About Linux support at marketplace

[Rubén Martín]

Hello,

I know that sometimes some features are more complicated to implement in
one OS than others, but if we want to keep the Linux users happy and
avoid them to think their OS is a second class OS for mozilla, we should
take care of these things.

Mailing all mozillians about testing marketplace before having support
under Linux is not good, most community members use Linux as their main
OS (including myself) and most of them would love to test apps and help.

Regards.

--
Rubén Martín [Nukeador]
Mozilla Reps Council Member
http://www.mozilla-hispano.org
http://twitter.com/mozilla_hispano
http://facebook.com/mozillahispano

[Rubén Martín]

I just want to add that I can totally understand that Linux support is
not ready yet, but let's take care of communication and explain why to
the future testers and/or link to the work in progress bug
<https://bugzilla.mozilla.org/show_bug.cgi?id=744193>. We don't want
people to feel undervalued just because they/we use Linux.

[Rubén Martín]

I've just written to dev-platform to get more information about it, but
it seems that Linux is P3 for Kilimanjaro event:

https://wiki.mozilla.org/Kilimanjaro/ProductDraft#You_will_be_able_to_install_and_use_your_apps_across_phones_and_PCs_where_WebRT_is_available

Let's wait for an explanation because I can't believe it's decided that
we will have a second class Firefox on Linux without apps.

Is someone more informed about this issue?

[Daniel Mills]

Hey buddy! ;)

I happen to be lurking on this list, but it's not really a good place to discuss this... I only read it by chance. However, since I'm here, I'll give you my $.02:

Linux support for apps is a nice to have because most of our users are not running Linux. I think we're supportive and absolutely willing to accept patches to make something work on Linux, but it's just not something that affects the 80% (I don't think it's even 10%, though I don't have any data handy). By definition, this is a nice to have, not a stop-ship feature.

Remember that we are making software for a lot of people, and staff and community are actually a tiny slice of the userbase. I know it's hard, but we need to focus on the userbase at large, not on us.

Dan

> _______________________________________________
> Interested in promoting Mozilla? Check out the Mozilla Community Marketing Guide: http://contribute.mozilla.org/Marketing
> mark...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/marketing

[Percy Cabello]

Hi

On Tue, May 8, 2012 at 7:12 PM, Daniel Mills <thu...@mozilla.com> wrote:

> Linux support for apps is a nice to have because most of our users are not
> running Linux. I think we're supportive and absolutely willing to accept
> patches to make something work on Linux, but it's just not something that
> affects the 80% (I don't think it's even 10%, though I don't have any data
> handy). By definition, this is a nice to have, not a stop-ship feature.
>
> Remember that we are making software for a lot of people, and staff and
> community are actually a tiny slice of the userbase. I know it's hard, but
> we need to focus on the userbase at large, not on us.
>
>

Don't have number either but I guess among "influencers" (the ones we are
going to need to help drive users and developers out of closed but well
established ecosystems) Linux share is more significant, not to mention the
very vocal open source community which is very Linux centric.

I also think these are marketing rather than technical concerns.

Percy

[Rubén Martín]

El 09/05/12 02:12, Daniel Mills escribió:

> Linux support for apps is a nice to have because most of our users are not running Linux. I think we're supportive and absolutely willing to accept patches to make something work on Linux, but it's just not something that affects the 80% (I don't think it's even 10%, though I don't have any data handy). By definition, this is a nice to have, not a stop-ship feature.
>
> Remember that we are making software for a lot of people, and staff and community are actually a tiny slice of the userbase. I know it's hard, but we need to focus on the userbase at large, not on us.

Then I think it not a good idea, Linux it's been a P1 OS always and one
of the reasons we have a big community of enthusiasts is due Linux users.

Linux is not another platform, it's the platform which shares our values
about being open and the reason most people gets involved with mozilla,
because they believe in libre software and in the open web. Not
supporting linux is not supporting a big group of people that empowers
mozilla, and not supporting them/us is not supporting mozilla.

From my point of view if we are lack of resources for linux (both
volunteers and paid staff), let's get more employees working on it or do
a call for help to the community, it shouldn't be that hard.

I choose marketing mailing list because is where most people interested
in how we communicate or the result of our actions are, what mailing
list do you suggest we have both technical and non-technical people?

[fabi1.c...@gmail.com]

Le mercredi 9 mai 2012 02:12:00 UTC+2, Daniel Mills a écrit :
> Linux support for apps is a nice to have because most of our users are not running Linux. I think we're supportive and absolutely willing to accept patches to make something work on Linux, but it's just not something that affects the 80% (I don't think it's even 10%, though I don't have any data handy). By definition, this is a nice to have, not a stop-ship feature.

I can’t believe I’m reading this. o_O

Are you saying that the Marketplace is unsupported on Linux because it has little market share? Then why do we even bother having Linux versions of Firefox at all??

There must be a *big* misunderstanding here. I must be missing something obvious.

[pascalc]

On 9 mai, 02:12, Daniel Mills <thun...@mozilla.com> wrote:
> Hey buddy! ;)
>
> I happen to be lurking on this list, but it's not really a good place to discuss this... I only read it by chance. However, since I'm here, I'll give you my $.02:
>
> Linux support for apps is a nice to have because most of our users are not running Linux. I think we're supportive and absolutely willing to accept patches to make something work on Linux, but it's just not something that affects the 80% (I don't think it's even 10%, though I don't have any data handy). By definition, this is a nice to have, not a stop-ship feature.
>
> Remember that we are making software for a lot of people, and staff and community are actually a tiny slice of the userbase. I know it's hard, but we need to focus on the userbase at large, not on us.
>
> Dan

Hi Dan,

We call it "Open" Web because the web page does not care if I run
Linux, Windows or Mac. Mozilla has always made sure that the Web
remains open to all, even those that are not Apple or Windows
customers. It would be easier to only support Windows and Mac for
Firefox, it would even be easier to only support Windows and drop Mac
as well, our market share would probably not be that impacted. But I
believe that we make decisions not only for the potential user base we
can ship a product to like any regular dotcom company that would only
target the profitable market, but also (and hopefully foremost)
because we are following the Mozilla Mission and its Manifesto. Don't
you think that there is a problem in telling that we are building a
marketplace that is built on the Open Web but closed to our own
volunteer community? We wouldn't support Linux on our own marketplace
while our new flagship product, B2G is based on a Linux kernel? We
wouldn't support Linux when even Chrome has a marketplace that is open
today to Linux users? That just doesn't make sense to me and I hope
this decision will be revised.

Regards,

Pascal

[Bill Walker]

Rubén,

Thanks for your interest in the Marketplace!

I can understand your frustration in our launching the Marketplace without Linux support. As you heard, Marco Castelluccio is working on the Web Runtime (WebRT) for Linux as we speak.

Unfortunately, we need to start getting feedback about the Marketplace now if we are going to meet our Kilimanjaro goals of making the Marketplace awesome for B2G. Given our finite engineering resources and schedule, we decided to make the announcement to Mozillians before WebRT for Linux was finished.

I'd be happy to try to answer any other questions, but I'll be on PTO for a week. Likely others on this list can chime in, too.

yours,
-Bill Walker, engineering manager, Mozilla Labs

----- Original Message -----
From: "Rubén Martín" <nuke...@mozilla-hispano.org>
To: "marketing" <mark...@lists.mozilla.org>
Cc: apps-f...@mozilla.com
Sent: Tuesday, May 8, 2012 1:09:13 PM
Subject: About Linux support at marketplace

Hello,

I know that sometimes some features are more complicated to implement in
one OS than others, but if we want to keep the Linux users happy and
avoid them to think their OS is a second class OS for mozilla, we should
take care of these things.

Mailing all mozillians about testing marketplace before having support
under Linux is not good, most community members use Linux as their main
OS (including myself) and most of them would love to test apps and help.

[Rodrigo Garcia]

On Tue, May 8, 2012 at 10:08 PM, Bill Walker <bwa...@mozilla.com> wrote:
> Rubén,
>
> Thanks for your interest in the Marketplace!
>
> I can understand your frustration in our launching the Marketplace without Linux support. As you heard, Marco Castelluccio is working on the Web Runtime (WebRT) for Linux as we speak.
>
> Unfortunately, we need to start getting feedback about the Marketplace now if we are going to meet our Kilimanjaro goals of making the Marketplace awesome for B2G. Given our finite engineering resources and schedule, we decided to make the announcement to Mozillians before WebRT for Linux was finished.
>
> I'd be happy to try to answer any other questions, but I'll be on PTO for a week. Likely others on this list can chime in, too.
>
> yours,

Hi Everyone,
I agree with Rubén Martin. IMHO is very, very, very sad that
MozillaLabs team has chosen Windows and OSX as the initial supported
platform, putting Linux support on the queue. I think is a very bad
signal either for the Market and for our volunteers. I hope this
decision will be revisited.
If the problem is related with resources or budget for the project,
I'm sure the organization can do an extra effort to correct this
decission.


rod

[Alex Jordan]

On Tue, May 8, 2012 at 7:19 PM, Rodrigo Garcia <r...@firefox.cl> wrote:

> Hi Everyone,
> I agree with Rubén Martin. IMHO is very, very, very sad that
> MozillaLabs team has chosen Windows and OSX as the initial supported
> platform, putting Linux support on the queue.

Did you even read the post you just replied to?
Linux WebRT support is ongoing.

I think is a very bad
> signal either for the Market and for our volunteers. I hope this
> decision will be revisited.
> If the problem is related with resources or budget for the project,
> I'm sure the organization can do an extra effort to correct this
> decission.
>

Has it occurred to you that it's possible that Mozilla actually can't do
this? Has it occurred to you that Mozilla is possibly already in overdrive
what with Rapid Release, B2G, launching the *beta* of the Marketplace, and
other things?
Mozilla doesn't have infinite resources.
Think of it this way: there would be a certain amount of time before WebRT
was on Linux. Mozilla could choose to launch it then, or it could launch it
earlier and get valuable feedback. There's no reason to deny Windows and OS
X users the ability to test it, and there's strong advantages in that
Mozilla can get feedback earlier and stay on track for the Kilimanjaro
event.

[chris hofmann]

I think its probably a premature optimization to be dropping platforms
at this stage of the requirements definition and it definitely doesn't
match the the initial context section of the document.
https://wiki.mozilla.org/Kilimanjaro/ProductDraft#Context_-_The_Web_is_the_Platform

especially this part....

A people-centered system will enable users to identify themselves to
websites and apps on their own terms*; to take their apps easily
from one OS to another;* to disclose only what they want about
themselves in proportion to the value they receive.


If we want to meet that requirement we will need to do tablets and linux.

I think the problem here is problem one of not having a deep enough
understanding of how we will do *any* of the OS integration for these
requirements:

P1: Launch from native standard launch points (start menu, dock, start
page, Home Screen etc.)
P1: Launch in chromeless window, outside the browser application (bug
740586 <https://bugzilla.mozilla.org/show_bug.cgi?id=740586>) (bug
743917 <https://bugzilla.mozilla.org/show_bug.cgi?id=743917>)

The other requirements appear to be largely cross platform.

Its always been easier to find windows and mac knowledgable people to
figure out the OS integration parts. We just need to find the right
engineer to figure out these requirements for Linux and Tablets, and as
the last line indicates iOS. That will be an even tougher integration
job.

That's the way get
https://wiki.mozilla.org/Kilimanjaro/ProductDraft#You_will_be_able_to_install_and_use_your_apps_across_phones_and_PCs_where_WebRT_is_available
fixed so its on track to ship with the other platforms.

Any ideas on people that could step up to figure out these areas on
those platforms? I think this goes beyond thunder's suggestion of
"patches accepted" to more like "patches highly desired and sought
after!" Eventually we will want this work done, so lets find the right
module owner to start tackling it now.

-chofmann

On 5/8/12 7:19 PM, Rodrigo Garcia wrote:
> On Tue, May 8, 2012 at 10:08 PM, Bill Walker<bwa...@mozilla.com> wrote:
>> Rubén,
>>
>> Thanks for your interest in the Marketplace!
>>
>> I can understand your frustration in our launching the Marketplace without Linux support. As you heard, Marco Castelluccio is working on the Web Runtime (WebRT) for Linux as we speak.
>>
>> Unfortunately, we need to start getting feedback about the Marketplace now if we are going to meet our Kilimanjaro goals of making the Marketplace awesome for B2G. Given our finite engineering resources and schedule, we decided to make the announcement to Mozillians before WebRT for Linux was finished.
>>
>> I'd be happy to try to answer any other questions, but I'll be on PTO for a week. Likely others on this list can chime in, too.
>>
>> yours,

> Hi Everyone,
> I agree with Rubén Martin. IMHO is very, very, very sad that
> MozillaLabs team has chosen Windows and OSX as the initial supported

> platform, putting Linux support on the queue. I think is a very bad

> signal either for the Market and for our volunteers. I hope this
> decision will be revisited.
> If the problem is related with resources or budget for the project,
> I'm sure the organization can do an extra effort to correct this
> decission.
>
>

> rod

[Rodrigo Garcia]

On Tue, May 8, 2012 at 10:32 PM, Alex Jordan <alexande...@gmail.com> wrote:

> On Tue, May 8, 2012 at 7:19 PM, Rodrigo Garcia <r...@firefox.cl> wrote:
>
>> Hi Everyone,
>> I agree with Rubén Martin. IMHO is very, very, very sad that
>> MozillaLabs team has chosen Windows and OSX as the initial supported
>> platform, putting Linux support on the queue.
>

> Did you even read the post you just replied to?
> Linux WebRT support is ongoing.

Alex,
Thanks for the heads-up. Of course I read it.

>
> I think is a very bad
>> signal either for the Market and for our volunteers. I hope this
>> decision will be revisited.
>> If the problem is related with resources or budget for the project,
>> I'm sure the organization can do an extra effort to correct this
>> decission.
>>

> Has it occurred to you that it's possible that Mozilla actually can't do
> this? Has it occurred to you that Mozilla is possibly already in overdrive
> what with Rapid Release, B2G, launching the *beta* of the Marketplace, and
> other things?

Beside ironies, I'm sure there are powerful reasons to take this
decision. My point, and seems the concerns of other people who wrote
also through this tread, is the impact this on the media and users.

> Mozilla doesn't have infinite resources.

Nobody has said that.

> Think of it this way: there would be a certain amount of time before WebRT
> was on Linux. Mozilla could choose to launch it then, or it could launch it
> earlier and get valuable feedback. There's no reason to deny Windows and OS
> X users the ability to test it, and there's strong advantages in that
> Mozilla can get feedback earlier and stay on track for the Kilimanjaro
> event.

Nobody is denying the rights of those users. The only point here is a
question and a strong suggestion of community members. We're allow to
disagree and communicate it in a proper way.

Thanks!

[Alex Jordan]

That's a valid concern, and one that I hadn't thought of.
I apologize.

>
> > Mozilla doesn't have infinite resources.
>
> Nobody has said that.
>
> > Think of it this way: there would be a certain amount of time before
> WebRT
> > was on Linux. Mozilla could choose to launch it then, or it could launch
> it
> > earlier and get valuable feedback. There's no reason to deny Windows and
> OS
> > X users the ability to test it, and there's strong advantages in that
> > Mozilla can get feedback earlier and stay on track for the Kilimanjaro
> > event.
>
> Nobody is denying the rights of those users.

My point was that holding this back just for Linux support would be denying
the rights of Windows and OS X users to test this.

[Jason Smith]

Hi Chris,

+Aaron Train (Android WebRT QA)
+Stephen Donner (k9o QA driver)
+Tony Chung (Mobile QA lead)
+Kevin Brosnan (Contributor turned employee)

Good feedback. Comments inline.

Sincerely,
Jason Smith

Desktop QA Engineer
Mozilla Corporation
https://quality.mozilla.org/

On 5/8/2012 7:44 PM, chris hofmann wrote:
>
> I think its probably a premature optimization to be dropping platforms
> at this stage of the requirements definition and it definitely doesn't
> match the the initial context section of the document.
> https://wiki.mozilla.org/Kilimanjaro/ProductDraft#Context_-_The_Web_is_the_Platform
>
> especially this part....
>
> A people-centered system will enable users to identify themselves
> to websites and apps on their own terms*; to take their apps
> easily from one OS to another;* to disclose only what they want
> about themselves in proportion to the value they receive.
>
>
> If we want to meet that requirement we will need to do tablets and linux.

Right, I agree that the vision reflects the need for tablets and linux.
From the tester's perspective, I'm a little less worried about the risk
of not having tablet support, as the web apps integration for android is
built into fennec native and reusing some of it's functionality. Fennec
native itself, can run on a tablet, so we might still be okay to get
some tablet support (I'll want to confirm with it with some testing
though, Aaron & Tony probably could provide more insight here).

For linux, I believe the original rationale for not going for support
initially was linux's low marketshare along with added engineering
resources needed for it (i.e. it doesn't get the same benefit tablet
gets above). However, a recent discussion I had with a contributor
turned employee (Kevin Brosnan) makes me rethink that the primary reason
for supporting linux is not marketshare. A rationale like you stated
above relates to openness - the user makes the decision to use apps
wherever they like. Another relates to paying attention to the needs of
contributors - What do they find important? Note that I know linux users
also happen to be loud, so not supporting their platform could hurt us
from PR perspective (e.g. blog posts bashing us). Overall, I definitely
believe we need to rethink our strategy on linux based on the rationale
provided.

>
> I think the problem here is problem one of not having a deep enough
> understanding of how we will do *any* of the OS integration for these
> requirements:

For linux, this is being thought about by one of our contributors,
Marco. He's written up a specification of what he thinks should be done
here - https://etherpad.mozilla.org/tzkVRnfvm1. Feel free to provide
feedback.

>
> P1: Launch from native standard launch points (start menu, dock, start
> page, Home Screen etc.)
> P1: Launch in chromeless window, outside the browser application (bug
> 740586 <https://bugzilla.mozilla.org/show_bug.cgi?id=740586>) (bug
> 743917 <https://bugzilla.mozilla.org/show_bug.cgi?id=743917>)
>
> The other requirements appear to be largely cross platform.
>
> Its always been easier to find windows and mac knowledgable people to
> figure out the OS integration parts. We just need to find the right
> engineer to figure out these requirements for Linux and Tablets, and
> as the last line indicates iOS. That will be an even tougher
> integration job.

Right, although I actually think keeping iOS out of scope is alright to
do for now, as we need to nail down the web apps UX flow correctly on
one mobile operating system and being successful, before we tackle
another operating system. For linux, this is probably a good opportunity
to form a community around web apps integration into linux, as even
today I've seen interest in this.

>
> That's the way get
> https://wiki.mozilla.org/Kilimanjaro/ProductDraft#You_will_be_able_to_install_and_use_your_apps_across_phones_and_PCs_where_WebRT_is_available
> fixed so its on track to ship with the other platforms.
>
> Any ideas on people that could step up to figure out these areas on
> those platforms? I think this goes beyond thunder's suggestion of
> "patches accepted" to more like "patches highly desired and sought
> after!" Eventually we will want this work done, so lets find the
> right module owner to start tackling it now.

Good call. Myk Melez is looking to get a module owner defined for the
web app runtime on desktop on the governance discussion thread. Let's
get that defined quickly and come up with a way to get patch support
both internally and externally to get support platforms such as linux. I
would be more than happy to help out with promoting community
involvement with this. Myk - What do you think?

>
> -chofmann
>
> On 5/8/12 7:19 PM, Rodrigo Garcia wrote:
>> On Tue, May 8, 2012 at 10:08 PM, Bill Walker<bwa...@mozilla.com> wrote:
>>> Rubén,
>>>
>>> Thanks for your interest in the Marketplace!
>>>
>>> I can understand your frustration in our launching the Marketplace without Linux support. As you heard, Marco Castelluccio is working on the Web Runtime (WebRT) for Linux as we speak.
>>>
>>> Unfortunately, we need to start getting feedback about the Marketplace now if we are going to meet our Kilimanjaro goals of making the Marketplace awesome for B2G. Given our finite engineering resources and schedule, we decided to make the announcement to Mozillians before WebRT for Linux was finished.
>>>
>>> I'd be happy to try to answer any other questions, but I'll be on PTO for a week. Likely others on this list can chime in, too.
>>>
>>> yours,

>> Hi Everyone,
>> I agree with Rubén Martin. IMHO is very, very, very sad that
>> MozillaLabs team has chosen Windows and OSX as the initial supported

>> platform, putting Linux support on the queue. I think is a very bad

>> signal either for the Market and for our volunteers. I hope this
>> decision will be revisited.
>> If the problem is related with resources or budget for the project,
>> I'm sure the organization can do an extra effort to correct this
>> decission.
>>
>>

[Ragavan Srinivasan]

Hi Rubén and others,

On Tuesday, May 8, 2012 4:36:35 PM UTC-4, Rubén Martín wrote:
> I just want to add that I can totally understand that Linux support is
> not ready yet, but let's take care of communication and explain why to
> the future testers and/or link to the work in progress bug
> <https://bugzilla.mozilla.org/show_bug.cgi?id=744193>. We don't want
> people to feel undervalued just because they/we use Linux.

I am the product manager for Apps and I wanted to address some of the concerns raised here.

First of all, I'll admit that we could have done a much better job of communicating the level of Linux support at this stage of the Apps project. Though we did call it out on the Kilimanjaro planning page and during other engineering meetings, we will be sure to keep this in mind for future communications.

Coming to the question at hand, lack of support for Linux at this stage is mainly due to two factors.
1. The technical feasibility of implementing support for Apps features in the various Linux distributions. and the resource and time constraints associated with doing so.
2. The market share of Linux desktop users.

I'd like to focus on #1 here. As you can see from the bug you link to above (and the etherpad linked to from the bug), this is a particularly hard technical problem. We don't have the right level of skills nor the time commitment required to do this justice. Which is why we are very excited about Marco's work as part of the Summer of Code project. If you or others have the skills and time to help move this forward, we'd love to get your help.

I just read chofmann's reply and his suggestion of finding the right technical help to move this forward is spot on!

Regards,
Ragavan

[Ragavan Srinivasan]

Hi Rubén and others,

On Tuesday, May 8, 2012 4:36:35 PM UTC-4, Rubén Martín wrote:

> I just want to add that I can totally understand that Linux support is
> not ready yet, but let's take care of communication and explain why to
> the future testers and/or link to the work in progress bug
> <https://bugzilla.mozilla.org/show_bug.cgi?id=744193>. We don't want
> people to feel undervalued just because they/we use Linux.

[Rodrigo Garcia]

I don't know who handles PR related with Engineering decisions, but I
would suggest this should be addressed in a blog post to explain to
the market this decision in terms doesn't hurt the image of openness
and neutrality of Mozilla. IMHO Chris and Jason details puts, the
things in a good perspective. Thanks!
Also, I think this has been a great exercise about communication
between the community and the employees.

rod

On Wed, May 9, 2012 at 12:02 AM, Jason Smith <jsm...@mozilla.com> wrote:
> Hi Chris,
>
> +Aaron Train (Android WebRT QA)
> +Stephen Donner (k9o QA driver)
> +Tony Chung (Mobile QA lead)
> +Kevin Brosnan (Contributor turned employee)
>
> Good feedback. Comments inline.
>
> Sincerely,
> Jason Smith
>
> Desktop QA Engineer
> Mozilla Corporation
> https://quality.mozilla.org/
>
>
> On 5/8/2012 7:44 PM, chris hofmann wrote:
>
>
> I think its probably a premature optimization to be dropping platforms at
> this stage of the requirements definition and it definitely doesn't match
> the  the initial context section of the document.
> https://wiki.mozilla.org/Kilimanjaro/ProductDraft#Context_-_The_Web_is_the_Platform
>
> especially this part....
>
> A people-centered system will enable users to identify themselves to

> websites and apps on their own terms; to take their apps easily from one OS
> to another; to disclose only what they want about themselves in proportion

> 740586) (bug 743917)

[ragavan.s...@gmail.com]

On Wednesday, May 9, 2012 12:02:43 AM UTC-4, Jason Smith wrote:

> > Its always been easier to find windows and mac knowledgable people to
> > figure out the OS integration parts. We just need to find the right
> > engineer to figure out these requirements for Linux and Tablets, and
> > as the last line indicates iOS. That will be an even tougher
> > integration job.
> Right, although I actually think keeping iOS out of scope is alright to
> do for now, as we need to nail down the web apps UX flow correctly on
> one mobile operating system and being successful, before we tackle
> another operating system. For linux, this is probably a good opportunity
> to form a community around web apps integration into linux, as even
> today I've seen interest in this.

I second the call to get more community contributions for apps integration on linux. That said, I'd also not view Linux support as being too different from the lack of iOS support you mention above. We have picked *one* mobile OS to support for a new product and we have picked *two* desktop OSs to support that same new product. I'd encourage you to not look at this as a feature of Firefox, but rather to look at this is a new product (Apps and more specifically, WebRT) we are bringing to the market. We need to start somewhere.

Regards,
Ragavan

[ragavan.s...@gmail.com]

On Wednesday, May 9, 2012 12:02:43 AM UTC-4, Jason Smith wrote:

> > Its always been easier to find windows and mac knowledgable people to
> > figure out the OS integration parts. We just need to find the right
> > engineer to figure out these requirements for Linux and Tablets, and
> > as the last line indicates iOS. That will be an even tougher
> > integration job.
> Right, although I actually think keeping iOS out of scope is alright to
> do for now, as we need to nail down the web apps UX flow correctly on
> one mobile operating system and being successful, before we tackle
> another operating system. For linux, this is probably a good opportunity
> to form a community around web apps integration into linux, as even
> today I've seen interest in this.

[Patrick Finch]

Two things, but maybe they are not obvious:

Linux is a P3 for this initial release. Linux support will come. I'm
disappointed too, but I find it very hard to disagree with the
prioritisation: if you want to test at volume or to make the product
credible in the market, Linux is unfortunately not a priority. It's
1-3% of the desktop market depending on the country.

You might argue that a prioritisation like this will only perpetuate
that situation...but when the Web wins, Linux wins. Even if you
disagree with the prioritisation for the initial release, you presumably
agree with the vision?

There's certainly no inconsistency with having supported Linux builds of
Firefox.


Patrick

--
Patrick Finch
Mozilla
pfi...@mozilla.com
Mobile: +46 768 444 833
IM: patric...@gmail.com

[Nukeador]

Hi,

Thanks for all the replies. I think we all agree on this first release
problem, it's OK not having Linux support if there are still some problems
to solve and it should be explained to the community.

What it worries me the most is not being sure about webapps on Linux for
beta or stable marketplace release, reading bugs and wiki is not clear, I
don't mind having a delay for this first closed release, but I really do
for the first public one and of course for the final stable.

Regards.
--
Rubén Martín (Nukeador)
Mozilla Reps Council member
http://mozilla-hispano.org
http://twitter.com/mozilla_hispano
http://facebook.com/mozillahispano

[Pascal Chevrel]

Le 09/05/2012 06:09, Ragavan Srinivasan a écrit :
> Hi Rubén and others,
>
> On Tuesday, May 8, 2012 4:36:35 PM UTC-4, Rubén Martín wrote:
>> I just want to add that I can totally understand that Linux support is
>> not ready yet, but let's take care of communication and explain why to
>> the future testers and/or link to the work in progress bug
>> <https://bugzilla.mozilla.org/show_bug.cgi?id=744193>. We don't want
>> people to feel undervalued just because they/we use Linux.
>
> I am the product manager for Apps and I wanted to address some of the concerns raised here.
>
> First of all, I'll admit that we could have done a much better job of communicating the level of Linux support at this stage of the Apps project. Though we did call it out on the Kilimanjaro planning page and during other engineering meetings, we will be sure to keep this in mind for future communications.
>
> Coming to the question at hand, lack of support for Linux at this stage is mainly due to two factors.
> 1. The technical feasibility of implementing support for Apps features in the various Linux distributions. and the resource and time constraints associated with doing so.

I would suggest starting by CCing to this bug our own Linux developers
and our distro packagers for Firefox/Thunderbird for at least Redhat,
Debian and Ubuntu

Cheers,

Pascal

[Percy Cabello]

On Wed, May 9, 2012 at 2:52 AM, Patrick Finch <pfi...@mozilla.com> wrote:

>
>
> Linux is a P3 for this initial release. Linux support will come. I'm
> disappointed too, but I find it very hard to disagree with the
> prioritisation: if you want to test at volume or to make the product
> credible in the market, Linux is unfortunately not a priority. It's 1-3%
> of the desktop market depending on the country.
>

I only hope the "lots of influencers use Linux" side of the argument has
been well weighted, as we are going to need as much traction as possible
here. At the very least, when going public maybe we can offer a set date on
when Linux support might be expected.

Patrick
>
>
Best,
Percy

> --
> Patrick Finch
> Mozilla
> pfi...@mozilla.com
> Mobile: +46 768 444 833
> IM: patric...@gmail.com
>

> ______________________________**_________________

> Interested in promoting Mozilla? Check out the Mozilla Community Marketing

> Guide: http://contribute.mozilla.org/**Marketing<http://contribute.mozilla.org/Marketing>
> mark...@lists.mozilla.org
> https://lists.mozilla.org/**listinfo/marketing<https://lists.mozilla.org/listinfo/marketing>
>

[chris hofmann]

On 5/8/12 9:09 PM, Ragavan Srinivasan wrote:
> I'd like to focus on #1 here. As you can see from the bug you link to above (and the etherpad linked to from the bug), this is a particularly hard technical problem. We don't have the right level of skills nor the time commitment required to do this justice. Which is why we are very excited about Marco's work as part of the Summer of Code project. If you or others have the skills and time to help move this forward, we'd love to get your help.
>
> I just read chofmann's reply and his suggestion of finding the right technical help to move this forward is spot on!
>
> Regards,
> Ragavan

There are two things that are needed as part of that plan. First is to
make as much of the code cross platform as possible, and to have a good
definition of the platform specific pieces. Maybe these should be
added as requirements for that feature or that section of the planning
doc should be broken down in that way. That would probably make it a
lot more clear for developers and testers on Windows Mac and Android as
well.

Having a good definition of the what the platform specific pieces look
like on Windows, Mac, and Android look like will allow linux hackers to
follow along and "translate" the requirements for that platform (same
for tablet and iOS as well).

Are there other pieces that are intended to be platform specific other
than these two that I listed?

P1: Launch from native standard launch points (start menu, dock,
start page, Home Screen etc.)
P1: Launch in chromeless window, outside the browser application

(bug 740586 <https://bugzilla.mozilla.org/show_bug.cgi?id=740586>)
(bug 743917 <https://bugzilla.mozilla.org/show_bug.cgi?id=743917>)

Can we get someone to go though the other bugs that are on file to make
sure that those are all going to be cross platform parts of the code?

https://bugzilla.mozilla.org/show_bug.cgi?id=740586 has been made into
an Android bug. Is the plan for this to also be supported on "Desktop
OSes"? If so we probably need to spin off other bugs for Widows, Mac,
and Linux, etc....

-chofmann

[Patrick Finch]

On 5/9/12 2:13 PM, Percy Cabello wrote:
> On Wed, May 9, 2012 at 2:52 AM, Patrick Finch<pfi...@mozilla.com> wrote:
>
>>
>>
>> Linux is a P3 for this initial release. Linux support will come. I'm
>> disappointed too, but I find it very hard to disagree with the
>> prioritisation: if you want to test at volume or to make the product
>> credible in the market, Linux is unfortunately not a priority. It's 1-3%
>> of the desktop market depending on the country.
>>
>
> I only hope the "lots of influencers use Linux" side of the argument has
> been well weighted, as we are going to need as much traction as possible
> here. At the very least, when going public maybe we can offer a set date on
> when Linux support might be expected.

I agree, the importance of Linux, and Linux users, is much greater than
the Linux market share.

Patrick

[Rubén Martín]

Also I forgot to mention, but this affects B2G directly too.

We have a lot of people interested in helping with B2G, and the B2G team
is trying to organize and work with the community of volunteers.

Guess which OS these people use? Right, Linux, and we already have some
complains and questions from frontend and backend devs around this and
what's going on.

[Majken Connor]

I just want to reiterate that traditionally it's been hard to get the
community contributions required to keep up. Maybe because people who are
passionate about linux are already using their spare time for linux? I can
say with some certainty that linux lagging behind has also frustrated the
people who ultimately make these decisions. Lack of marketshare hasn't been
an excuse to support apathy, though it can seem like that because the
people making that call have come to terms with it.

Is there any sort of initiative similar to the enterprise working group
where contributors can help coordinate efforts to help out with the
problems chofmann and others have identified? Maybe Linux dev would be a
good SIG for ReMo?

On Wed, May 9, 2012 at 4:11 PM, Rubén Martín
<nuke...@mozilla-hispano.org>wrote:

> _______________________________________________

> Interested in promoting Mozilla? Check out the Mozilla Community Marketing

> Guide: http://contribute.mozilla.org/Marketing
> mark...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/marketing
>

[Asa Dotzler]

On 5/8/2012 1:09 PM, Rubén Martín wrote:
> Hello,
>
> I know that sometimes some features are more complicated to implement in
> one OS than others, but if we want to keep the Linux users happy and
> avoid them to think their OS is a second class OS for mozilla, we should
> take care of these things.
>
> Mailing all mozillians about testing marketplace before having support
> under Linux is not good, most community members use Linux as their main
> OS (including myself) and most of them would love to test apps and help.
>
> Regards.

I think you overestimate the number of community members using Linux. At
the last summit we had, I did some informal polling and Linux was a tiny
minority of our active community represented there. Mac was most used,
followed by Windows, and then, in a distant third, Linux.

The proportion of Linux users within the Mozilla community is not, as I
see it, an issue here. What matters is what portion of our community is
capable and available to help make the Maketplace and Firefox apps
integration (and any other relevant parts of our Apps effort) work well
for Linux users.

What we need most, I suspect, is available Linux coders, people who know
Gnome, Unity, GTK, etc. to do the platform integration work. I don't
know who those people are. Looking around the sub-set of community
members employed by Mozilla who could help on this, I don't see any
available resources or even any resources I would move from their
current work to this work.

If there are even half as many Linux users in our community as you
suspect, then it shouldn't be hard to find some of them with the skills
to work on the Firefox App integration problems. Assuming that
Mozilla-paid Linux hackers are busy with higher priority items, then we
need to find some volunteers to help.

So who can help?

- A

[Santiago Hollmann]

El Wed May 9 19:15:02 2012, Asa Dotzler ha escrito:

>
> On 5/8/2012 1:09 PM, Rubén Martín wrote:
>>
>> Hello,
>>
>> I know that sometimes some features are more complicated to implement in
>> one OS than others, but if we want to keep the Linux users happy and
>> avoid them to think their OS is a second class OS for mozilla, we should
>> take care of these things.
>>
>> Mailing all mozillians about testing marketplace before having support
>> under Linux is not good, most community members use Linux as their main
>> OS (including myself) and most of them would love to test apps and help.
>>
>> Regards.
>
>
> I think you overestimate the number of community members using Linux.
> At the last summit we had, I did some informal polling and Linux was a
> tiny minority of our active community represented there. Mac was most
> used, followed by Windows, and then, in a distant third, Linux.
>

I think that poll doesn't reflect what is happening in Latin America at
least . Many of the contributors here use Linux and they also contribute
spreading free software projects. I think Rubén was talking with that on
mind.

At Mozilla many people use Mac, because of that we can see a high number
of mac users on Summit (where we have lots of employees). It would be a
very interesting poll to do to our community. We can start asking to our
reps.

> The proportion of Linux users within the Mozilla community is not, as
> I see it, an issue here. What matters is what portion of our community
> is capable and available to help make the Maketplace and Firefox apps
> integration (and any other relevant parts of our Apps effort) work
> well for Linux users.
>
> What we need most, I suspect, is available Linux coders, people who
> know Gnome, Unity, GTK, etc. to do the platform integration work. I
> don't know who those people are. Looking around the sub-set of
> community members employed by Mozilla who could help on this, I don't
> see any available resources or even any resources I would move from
> their current work to this work.
>
> If there are even half as many Linux users in our community as you
> suspect, then it shouldn't be hard to find some of them with the
> skills to work on the Firefox App integration problems. Assuming that
> Mozilla-paid Linux hackers are busy with higher priority items, then
> we need to find some volunteers to help.

The number of Linux users in the community don't necessarily determine
the number of Linux hackers.

If we need Linux hackers contributing on a project we just have to
mention that and every mozillian will do their best to find them. Also,
we can use a mozillian.org tag to identify them and then ask to them to
code on a project.

Definitely this is something we should discuss and see how we can fix it.

--
Santiago Hollmann

t: @santihollmann <http://twitter.com/santihollmann>
Piedra Libre! <http://piedralibre.wordpress.com>

[Rubén Martín]

El 10/05/12 00:15, Asa Dotzler escribió:

>
> I think you overestimate the number of community members using Linux.
> At the last summit we had, I did some informal polling and Linux was a
> tiny minority of our active community represented there. Mac was most
> used, followed by Windows, and then, in a distant third, Linux.

Summit is a meeting for 500 people, and our community a way bigger. Most
communities are creating around open source/linux enthusiasts, I can
speak about the Mozilla Hispano and European communities I know where
the vast majority uses linux.

> What we need most, I suspect, is available Linux coders, people who
> know Gnome, Unity, GTK, etc. to do the platform integration work. I
> don't know who those people are. Looking around the sub-set of
> community members employed by Mozilla who could help on this, I don't
> see any available resources or even any resources I would move from
> their current work to this work.
>
> If there are even half as many Linux users in our community as you
> suspect, then it shouldn't be hard to find some of them with the
> skills to work on the Firefox App integration problems. Assuming that
> Mozilla-paid Linux hackers are busy with higher priority items, then
> we need to find some volunteers to help.

As Santiago mentioned, "The number of Linux users in the community don't
necessarily determine the number of Linux hackers", we can ask for help
finding these hackers to the community.

Also I think that one of the great things about mozilla is that now we
have money to spend hiring people in areas we think are important and
where we need people working full-time to help volunteers.

The problem here is that some people don't consider that spreading money
or resources on Linux support it's good due it's market share and I
think that's not good. If we have the possibility to do whatever we can
to support an OS which share our values and that most people that
empowers mozilla use, we should do it, as we did in the past supporting
for example Theora and Xiph.org even its market share was tiny.

This seems to most volunteers as a Business-minded strategy and
completely opposite to our values. Linux is very important for mozilla
and for the volunteers in particular. How are we suppose to defend or
present a technology we can't run on our laptops?

If we lose linux supporters, we'll lose the passion and mozillians which
are behind, and that would have a big and negative impact to the whole
project. Can you imagine mozilla without all open source and linux
supporters? I can't because that wouldn't be mozilla as we know it anymore.

[Brendan Eich]

Thanks for bringing this up. I think it's an avoidable conflict, because
we have an owner for the bug to implement on Linux. The prioritization
may be ok for a staged release. The "nice-to-have" wording was regrettable.

Indeed the whole apps, marketplace and web runtime plan is too large to
do at one step, or even with platform parity at the first step. That
does not mean we give up our cross-platform commitments.

We support Linux as you say, because of our cross-platform principles
first, and because of lead users in the Linux community and among our
top Gecko hackers. There's a nexus: B2G is based on Linux and Gecko, but
of course without any Linux desktop (and without X-Windows. This is a
good thing!).

My two cents,

/be

Rubén Martín wrote:
> El 10/05/12 00:15, Asa Dotzler escribió:

>> I think you overestimate the number of community members using Linux.
>> At the last summit we had, I did some informal polling and Linux was a
>> tiny minority of our active community represented there. Mac was most
>> used, followed by Windows, and then, in a distant third, Linux.

> Summit is a meeting for 500 people, and our community a way bigger. Most
> communities are creating around open source/linux enthusiasts, I can
> speak about the Mozilla Hispano and European communities I know where
> the vast majority uses linux.

>> What we need most, I suspect, is available Linux coders, people who
>> know Gnome, Unity, GTK, etc. to do the platform integration work. I
>> don't know who those people are. Looking around the sub-set of
>> community members employed by Mozilla who could help on this, I don't
>> see any available resources or even any resources I would move from
>> their current work to this work.
>>
>> If there are even half as many Linux users in our community as you
>> suspect, then it shouldn't be hard to find some of them with the
>> skills to work on the Firefox App integration problems. Assuming that
>> Mozilla-paid Linux hackers are busy with higher priority items, then
>> we need to find some volunteers to help.

[jamesmi...@googlemail.com]

On Wednesday, May 9, 2012 2:12:00 AM UTC+2, Daniel Mills wrote:
> Linux support for apps is a nice to have because most of our users are not running Linux. I think we're supportive and absolutely willing to accept patches to make something work on Linux, but it's just not something that affects the 80% (I don't think it's even 10%, though I don't have any data handy). By definition, this is a nice to have, not a stop-ship feature.
>

> Remember that we are making software for a lot of people, and staff and community are actually a tiny slice of the userbase. I know it's hard, but we need to focus on the userbase at large, not on us.

I am willing to help out on the linux porting, please provide me with details on what I need to do.

mike

[jamesmi...@googlemail.com]

[Daniel Mills]

Excellent! There were some bugs linked to elsewhere on this thread - that's a start. Also, you might want to hang out on the #openwebapps irc channel and ask questions.

Dan

[James Michael Dupont]

This bug here seems to be actively worked on by Marco Castelluccio,
https://bugzilla.mozilla.org/show_bug.cgi?id=744193

please let me know if this topic still needs work and what needs to be done so that I dont conflict with marco.

mike

[James Michael Dupont]

[Mike Dupont]

Ok, I will do some testing. I have your patches here, and will put
them in my github branch.

To g...@github.com:h4ck3rm1k3/mozilla-central.git
83eb825..47a2926 master -> master

thanks
mike

On Sat, May 12, 2012 at 10:36 AM, <mar.cast...@studenti.unina.it> wrote:
> James Michael Dupont <jamesmi...@googlemail.com> ha scritto:

> Yes, I'm already working on Linux support as a Google summer of code
> project: https://wiki.mozilla.org/SummerOfCode/2012/LinuxNativeWebApps
>
> If you want, you could help me with testing my work (try builds will be
> available in the coming days).
>  As you may know there are a lot of differences between Linux desktop
> environments, so we need to test thoroughly.
>
> Thank you,
> Marco Castelluccio.
>
>



--
James Michael DuPont
Member of Free Libre Open Source Software Kosova http://flossk.org

[Mike Dupont]

I have build the latest version of Rhino but it cannot compile the code,

I defined the XP_UNIX and then preprocessed the code :
python ../../config/Preprocessor.py -DXP_LINUX=1 WebappsInstaller2.jsm
> WebappsInstaller2.js

Then ran the
java -jar build/rhino1_7R4pre/js.jar
~/experiments/mozilla/mozilla-central/browser/modules/WebappsInstaller2.js
js: "/home/mdupont/experiments/mozilla/mozilla-central/browser/modules/WebappsInstaller2.js",
line 276: missing { before function body
js: NetUtil.asyncCopy(istream, ostream, function(x) aCallback(x));
js: ...........................................................^
js: "/home/mdupont/experiments/mozilla/mozilla-central/browser/modules/WebappsInstaller2.js",
line 378: missing { before function body
js: iconSizes.sort(function(a, b) a - b);
js: .................................^
js: "/home/mdupont/experiments/mozilla/mozilla-central/browser/modules/WebappsInstaller2.js",
line 1: Compilation produced 2 syntax errors.

Here is one patch for the first error,
To g...@github.com:h4ck3rm1k3/mozilla-central.git
47a2926..e63017e master -> master

and the second was also the same, missing braces {}

mike

On Sat, May 12, 2012 at 10:40 AM, Mike Dupont
<jamesmi...@googlemail.com> wrote:
> Ok, I will do some testing. I have your patches here, and will put
> them in my github branch.
>
> To g...@github.com:h4ck3rm1k3/mozilla-central.git
>   83eb825..47a2926  master -> master
>
> thanks
> mike
>
> On Sat, May 12, 2012 at 10:36 AM,  <mar.cast...@studenti.unina.it> wrote:
>> James Michael Dupont <jamesmi...@googlemail.com> ha scritto:
>>

[Rubén Martín]

Do we have a wiki page with all the relevant information in one place?

I would be nice to know the current planning and status for linux
support, bugs that need attention...

This way we can ask our local communities to help with specific tasks
and speed up the process to be ready at the same time as other OS when
marketplace goes public.

[Nukeador]

I just want to point out that we are now on slashdot frontpage:

http://linux.slashdot.org/story/12/05/15/1316229/mozilla-leaves-out-linux-for-initial-web-app-support

Some of our comments also quoted on the source article.

I would like PR to chime in and tell us how to deal from this point.

Regards.

--
Rubén Martín [:Nukeador]
Mozilla Reps Council Member

[chris hofmann]

On 5/15/12 6:35 AM, Nukeador wrote:
> I just want to point out that we are now on slashdot frontpage:
>
> http://linux.slashdot.org/story/12/05/15/1316229/mozilla-leaves-out-linux-for-initial-web-app-support
>
> Some of our comments also quoted on the source article.
>
> I would like PR to chime in and tell us how to deal from this point.
>
> Regards.
>

If we follow though on my suggestions of getting a good spec on what
platform integration looks like in general, and what parts are cross
platform, and we get some owners to step up for linux part of the code
and keep up the progress on
https://bugzilla.mozilla.org/show_bug.cgi?id=744193 (seems like this is
has happened - thanks Marco Castelluccio and Mike Dupont and others)
there is really nothing to get worked up about anymore. I'd think we
just tell people "no story here, move along."

Work seems to be progressing. Nukeador, what do you think needs to be done?

-chofmann

[Marco]

Hello,
I'm working on the support for Linux, as already pointed out a cross-distribution implementation isn't so simple.
I don't know if we can ship the feature in Firefox 15, but I can provide try builds for testing purposes, so that the Linux community can test this new feature on Linux just like the Mac and Windows users are now doing.
Anyway, there will surely be support for Linux in time for the "Kilimanjaro Event".

You can follow the development here: https://wiki.mozilla.org/SummerOfCode/2012/LinuxNativeWebApps
I'll post links to the try builds on that wiki page as soon as possible.

Let me know if you have any question.

Thank you,
Marco Castelluccio.