And the other thing that they didn't reckon on was the Flash video format, which is beginning to
crush the old Windows media format files, because it's a better streaming format. Although,
you know, this again, this is a proprietary application, and I'm hoping what there will
tell us much more about this later on, because I was really hoping you were going to be the audience
for this slide, because I'm thinking, God, I know nothing about this, I need someone who does.
Yeah, well, please do, and I believe you're giving us a talk on that later on,
so which I'm looking forward to.
Okay, well, there's no detail here either, so.
But the thing to remember is that this isn't really good for free software either,
and silver light is, oh, we've got a question there.
Yes, yes, yes, yes, so the comment was that the X in Ajax is actually a Microsoft invention.
I didn't say that we're all bad.
You know, they do some good things. That was a good thing. Yeah, I agree, and
there was a comment there.
Yeah, I suppose it's a smart slide to ask you since you're having some trouble in your Google
hat, but I can't put a phrase last night to the plane, the plane to KL is in the SDD, and Google
is in the market as well.
The slide we're talking about this, these issues, Google is really the figure threat of
upon the Ajax issue and the rich applications based on Microsoft.
I'm going to take the fifth on that.
Because somebody here is going to report back to my boss. I just know this.
So it's, as I said, Microsoft doesn't control the server here. This is hard of them to win.
I actually am mentioning this that I'm done. Yes, so my prediction for this long term is
fail on this one for Microsoft. I think this is the first one that they're going to really get
a black eye. IE8 is not going to regain the past glories of IE6, including the proprietary mess.
Silverlight made what I'm really hoping is that pressure from Silverlight from Microsoft is
going to force Adobe to completely open up Flash and make it free software friendly.
Oh, okay. Well, as of last week, it's sort of like, go swivel, but let's see what happens when
Microsoft bribe a lot more media corporations to go silverlight or, you know, because remember,
I know absolutely nothing about Google's internal processes. I'm free software facing or whatever.
So if Google adopted Silverlight for something, I mean, at that point, you might find
Adobe coming to the table really damn quickly. Well, okay, maybe free software is not such a bad idea,
but this is kind of politics. So the move to cloud computing, as Bradley points out,
will further rest control away from Microsoft, but this is not good for free software,
and this is where the ATPL, I think, is increasingly important and, you know, we should support it
as much as possible. So, ah, question.
Oh, I just was saying that, IE80 doesn't really support scalable vector graphics as well as
weather clouds. And I think that's the greatest improvement because the graphics oil is
oh, that's cool. Oh, sorry.
Yeah, and I'm not going to don't even see as far up for it. It's just creating all these
points and maybe it's the code to make it faster on your machine, and they also added a lot
of the tools and capacities of SDG.
Right.
With all these new like online gaming layout and virtual world-type environments,
and world-emergent environments, I'm wondering if Apple is going to turn around by
Oh, so the, so the, so the, so the comment is that SVG might be the tipping point that pushes
people away from IE80, yeah, because I don't know about Apple buying Adobe because Apple is,
Apple in the position of Microsoft terrifies me because Apple, Apple is, you know,
Apple is the Borg in terms of, they're one unified.
Well, I hate each other, which is, which is good for us. I think, yes. So, so as I point out,
trade secrets don't work anymore. The antitrust authorities are really breathing down the neck,
standards are gaining traction in the world. So, you know, what the hell are they going to do?
Well, what do you always do in a game when you're back against the wall?
You push the button. You go for the nuclear war option.
Or as I like to call this slide, global thermonuclear monopoly.
So, patents. So, patents are a big problem. And in fact, as I don't know how many of you know this,
but Bill Gates actually believes that patents were a massive problem. In 1991, he understood
perfectly. And in fact, he was quoting, and I was amazed by this when I went back and read
these references, again, from the Microsoft antitrust case notes. He was actually quoting here
from a document produced by the League for Programming Freedom, if anyone remembers that,
which was the first anti-pattern organization associated with the FSF. So, Bill Gates was a fan.
Of course, probably never gave us any money, but, you know, he understood completely the danger
that software patents would cost a Microsoft and to software in general. Of course, the view now
is somewhat different. Well, you know, well, well, we honor and support the honoring of intellectual
property, the Weasel word of our time. They're going to have to play by the same rules in the rest
of business. What's fair is fair, which really does sound like a petulant jar for me, but that's
that's just me. And then, of course, there's the more hard-nosed, well, open-source software of
I-8, 235 Microsoft patents. Okay, which ones? Well, they finally showed us one of them, which
was the Tom Tom one. So why are they finally turning to the nuclear option? And I think this is an
admission of failure. And they finally have a strategy that they think, I think, will work
against copy-left and GPL software, because they patents and patent licensing agreements
have the benefits to them of being completely incompatible with GPL and copy-left, but completely
ignored by, let's be honest here, dumbass licenses like MIT and BSD. So by making patent threats,
they are really giving two options. And they have been doing this for years. Don't ask me how I
know this, but I do know this for a fact. What they do is they go around to a company that's using
GPL software, and they say, nice products you have there. Wouldn't that be a shame if you
violated some of our patents? And company X says, well, what patents? And Microsoft says,
well, we could tell you that, but then if you violate and we win, it's triple damages. So
why don't you just sign a license on these very reasonable terms, which include violating the GPL?
And let's not upset those nasty free software people who make it first. Let's do it under an
NDA. And they have been doing that for years, and there have been many companies who have signed up
that. So what this is is an attempt to split our community. And, you know, I have some
personal experience with this. People who do patent crossed licenses are split in the community.
And I know they feel very sorry about it now. I'm not trying to make amends, but
very old cynical smart corporations like IBM and HP. They also do cross license with Microsoft,
and many other companies over patents, but they do not include free software. And that's the
difference. They're still done under NDA, but they're not, they do, the cross licenses that they
have. I don't know about some, but IBM to HP and HP, they do not include free software in the
patent cross licenses, which is why, you know, they're still good members of our community.
And so the interesting thing is, and we will not know this unless there's another antitrust suit
and a bunch more stuff gets released under evidence laws, but many other companies have agreed to
things under NDAs, and we don't know what they are. And my guess is that many of them are in
fact violating the GPL here. So the Tom Tom lawsuits, let's get up today. Of course, you know,
no one is a good guy here, because Tom Tom of turn around and cross the supermarkets off
of a violating their patents. So I mean, patents just, I mean, let's be honest, patents just kill
people, all right? I know they're medical patents, but in general, patents are a very bad thing.
So there are workarounds for the violations in the New Linux fat file system. Trich actually came
up with some interesting ones. You can make it do it read only. Tom Tom don't actually need to make
it read right. But they're not for this one, because Tom Tom can get out of this at least on the
free software, so I'm relatively simply, is this is the first openly aggressive move that they've
made. This is the first time they, you know, they did the, they shot the missile over their borders
kind of thing. They made the first aggressive move. And my guess is, looking at the Microsoft
Warring Tribes model, the legal department thought this was a good idea, rested control of
Balma, and finally said, yeah, we're going to do something about this. And the open source guys
are probably going, oh my god, what did they do? Because they just wrecked every single one of
some round uni guys at outreach efforts. They, they've, so there's a war within Microsoft over
this, I'm sure. As a little aside, actually, slide doesn't kind of fit here, but I couldn't think
of anyone else to put it, so probably should have moved it back a bit. The network war, the,
the last front that they're opening is really on netbooks, which were, which was a product that
came out of blue that they really didn't expect. And I think we started out with more than 10%
market share, we're now down, down to 10% market share. And they are aggressively moving to quash
this market for GNU Linux. And the way they do it is, yeah, you can offer GNU Linux, but it's not
going to be prominent on your website. You're going to have to sort of go down the stairs, open the
file cabinet in the lock door, and we're aware of the leopard before you can submit a web form to,
to buy a GNU Linux network. Because they no longer use the, what was determined illegal in the
first antitrust case in the 90s, in the 90s in the US, which were the per-processor license
agreements, which they used for dots. Now what they do is co-marketing dollars, which is like,
well, you know, you're advertising is very expensive, we could help offset some of that,
you know, which is why this talk for instance recommends use of Microsoft Vista for all your
business needs, yeah. So you will find on, you know, even the web pages advertising, you know,
GNU Linux systems, you will find, oh, we recommend the use of, because they have to put that crap
on there, because that's how they get the co-marketing dollars, which, you know, in the cuts for
business of PCs, they depend on. So netbooks, mobile phones, and appliances are really the breakout
markets, but we're in right now and winning. And I think this is where there's going to be the
most conflict between Microsoft and the free world. So what are our options, because I'm kind
of running out of time here, what are our options to the elephant in the room? Ah, yes, somebody's
waving something at me, I have no idea what it means, but I'll hurry up. So our first option is
to ignore it, which actually is a pretty good option. So, you know, we've created a wonderful system,
it's running everything that you're watching now, which is, you know, I've seen worse presentations
at Microsoft conferences. So we should just keep creating and sharing free software and free
content under our licenses. And to be honest, in the long run, this is how we will win. We must
concentrate on the competition too much. We have to keep doing what we're good at. Keep our eyes
on the prize, as they say, and create a free software world and make others, you know, communicate
with others that they can share this world with us. Basically, extend the helping hand to get
them out of the sewer. The other thing that we can do, which I also think is important, is we can
corral the elephant, which is harder, because elephants are, you know, kind of heavy than the hard
to corral. But if we can keep up, and this is where lobbying organizations and the anti-patent
stuff in the EU works, is keep pressure on governments to adopt open and uncorrupted standards
and to keep investigating monopolies, and to help out as much as we can. And transparency and
openness are really what we need here. As I put it, elephants do like to work in the dark. I
remember going to a set of government meetings about open standards, which Microsoft were there,
and they kind of put up a completely token presence. They turned up, they sort of read a statement
and they left, and I was like, wow, that was great, we won. And I spoke to someone who was a little
older and a bit more cynical than me. I told him what happened. He said, yeah, you know why? He said
they already had their meetings with the government officials before you turned up. And I'm like, oh
yeah, they probably did, didn't they? Yes. So they do like to work in the dark. And this one is
really important, because this is a thing that could sink all of it. Software patents, just lobby
against them wherever they are in it proposed. They are an evil and they must be destroyed,
you know, can't to quote one of the Roman senators and we should plow salt into their fields.
Eternal vigilance is a priority and we have to keep the US patent system out of the rest of the
world, which is kind of hard because the US is visually trying to push its patent system around
the rest of the world. But I believe actually patent trolls are our friends here, because
I was once in a very long phone call with some lawyers for a very large company and we were talking
about software patents and how word we were. And at the end of it, one of the lawyers finally got
an opening and said, look, you idiot, you're not in any danger from software patents, you haven't
done any money. Nobody's going to sue you, they're going to sue us. Oh yeah, kind of, yeah, that's
where the money is. So, so patent trolls tend to sue Microsoft because that's where the money is.
So, you know, eventually they've got to balance that the patent trolls are going to be more
damaging to them than the defense that they get out of threatening us with patents. So, I still
have the hopes that they will eventually see sense on that. So, the way we respond these
prior are, I mean, we have the community out there to do this in a distributed way,
grot law proved how it could be done with a skill case. Just destroy any patents that they pop
up over the paper, any patents that we can't destroy rewrite. It's, if you've actually readily
software patents, they're incomprehensible and it's, it's very hard to write a patent that you
can't actually work around. It's not impossible. You can either knock them down or work around.
The trouble is this is very expensive to do. So, the thing to do is anyone who pops up with these
dumbass patents, like the Tom Tom one, for instance, is to knock them down as quickly as possible
and make them understand that the more they do this, the more they will lose what they consider
their valuable intellectual property. And the other thing, I mean, I'm going to personally do is,
I'm going to get a beef with Sam Ramji because he's a nice guy, but I'm not going to let him
be hard behind company skirts on this one. And, you know, I'm not going to be nasty or anything,
but I'm going to say, looks, Sam, you're full of shit. Sorry. You know, you say all these things
in you minute, but look what your bosses do. And, you know, working for a company like that
is an individual moral choice. And, you know, hey, maybe someone else is still hiring.
And eventually, if they lose enough good free software people, because they do have some good
free software people inside, they will realize that this kind of aggressive tactic is counterproductive
on their own employees. So, I'm going to end on a positive note. So, elephants can learn new tricks
and eventually, and this might take as a take a while, but I don't see it impossible that will have
Microsoft engineers and representatives in an FSF general meeting. You know, I have five minutes,
but I'm finished already. So, we've got questions. Question there, yeah.
So, so Neely Krause is the EU antitrust commissioner. There were two points there. So, that's
about Neely Cruz leaving. I didn't know about that. So, that's news to me. I thought Neely Cruz was
doing a great job in the EU antitrust. I hope that it depends who they hire to replace, to replace
Neely. If their hire, if their hire, Microsoft exact, which is, you know, possible given the
politics of governments, then I think that would be a bad thing. So, it really depends who
replaces her. But I think, I think she set up the EU on a track that I hope that will continue
to follow. And, you know, they are essentially, they are becoming the premier antitrust watchdogs
in the world. And I think that's a good thing for us, you know. I think that's a good thing for
everybody. So, let's hope they continue doing that. And as for a sun buying, IBM buying sun,
man, that's just weird. It's kind of like, that means that they buy another SIF server.
So, I'm thinking with my stand-by hat on here, you know, because there's some of a SIF server in
the kernel, which I think is rubbish. But, yeah, I don't know. But it's only Wall Street rumours
right now, Wall Street Journal rumours. And, you know, something, these kind of rumours have a
way of actually destroying any actual reality behind it. So, even if it was happening,
there's been so much opinion about it, it might now not. But I don't think we ever had to worry
about Microsoft buying sun, because antitrust competition authorities would simply disallow that.
Microsoft will never be allowed to buy a PC maker, I think, and sun is a PC maker. So,
you know, because at that point, then you really do have standard oil controlling everything.
So, as I point out there, elephants can learn new tricks. And, you know, if you're old enough,
IBM used to be hated and feared just as much as Microsoft did today. And now IBM is a very,
you know, respected and happy and friendly member of our community. And they do a lot of great
things. They pay for our hosting members, for instance. So, yeah, they do a lot of scummy things
with patents, but who doesn't, right? I mean, that's what corporations do. So, yeah, I really do hope
that Microsoft, my feeling is that Microsoft have to go through the same kind of pain that IBM
went through in the early to mid-90s. But eventually, they'll come out on the other side.
They might not be a completely free software company, but I think there'll be a lot more free
software friendly. So, it is, our job was good understanding citizens to help them through this
pain and to keep kicking them when they do dumb things, which I hope I've done today.
So, any other questions or comments? Yeah.
I think I'll come into the details. Sure. She will be leaving in November because her
term ends and it's terrible last, because she's great at displaying about hindering monopolies,
but we don't have to worry too much because monopolies, they hate her whole department.
So, I don't know if you heard that, but Nellie is leaving, but the whole department is just as
despised. So, that's wonderful. Yeah. So, any other comments, questions?
All right. Well, thank you very much.
Thank you for listening to Half of the Public Radio.
HPR is sponsored by Kero.net, so head on over to C-A-R-O-O-D-E-C for all of us here.