Health of Open-Source Application Communities

[Shane Isbell]

I'm trying to gauge the health of any community moving in to support open-source alternative implementations of the close-source Google applications. I'd like to go ahead and raise this early. I read Cyanogen's latest post and it appears the plan would be to have a minimal ROM with a sync of the closed-source applications back to the device: http://www.cyanogenmod.com/home/the-current-state

While good for the user community, this strikes me as a blow against any developer community for open-source implementations of replacement apps. It appears the real motivation for wanting alternatives is that people can't have both improved ROMs AND the Google apps. Once the itch is gone from the user/developer community, what motivation is there for open-source alternatives or alternatives at all?

--
Shane Isbell (Co-founder of SlideME - The Original Market for Android)
http://twitter.com/sisbell
http://twitter.com/slideme

[Dan Ballance]

Surely this is just the nature of the world in which we live: users are seeking the best applications they can get and will gravitate to the device or platform that offers them the functionality they require.

When it seems that the bare aosp rom is so basic in functionality that users will quickly drift back to the official google rom, then surely the pragmatic approach is to keep developing the aosp platform alongside the closed source apps and keep the community on board.

Maybe once aosp is a healthier beast, strong enough to survive in the wilderness alone,then perhaps it might be appropriate to talk about seeking pure open source alternatives,

Sincerely,

Dan ( @crazywizdom )

[Mark Murphy]

Shane Isbell wrote:
> I'm trying to gauge the health of any community moving in to support
> open-source alternative implementations of the close-source Google
> applications. I'd like to go ahead and raise this early. I read
> Cyanogen's latest post and it appears the plan would be to have a
> minimal ROM with a sync of the closed-source applications back to the
> device: http://www.cyanogenmod.com/home/the-current-state

For Cyanogen's audience, that's probably a fine strategy, particularly
if it has some measure of a seal of approval from The Powers That Be.

> While good for the user community, this strikes me as a blow against any
> developer community for open-source implementations of replacement apps.
> It appears the real motivation for wanting alternatives is that people
> can't have both improved ROMs AND the Google apps. Once the itch is gone
> from the user/developer community, what motivation is there for
> open-source alternatives or alternatives at all?

There are multiple motivations, and the importance of replacement apps
varies by motivation.

For Cyanogen's audience, "the current state" may suffice, and their
motivation probably wanes as a result.

I'm here for liberty. I am hoping that either the Android or Symbian
ecosystems can build a commercially-viable end-to-end open source
firmware, that can be vetted as secure. Then, anyone interested in a
device that cannot be messed with (see UAE/Blackberry incident from
July) has some place to go and have a platform upon which to build
secure communications services. For me, the only closed-source Google
application that really needs to be in this firmware is a market;
everything else can be treated as a regular Android app AFAICT.

Others will have their own motivations. I doubt it's "one size fits all".

--
Mark Murphy (a Commons Guy)
http://commonsware.com | http://twitter.com/commonsguy

_Beginning Android_ from Apress Now Available!

[Jean-Baptiste Queru]

I think that you're hitting the nail on the head here: for most users,
"open-source" is not a feature (though free-as-in-beer is). Users will
go for whichever option works best for them.

That means two things:
-indeed if there's a way for "regular" users to run with the official
Google apps or with less-functional open-source replacements, all else
being equal, they'll use that. The good news is that they typically
won't mind switching from a non-open-source version to an open-source
one if the latter offers all the functionality of the former.

-by extension of not caring about licenses, most users won't care
about piracy. Given two equally available versions of similar systems,
there's a good chance that those users will go for a version that
contains unlicensed software if it's more functional.

As a consequence, I believe that we need to "play" on two fronts:

-we need to find ways to continue offering a high level of
functionality without getting on the wrong side of copyright law and
licenses. I think that we're making some progress on that front.

-we need to be sure that the properly licensed builds are highly
available and visible to users. We might be able to get there by
giving a higher visibility to those properly licensed builds.

I guess that we could achieve some success on both fronts by reducing
fragmentation: combining our efforts on fewer variants would avoid
having to duplicate some effort (and would also save some engineering
overhead), and would allow making those variants more visible. I have
some ideas about that, let me explore them a bit deeper.

JBQ

--
Jean-Baptiste M. "JBQ" Queru
Software Engineer, Android Open-Source Project, Google.

Questions sent directly to me that have no reason for being private
will likely get ignored or forwarded to a public forum with no further
warning.

[Disconnect]

You're saying "we" a lot, but it is Sunday. How much is google willing
to devote (people, time, etc) to helping clean up their mess?

[Jer Warren]

I think that "functioning at least as well" is kind of a poor selling point, especially in light of some of the shortcomings of these apps.  For instance, the GTalk app drives me nuts on a fairly regular basis and could really use some attention.  Likewise, if the open source GTalk replacement functioned the same, but ALSO worked with any Jabber address, it'd then be something people might actually choose over the closed version because it is better -- and not just because the law has their hands tied and it's "just as good."

The way I see it, the important thing that this project fulfills is not really meant to be measured tomorrow; the real benefit will be at the manufacturer/carrier level, when any new device deployment allows carriers to forgo the "Google Experience" costs, thus putting even more low-cost Android devices on the market.  More devices means more users means more app sales means more developers means better apps.  Everyone wins.

[Disconnect]

More app sales, but not through the depressingly-buggy incredibly
dominant market, since "android market" is for google devices only..

[Shane Isbell]

On Sun, Sep 27, 2009 at 10:06 AM, Disconnect <dc.dis...@gmail.com> wrote:

More app sales, but not through the depressingly-buggy incredibly
dominant market, since "android market" is for google devices only..

This is a valid point. In the case of Market, a superior alternative marketplace won't cut general acceptance. It is primarily based on number of apps in Market (10K vs say 330 at SlideME). This has been the sticking point for us to get operator acceptance (they like SAM, the marketplace app).

The problem with a Market is developers won't upload if there aren't users and users won't use a Market if developers don't upload the applications. In niche markets we have great growth, but the general community will go for "most apps" not best Market.

[Jean-Baptiste Queru]

Sadly, I'm sitting between two chairs here, so it's never quite clear
whether "we" means "the community" or "Google". In this specific post
it was all "the community".

I can make a case within Google that making AOSP more usable "out of
the box" is good for the entire Android ecosystem (by letting people
who work from the open-source tree spend time on making actual
progress instead of creating workarounds), and therefore good for
Google. Google has limited resources (yes, I know, it sounds crazy),
which means that such work needs to be prioritized against other
goals. I can't imagine that Google will be able to immediately put
dozens on engineers on helping AOSP, but I think that Google will
attempt to stay responsive to review requests as long as they're not
gigantic.

JBQ

[Jean-Baptiste Queru]

Yup, those chicken-and-egg are tough to crack. I'd imagine that some
community activism could help chip at the issues on both sides of the
vicious circle. I'll guess that your presence on this group at this
very moment will raise awareness about your solution in the eyes of
people who find themselves thinking about non-Google solutions,
especially in environments where Google's solutions aren't available.

JBQ

[Jer Warren]

On Sun, Sep 27, 2009 at 10:16 AM, Shane Isbell <shane....@gmail.com> wrote:

This is a valid point. In the case of Market, a superior alternative marketplace won't cut general acceptance. It is primarily based on number of apps in Market (10K vs say 330 at SlideME). This has been the sticking point for us to get operator acceptance (they like SAM, the marketplace app).

The problem with a Market is developers won't upload if there aren't users and users won't use a Market if developers don't upload the applications. In niche markets we have great growth, but the general community will go for "most apps" not best Market.

What if there were an "open" market repository to which developers could upload, which had an interface that could then push updates to any markets they wish to utilize?  This could be an attractive way to solve the problem of developers who don't want to expend the effort to publish to markets they deem too small, but would also have a wonderful side-effect of giving us a place for an Open Market client to get its data.

I for one am in the "too lazy" camp, only putting my apps in the Google Android Market, but if there were someplace I could upload them that could then push them out to any other markets with minimal effort, I would totally use that.

Brainstorming along these lines has me imagining incorporating the apt functionality previously discussed, which could enable an extremely simple mechanism for other markets to stay abreast of package updates.  I confess to not having an inkling of how to handle registration/payments/etc on the "Open Market," though.

[Shane Isbell]

I'd support such an idea. We proposed something to Archos about pushing apps to them, but they never bothered to respond. It's always one thing I've hated about the mobile industry - this cut-throat attitude of duking it out, last man standing. It really wore on me and I left behind the mobile industry for two-years before Android brought me back in. Seeing community effort in the content space is refereshing.

I think it would have to be a content repository for stocking apps, not for delivery or payment of applications (that would be up to individual markets). It clearly makes sense that this open-repository would be open-source and community run. The paid content would have to be protected so the actual production systems would not be as open.

At SlideME, we can provide servers, monitoring and scaling for this, if that is the way people want to go.

[Brian F]

On Sep 27, 12:20 pm, Jean-Baptiste Queru <j...@android.com> wrote:
> Sadly, I'm sitting between two chairs here, so it's never quite clear
> whether "we" means "the community" or "Google". In this specific post
> it was all "the community".
>
> I can make a case within Google that making AOSP more usable "out of
> the box" is good for the entire Android ecosystem (by letting people
> who work from the open-source tree spend time on making actual
> progress instead of creating workarounds), and therefore good for
> Google. Google has limited resources (yes, I know, it sounds crazy),
> which means that such work needs to be prioritized against other
> goals. I can't imagine that Google will be able to immediately put
> dozens on engineers on helping AOSP, but I think that Google will
> attempt to stay responsive to review requests as long as they're not
> gigantic.
>
> JBQ

I think this is a reasonable approach.

End users are going to demand functionality in most cases. Even I, as
a huge supporter of OSS software, would run the Google binaries until
suitable replacements were available. I have the licenses to run just
about ANY Microsoft product XP/Vista/W7, Server, Exchange, Office etc.
through a MSDN subscription, which is needed for supporting my
customers, and have bought PeachTree Accounting for billing in the
past, but I choose to run Linux on all my work, home, and family
computers. I choose to run Postfix for mail services instead of
Exchange Server, OpenOffice.org instead of Office, and SMBLedger in
place of PeachTree for billing.

I switched to the Open Source alternative as each became available and
able to do what I wanted/needed it to.

Freedom, to me, is much more important than Free.

My wife on the other hand doesn't see it that way. She is much more a
typical end-user, maybe even more so than average. She is the end
user who can not accept "change" well. If the program doesn't work
like she is used to it working, she kicks and screams until she gets
used to it. She still runs her years-old version of Photo Shop under
Wine on her Linux netbook. (It came with Windows installed, as did my
own laptop and my daughter's. All 3 were formatted and had Linux
installed on them before ever booting into Windows the first time.

From my experience you are going to find 4 types of users:

1. Ones who don't care where the app comes from, if it is open or
closed source, they want what they are familiar with.
2. Ones who don't care where the app comes from, if it is open or
closed, but want the BEST user experience. (Most people I think, but
#1 is probably a close 2nd)
3. Ones who would prefer Open Source, but are not willing to
sacrifice *too much* functionality in doing so. (The majority of Open
Source users I think - After all, how many Linux users want to use the
open source equivalents for their video card drivers on ATI video
cards.)
4. The OSS purists who demand everything Open Source, regardless of
the functionality. (There are not many of these around. (Richard
Stallman types.)

By allowing Cyanogen to satisfy the immediate needs of #1,2 & 3 above
(3 out of 4 types using the work-around), and working with the #3
types, we can eventually satisfy the #2 and #4 we can reverse that to
the point that 3 out of 4 groups will be using the Open solution on
Android.

We will never be able to convince the first two groups to install a
ROM that takes away much of the functionality they expect from a
modern "smart phone".

I see the app "markets" as a problem, a source for a lot of
fragmentation to occur. If a single alternative to Google Market can
emerge as a "leader" of the alternatives to the Google-owned
Marketplace, I think it can work. Going to 2 different possible
sources, while not preferable to a single source for applications, is
far preferable to having half a dozen different locations to search
for an app. If the alternative markets can somehow entice a user with
lower prices or other "bonus", an app may be able to be created to
serach "both markets" show the prices or "special offers" at both the
Google market, and the alternative, and allow the person to choose
which market to click on to get their app. (I know, if it's a free
app, the "enticement" might be harder to come up with, maybe better
reviews etc.)
If developers had to deal with only 2 markets, instead of the
potential of dozens of different markets, and the "price to entry" was
low enough for the alternative market, I don't see why they would not
support both of the markets with their apps. We would need to make
sure that the licensing of the alternative app market was created in
such a way that phone manufacturer's ad/or carriers could pre-install
the app on phones without cost to them, even make it possible for them
to put their own logo on it in some way, but NOT restrict the apps
available.

[Brian F]

On Sep 27, 12:37 pm, Jer Warren <j...@nyquil.org> wrote:

As to the payment system, maybe take a smaller percentage than Google
does to entice the app developers, and any extra money beyond
operating costs go to the Open Android developers in some way. (As to
how the money is distributed, I'm not going to guess at this point.)

Google is apparently taking the same percentage as Apple, but Apple
with their "manual review" process has to have a much higher cost to
operate. They "claim" they are just "breaking even" on this kind of
stuff. I personally find that REALLY hard to believe, myself.

I would think that an Open Market would be able to actually make a
profit from paid apps, IF the platform was popular enough. If the end-
users knew more of their purchase price was going to the developers
themselves to help improve the next release than if they purchased at
the Google Market, I think many would support the Open Market over the
Google market, even if the prices of the apps were the same.

[schwiz]

I can only assume cyanogen will move his efforts to modding this
release once its ready...

> Mark Murphy (a Commons Guy)http://commonsware.com|http://twitter.com/commonsguy

[Shane Isbell]

At SlideME, the developer gets about 95% of the cut (rest goes to cover costs of transaction).  We do allow free-distribution to handset manufacturers (if they don't white-label or brand the SAM client). Neither of these has been sufficient to drive massive adoption, as you suggest. It's strictly numbers of users and numbers of apps. Devs don't care if it is a %5 or 30% cut, it's a matter how much money they can put in their pocket.

[Dianne Hackborn]

On Sun, Sep 27, 2009 at 9:06 AM, Mark Murphy <mmu...@commonsware.com> wrote:

For me, the only closed-source Google
application that really needs to be in this firmware is a market;
everything else can be treated as a regular Android app AFAICT.

I think that is a fine goal, and ultimately Google's as well. :)  There is just a lot of work to get there, and a lot of other things to do at the same time as well.

The Market app itself is a difficult case because while it is a key part of the platform, at this point it would also be tremendously detrimental to the overall platform to have it open-source.

Market is increasingly the main leverage we have for defining what Android is.  This is important for such key things as maintaining the compatibility and quality of Android devices, ensuring that all Android devices support the expected developer tools, and retaining key platform features such as the ability to freely side-load applications.

Consider a concrete example of how a non-open Market helps the open-source platform: there were manufacturers who were ready to release 1.5-based QVGA devices.  I held one in my hand, and their software worked...  but the way they did this was by brute-force fiddling with the layouts of all of the built-in apps to work on such a device.  When it came to installing a third party app?  Well good (actually mostly bad) luck on how -that- works.

Because Market requires a license to distribute, we can ensure that it doesn't appear on a device like that and wreak havoc with third party developers.  Then once we are ready we can allow Market to be distributed on those devices, doing whatever is appropriate to provide a good experience for the existing applications.

(A side note: the development time from 1.5 to 1.6 consisted of 1 or 2 full time engineers implementing the backwards compatibility code for existing applications as well as finishing the screen density support that had been started in 1.5.  In the case of QVGA devices we actually require that they report they are QVGA, so Market can filter out existing applications, since we found they generally are not directly compatible with such a screen; this itself was a fairly late decision as we looked at the best we could do for unmodified apps running on a QVGA screen and decided "not good enough.")

So having control over the Market app is good for developers because they don't get stuck with their applications running on device configurations they had no way to prepare for, or with serious modifications that break a significant number of market apps.

It is also good for users because it provides some basic quality bar for the experience they will have with Market.

And it is good for the platform because it is one of the few ways we can directly attack the dreaded "fragmentation" problem that so many people are (understandably) worried about.

Right now is a very delicate time in Android's life, where we are trying to carefully balance between growing the device ecosystem as fast as possible while not letting the thing spin out of control.  Without the ability to impose some constraints on what products people are releasing, I am pretty confident it -would- spin out of control.

--

Dianne Hackborn
Android framework engineer
hac...@android.com

Note: please don't send private questions to me, as I don't have time to provide private support, and so won't reply to such e-mails.  All such questions should be posted on public forums, where I and others can see and answer them.

[nEx.Software]

I have just started to look at moving my apps into "alternative"
markets. Basically, it was fragmentation that has kept me from doing
so... I, as a user, don't want to have to go to many different
markets, and as a developer don't want to have to update many
different markets. If I could push once and have that one copy make it
to the (1, 2, 5, etc...) markets I have chosen, that'd be great.

[Fred Grott]

You mean federated markets with Google having control over both Market app and maybe protocol?

[Shane Isbell]

On Sun, Sep 27, 2009 at 1:55 PM, nEx.Software <email.nex...@gmail.com> wrote:

I have just started to look at moving my apps into "alternative"
markets. Basically, it was fragmentation that has kept me from doing
so... I, as a user, don't want to have to go to many different
markets, and as a developer don't want to have to update many
different markets. If I could push once and have that one copy make it
to the (1, 2, 5, etc...) markets I have chosen, that'd be great.

Right, to me this makes perfect sense. It would be a community site that handles stocking and pushing of apps to various alternative markets (keep in mind not all markets would take all content, for instance a game portal). If Google were an end-point for such a feed, it could really drive adoption of Android content even further.

--

[nEx.Software]

Personally, I wouldn't quite care who had control of it (per se) as
long as I didn't have to upload to a bunch of different markets. And
as long as I could choose which markets I was pushing applications
into, I'd be a happy camper. I am pretty easy to please.

On Sep 27, 2:06 pm, Fred Grott <fred.gr...@gmail.com> wrote:
> You mean federated markets with Google having control over both Market app
> and maybe protocol?
>
> On Sun, Sep 27, 2009 at 3:55 PM, nEx.Software

> <email.nex.softw...@gmail.com>wrote:

[Joe LaPenna]

Its also worthwhile to note that many Googlers are already and have been working in the open, the kernel and gerrit folks, for example. Some Googlers are less vocal than others and like it that way.

On Sep 27, 2009 10:20 AM, "Jean-Baptiste Queru" <j...@android.com> wrote:


Sadly, I'm sitting between two chairs here, so it's never quite clear
whether "we" means "the community" or "Google". In this specific post
it was all "the community".

I can make a case within Google that making AOSP more usable "out of
the box" is good for the entire Android ecosystem (by letting people
who work from the open-source tree spend time on making actual
progress instead of creating workarounds), and therefore good for
Google. Google has limited resources (yes, I know, it sounds crazy),
which means that such work needs to be prioritized against other
goals. I can't imagine that Google will be able to immediately put
dozens on engineers on helping AOSP, but I think that Google will
attempt to stay responsive to review requests as long as they're not
gigantic.

JBQ

On Sun, Sep 27, 2009 at 9:42 AM, Disconnect <dc.dis...@gmail.com> wrote: > > You're saying "we"...

[Al Sutton]

We, and AndAppStore, recently started to get developers approvals to
redistribute application information to other sites. We're currently
sitting at around 60% agreeing to it and if any other site wants to do
a mutual exchange we'd be happy to discuss it, but the main condition
is that it's a two way thing (they get our apps we get theirs).

The reason we want a two way exchange is because we don't believe it's
in developers interests to have a single site which feeds out to
everyone else because it'll create a naturally dominant site and we've
all seen what can happen in that situation with the features Market
has been lacking (yes, some are being addressed, but has anyone seen
developer replied to comments mentioned?, this is something
AndAppStore has had for months). What we're aiming to create is more
of a peer-to-peer structure where developers can submit apps to
whichever site they prefer to use and then it appears on the other
sites.

Its' with this goal in mind that we've created an open licensing
solution where other markets can generate license keys on our system,
this means that developers can implement a single solution and other
markets can create the necessary licenses (as long as the developer
gives them their API key of course).

We are also working on taking application updates from an RSS feed so
that we feed from the developers site, which should reduce the amount
of work developers need to do (they submit the feed URL once and then
we automatically list all the updates).

At the moment app directories is a very cut throat area to be involved
in. To give you an idea of the sort of thing going on; We were in
discussions with one OEM who wanted to take a feed from our database,
license the AndAppStore architecture and a customised client, but, a
month before launch, produced a contract which excluded us from
listing at AndAppStore any applications that the OEM made available
via their servers and handed all IP rights to our application over to
them, which was a long way from what was agreed. We didn't agree to
those terms and they now have launched their own store using revenue
models and data formats we supplied to them implemented on a server
architecture we designed but with an ISP we had concerns about, and
they have significantly less applications then we've listed, oh, and
they made no indication that they intend to pay us for the work we did.

We've also had approaches from sites saying they want a feed from us
and when I've asked them for a reciprocal feed they've lost interest.
This has to be a two way thing for it to work for everyone, it can't
be a case of one site gets all the feeds, shows all the apps, and
takes all the money from advertising and payment revenues, whilst
others do all the work with developers.

Personally I believe Android needs multiple market sites, if for no
other reason than to drive innovation and provide users with choice,
but hey, that's just my personal opinion :).

Al.

--

* Looking for Android Apps? - Try http://andappstore.com/ *

======
Funky Android Limited is registered in England & Wales with the
company number 6741909. The registered head office is Kemp House,
152-160 City Road, London, EC1V 2NX, UK.

The views expressed in this email are those of the author and not
necessarily those of Funky Android Limited, it's associates, or it's
subsidiaries.

[Eric F]

I couldn't agree with this any more, Mark. I think part of the
insanity is how Google Maps is a free download for Windows Mobile
phones but contraband on Android. I think if it weren't a part of the
the Android "platform", and was an App listed in the market, we
wouldn't be nearly as deep in this mess.

I think I understand why some of these components seem to be in the
"platform". Because so many early apps were just Maps mashups,
treating Maps like an application implies that it might not be there
and all these apps don't work. But I think the right solution would be
to have "pick a location on a map" intent, cause the user to realize
"you don't have an app for that.... yet, search _____ Market for apps
that handle this intent". Maybe that's not as slick as everything
coming loaded out of the box. But in my experience, as long as users
always have a fairly simple "next" button to get an acceptable
solution they'll be willing to tolerate a little inconvenience.

-E

On Sep 27, 9:06 am, Mark Murphy <mmur...@commonsware.com> wrote:
> For me, the only closed-source Google
> application that really needs to be in this firmware is a market;
> everything else can be treated as a regular Android app AFAICT.
>
> Others will have their own motivations. I doubt it's "one size fits all".
>
> --

> Mark Murphy (a Commons Guy)http://commonsware.com|http://twitter.com/commonsguy

[Al Sutton]

I think it's more business constraints' than insanity.

Google have to provide a tangible product for the "Google Experience"
brand which they want OEMs to sign up for, and on Android the tangible
benefits include the right to use certain apps from Google which
improve the user experience.

Take, for example, the Archos device. It uses an officially
unsupported resolution on cupcake (WVGA), uses software buttons
instead of hardware buttons for home, back, and menu, has no cellular
functionality, and no compass. I'm sure Google don't like some (if not
all) of these points and so if Archos could ship all the apps a Google
Experience device has then it would leave users confused as to what
the Google experience actually is.

On WinMo, Blackberry, Symbian, etc. there is no Google experience
concept, and thus nothing that needs protecting.

Don't get me wrong, I'm not saying it's an approach I agree with, I'm
just saying I can understand where Google is coming from.

Al.


--

* Looking for Android Apps? - Try http://andappstore.com/ *

======
Funky Android Limited is registered in England & Wales with the
company number 6741909. The registered head office is Kemp House,
152-160 City Road, London, EC1V 2NX, UK.

The views expressed in this email are those of the author and not
necessarily those of Funky Android Limited, it's associates, or it's
subsidiaries.

[Mark Murphy]

Eric F wrote:
> But I think the right solution would be
> to have "pick a location on a map" intent, cause the user to realize
> "you don't have an app for that.... yet, search _____ Market for apps
> that handle this intent".

That might even be just an extension of the current
Intent.createChooser() logic. That's the code that pops up the menu of
apps that, say, support ACTION_SEND when you want to send something.
Stick a standard entry on there for "search for another app". The
manifest already contains the information needed to determine many apps
that would support that action or MIME type.

--

Mark Murphy (a Commons Guy)
http://commonsware.com | http://twitter.com/commonsguy

Need Android talent? Ask on HADO! http://wiki.andmob.org/hado

[Dianne Hackborn]

On Mon, Sep 28, 2009 at 1:01 AM, Eric F <ericf...@gmail.com> wrote:

I couldn't agree with this any more, Mark. I think part of the
insanity is how Google Maps is a free download for Windows Mobile
phones but contraband on Android. I think if it weren't a part of the
the Android "platform", and was an App listed in the market, we
wouldn't be nearly as deep in this mess.

The maps application has been available on market as a separate download for a while.  As we've said, this is where we would eventually like to be with all of the Google-specific stuff.  It just isn't something that will happen over-night.

If someone were to take the Windows Mobile maps application and start distributing it themselves without some kind of license, I'm not too sure Google would be happy about that either.

[Dianne Hackborn]

On Mon, Sep 28, 2009 at 1:42 AM, Mark Murphy <mmu...@commonsware.com> wrote:

That might even be just an extension of the current

Intent.createChooser() logic.

Yep, this is definitely something we want to do once some other higher priority things are out of the way.  (Also just allow this to happen for any Intent an app starts where there isn't something already installed on the device that can handle it.  Like say you get a PDF file as a mail attachment, click to view it, and there is no activity installed that can view that MIME type.)

[Mark Murphy]

Dianne Hackborn wrote:
> On Mon, Sep 28, 2009 at 1:42 AM, Mark Murphy <mmu...@commonsware.com
> <mailto:mmu...@commonsware.com>> wrote:
>
> That might even be just an extension of the current
> Intent.createChooser() logic.
>
>
> Yep, this is definitely something we want to do once some other higher
> priority things are out of the way. (Also just allow this to happen for
> any Intent an app starts where there isn't something already installed
> on the device that can handle it. Like say you get a PDF file as a mail
> attachment, click to view it, and there is no activity installed that
> can view that MIME type.)

*That* will be seriously sweet. I just went through an hunt for a pair
of customers, trying to find a Twitter app that supports ACTION_SEND
(FWIW, Twidroid does). Having that sort of research be effectively
automatic for all users would be magnificent.

[Disconnect]

If "android" is defined by the market, then it is a very closed
system. Google hasn't published any information about getting a device
certified for the market, other than the insistence that I can't prove
its not free.. (Look, its my n800: It runs android, it has a
reasonable screen size, it supports the OSK. What does it take to get
market - non-protected apps only - onto it?)

If Android is open, and its defined by market, there needs to be a way
for the community to get some variant of the market onto supported
devices. I feel the pain of the qvga saga (and in fact, we were going
to buy a couple of those before they got cancelled) but it is time to
let go a little. If - eg - cyanogen's roms come with andappstore
(sorry Shane :) ..) and suddenly 20k users move over for their primary
purchasing, the devs are going to follow. Due to the nature of the
beast, it is certainly possible that google could lose the market
relevance entirely, for 'social' android users. ("Now, the first thing
you want to do is go to go.to/andapps and get that client. Thats where
all the good apps are, and they're cheaper than google.")

The federated update plan is only going to make that worse (from
google's perspective) because now the developer's barrier to entry is
-very- low - upload to google, upload to federation, collect checks.
(And enough time has passed to give confidence in the new stores
actually paying on time; that seemed to be a big concern back when
they got started.) And the users don't have to deal with chasing down
app home pages, etc - they either get it pre-shipped or it is a single
quick install.

(Forgive any incoherence in this one, its a bit early.)

[gjs]

Hi,

After nearly a year, the facilities of the Market app are still not
yet provided uniformly to the worldwide Android developer community,
with paid apps only allowed to be sold from developers in 9 countries/
regions, but capable of being purchased by users in more regions,
particularly with rooting.

This keeps a lot (?) of developers out of the Market (app) which does
not serve Android, particularly where free apps are restricted from
referring to paid app's on other markets or other payment means.

The Market app seems to be part of the Google experience which seems
to be tied to the deals with telcos and handsets, and this seems to be
at the expense of developers in regions where deals are still pending
(or can't be fully agreed).

The Market app would be better if it at least treated the Android
developer community uniformly ( Open Market for developers with paid
apps ) regardless of their region and separate to other deals and
initiatives.

Regards

On Sep 28, 6:51 am, Dianne Hackborn <hack...@android.com> wrote:

> hack...@android.com

[Chris Stratton]

On Sep 27, 11:27 am, Shane Isbell <shane.isb...@gmail.com> wrote:

> While good for the user community, this strikes me as a blow against any
> developer community for open-source implementations of replacement apps. It
> appears the real motivation for wanting alternatives is that people can't
> have both improved ROMs AND the Google apps. Once the itch is gone from the
> user/developer community, what motivation is there for open-source
> alternatives or alternatives at all?

I would suggest maybe taking a look at the spectrum of things that
happen on a phone, and trying to carefully identify the specific band
where open solutions really make sense.

On the one side, you have hardware drivers. Open solutions would be
welcome, but since the hardware itself comes with some many strings
attached (only a minority of devices officially permit reflashing) I
start to wonder what the point is - if the binary blobs work, why not
just retain them and only distribute the open components?

On the other end, a lot of the nice features of android phones are
interfaces to servers... specifically, google's search, maps, gmail,
etc servers. Again, is it really worthwhile to develop open
solutions to what could fairly be called closed problems, if closed
solutions can simply be retained?

My feeling is that the prime area for open efforts would really be the
middle ground between those two - things that aren't too hardware
specific, and aren't front ends for servers where protocols would have
to be reverse engineered. I'm not actually thinking of good examples
of such middle ground problems off the top of my head, but I'm sure
they are out there.

[Chris Stratton]

On Sep 28, 4:50 am, Dianne Hackborn <hack...@android.com> wrote:

> If someone were to take the Windows Mobile maps application and start
> distributing it themselves without some kind of license, I'm not too sure
> Google would be happy about that either.

The real irony would be to make an android sort of wine, so you could
download the winmo executable from google and run that !

[Mike Hearn]

> The Market app seems to be part of the Google experience which seems
> to be tied to the deals with telcos and handsets, and this seems to be
> at the expense of developers in regions where deals are still pending
> (or can't be fully agreed).

Actually, when I asked about this I was told the main problem is
getting the necessary paperwork in place. Apparently [re]selling
software in the way the Market does has all kinds of tax implications.
Regional limitations like that are really not much about business
deals and more about government regulation.

BTW I'd like to thank Diane for her post explaining the market
situation. It's a refreshingly clear and useful explanation of what's
going on. I think many of us are here supporting Android exactly
because it's a pragmatic approach to open source that is intelligently
balancing many different things. I worked on desktop Linux for a long
time and it never made progress because of that lack of balance.

[gjs]

> Actually, when I asked about this I was told the main problem is
> getting the necessary paperwork in place. Apparently [re]selling
> software in the way the Market does has all kinds of tax implications.
> Regional limitations like that are really not much about business
> deals and more about government regulation.

Hi,

I have seen that mentioned before and used to just accept that may be
the case. But it gets harder to continue accept as time goes by when
it is considered with -

* I can go to the alternate Markets to try to sell apps, so they seem
to be able to short out the paper work and regulations.
* There is 2 telcos selling Android google experience phones in my
region since Feb & Jun 2009, at least one of which is allowing users
to purchase apps through the Market AFAIK, both allow free apps.
* Developers in my region can be use the Market to distribute fee apps
since around Feb / March 2009.
* I already use Google Checkout for other purchases in my region.

IANAL but I did not think that the Market [re]sells apps, Developers
(own &) sell (licenses to use) their apps, the Market facilitates that
by accepting, listing and distributing apps as well as providing the
payment infrastructure and disbursing the proceeds.

I really appreciate what Android is and the apparent care with which
it is being managed, particularly Dianne explanations and the recent
hard work being done by Google and the community to make it acceptably
AOSP for everybody.

I've been using, developing (and advocating) for Android since day 1,
investing my energy for nearly 2 years now and I'm still waiting for
the Market to allow me the option of selling apps. So its somewhat a
smoldering and ongoing frustration so I continue to mention it not
only for myself but for the rest of the developers outside the
supported regions.

Market support for selling app from Android developers from any region
can only benefit Android and I wish we could be helped more to achieve
that, the Market slogan is 'Come one, Come All'.

Android has lowered the barriers to entry in the Market - provided
your allowed in and with no (concrete) ideas about when that might be
if your not.

Not all of us want or can afford to just give away our efforts or can
afford to run other business models where you can give away your free
client app.

Regards

[gjs]

should be -

* Developers in my region can be use the Market to distribute FREE

apps
since around Feb / March 2009.