Scenario: AOLTW reduces subsidization of moz development

1 view
Skip to first unread message

Ralph Mellor

unread,
Apr 8, 2002, 8:42:12 AM4/8/02
to
AOLTW is currently spending a lot of money subsidizing mozilla
development. [1]

This could stop.

So, my main question is, how would AOLTW ensure that any reduction of
subsidy did not unduly harm moz?

Has this question been discussed before? Have AOLTW got an official
position regarding how and when they would reduce their role?

--
ralph
Remove XXX if emailing me.

[1] The cost of mozilla development must surely far exceed $10m a year.
This is small change relative to AOL's ten billion dollars of earnings a
year. It's also arguably small relative to the strategic importance of a
browser to AOLTW. However, it might be enough to upset impatient
shareholders (not to mention any AOLTW executives that do not buy into
the mozilla strategy) if times are hard and especially if significant
Netscape revenue does not materialize in the next 6 months.

Christian Biesinger

unread,
Apr 8, 2002, 10:25:58 AM4/8/02
to
Ralph Mellor wrote:
> [...] especially if significant

> Netscape revenue does not materialize in the next 6 months.

You talk like you know how much revenue Netscape does materialize - how
much does it?


--
"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety."
-- Benjamin Franklin

Jay Garcia

unread,
Apr 8, 2002, 12:01:04 PM4/8/02
to

Since AOL is highly dependant on Gecko development, there is no way this
will happen. You HAVE been 'reading' the news ?? yes ??

--
Jay Garcia - Netscape Champion
Novell MCNE-5/CNI-Networking Technologies-OSI
UFAQ - http://www.UFAQ.org
** Post To Group ONLY, do NOT email **

JTK

unread,
Apr 8, 2002, 1:17:15 PM4/8/02
to
Ralph Mellor wrote:
> AOLTW is currently spending a lot of money subsidizing mozilla
> development. [1]
>
> This could stop.
>
> So, my main question is, how would AOLTW ensure that any reduction of
> subsidy did not unduly harm moz?
>

Huh?:

1. Why would AOL *care*? If they were to withdraw their support for
Mozilla development, the reason would be that they no longer cared about
the project or what their withdrawl would mean to it.

2. If AOL withdrew support from Mozilla, development would cease
completely. Windows users don't need a half-assed
non-compliant-to-defact-standards browser, and there isn't a *nix
developer alive or dead that has the first notion about how to do a
decent GUI. The best you could hope for would be (assuming AOL released
everything under the GPL, which won't ever happen) that somebody rips
out the newsreader and gets that running on Windows standalone. That
might possibly be a viable project.

Christian Biesinger

unread,
Apr 8, 2002, 11:58:43 AM4/8/02
to
JTK wrote:

> Ralph Mellor wrote:
>> So, my main question is, how would AOLTW ensure that any reduction of
>> subsidy did not unduly harm moz?
>
> 1. Why would AOL *care*? If they were to withdraw their support for
> Mozilla development, the reason would be that they no longer cared about
> the project or what their withdrawl would mean to it.

JTK, your opinions get more and more reasonable.

> 2. If AOL withdrew support from Mozilla, development would cease
> completely.

1. Current developers can continue developing Mozilla (maybe less than
before, but they can)

> Windows users don't need a half-assed
> non-compliant-to-defact-standards browser

There are enough windows users who are using Mozilla.

> The best you could hope for would be (assuming AOL released
> everything under the GPL, which won't ever happen)

In this case, MPL is just as good as GPL.

As you might know, the relicensing to triple-license [MN]PL/GPL/LGPL is
nearing completion, I was told.

Ralph Mellor

unread,
Apr 8, 2002, 5:41:52 PM4/8/02
to
Christian Biesinger wrote:
> Ralph Mellor wrote:
> > [...] especially if significant
>> Netscape revenue does not materialize in the next 6 months.

Hmm. On reflection, the above is irrelevant to my point. It does not
matter *why* AOLTW might choose to reduce subsidizing moz development,
it merely matters that they might.

Also, I was a little loose with language there. Although I now consider
it irrelevant, I'll note that what I meant was something more like:

"especially if significant extra revenue attributed by AOLTW to having
developed moz to this point does not materialize in the next 6 months."

---

As I said above, I now consider this a red herring. However, you posted
so I'll respond.

> You talk like you know how much revenue Netscape does materialize - how
> much does it?

Perhaps I talked like that, but I did not mean to do so. I certainly do
not know how much revenue is accruing at the moment.

I do recall that Netscape revenues, around the time the browser was
freed, were very roughly $600m a year, and that, at peak, revenues
directly attributed to the browser were very roughly $150m a year, and
that AOLTW paid very roughly $4b for Netscape. These numbers might set
the parameters for what AOLTW is expecting. Or they may not.

--
ralph

Ralph Mellor

unread,
Apr 8, 2002, 6:04:57 PM4/8/02
to
> Since AOL is highly dependant on Gecko development,
> there is no way this will happen.

Afaik, AOLTW has not yet officially committed to having
Gecko be the default browser in AOL8, but it's clear
they would very much like to do that. So I'll allow the
first bit. (Though it's still theoretically plausible
that, in a fit of madness, AOLTW execs could decide to
give up on Gecko and go with IE.)

But I don't agree with your conclusion. For starters,
Gecko is not the same as moz. If one reduced the scope
of moz development to, say, just delivering Gecko,
then that would itself be a huge reduction.

--
ralph

Dan Howard

unread,
Apr 8, 2002, 6:10:35 PM4/8/02
to
JTK wrote:
Windows users don't need a half-assed
> non-compliant-to-defact-standards browser,


That's because they've already got one.

us...@domain.invalid

unread,
Apr 8, 2002, 6:24:36 PM4/8/02
to
>> So, my main question is, how would AOLTW ensure that any
>> reduction of subsidy did not unduly harm moz?
>
> Huh?:
>
> 1. Why would AOL *care*? If they were to withdraw their
> support for Mozilla development, the reason would be that
> they no longer cared about the project or what their
> withdrawl would mean to it.

I wasn't asking about the black and white case of withdrawing
ALL support. I'm asking about what happens if they decide to
cut back even a little.

> 2. If AOL withdrew support from Mozilla, development would
> cease completely.

Would it? Perhaps it would. Perhaps IBM, Sun, HP and others
would take up the reins. Perhaps the infrastructure and doc
of existing code architecture is good enough that the project
could survive an instant drop of corporate sponsorship. (Note
that I said infrastructure and architecture doc, not code; I
believe at this point that the code is good enough, so the
only issue is whether new people could pick up the pieces
(always assuming that most of the current paid coders would
not remain available for long if the money went away).)

But anyhow, total withdrawal isn't the scenario I am asking
about, partly because I think it a very remote possibility.

> Windows users don't need a half-assed non-compliant-to-
> defact-standards browser

Perhaps. For the most part, I don't think they care.

> and there isn't a *nix developer alive or dead that has
> the first notion about how to do a decent GUI.

Oh I don't think that's true. Many people, including Bill
Gates, feel that the Mac incorporates some of the best GUI
tech available in a mainstream produc, and many of those
people think Mac OS X is the best in Mac GUI.

So I'd say instead that the best gui today was produced
by someone (Jobs and crew) who recently spent 10 years
in the *nix space.

> (assuming AOL released everything under the GPL, which
> won't ever happen)

Huh? ALL of Moz is GPLed. I'm beginning to wonder if
you have any clue what's going on.

--
ralph

us...@domain.invalid

unread,
Apr 8, 2002, 6:39:55 PM4/8/02
to
>>> So, my main question is, how would AOLTW ensure that
>>> any reduction of subsidy did not unduly harm moz?

Christian Biesinger wrote:
> 1. Current developers can continue developing Mozilla
> (maybe less than before, but they can)

Perhaps. But I've watched several open source projects
really struggle when they merely lose some "momentum",
let alone a reduction in corporate sponsorship.

I recall quite a stink when AOLTW dropped Mitchell Baker
off the payroll. What if they dropped 10 people?

My concern here is for AOLTW to present a credible case
that moz momentum is not completely tied to their good
graces.

--
ralph

DeMoN LaG

unread,
Apr 8, 2002, 8:01:55 PM4/8/02
to
us...@domain.invalid wrote in news:3CB218A4...@domain.invalid, on
08 Apr 2002:

> Huh? ALL of Moz is GPLed. I'm beginning to wonder if
> you have any clue what's going on.

No, he doesn't. Seems most of the time he never did

--
AIM: FlyersR1 9
email: de_on-lag@co_cast.net
_ = m

Garth Wallace

unread,
Apr 8, 2002, 8:04:25 PM4/8/02
to

AOL/Netscape isn't the only major corporate contributor. Red Hat and IBM
have also contributed quite a bit of time and effort into the project,
and they're hardly insubstantial entities.

Jay Garcia

unread,
Apr 8, 2002, 9:36:37 PM4/8/02
to
On 04/08/2002 7:04 PM, Garth Wallace wrote:
>
> AOL/Netscape isn't the only major corporate contributor. Red Hat and IBM
> have also contributed quite a bit of time and effort into the project,
> and they're hardly insubstantial entities.
>

Also, Sun Microsystems has been a major player.

Emlyn

unread,
Apr 9, 2002, 3:09:23 AM4/9/02
to
> The best you could hope for would be (assuming AOL released
> everything under the GPL, which won't ever happen) ....

Okay. To be clear about this, I am not in favour of the GNU GPL.
Nevertheless, I must point out that almost the whole of the Mozilla
codebase (including dust-covered, long-forgotten projects such as
Rhino) is under a license that allows distribution under the GNU LGPL
or GNU GPL. And - and pay attention - the bits that do not yet conform
to such a scheme are the bits under the Mozilla Public Licence. All
*N*PL'd files are disjunctively NPL/LGPL/GPL'd. Dispite the best
efforts of Mozilla.org to track down and get agreement from outside
contributors, a few MPL'd files are still not moved over to the new
scheme. It is in no way AOLTW's responsibilty that we don't yet have
complete GNU GPL compatability. They have done everything that they
can do.

Somedays, JTK is as bad as the Mozillaquest guy. And the rest, he's
but marginally better. At least get your facts straight.


To be clear, here are some valid criticisms of Mozilla:

Lack of Polish
Icky UI in places
Does not act like a Macintosh application under Mac OS (9 or X)
Memory Consumption


Here are some invalid criticisms of Mozilla:

Crashes all the time - this is not the case. We (that is, the people
who download mozilla.org binaries) get the idea that mozilla is
unstable and crash-y because we're always living on the edge of
unstable development. Branches have been rock solid in the past, and
no doubt in a few months the mozilla 1.0 branch will be more stable
than IE (whose approach to stability seems to be "display a dialog box
apologising for the crash")

Rendering bugs - there are few of these. We've all seen lots, but what
may not occur to you is that you only get a few rendering bugs per
build. Because we build-sluts use builds and throw them away by the
dozen, we get an impression of more rendering bugs than the user who
uses a single version of a mozilla-derived program would. This applies
to bugs of all stripes to some extent.

Speed - Just get a faster computer. I mean, really, apart from
new-window performance and (this is the big one) slow-down related to
swapping mozilla in and out of memory, Mozilla on my two-year old G4
Cube 450mhz is just as fast or faster than IE5 at everything bar
DHTML. But we were on top of the DHTML thing at one stage, and
hopefully we can be again. If mozilla works fast enough on a two year
old computer that was by no means top of the line when I bought it, it
should be good enough for most people (or at least, most people with
more than 128mb of ram)


Valid Criticisms of Mozilla.org

They don't push people enough towards writing documentation
Too hard for newbies to enter the project


Invalid criticisms of Mozilla.org

Puppet of AOL - uh huh. Sure. But did you know that AOL was a puppet
of Starbucks, and StarBucks is a subsidy of Virtuacon, and Dr Evil
owns Virtucon? So, really, it's all Dr Evil's fault.

Only interested in making Proprietary Software by capitalising on the
Open Source methodology - I don't think so. Why, in that case, do they
release the codebase under the GNU GPL?


Valid Criticsms of Netscape

Don't always work on the top priorty bugs
They make links on the mozilla.org status page like
"http://status/..." which only work for people on a particular side of
the Netscape firewall. The side that I'm not on
Generally act like they run the show
Released Netscape 6.0


Actually, maybe there are lots of criticsms of Netscape. But few of
Mozilla.org, or the code they distribute.


Now, in future, could you please stick to the above lists of
criticisms that have some basis in reality. Thank you.

-Emlyn

Ralph Mellor

unread,
Apr 9, 2002, 6:05:02 AM4/9/02
to
I would like to refocus this thread if possible and if
y'all still have time.

I am preparing an analysis for a client. That analysis is
whether they should adopt mozilla as a platform on which
to base a product.

That's where I'm coming from. I thought I'd use this group
as a way to feel my way through the issue in the subject
line.

So...

Mozilla is free, in the constitutional/legal sense. [1]
Freedom in this sense, more so than access to the source,
reduces reliance on a platform vendor. This is a HUGE plus.

But moz is not yet in the sort of position enjoyed by linux
or apache. For both of these, it is very clear that they
will survive and prosper regardless of what any specific
player might do. This gives a warm fuzzy feeling for any
party considering building on linux of apache. The same
is not yet true, imo, of moz.

First, the majority of the funding for moz currently comes
from a single player. [2]

Second, moz was born from an existing ambitious proprietary
project, and in its infancy became an even more ambitious
project which was only made possible because the money was
there. That's not the way most open source projects evolve.

The above two factors remind me of several GPLed platform
type projects from which the principal sponsor withdraws
with ill effect. [3] These situations leave those who built
on them in a position that is at least difficult. (Even
though it is certainly no where near as technically
untenable as it would be with a proprietary platform.)

Interestingly, moz may *never* arrive at the same position
as linux and apache, or at least be perceived to be at that
position, unless ALOTW change their role, very possibly by
*reducing* their funding! So, paradoxically, it may be that
it is in the best interests of mozilla (as a platform not a
product) for AOLTW to reduce their funding of moz, even if
they do not otherwise want to!

Anyhow, my concern here is to consider the scenario in
which AOLTW reduces subsidization of moz development.

--
ralph

[1] I view Mozilla.org and the Mozilla codebase as being
free now and forevermore in the constitutional/legal sense.
(I realize that the GPL could be challenged and all that,
but I am not interested in that debate in the context of
this thread.)

[2] Things *do* seem to be headed in the right direction.
Last I looked, there were about 5 contributors of record
that were not funded by AOLTW for each one that was funded
by AOLTW. Obviously the AOLTW funded coders are, on
average, putting in a lot more time than the others, but
the quality of their collaborative tools and methods, and
the trends, are clearly in favor of this contribution ratio
continuing to improve. (In this they have, imo, triumphed
in a major way. Microsoft are many years behind on this,
even as they dabble with "shared source".)

[3] There's been several web application servers that have
died this way. Or consider plex86, the open source project
to compete with vmware.

Gervase Markham

unread,
Apr 9, 2002, 8:24:19 AM4/9/02
to
> But moz is not yet in the sort of position enjoyed by linux
> or apache. For both of these, it is very clear that they
> will survive and prosper regardless of what any specific
> player might do. This gives a warm fuzzy feeling for any
> party considering building on linux of apache. The same
> is not yet true, imo, of moz.

As you say, the long term future of Mozilla depends on how many
different companies have an interest in doing things with the codebase.
An incomplete list might be:
- AOL/TW, who appear to be making it the basis of their next-generation
AOL client. In this scenario, you have to ask yourself what the
likelihood is of them reducing their commitment.
- Red Hat, and other Linux vendors which ship Mozilla-based browsers
Any other company which would like to see Linux on the desktop
Sun, who are ramping up a large development effort to make a browser for
Solaris
OEone
IBM, who make the web browser for OS/2
A host of smaller companies (Crocodile Clips, Beonex, Worldgate etc.)

> First, the majority of the funding for moz currently comes
> from a single player. [2]

This is true. One reason is that this player is currently happy to fund
at this level, and there has been no need to seek alternative sources of
funding for mozilla.org infrastructure.

Gerv

Florian Ramillien

unread,
Apr 9, 2002, 8:31:32 AM4/9/02
to
Gervase Markham wrote:
>> But moz is not yet in the sort of position enjoyed by linux
>> or apache. For both of these, it is very clear that they
>> will survive and prosper regardless of what any specific
>> player might do. This gives a warm fuzzy feeling for any
>> party considering building on linux of apache. The same
>> is not yet true, imo, of moz.
>
>
> As you say, the long term future of Mozilla depends on how many
> different companies have an interest in doing things with the codebase.
> An incomplete list might be:
> - AOL/TW, who appear to be making it the basis of their next-generation
> AOL client. In this scenario, you have to ask yourself what the
> likelihood is of them reducing their commitment.
> - Red Hat, and other Linux vendors which ship Mozilla-based browsers
> Any other company which would like to see Linux on the desktop
> Sun, who are ramping up a large development effort to make a browser for
> Solaris
> OEone
> IBM, who make the web browser for OS/2

We can had :
HP, who make the web browser for HP/UX

Christian Biesinger

unread,
Apr 9, 2002, 8:45:07 AM4/9/02
to
us...@domain.invalid wrote:
> Perhaps the infrastructure and doc
> of existing code architecture is good enough that the project
> could survive an instant drop of corporate sponsorship.

Infrastructure perhaps is - documentation is definitely not.

> Huh? ALL of Moz is GPLed. I'm beginning to wonder if
> you have any clue what's going on.

That's very wrong.
See for example:
http://lxr.mozilla.org/seamonkey/source/config/Makefile.in
http://lxr.mozilla.org/seamonkey/source/dom/src/base/nsWindowRoot.cpp

And probably many others. These are just two examples.
Many others are NPL or MPL only. (though soon, after contacting all
contributors to MPL files for their agreement, all files will be
relicensed to a NPL/MPL + GPL + LGPL triple-license).

Christian Biesinger

unread,
Apr 9, 2002, 8:53:43 AM4/9/02
to
Emlyn wrote:
> And - and pay attention - the bits that do not yet conform
> to such a scheme are the bits under the Mozilla Public Licence. All
> *N*PL'd files are disjunctively NPL/LGPL/GPL'd.

Wrong... I just stumpled over a NPL only file (though I also assumed
that all NPL files were triple-licensed) (the file is config/Makefile.in)

> Valid Criticisms of Mozilla.org
>
> They don't push people enough towards writing documentation

Well, you can't force people to write docs.

> Too hard for newbies to enter the project

True - to some extent due to lack of documentation.

us...@domain.invalid

unread,
Apr 9, 2002, 10:41:37 AM4/9/02
to
Gerv, the following is all devil's advocate stuff. I
am so pro moz the platform from a gut instinct point
of view that I have to do this to keep my wits about
me. I've been trying to convince my client to consider
a plan for building on moz for 2 years, and I want my
arguments to be super sharp. I've always told them
moz wasn't ready for prime time, so they didn't need
to make a decision yet, but I've recently been guiding
them to think in terms of the next few months as being
the time to get off the fence or miss the boat (mmm,
nice mixed metaphor).

Anyhow, thanks for input thus far and any further.


Gervase Markham wrote:
> - AOL/TW, who appear to be making it the basis of their next-generation
> AOL client. In this scenario, you have to ask yourself what the
> likelihood is of them reducing their commitment.

When you say "next-generation AOL client", do you mean
AOL 8, or something later? If you just mean AOL 8, then
I think we're just talking Gecko. Gecko is a subset of
moz the suite, in turn a subset of moz the platform.

I would assume AOLTW are committed to continuing, for the
forseeable future, their practice of selling custom AOL
client software. As such, they need a codebase for that,
and must be at least open to the idea of using moz the
suite as the basis for that. Hmm. I guess you can not
comment, right?

Anyhow, their use of Gecko, and even their interest in
moz the suite (as against moz the platform), still does
not make a compelling argument to me that AOLTW will
want to continue to fully fund the entire platform
effort.

And, to apparently contradict myself, I'm also worried
about several negative aspects of AOLTW continuing to
fully fund the project. For example, my client would
be extending a piece of moz using a mix of open and
proprietary extensions. It could be disastrous if some
other party were to get some unfair advantage due to
a closer relationship with the "powers-that-be" at the
project, especially if that party were AOLTW themselves.
That argues against having dominant funding players.
Basically, the notion of a platform doesn't play well
with the notion of dominant players being in de facto
control of the platform.

> - Red Hat, and other Linux vendors which ship
> Mozilla-based browsers

I just recently read that RH have about $50m in the bank,
and lost about $25m last year. I really don't think they
can be relied upon to step up to the plate in anything
but a relatively minor capacity. The same sort of thing
applies to pretty much all the players I can think of,
short of the likes of Sun, IBM, HP.


> Sun, who are ramping up a large development effort
> to make a browser for Solaris

Is contributing to moz, the platform, necessarily a good
thing for Sun?


> OEone

OEone tout the notion that it's good for them to put their
efforts back in to the public codebase, because a key thing
they offer is the reverse of vendor lock in. I happen to
think that this IS in the long run, the smart thing to do.
But I've seen many vendors reverse themselves on this.
The long term ArsDigita situation still isn't clear, but
it stands as a lesson, for example.

So, would OEone contribute their moz, the platform, efforts,
to everyone, in the event that AOLTW backed off of moz the
platform? Maybe. Maybe their investors would force them to
not do so. And maybe this pattern would repeat for pretty
much all the smaller players that are struggling to make
money and have technical strategy subject to investor
pressure.


> IBM

Yeah. I think this would be a no brainer for them. Plus
I would love it if IBM became a really major player in
developing moz the platform. I'm pretty sure IBM won't
compete in the space my client is in. Interesting.


> A host of smaller companies (Crocodile Clips, Beonex,
> Worldgate etc.)

Oh I'm sure they'd keep things moving along a little.
I just don't know whether their collective efforts,
including those of the bigger boys, would sustain
moz-as-platform vitality.


>> First, the majority of the funding for moz currently comes
>> from a single player. [2]
>
> This is true. One reason is that this player is currently
> happy to fund at this level, and there has been no need
> to seek alternative sources of funding for mozilla.org
> infrastructure.

I always thought AOLTW would see this thing out till at
least a while (a year?) after 1.0. But I'm consulting for
a client whose products have a 10-15 year product lifetime.
I have to think ahead and present a compelling case for the
likely longevity and vitality of a given platform.

--
ralph

Garth Wallace

unread,
Apr 9, 2002, 1:13:22 PM4/9/02
to
us...@domain.invalid wrote:

>
> Gervase Markham wrote:
> >
> > Sun, who are ramping up a large development effort
> > to make a browser for Solaris
>
> Is contributing to moz, the platform, necessarily a good
> thing for Sun?

Can't see why not. Judging by the fact that they created Java, they're
not averse to cross-platform interoperability.

> > IBM
>
> Yeah. I think this would be a no brainer for them. Plus
> I would love it if IBM became a really major player in
> developing moz the platform. I'm pretty sure IBM won't
> compete in the space my client is in. Interesting.

They've already contributed much of the work on bidirectional text
rendering, among other things.

Gervase Markham

unread,
Apr 9, 2002, 1:25:21 PM4/9/02
to
> Huh? ALL of Moz is GPLed. I'm beginning to wonder if
> you have any clue what's going on.

That's not true. mozilla.org is moving towards an [M|N]PL/GPL/LGPL
tri-license, but we aren't there yet. See http://www.mozilla.org/MPL .

Gerv

Gervase Markham

unread,
Apr 9, 2002, 1:34:45 PM4/9/02
to
us...@domain.invalid wrote:

Please could you use a vaguely-valid address, even if it's spamproofed?

> > - AOL/TW, who appear to be making it the basis of their next-generation
> > AOL client. In this scenario, you have to ask yourself what the
> > likelihood is of them reducing their commitment.
>
> When you say "next-generation AOL client", do you mean
> AOL 8, or something later? If you just mean AOL 8, then
> I think we're just talking Gecko. Gecko is a subset of
> moz the suite, in turn a subset of moz the platform.

I'm not privy to AOL's plans, but I'm drawing inference from the sound
business case that I see for basing the AOL client on Gecko. You are
right - it's not the entire suite.

> Anyhow, their use of Gecko, and even their interest in
> moz the suite (as against moz the platform), still does
> not make a compelling argument to me that AOLTW will
> want to continue to fully fund the entire platform
> effort.

That's a good point. If AOL/TW stopped being interested, other parties
would have to step in to take up the slack.

However, I think it's fair to say that it's unlikely that Moz will ever
get to the stage that it's so ubiquitous that you don't need to ship all
the components you need with your custom app. So wide distribution of
the browser suite should have little effect on the business case for
building your product on Mozilla technology.

> And, to apparently contradict myself, I'm also worried
> about several negative aspects of AOLTW continuing to
> fully fund the project. For example, my client would
> be extending a piece of moz using a mix of open and
> proprietary extensions. It could be disastrous if some
> other party were to get some unfair advantage due to
> a closer relationship with the "powers-that-be" at the
> project, especially if that party were AOLTW themselves.
> That argues against having dominant funding players.

If you want to avoid dominant funding players, you need to find
alternative sources of funding. If your company is willing to fund a
mozilla.org staff position, that would help. :-)

> > - Red Hat, and other Linux vendors which ship
> > Mozilla-based browsers
>
> I just recently read that RH have about $50m in the bank,
> and lost about $25m last year. I really don't think they
> can be relied upon to step up to the plate in anything
> but a relatively minor capacity. The same sort of thing
> applies to pretty much all the players I can think of,
> short of the likes of Sun, IBM, HP.

Red Hat have employed Chris Blizzard to hack on Mozilla for quite a
while now.

> > Sun, who are ramping up a large development effort
> > to make a browser for Solaris
>
> Is contributing to moz, the platform, necessarily a good
> thing for Sun?

You'd have to ask them. But you need Moz, the platform, to make the
browser suite.

> I always thought AOLTW would see this thing out till at
> least a while (a year?) after 1.0. But I'm consulting for
> a client whose products have a 10-15 year product lifetime.
> I have to think ahead and present a compelling case for the
> likely longevity and vitality of a given platform.

If your business model will depend on other companies doing work on Moz
for you, then that's definitely a risk. But the code is Free - it's not
going to suddenly become unavailable or stop working as well as it does
now if some contributors pull out. And this will be true in 15 years.

What are your alternative options to Mozilla? I certainly wouldn't rely
on IE being around with backwardly-compatible APIs in 15 years time. But
the Mozilla code you ship with your product will be.

Gerv

Stuart Ballard

unread,
Apr 9, 2002, 1:58:22 PM4/9/02
to
us...@domain.invalid wrote:
>
> When you say "next-generation AOL client", do you mean
> AOL 8, or something later? If you just mean AOL 8, then
> I think we're just talking Gecko. Gecko is a subset of
> moz the suite, in turn a subset of moz the platform.
>
> I would assume AOLTW are committed to continuing, for the
> forseeable future, their practice of selling custom AOL
> client software. As such, they need a codebase for that,
> and must be at least open to the idea of using moz the
> suite as the basis for that. Hmm. I guess you can not
> comment, right?
>
> Anyhow, their use of Gecko, and even their interest in
> moz the suite (as against moz the platform), still does
> not make a compelling argument to me that AOLTW will
> want to continue to fully fund the entire platform
> effort.

I disagree with this analysis right from the top... I'm not *100%* sure
(since everyone seems to have a different idea of what "moz the
platform" means) but here's how I'd draw a diagram of Moz as a platform:

Moz the suite Your app <dream on>AOL9?</dream on>
\ | /
\ | /
AOL8? "Moz the platform"
\ / | \
\ / | \
\ / | \
Gecko XPCOM NetLib etc
\ | /
GFX | /
\ | /
NSPR


This is simplified because drawing ascii graphics is making my brain
hurt :) XPCOM should probably be below NetLib-etc, and there ought to be
nodes for XUL, XBL, JS, etc (they'd probably live directly under "Moz
the platform" - XUL and XBL are built on Gecko, XBL is built on JS, etc.

By this kind of analysis, "Moz the suite" is a *superset* of "Moz the
platform". Not because the suite is a platform in itself, but because
the suite is *built on the platform*. AOL can certainly direct funding
to the parts of the platform that their specific suite relies on, rather
than other parts, but they need to keep funding the vast majority of the
platform just to be able to run the suite. The bits that are left can
comfortably be picked up by hobbiest hackers and other companies - to
some extent this already happens.

So if you take as given AOL's interest in "moz the suite", you don't
really need to worry about the status of the platform. You suggest that
AOL might want to build a future version of AOL on top of "moz the
suite". I disagree, but keep reading. None of us (probably even most of
the AOL/NS employees) know what AOL's plans actually are, but this is
what *I'd* do if I were AOL.

AOL 8 is (allegedly - but I'm taking this as fact) going to use Gecko as
its renderer for web pages, replacing IE. This is a fairly minimal
change, since they already use IE only in a limited aspect. However, it
seems to me that AOL is already 3/4 of the way to being a mozilla-like
application - it uses its own toolkit with a fluffy, image-based "skin",
it is a monolithic application (no need for fast spawning of new
processes based on it) and it's largely based on laying out text and
images, with some network functionality going on under the hood.

So, if *I* were AOL, I'd be thinking that AOL 9 or 10 would replace
their existing proprietary toolkit and infrastructure with one based on
"moz the platform". Then Gecko wouldn't just be a bolted-on module, but
the heart of the system. I don't know much about their existing
infrastructure, but the mozilla platform is certainly newer and probably
more flexible. So this seems like a very real possibility to me.

What I *wouldn't* do is base a future AOL client on "Moz the suite". AOL
users are pretty resistent to change - many still use AOL 4 or 5 when 7
is current, and none of the interim versions have made any major
user-visible changes. Their email program doesn't do mailboxes, at least
by default. Their interface is MDI. In short, it's pretty stable and
mature, but *doesn't look a lot like moz the suite*. Therefore, I'd
build a mozilla application that looks exactly like the existing AOL
client, but I'd do it from scratch, not based on the browser suite.

So we've established a (plausible, at least) scenario for AOL's interest
in moz the platform, but have we suggested that their interest in the
suite itself is limited? I don't think that's necessarily true either.

AOL recognizes that not everyone wants to use their variety of
simplified, sandboxed web browsing. They provide AIM and ICQ IM clients,
because they knew that "one size didn't fit all". As such, I think they
recognize an advantage in having a browser suite for the more, um,
technically sophisticated web user. I see the Netscape web browser suite
continuing - 6.0 hurt it a lot, but in my experience people are
beginning to give it a fresh look based on 6.2, and the 1.0-based
(allegedly) upcoming 6.5 release will be orders of magnitude better
again.

I see the Netscape brand being what they push to corporate users,
technical users and people who want a little bit more flexibility than
they can get from the AOL client. Keeping the two products separate, but
based on the same platform, will allow AOL to serve both markets at a
much lower cost than developing two clients separately, and much more
effectively than trying to shoehorn both into a single user interface.

And if they're developing *two* products based on the platform, they
have all the *more* interest in making sure the platform itself is
flexible and able to be used for more than just a single suite.

Of course, that's just my opinion, I could be wrong.

Stuart.

--
Stuart Ballard, Programmer
NetReach - Internet Solutions
(215) 283-2300, ext. 126
http://www.netreach.com/

Ralph Mellor

unread,
Apr 9, 2002, 2:11:34 PM4/9/02
to

Urg. That was sloppy of me. I knew it wasn't 100% complete.

Otoh, my assumption was that it was already, say, 90% plus
complete (measured by SLOC). Is that a fair assessment?

--
ralph

Ralph Mellor

unread,
Apr 9, 2002, 2:15:54 PM4/9/02
to
Garth Wallace wrote:
>> Is contributing to moz, the platform, necessarily a good
>> thing for Sun?
>
> Can't see why not.

Well...

> they created Java

The context of this thread is Mozilla as a cross
platform environment. In some respects, Java and
Moz are alternatives/competitors. It's not clear
to me if Sun would currently choose to take the
lead in driving Moz forward.

--
ralph

Ralph Mellor

unread,
Apr 9, 2002, 3:37:49 PM4/9/02
to
Gervase Markham wrote:
> us...@domain.invalid wrote:
>
> Please could you use a vaguely-valid address, even if it's spamproofed?

Sorry. [1]


> ... the sound business case that I see for basing the AOL
> client on Gecko.

What is the business case that you see? In particular,
why doesn't it make more sense for AOLTW to focus on
their business and let M$ focus on theirs, and therefore
use IE for the AOL client browser window and forget this
moz thing already?


> If AOL/TW stopped being interested, other parties
> would have to step in to take up the slack.

Yes, and, partly reiterating points I've made earlier, three
things follow from this:

1. The mere fact it is *plausible* that AOLTW could back off
from moz the platform means that this will be a concern,
even if unconscious, for pretty much anyone or any company
that is considering moz as a platform. As such there should
be information and marketing that addresses this issue, or
many would-be adopters might pass on moz.

2. If AOLTW were to back off, unless the announcement was
handled with great care, many members of the press and
public could view the change negatively, regardless of
whether the slack was thoroughly taken up.

3. Would the slack be adequately taken up?

> However, I think it's fair to say that it's unlikely that Moz will ever
> get to the stage that it's so ubiquitous that you don't need to ship all
> the components you need with your custom app. So wide distribution of
> the browser suite should have little effect on the business case for
> building your product on Mozilla technology.

In general, I agree. As far as the specific product I am
proposing my client builds, distribution IS relevant, but
that's not something I want to get into at this point.

> If you want to avoid dominant funding players, you need to find
> alternative sources of funding. If your company is willing to fund a
> mozilla.org staff position, that would help. :-)

Hmm. I doubt they would spring for the full cost of a position.
I could imagine them being prepared to cough up, say, $20K a
year as a contribution. There's a lot of logic to that. Hmm.
I could see them getting benefits like A) help getting up to
speed on the project structure and collaborating with netscape
employees etc.; B) getting their name out there as a corporate
sponsor of the whole shebang; C) having some representation of
their interests. Hmm. Does that sound remotely plausible?


> If your business model will depend on other companies doing
> work on Moz for you

Absolutely. A key part of the pitch is, adopt a more competitive
development cost structure by building on something that already
incorporates many man years worth of relevant work and is set to
have many more done for you as time passes.


> then that's definitely a risk. But the code is Free - it's not
> going to suddenly become unavailable or stop working as well
> as it does now if some contributors pull out. And this will
> be true in 15 years.

Well, not immediately. But it might begin to rot, unless my
client takes over sufficient development responsibilities,
and that may well defeat the key point just described above.


> What are your alternative options to Mozilla?

The current product is a Windows desktop application.


--
ralph

[1] I occasionally forget to select a valid address from
the moz news composer From: drop down. us...@domain.invalid is
the default moz puts in so I have to remember to change it.
I've noticed quite a few us...@domain.invalid posts in this ng
and I know I didn't post the others so there's at least two of
us with something wrong in our moz setup...

JTK

unread,
Apr 9, 2002, 4:07:03 PM4/9/02
to
Jay Garcia wrote:
> On 04/08/2002 7:04 PM, Garth Wallace wrote:
>
>>AOL/Netscape isn't the only major corporate contributor. Red Hat and IBM
>>have also contributed quite a bit of time and effort into the project,
>>and they're hardly insubstantial entities.
>>
>
>
> Also, Sun Microsystems has been a major player.
>

Well it's not like you need to mention that - why just look at the
wonderful not-compliant-to-any-known-standards GUI!

JTK

unread,
Apr 9, 2002, 4:14:41 PM4/9/02
to
Christian Biesinger wrote:
> Emlyn wrote:
>
>> And - and pay attention - the bits that do not yet conform
>> to such a scheme are the bits under the Mozilla Public Licence. All
>> *N*PL'd files are disjunctively NPL/LGPL/GPL'd.
>
>
> Wrong... I just stumpled over a NPL only file (though I also assumed
> that all NPL files were triple-licensed) (the file is config/Makefile.in)

It's high time for ol' JTK to break out the cool Perl licence checker
again, to see how much Open Source lip-service is left.

Oh joy, another billion GB CVS checkout just to prove what we all
already know....

Ralph Mellor

unread,
Apr 9, 2002, 4:17:51 PM4/9/02
to
Stuart Ballard wrote:
> (since everyone seems to have a different idea of what "moz the
> platform" means) but here's how I'd draw a diagram of Moz as a platform:
>
> Moz the suite Your app <dream on>AOL9?</dream on>
> \ | /
> \ | /
> AOL8? "Moz the platform"
> \ / | \
> \ / | \
> \ / | \
> Gecko XPCOM NetLib etc
> \ | /
> GFX | /
> \ | /
> NSPR
>
>
> This is simplified because drawing ascii graphics is making my brain
> hurt :) XPCOM should probably be below NetLib-etc, and there ought to be
> nodes for XUL, XBL, JS, etc (they'd probably live directly under "Moz
> the platform" - XUL and XBL are built on Gecko, XBL is built on JS, etc.

I agree with the above as a current technical arrangement.

But what if AOLTW find issues with using the moz codebase
as it stands, but believe they could resolve those issues
if they removed some layers to gain performance? What's to
stop AOLTW deciding that they need to cut out some of the
platform to boost performance and cut development costs?
A decision, a fork, a few months work and boom, disaster
has struck the platform dependents. (I don't for a moment
personally believe any of this, but, as I said in another
email, I have to sharpen my views on this before I write
this up or discuss it with my client, so I will continue
to play devil's advocate and press these issues.)


> You suggest that AOL might want to build a future version of AOL
> on top of "moz the suite".

Sorry, I didn't mean that. I meant they might want more than
just Gecko.


> I disagree, but keep reading.

But of course. ;>


> seems to me that AOL is already 3/4 of the way to being a mozilla-like
> application - it uses its own toolkit with a fluffy, image-based "skin",
> it is a monolithic application (no need for fast spawning of new
> processes based on it) and it's largely based on laying out text and
> images, with some network functionality going on under the hood.
>
> So, if *I* were AOL, I'd be thinking that AOL 9 or 10 would replace
> their existing proprietary toolkit and infrastructure with one based on
> "moz the platform".

Well I almost agree with this. If *I* were AOLTW, I'd be thinking
that I *might* instead use selected parts of the moz project, not
*necessarily* staying faithful to the broader platform, and quite
possibly significantly changing pieces.

Please understand that I doubt this would happen; I think you guys
built the right levels of abstraction as far as I can see, Jamie
(lovable though he is), Joel, et al be damned. In fact that is
understating things. I think you guys did a quite breathtakingly
brave and brilliant job. But I have to put that aside for now. I
want to explore the risks my client might have to face despite
my optimism.

> I see the Netscape [suite] being what they push to corporate users,


> technical users and people who want a little bit more flexibility

I have a different class of concerns in relation to this. Sort of
the opposite of AOLTW not sufficiently supporting the platform.
My concern here would be that AOLTW decide to compete with my
client's product. My client's product would be an extension of
moz that would be somewhat broadly popular. AOLTW might decide
to compete. But I'll explore that in another thread.


--
ralph

Christian Biesinger

unread,
Apr 9, 2002, 5:23:22 PM4/9/02
to
JTK wrote:

> Christian Biesinger wrote:
>> Wrong... I just stumpled over a NPL only file (though I also assumed
>> that all NPL files were triple-licensed) (the file is config/Makefile.in)
>
> It's high time for ol' JTK to break out the cool Perl licence checker
> again, to see how much Open Source lip-service is left.

That might actually be interesting indeed.

> Oh joy, another billion GB CVS checkout just to prove what we all
> already know....

No... it's like 200-300 MB... not sure how much exactly. Or you could
download the source .tar.gz / .tar.bz2

Chuck Simmons

unread,
Apr 9, 2002, 8:35:02 PM4/9/02
to

What standards did you have in mind? I use CDE on a Sun but the GUI is
changable like the weather in Unix. You can change it without restarting
the machine. On my Linux boxes at home, I can run several X sessions
each having a different window manager (GUI) and all of the window
managers allow a great deal of customization. I use fvwm at home mostly
because it is very simple and successfully restarts without disturbing
running X applications. So what's standard with the whole GUI up for
grabs?

Chuck
--
... The times have been,
That, when the brains were out,
the man would die. ... Macbeth
Chuck Simmons chr...@webaccess.net

Emlyn

unread,
Apr 10, 2002, 12:03:21 AM4/10/02
to
> > And - and pay attention - the bits that do not yet conform
> > to such a scheme are the bits under the Mozilla Public Licence. All
> > *N*PL'd files are disjunctively NPL/LGPL/GPL'd.
>
> Wrong... I just stumpled over a NPL only file (though I also assumed
> that all NPL files were triple-licensed) (the file is config/Makefile.in)
>

I can't think how I didn't notice this before in my Rhino embedding
project, but Rhino (which rocks, incidentally), is only NPL/GPL'd, not
NPL/LGPL/GPL'd. Perhaps the relicencing project missed a few things
(or perhaps I was misinterpreting things, and they never intended to
relicence Rhino).

Hmm. The relicensing page goes on about "Mozilla source files", and I
sort of assumed that meant everything that Mozilla.org hosts. But I
could have been wrong. Or perhaps there has been a huge blanket
relisencing and the copyright notices merely haven't been changed, or
perhaps poor little Rhino was overlooked.

Not that it bothers me particularly to use Rhino as is, but LGPL might
be a nice option to have. Does anyone know whether the NPL'd
mozilla.org side-projects will be relicenced?

-Emlyn

Gervase Markham

unread,
Apr 10, 2002, 5:26:01 AM4/10/02
to
> Urg. That was sloppy of me. I knew it wasn't 100% complete.
>
> Otoh, my assumption was that it was already, say, 90% plus
> complete (measured by SLOC). Is that a fair assessment?

In terms of files where the license headers have been changed, I'm not
sure, but something like 40-50%.

Gerv

Gervase Markham

unread,
Apr 10, 2002, 5:26:54 AM4/10/02
to
> I can't think how I didn't notice this before in my Rhino embedding
> project, but Rhino (which rocks, incidentally), is only NPL/GPL'd, not
> NPL/LGPL/GPL'd. Perhaps the relicencing project missed a few things
> (or perhaps I was misinterpreting things, and they never intended to
> relicence Rhino).

Rhino (JS in Java) is not on the list of code to relicense.

> Hmm. The relicensing page goes on about "Mozilla source files", and I
> sort of assumed that meant everything that Mozilla.org hosts. But I
> could have been wrong. Or perhaps there has been a huge blanket
> relisencing and the copyright notices merely haven't been changed, or
> perhaps poor little Rhino was overlooked.

It defines what we are relicensing most carefully.

Gerv

Gervase Markham

unread,
Apr 10, 2002, 5:59:05 AM4/10/02
to
>> ... the sound business case that I see for basing the AOL
>
> > client on Gecko.
>
> What is the business case that you see? In particular,
> why doesn't it make more sense for AOLTW to focus on
> their business and let M$ focus on theirs, and therefore
> use IE for the AOL client browser window and forget this
> moz thing already?

No, for several reasons.
1) A company likes to be in control of its own destiny. MS is moving
into AOL's markets (MSN is the nearest competitor to AOL) and being
dependent for components on a competitor is bad.
2) Non-Windows devices are going to play more of a role in the future,
and IE is tied to Windows (and, if the AOL client depends on IE, so is
the AOL client.) I'm sure AOL would love an "AOL Computer" which ran
e.g. Linux underneath but AOL on top, and was controlled by them.
Captive eyeballs. The computer also won't have to pay the MS tax, and so
can be cheaper.
3) Access. If they want improvements made to IE, they have to ask
Microsoft and hope. If they depend on Gecko, they can get them.
4) Licensing. IE has a reasonably nasty contract you have to agree to in
order to bundle it.
5) Strategic. MSHTML becoming the de facto standard benefits only MS.
The more open the web stays, the more level the playing field is.
6) Steve Case doesn't like Bill Gates. ;-)

I can't assure you that AOL/TW will not abandon Mozilla development, but
I personally think it's very unlikely. What is more likely is a focus
shift towards embedding. But with Sun, who have a large number of
engineers (not sure if the number is confidential) coming on stream at
Sun China, IBM and HP all making products based on the browser suite,
the future of the UI looks secure - although maybe not quite in its
present form.

>> If you want to avoid dominant funding players, you need to find
>> alternative sources of funding. If your company is willing to fund a
>> mozilla.org staff position, that would help. :-)
>
> Hmm. I doubt they would spring for the full cost of a position.
> I could imagine them being prepared to cough up, say, $20K a
> year as a contribution. There's a lot of logic to that. Hmm.
> I could see them getting benefits like A) help getting up to
> speed on the project structure and collaborating with netscape
> employees etc.; B) getting their name out there as a corporate
> sponsor of the whole shebang; C) having some representation of
> their interests. Hmm. Does that sound remotely plausible?

Very. Email mitc...@mozilla.org - she's the person to contact about
such things.

Gerv

Gervase Markham

unread,
Apr 10, 2002, 6:26:19 AM4/10/02
to
> But what if AOLTW find issues with using the moz codebase
> as it stands, but believe they could resolve those issues
> if they removed some layers to gain performance? What's to
> stop AOLTW deciding that they need to cut out some of the
> platform to boost performance and cut development costs?
> A decision, a fork, a few months work and boom, disaster
> has struck the platform dependents. (I don't for a moment
> personally believe any of this, but, as I said in another
> email, I have to sharpen my views on this before I write
> this up or discuss it with my client, so I will continue
> to play devil's advocate and press these issues.)

IF AOLTW's technical direction diverges significantly from that of
everyone else, then mozilla.org is chartered with trying to provide
technical solutions that enable everyone to get along. Exactly how this
would be done cannot be worked out until, and if, it actually happens.

I doubt that AOL/TW will ever completely fork (as opposed to abandoning
the platform); I think that even the most open-source-unfriendly person
can see that, with approx. 1/3 of their engineers (including a
proportion of their uber-hackers) being recruited through the project,
and with all the outside help they are getting, the game is definitely
worth the candle.

I also disagree that a reduction of AOL/TW involvement will mean
"disaster strikes the platform dependents" in a few months. I still use
Windows95 OSR2, even though it's a closed-source product and MS
abandoned it years ago. It still works fine. And so would the Mozilla code.

If _all_ development stopped at 1.0 and the source were magically made
unavailable and closed (not that this is possible), we'd still have a
product that people could use for at least five years.

Gerv

Ralph Mellor

unread,
Apr 10, 2002, 8:03:37 AM4/10/02
to
Gervase Markham wrote:
>> What is the business case that you see [for AOL using Gecko]?
>
> [snip]

Snap! Your first five points more or less match the way I have
thought about this since I found out AOL had bought Netscape.
Excellent stuff, thanks.


> I can't assure you that AOL/TW will not abandon Mozilla development,

To bend a well known saying, "Assurances make an ASS out of
U and, er, RANCES".

Anyhow, the main thing I'm looking for here is logic and facts
that can help me understand things better and/or which I can
present to my client rather than opinion. They will draw their
own conclusions.


> But with Sun, who have a large number of
> engineers (not sure if the number is confidential) coming on stream at
> Sun China, IBM and HP all making products based on the browser suite,
> the future of the UI looks secure - although maybe not quite in its
> present form.

I don't know of the news you are talking about. When I
just searched Sun.com for 'mozilla', I got nothing that
I would consider relevant to Sun building on mozilla the
platform. And I didn't see anything useful in the first
few pages of matches of 'sun china mozilla' in google
and teoma.

Is there anywhere I could go that focuses on moz platform
news?


> Email mitc...@mozilla.org - she's the person to contact about
> such things.

Cool. I'll contact her if/when I have something more solid.

--
ralph

Ralph Mellor

unread,
Apr 10, 2002, 8:42:10 AM4/10/02
to
Thanks to everyone for their time and thought provoking
responses. I think I now have a decidedly better handle
on the topic in the subject line than I originally had.
Thanks very much! :>

A few more points:

Gervase Markham wrote:
> IF AOLTW's technical direction diverges significantly from that of
> everyone else, then mozilla.org is chartered with trying to provide
> technical solutions that enable everyone to get along. Exactly how this
> would be done cannot be worked out until, and if, it actually happens.

I initially wrote a rebuttal to this based on the idea
that AOLTW, if it chose to fork, might also reduce its
sponsorship of mozilla.org. But I've concluded that that
is a silly notion. If AOLTW forks, that's one thing, but
it would clearly not make sense to then unnecessarily
further aggravate the situation in terms of alienating
the rest of the moz community by pulling its funding,
probably quite small relative to the total Netscape
related spend, of mozilla.org. Otoh, if they fork, it
would be because they have gone mad, so, who knows...

Hmm. Even though I can see that AOLTW just ain't gonna
forsake mozilla.org, at least not in the next few years,
because it just won't make sense, I'd still prefer to
see mozilla.org's funding guaranteed. Has AOLTW set up
something akin to a trust fund for funding mozilla.org,
and if not, would they?


> I doubt that AOL/TW will ever completely fork (as opposed to abandoning
> the platform); I think that even the most open-source-unfriendly person
> can see that, with approx. 1/3 of their engineers (including a
> proportion of their uber-hackers) being recruited through the project,
> and with all the outside help they are getting, the game is definitely
> worth the candle.

If I read you correctly here, you're saying that there
is one outside coding contributor for each two AOLTW
funded coding contributors.

One issue here is that a permanent full time coder is,
all other things being equal, worth N times a possibly
temporary, part time coder.

The only stats I've gotten on this important issue are
from my own informal searches for info on mozilla.org,
especially bugzilla, and an article on a mozilla related
site (that shall remain nameless for fear I'd be promoting
it).


> I also disagree that a reduction of AOL/TW involvement will mean
> "disaster strikes the platform dependents" in a few months. I still use
> Windows95 OSR2, even though it's a closed-source product and MS
> abandoned it years ago. It still works fine. And so would the Mozilla code.

The platform dependent in the case you describe is a user
that can continue functioning, does not mind a failure to
advance with the state of the art, and who can migrate in
their own good time. The platform dependents I had in mind
are ISVs, for whom the situation is somewhat different.


--
ralph

Gervase Markham

unread,
Apr 10, 2002, 10:01:54 AM4/10/02
to
>> I doubt that AOL/TW will ever completely fork (as opposed to
>> abandoning the platform); I think that even the most
>> open-source-unfriendly person can see that, with approx. 1/3 of their
>> engineers (including a proportion of their uber-hackers) being
>> recruited through the project, and with all the outside help they are
>> getting, the game is definitely worth the candle.
>
> If I read you correctly here, you're saying that there
> is one outside coding contributor for each two AOLTW
> funded coding contributors.

No; I mean 1/3 of the engineers currently working on Mozilla for
Netscape were at one time contributors from outside Netscape. (One of
the reasons why Mozilla has a smaller-than-expected number of
non-Netscape contributors is that Netscape hires them all.)

Gerv

Gervase Markham

unread,
Apr 10, 2002, 10:00:12 AM4/10/02
to
> I don't know of the news you are talking about. When I
> just searched Sun.com for 'mozilla', I got nothing that
> I would consider relevant to Sun building on mozilla the
> platform. And I didn't see anything useful in the first
> few pages of matches of 'sun china mozilla' in google
> and teoma.
>
> Is there anywhere I could go that focuses on moz platform
> news?

Quite a few of them are appearing on IRC. I don't know much more than that.

Gerv

JTK

unread,
Apr 10, 2002, 1:38:55 PM4/10/02
to
>> It's high time for ol' JTK to break out the cool Perl licence checker
>> again, to see how much Open Source lip-service is left.
>
>
> That might actually be interesting indeed.
>
>> Oh joy, another billion GB CVS checkout just to prove what we all
>> already know....
>
>
> No... it's like 200-300 MB... not sure how much exactly. Or you could
> download the source .tar.gz / .tar.bz2

Well, got it half done last night. Strange that I happened to notice
some files with "AOL" in the name go by. Gave me time to improve the
cool perl script I guess, so it wasn't a total waste. Stay tuned, the
true story on the licensing is on its way!

Ralph Mellor

unread,
Apr 10, 2002, 3:36:29 PM4/10/02
to
JTK wrote:
>>> It's high time for ol' JTK to break out the cool Perl licence checker
>>> again, to see how much Open Source lip-service is left.
>
> Well, got it half done last night.

Just in case you missed it, here's some relevant posts
by the guy who is supervising the relicensing in reply
to some things I said:

I wrote:
> ALL of Moz is GPLed.

Gervase Markham wrote:
> That's not true. mozilla.org is moving towards an [M|N]PL/GPL/LGPL
> tri-license, but we aren't there yet. See http://www.mozilla.org/MPL .

I wrote:
> my assumption was that it was already, say, 90% plus
> complete (measured by SLOC). Is that a fair assessment?

Gervase Markham wrote:
> In terms of files where the license headers have been changed, I'm
not > sure, but something like 40-50%.

hth.

--
ralph

Chris Hoess

unread,
Apr 10, 2002, 5:19:32 PM4/10/02
to
In article <3CB42A19...@dimp.com>, Ralph Mellor wrote:
>
> I don't know of the news you are talking about. When I
> just searched Sun.com for 'mozilla', I got nothing that
> I would consider relevant to Sun building on mozilla the
> platform. And I didn't see anything useful in the first
> few pages of matches of 'sun china mozilla' in google
> and teoma.
>
> Is there anywhere I could go that focuses on moz platform
> news?
>

See <URL:http://groups.google.com/groups?selm=3CB3CF8A.F5FACB0A%40sun.com>
for news on Sun's involvement in Mozilla. I'd characterize their work as
focused on accessibility and embedding, however.

--
Chris Hoess

Chris Hoess

unread,
Apr 10, 2002, 5:32:18 PM4/10/02
to
In article <3CB2FDA...@domain.invalid>, us...@domain.invalid wrote:
>
> Anyhow, their use of Gecko, and even their interest in
> moz the suite (as against moz the platform), still does
> not make a compelling argument to me that AOLTW will
> want to continue to fully fund the entire platform
> effort.
>

This is correct. However, consider that one of the advantages of
Mozilla-as-platform is that it makes use of XML, CSS, and JavaScript, all
of which need to be part of Gecko or closely associated with it. That
leaves XBL, XUL, XPCOM and XPConnect; these technologies could be
maintained by a signficantly smaller development effort.



>
> > IBM
>
> Yeah. I think this would be a no brainer for them. Plus
> I would love it if IBM became a really major player in
> developing moz the platform. I'm pretty sure IBM won't
> compete in the space my client is in. Interesting.
>

While IBM hasn't become a major contributor to the platform *yet*, AFAIK
(what I"ve seen them contribute is BiDi rendering and the OS/2 port), take
a look at
<URL:http://oss.software.ibm.com/developerworks/opensource/sashxb/>. This
is very heavily dependent upon the technologies I've mentioned above.
Even were AOL-TW to widthdraw from the platform technologies, I strongly
suspect IBM would pick up the slack to keep SashXB working.

--
Chris Hoess

JTK

unread,
Apr 11, 2002, 12:13:03 AM4/11/02
to
Chuck Simmons wrote:
>
> JTK wrote:
> >
> > Jay Garcia wrote:
> > > On 04/08/2002 7:04 PM, Garth Wallace wrote:
> > >
> > >>AOL/Netscape isn't the only major corporate contributor. Red Hat and IBM
> > >>have also contributed quite a bit of time and effort into the project,
> > >>and they're hardly insubstantial entities.
> > >>
> > >
> > >
> > > Also, Sun Microsystems has been a major player.
> > >
> >
> > Well it's not like you need to mention that - why just look at the
> > wonderful not-compliant-to-any-known-standards GUI!
>
> What standards did you have in mind?

Whatever the local standards happen to be. E.g. a Windows "look and
feel" on Windows. Mac on Macs. A freakshow on the various Unii. You
get the idea.

> I use CDE on a Sun but the GUI is
> changable like the weather in Unix. You can change it without restarting
> the machine. On my Linux boxes at home, I can run several X sessions
> each having a different window manager (GUI) and all of the window
> managers allow a great deal of customization. I use fvwm at home mostly
> because it is very simple and successfully restarts without disturbing
> running X applications.

All that and you can't hardly give it away. Go figure.

> So what's standard with the whole GUI up for
> grabs?
>

Like I said, a freakshow on Linux is fine. The same freakshow on
Windows and Mac isn't.

JTK

unread,
Apr 11, 2002, 12:19:09 AM4/11/02
to
Ralph Mellor wrote:
>
> JTK wrote:
> >>> It's high time for ol' JTK to break out the cool Perl licence checker
> >>> again, to see how much Open Source lip-service is left.
> >
> > Well, got it half done last night.
>
> Just in case you missed it,

No, I've been watching closely. I got AOL to pay some lip service to
this initially, so it'll be interesting to see if they've done anything
beyond said lip service.

> here's some relevant posts
> by the guy who is supervising the relicensing in reply
> to some things I said:
>
> I wrote:
> > ALL of Moz is GPLed.
>
> Gervase Markham wrote:
> > That's not true. mozilla.org is moving towards an [M|N]PL/GPL/LGPL
> > tri-license, but we aren't there yet. See http://www.mozilla.org/MPL .
>
> I wrote:
> > my assumption was that it was already, say, 90% plus
> > complete (measured by SLOC). Is that a fair assessment?
>
> Gervase Markham wrote:
> > In terms of files where the license headers have been changed, I'm
> not > sure, but something like 40-50%.
>
> hth.

Thanks. IIRC, 40-50% would indeed be much better than what it was when
I first wrote the license checker explicitly for the Mozilla project
(like 0%), but we'll let Perl decide that and let the bits fall where
they may.

Christian Biesinger

unread,
Apr 14, 2002, 4:19:47 AM4/14/02
to
JTK wrote:
> Strange that I happened to notice
> some files with "AOL" in the name go by.

./mailnews/base/ispdata/aol.rdf

This is just an example file for ISPs to add default Accounts to own
distributions of Mozilla (and NS6) (afaik).

./editor/libeditor/text/nsAOLCiter.cpp
./editor/libeditor/text/nsAOLCiter.h

For citing AOL-style quotes (">> a quote <<")

JTK

unread,
Apr 14, 2002, 5:31:24 PM4/14/02
to
Christian Biesinger wrote:
>
> JTK wrote:
> > Strange that I happened to notice
> > some files with "AOL" in the name go by.
>
> ./mailnews/base/ispdata/aol.rdf
>
> This is just an example file for ISPs to add default Accounts to own
> distributions of Mozilla (and NS6) (afaik).
>

Dindn't catch that one.



> ./editor/libeditor/text/nsAOLCiter.cpp
> ./editor/libeditor/text/nsAOLCiter.h
>
> For citing AOL-style quotes (">> a quote <<")
>

Yep, these are ones that I noticed in particular. Didn't know there was
an "AOL-style quote".

Christian Biesinger

unread,
Apr 14, 2002, 5:50:51 PM4/14/02
to
JTK wrote:
> Christian Biesinger wrote:
>>./editor/libeditor/text/nsAOLCiter.cpp
>>./editor/libeditor/text/nsAOLCiter.h
>>
>>For citing AOL-style quotes (">> a quote <<")
>
> Yep, these are ones that I noticed in particular. Didn't know there was
> an "AOL-style quote".

Me neither until I read that files :)

Phillip M. Jones, C.E.T.

unread,
Apr 14, 2002, 6:57:30 PM4/14/02
to

Christian Biesinger wrote:
>
> JTK wrote:
> > Christian Biesinger wrote:
> >>./editor/libeditor/text/nsAOLCiter.cpp
> >>./editor/libeditor/text/nsAOLCiter.h
> >>
> >>For citing AOL-style quotes (">> a quote <<")
> >
> > Yep, these are ones that I noticed in particular. Didn't know there was
> > an "AOL-style quote".
>
> Me neither until I read that files :)

I thought everyone knew that >> were AOL style Quotes. <<GRIN>>
--
---------------------------------------------------------------------------
Phillip M. Jones, CET |MEMBER:VPEA (LIFE) ETA-I, NESDA,ISCET, Sterling
616 Liberty Street |Who's Who. PHONE:276-632-5045, FAX:276-632-0868
Martinsville Va 24112-1809 |pjo...@kimbanet.com, ICQ11269732, AIM pjonescet
---------------------------------------------------------------------------

If it's "fixed", don't "break it"!

mailto:pjo...@kimbanet.com

<http://www.kimbanet.com/~pjones/default.htm>
<http://www.kimbanet.com/~pjones/america/default.htm>
<http://www.kimbanet.com/~pjones/message/default.htm>
<http://home.kimbanet.com/~pjones/birthday/index.htm>

Ralph Mellor

unread,
Apr 25, 2002, 12:14:18 AM4/25/02
to Stuart Ballard
Stuart Ballard wrote:

> I'm not *100%* sure


> (since everyone seems to have a different idea of what "moz the
> platform" means) but here's how I'd draw a diagram of Moz as a platform:
>
> Moz the suite Your app <dream on>AOL9?</dream on>
> \ | /
> \ | /
> AOL8? "Moz the platform"
> \ / | \
> \ / | \
> \ / | \
> Gecko XPCOM NetLib etc
> \ | /
> GFX | /
> \ | /
> NSPR

Is there an official mozilla document / diagram that helps
explain "mozilla the platform"?


> So, if *I* were AOL, I'd be thinking that AOL 9 or 10 would replace
> their existing proprietary toolkit and infrastructure with one based on

> "moz the platform". ... I'd
> build a mozilla application that looks exactly like the existing AOL
> client, but I'd do it from scratch, not based on the browser suite.

So the expectation would, in theory, be that there would be
little change to C/C++ parts of Mozilla (ignoring bug fixing),
right?


--
ralph

Stuart Ballard

unread,
Apr 25, 2002, 10:03:38 AM4/25/02
to
Ralph Mellor wrote:
>
> Stuart Ballard wrote:
>
> > I'm not *100%* sure
> > (since everyone seems to have a different idea of what "moz the
> > platform" means) but here's how I'd draw a diagram of Moz as a platform:

> Is there an official mozilla document / diagram that helps


> explain "mozilla the platform"?

I don't think so. As I said above, everyone seems to have a different
idea of what the heck "moz the platform" is. I drew that little ascii
diagram in hopes of explaining what *I* think it means, in the spirit of
making sure all terms were defined before talking about them.

> So the expectation would, in theory, be that there would be
> little change to C/C++ parts of Mozilla (ignoring bug fixing),
> right?

Right, except for perhaps adding features to support things that mozilla
right now can't do at all. For example, AOL's client incorporates a
phone dialer and a TCP/IP tunneler. Those things couldn't be done in
XUL+js :) Those particular pieces probably wouldn't be much use to
release as open source, so I doubt we'd see them in Mozilla, but there
might be other places where the platform as it stands today couldn't
quite meet their needs. If that happened, I wouldn't be surprised to see
new features added to the platform (and released to Mozilla, if they
have general-purpose value).

Basically, I'd see that scenario as adding a new major client
application of the Mozilla platform. Wider use in clients is always a
good thing for a platform, as it reveals areas that need improving, and
exercises the code in new ways causing more bugs to be found and fixed.

In other words, that'd be good for your company (building a
mozilla-based product, IIRC) because AOLTW would find (some of) the bugs
and limitations and get them fixed before you get bitten by them :)

Stuart.

--
Stuart Ballard, Programmer
NetReach - Internet Solutions
(215) 283-2300, ext. 126
http://www.netreach.com/

Ralph Mellor

unread,
Apr 25, 2002, 10:48:35 PM4/25/02
to Stuart Ballard
>>[to recreate the AOL client software using the moz platform]

>>little change to C/C++ parts of Mozilla (ignoring bug fixing)
>
> Right, except for ... [eg] phone dialer and a TCP/IP tunneler.

OK, it is as I understood then.


> Those things couldn't be done in XUL+js :)

But, but, but... I was lead to believe that all the talk about
XUL being just to do with the user interface was a facade; that
XUL was a complete general purposes programming language with
built in AI; and that it was really created by aliens as a
substrate through which they would reconstitute themselves as
virtual entities that lived in the web. And now you tell me it
can't be used to program everything the AOL client does. I'm
now trying to work out if this news is inconsistent with the
alien reconstitution thing...


> [AOL basing their client suite on moz tech is good for moz]

Right.


> everyone seems to have a different idea of what the heck
> "moz the platform" is.

Yes. Here's my current thinking.

First, from the perspective of my client *technically*
*depending* on mozilla, I think that focusing on Mozilla
as a "platform" and betting the farm on that "platform"
makes the proposition look far too risky. So, instead of
an overall "platform" concept, I am starting to view
Mozilla as a number of discrete technologies and
components, each of which I am considering separately,
(or in groups, actually) insofar as what might be the
impact if current (AOLTW) funding is reduced.

Second, from the perspective of my client depending on
*distribution*, I have estimated, conservatively, that
AOL will have deployed 50 million Geckos by the end of
2007.

Third, from the perspective of my client depending on
use of Mozilla as a programming platform by individuals
and corporations in web and desktop software development,
I think the jury is still out as to which bits will work
out.

A much more complex proposition than I was originally
thinking in terms of, but such is life.

--
ralph

Gervase Markham

unread,
Apr 26, 2002, 5:31:20 AM4/26/02
to
> First, from the perspective of my client *technically*
> *depending* on mozilla, I think that focusing on Mozilla
> as a "platform" and betting the farm on that "platform"
> makes the proposition look far too risky. So, instead of
> an overall "platform" concept, I am starting to view
> Mozilla as a number of discrete technologies and
> components, each of which I am considering separately,
> (or in groups, actually) insofar as what might be the
> impact if current (AOLTW) funding is reduced.

Well, that's up to you - but http://www.oeone.com have bet the firm on
Mozilla-the-platform, and they have the most amazing kick-ass product
I've seen in years. Moreover, they developed it in a pretty short amount
of time. It's currently not being extensively-marketed, though - this
may be the fault of the current economic conditions.

> Second, from the perspective of my client depending on
> *distribution*, I have estimated, conservatively, that
> AOL will have deployed 50 million Geckos by the end of
> 2007.

My prediction: by the end of 2007, there will be far more Geckos in use
in other products, and in browsers in the 2nd and 3rd worlds on free
OSes, than are distributed by AOL.

Gerv

Ralph Mellor

unread,
Apr 26, 2002, 5:01:27 PM4/26/02
to Gervase Markham
>> Second, from the perspective of my client depending on
>> *distribution*, I have estimated, conservatively, that
>> AOL will have deployed 50 million Geckos by the end of
>> 2007.
>
> My prediction: by the end of 2007, there will be far more
> Geckos in use in other products, and in browsers in the
> 2nd and 3rd worlds on free OSes, than are distributed by AOL.

So, if you think my AOL estimate is reasonable, and I
think it is (I think AOLTW has around 35 million users
now, including AOL, Compuserve, Netscape, and they are
currently adding around 5 million users a year, which
means another 30 million if they maintain current rates,
for a total of 65 million users; hence 50 million is
fairly conservative), then that means over 100 million
Geckos. Clearly it could easily be more like 200-500
million. It may be a niche, but I'll take it! :>

--
ralph

Reply all
Reply to author
Forward
0 new messages