Link from themes manager to personas gallery - Theme discrimination

209 views
Skip to first unread message

Aronnax

unread,
Nov 19, 2009, 6:31:33 PM11/19/09
to
Hi,
this is a follow up to Bug 522571 - Link from themes manager to
personas gallery
https://bugzilla.mozilla.org/show_bug.cgi?id=522571

The developer want the diccussion here.
Ok, here we go again ;-)

> It is intended to have a page that is primarily personas, as you said.

Hi,

Why, please tell me why ;-)

I think a lot of themer are interested to know, why their stuff should
be hidden and why this limited background image change system should be
highlighted. Why?

Is it then as well intended to hide the normal extensions, when Jetpack
is later a part of the default installation?

And when you are still on the way to tell some reasons.
Why works Personas on 3.6 by default only with the default themes. Is it
still intended to change it with 3.7 or later. Why works the Personas
1.4 extension still only on Mac OS X only with the default theme and why
is there absolutely no progress in the last months to improve the
situation.

By the way,
i know a lot of themer, who think Mozilla want to kill the old theme
system. Some words as well about it would be very nice. It is hard to
find an example that something else is intended. This theme page
primarily for Personas is then only one example more.

Cheers

Mike Beltzner

unread,
Nov 19, 2009, 8:30:55 PM11/19/09
to dev-apps...@lists.mozilla.org
Hi Aronnax,

I'll try to answer your questions in order; you asked many.

We believe that Personas are a better user experience than XPI Themes.
While not as robust or powerful, they are quick and easy for users to
install, can be activated without requiring a restart, and are simple to
create. Although not very old, the Personas Add-on has been downloaded
by more than 10M people and is very popular.

For that reason, we will be promoting Personas as a new, different type
of theme that requires less commitment to install and use. They will be
displayed on the first run page for Firefox 3.6 users, and the link in
the Themes panel will go to the Personas gallery.

Personas are still themes, though, just a different type. We hope to add
more capability (ability to change the appearance of buttons, work with
other XPI themes, etc) over time. XPI themes will eventually be
installed like extensions, which better represents their capability and
function; an XPI theme can totally change the "face" of the browser, not
just the skin.

We do not have any plans to remove support for XPI themes. We expect
that most users will prefer to use Persona-style themes and will be
investing in making those the preferred mechanism for quickly changing
the appearance of the browser. However it would be foolish and against
our values to remove the capability to customize the full appearance of
the browser.

I hope this helps,
mike

> _______________________________________________
> dev-apps-firefox mailing list
> dev-apps...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-apps-firefox

Pardal Freudenthal

unread,
Nov 19, 2009, 11:43:11 PM11/19/09
to
Hi Mike,

thank you very much for answering here and clarifying things. I'm sure
the theme developers community, like me, very appreciate to know that
third party themes or "old-style" themes will not be deprecated.

I think and have stated that Personas is a nice feature and the way it
works now (on the user's side) is almost perfect (technically it has
some problems to be solved). Easy to install, to manage, to try and
see the results "on the fly". Users have the opportunity to choose
which persona fits well with the theme they are using making the
customization an easy and nice experience. Many authors are making
efforts to support Personas and the reviews of my theme pointing this
support are very positive.

My concerns are about how the integration on Firefox is going to be
implemented.

I can see two constellations depending if Bug 520124 - Enable to
activate a 'heavy weight theme' together with 'lwtheme':
https://bugzilla.mozilla.org/show_bug.cgi?id=520124 will be fixed or
not:

1) Personas will work with third-party themes:
In this case, IMHO, it would be much better to have it where it is
now: on the menu. Independently of which theme are in use at the
moment, for the user is a lot easier to just click on the menu and see
immediately which effect a persona will have on the browser. The
persona's change on mouse over in the sub-menus is simply amazing! And
without needing to visit a site to get it..
The menu is also a lot more "discoverable" than the Addons Manager (I
know a couple people that has never opened the Addons Manager, but for
sure have opened the menus...).
In this constellation, although, I can't see how Personas would
implement buttons changes. This would not work with third party themes
at all...

2) Personas will only work with the default theme:
In this case I can imagine it as a feature from the default theme,
making sense to be listed at the Addons Manager as sub-items from the
default theme (see https://bugzilla.mozilla.org/show_bug.cgi?id=529729).
In this constellation I can imagine that Personas could very well
offer the ability to change buttons. In this case, the Personas
extension could be further offered to be used with third-party themes.

I personally prefer the first option...

Actually I don't agree that Personas is a better user experience than
XPI Themes. For me it's more a different user experience (with the
same target), a kind of complement customization. The user choose the
preferred theme and "change" its background.
So, I can't see why compare one against another. They can live
together without (too much) problems just like they do now. This
document: https://support.mozilla.com/en-US/kb/Using+themes+with+Firefox
explains very well the differences from Personas and "full themes"
without any discrimination or treating Personas as a better user
experience as themes...

Sorry, but I think that, in the way it is being implemented so far, it
is more difficult to use both, Personas and third party themes.
Just my two cents.

Aronnax

unread,
Nov 20, 2009, 12:53:30 PM11/20/09
to
Hi,
thanks but i think it it helps only to start a new discussion.

First of all:
Who is We? Who decide it? Who had talked for the theme system?

I can�t remember only one Mozilla developer, who ever talks about his
theme experiences and i read now many years Mozilla blogs, discussions,
wikis and so on - not one. I have my doubt that many even care about
theme stuff and the few who cares works then probably only for the
default theme(s) ;-)
Another example: The first edition of the last redesign from
addons.mozilla.org had not even a theme listing. They removed the point
and added themes below Appearance together with tab stuff, new toolbars
and everything else. The themes were completely hidden and obviously
thought nobody that this is a problem.
A Mozilla guy for example, who has full time job to advertise Firefox,
compares theme systems from different browser talks then only about
Personas - can�t be true
http://weblogs.mozillazine.org/asa/archives/2009/08/29_vs_7000.html

I don�t think that all these We, who obviously don�t care about themes
stuff should alone decide, what happens with the theme system, behind
closed doors somewhere in Mountain View California.

----

Who cares about themes?
We have many many themes with 2 millions, 3 millions � one (very old)
with 23 millions or the current most popular with 10 millions downloads ...
Personas on the other hand has after some years and a lot of advertising
in the last mounts only 10 millions. Just for the record: 10 millions
are not a high number, when we consider the hype behind it and the
themes has nobody who hype them. They are actually more or less hidden.
I think a lot of user still care , what happens with themes.

----

Mike Conner talks the first time many months ago about these themes
system changes leaves many questions open. Your post answer not really
these questions and added actually a few more. For example: How should
it works, when Personas change the buttons and a theme do the same. It
is exactly the reason, why Personas on Mac OS X works only with the
default theme. Ok, stuff like this can be fixed, but who works now on a
concept to prevent bugs like this. Likely nobody. I think nobody cares -
only Personas counts now ;-)

-----

XPI themes will eventually be installed like extensions.
- It is then only logical to remove these extension-themes from the
theme selection of the Add-ons manager.
- They will be hidden somewhere in addons.mozilla.org. and many user
will not even know that these themes exist.
- Under this conditions will be many themes aborted � the theme
community will slowly die .. the documentation will die .. bugs will be,
sooner or later, fixed without a consideration of theme needs.

Without a doubt would it be the end of the current themes.

-----

"We believe that Personas are a better user experience than XPI Themes."

Without a restart is nice ;-) - naturally - And? Is this all?

Personas can do some things better and themes can do other things much
better.
This will be never changed. Personas can have many new user and themes
have already much more � i still don't understand why themes should be
hidden.

-----

I think Mozilla treated currently themer like stupid fogs in a cooking
pot .. and slowly increase the temperature .. everyone knows what will
happen ;-)
Suggestions like the XPI extensions-themes are for example maybe
something for frogs, but surely nothing for the themes community.

Cheers

J M R

unread,
Nov 20, 2009, 2:42:06 PM11/20/09
to
Mike Beltzner wrote:

>
> Personas are still themes, though, just a different type.

That's just rubbish. They're nothing alike. Themes are whole house
decorations and remodeling and Personas are just window blinds.

Personas all in all is just a vanity project. Bragging about millions of
downloads but not about how many of them are really in use.

Personas images can be previewed by just hovering the mouse over them.
Does this count as a download?

Do users continue to use the images or they revert back?

How frequently do they replace their images?

90% of Personas images are an usability nightmare, do you have seen a
rise in usability problems?

How will Personas images fit with the new interface that has a reduced
surface in v3.6 and then further reduced in v3.7? How will Personas
images be justified if users can hardly see them?

This is how i used them: in the personas web site I experimented with 20
to 30 of them, concluded that most of them are just rubbish and the
other make using Firefox more difficult I reverted back. This alone
counted as 20 to 30 downloads.

I understand the appeal. It's easy and faster to just to loose 30
minutes of time with that but, it's just not that rewarding. It isn't.
It's a fad.

Dave Townsend

unread,
Nov 20, 2009, 3:19:58 PM11/20/09
to
On 2009-11-20 11:42:06 -0800, J M R <j...@jmr.eu> said:

> Mike Beltzner wrote:
>
>>
>> Personas are still themes, though, just a different type.
>
> That's just rubbish. They're nothing alike. Themes are whole house
> decorations and remodeling and Personas are just window blinds.

They both do basically the same thing, they change how Firefox looks.
Whether you want a low impact aesthetic change or a complete reworking
of the UI is a personal choice. The fact that xpi style themes can in
fact do a little as a persona means that they are lumped in the same
boat in my head.

> Personas all in all is just a vanity project. Bragging about millions of
> downloads but not about how many of them are really in use.
>
> Personas images can be previewed by just hovering the mouse over them.
> Does this count as a download?

No. Personas do update checks the same as regular themes do so we will
be able to get good stats on how many are in use, not just how many
have been downloaded or previewed.

> 90% of Personas images are an usability nightmare, do you have seen a
> rise in usability problems?

I've certainly seen more bug reports filed due to issues with xpi-style
themes than with personas, but that is only to be expected since the
xpi-style themes are such high impact it is easier to cause problems
with them.

> How will Personas images fit with the new interface that has a reduced
> surface in v3.6 and then further reduced in v3.7? How will Personas
> images be justified if users can hardly see them?

An excellent question, but I imagine personas will evolve to support
small areas pretty easily, they are after all just images. As for
justification, well we will have the usage stats to indicate if users
are no longer finding the feature compelling and can use that as a
basis for deciding whether maintaining support is important or not.

> This is how i used them: in the personas web site I experimented with 20
> to 30 of them, concluded that most of them are just rubbish and the
> other make using Firefox more difficult I reverted back. This alone
> counted as 20 to 30 downloads.
>
> I understand the appeal. It's easy and faster to just to loose 30
> minutes of time with that but, it's just not that rewarding. It isn't.
> It's a fad.

This is of course all subjective and has no real bearing on this
discussion I think so I won't bother to respond to it.

DonGato

unread,
Nov 20, 2009, 6:37:46 PM11/20/09
to
I find funny how people say something is subjective while what they
say supposedly isn't. :P
Firefox is becoming bloated, slow and buggy; things that made me move
from IE to Firefox on the first place.
Instead of fixing bugs, making it more stable and improving
performance and resource usage the developers are wasting time in
vanity things like Personas or making it a clone of Chrome.

A lot of theme bugs are because Firefox doesn't work properly with
themes. There are many bug reports with 2 years of existence and still
no fix. If you want themes to work 100% fix those.

Shawn Wilsher

unread,
Nov 20, 2009, 6:55:38 PM11/20/09
to dev-apps...@lists.mozilla.org
On 11/20/09 3:37 PM, DonGato wrote:
> Firefox is becoming bloated, slow and buggy; things that made me move
> from IE to Firefox on the first place.
What's slow? Have you filed bugs about it? Bloat is, of course,
completely subjective...

> Instead of fixing bugs, making it more stable and improving
> performance and resource usage the developers are wasting time in
> vanity things like Personas or making it a clone of Chrome.

Right, we haven't fixed a single bug and are only adding features and
making it a clone of Chrome...

The reality, of course, is that we have fixed bugs, and we have improved
performance. http://graphs.mozilla.org/dashboard/snapshot/ shows that
we are improving in terms of performance in general. We've also fixed
nearly 5000 bugs since Firefox 3.5. None of this is subjective.

/sdwilsh

Ed Hume

unread,
Nov 20, 2009, 7:50:15 PM11/20/09
to
> > dev-apps-fire...@lists.mozilla.org
> >https://lists.mozilla.org/listinfo/dev-apps-firefox
>
>

OK. I see the reference to "XPI themes." It's been a long time since
XPI packaging was allowed for themes. Did you have something else in
mind?

Despite your lip service to third-party themes, I can see what you
have in mind for us. I just have one request: please make sure that
visually-impaired users can get big icons, and that your default theme
allows for extra-large text. Right now you are optimized for guys who
prefer tiny text and icons.

Ed

Jeff.tet

unread,
Nov 20, 2009, 8:26:17 PM11/20/09
to
I just want to make note of a couple of things:

10 million downloads. If I had the resources or the ability to
advertise my theme on addons.mozilla.org start page my theme would
also have a significant download number. This brings me to one of
Arronax's comments that making it more difficult to view a variety of
heavy weight themes they will slowly die. I and probably most people
don't want to spend allot of time "searching" for a theme vs clicking
on a link and being brought to a page that lists themes that you can
scroll through. I don't know about everyone else but that along to me
is the beginning of the end of the use of heavy weight themes.

If Mozilla wants to promote Personas why not just add them to the
current themes section if they are just themes like heavy weight
themes?

How does one get involved with making an intelligent opinion about a
significant change such as the one we are discussing here?

I am also going to ask Why? I don't want a politically correct answer
or a comment that defends the current ideals instead of truly
answering the question.

Please re-think and consider that by including Personas with heavy
weight themes together is giving users more choices for their Firefox
Experience and not will say "masking" a choice to give the illusion
that we're starting to phase out Themes as users currently know them.

Remember its about progress not regression!

Thank
Jeff

Chris Ilias

unread,
Nov 20, 2009, 9:38:31 PM11/20/09
to
On 09-11-20 6:37 PM, DonGato wrote:
> I find funny how people say something is subjective while what they
> say supposedly isn't. :P
> Firefox is becoming bloated, slow and buggy; things that made me move
> from IE to Firefox on the first place.
> Instead of fixing bugs, making it more stable and improving
> performance and resource usage the developers are wasting time in
> vanity things like Personas or making it a clone of Chrome.

Hi DonGato,
Have you tried Firefox 3.6? -
http://www.mozilla.com/en-US/firefox/all-beta.html

Pardal Freudenthal

unread,
Nov 20, 2009, 11:08:33 PM11/20/09
to
> If Mozilla wants to promote Personas why not just add them to the
> current themes section if they are just themes like heavy weight
> themes?
>

Can you imagine how the user reviews of Personas would look like? :-)

Looking at how things are being done, I understand this: "We expect
that most users will prefer to use Persona-style themes..."
like: "We are doing our best to integrate Personas in Firefox in a way
that we expect
that most users will prefer to use Persona-style themes..."

Dave Townsend

unread,
Nov 21, 2009, 1:29:12 PM11/21/09
to
On 2009-11-20 16:50:15 -0800, Ed Hume <edh...@gmail.com> said:
> Despite your lip service to third-party themes, I can see what you
> have in mind for us. I just have one request: please make sure that
> visually-impaired users can get big icons, and that your default theme
> allows for extra-large text. Right now you are optimized for guys who
> prefer tiny text and icons.
>
> Ed

I believe the default theme's text changes size based on the OS font
sizing, if not we should certainly fix that. I guess the icons probably
don't right now but we should probably try to do that too.

Mike Connor

unread,
Nov 22, 2009, 7:52:16 PM11/22/09
to Aronnax, dev-apps...@lists.mozilla.org

On 20-Nov-09, at 9:53 AM, Aronnax wrote:

> Hi,
> thanks but i think it it helps only to start a new discussion.
>
> First of all:
> Who is We? Who decide it? Who had talked for the theme system?

This was proposed a _long_ time ago, and I blogged about the ideas and
linked to the wiki doc you mention. Feedback was minimal.

> Mike Conner talks the first time many months ago about these themes
> system changes leaves many questions open. Your post answer not
> really these questions and added actually a few more. For example:
> How should it works, when Personas change the buttons and a theme do
> the same. It is exactly the reason, why Personas on Mac OS X works
> only with the default theme. Ok, stuff like this can be fixed, but
> who works now on a concept to prevent bugs like this. Likely nobody.
> I think nobody cares - only Personas counts now ;-)

Ultimately, the reason why we don't have <choose theme> + <choose
Persona> is that the UI starts getting harder to understand, and it's
largely a technical distinction. It also is harder to ensure a good
experience for users (many themes were poorly implemented).

If you had questions, did you raise them then?

> XPI themes will eventually be installed like extensions.
> - It is then only logical to remove these extension-themes from the
> theme selection of the Add-ons manager.
> - They will be hidden somewhere in addons.mozilla.org. and many user
> will not even know that these themes exist.

> - Under this conditions will be many themes aborted … the theme

> community will slowly die .. the documentation will die .. bugs will
> be, sooner or later, fixed without a consideration of theme needs.
>
> Without a doubt would it be the end of the current themes.

I think that we erred greatly in treating themes as somehow different
from extensions. They're just a type of customization, and I do not
believe that they are more or less interesting than any other type of
customization that users will do with current-style addons, except
that they're harder to build and maintain, in many ways.

> "We believe that Personas are a better user experience than XPI
> Themes."
> Without a restart is nice ;-) - naturally - And? Is this all?

Easy install/update, easy to create and tweak, very very sandboxed (no
chrome access, no trust decision required).

> Personas can do some things better and themes can do other things
> much better.
> This will be never changed. Personas can have many new user and

> themes have already much more … i still don't understand why themes
> should be hidden.

To turn it around, why should current themes get primacy over another
type of customization like tab addons, or toolbars?

-- Mike

Mike Connor

unread,
Nov 22, 2009, 7:35:31 PM11/22/09
to Ed Hume, dev-apps...@lists.mozilla.org

On 20-Nov-09, at 4:50 PM, Ed Hume wrote:

> OK. I see the reference to "XPI themes." It's been a long time since
> XPI packaging was allowed for themes. Did you have something else in
> mind?

Nope. The idea of "XPI themes" was shorthand for taking the
limitations off what a "theme" can do and have them simply be a type
of extension, with all of the same privs for authors. Rather than
simply styling the elements (alternate CSS+images) we would enable
theme authors to do more radical changes, and even provide alternate
widgets (i.e. separate back/forward buttons, or even different UIs).
It's really more "stop treating extensions and themes as separate
things in the UI" as a concept.

-- Mike

Mike Connor

unread,
Nov 22, 2009, 7:23:39 PM11/22/09
to Dave Townsend, dev-apps...@lists.mozilla.org

There's a bug, somewhere, on solving this better as higher-dpi screens
start coming out. One shouldn't need to install a third-party addon
to get sanely-scaled icons.

We didn't blow up with Extra Large fonts on Windows last I mucked with
themes, but that's a good 18 months ago. If we don't now, a bug
should be filed.

-- Mike

Aronnax

unread,
Nov 23, 2009, 5:49:36 AM11/23/09
to
Hi,

Am 23.11.09 01:52, schrieb Mike Connor:


>
> On 20-Nov-09, at 9:53 AM, Aronnax wrote:
>
>> Hi,
>> thanks but i think it it helps only to start a new discussion.
>>
>> First of all:
>> Who is We? Who decide it? Who had talked for the theme system?
>
> This was proposed a _long_ time ago, and I blogged about the ideas and
> linked to the wiki doc you mention. Feedback was minimal.

Some themer asked therefore the same questions (i as well) some months
ago, if it is intended to replace themes and the answer was no.
From a themer point of view was this minimal feedback enough. And we
have still no detailed plans, how themes and Personas should work in the
future together, or?

>
>> Mike Conner talks the first time many months ago about these themes
>> system changes leaves many questions open. Your post answer not really
>> these questions and added actually a few more. For example: How should
>> it works, when Personas change the buttons and a theme do the same. It
>> is exactly the reason, why Personas on Mac OS X works only with the
>> default theme. Ok, stuff like this can be fixed, but who works now on
>> a concept to prevent bugs like this. Likely nobody. I think nobody
>> cares - only Personas counts now ;-)
>
> Ultimately, the reason why we don't have <choose theme> + <choose
> Persona> is that the UI starts getting harder to understand, and it's
> largely a technical distinction. It also is harder to ensure a good
> experience for users (many themes were poorly implemented).

I read for example some blogs from Personas guys and they talked in
particular about time problems to get everything ready for 3.6.
Therefore only now with the default themes and later will it be maybe ..
probable .. who knows .. expanded. What should someone say against
(about) these time arguments? Not easy ;-)


>
> If you had questions, did you raise them then?

I talked with Myk M. about special Mac Personas problems and asked D�o
G. from time to time, how he build it for 3.6 (He build it in a kind of
way, that every theme could use it - now, or later)
When i read then on the other hand only about time problems .. What
should i asked more? I thought there are now only other questions for
the time after 3.6.

>
>> XPI themes will eventually be installed like extensions.
>> - It is then only logical to remove these extension-themes from the
>> theme selection of the Add-ons manager.
>> - They will be hidden somewhere in addons.mozilla.org. and many user
>> will not even know that these themes exist.

>> - Under this conditions will be many themes aborted � the theme


>> community will slowly die .. the documentation will die .. bugs will
>> be, sooner or later, fixed without a consideration of theme needs.
>>
>> Without a doubt would it be the end of the current themes.
>
> I think that we erred greatly in treating themes as somehow different
> from extensions. They're just a type of customization, and I do not
> believe that they are more or less interesting than any other type of
> customization that users will do with current-style addons, except that
> they're harder to build and maintain, in many ways.

Yes, correct.
But there is one difference.
Themes have a very weak lobby and will be from time time forgotten.

>
>> "We believe that Personas are a better user experience than XPI Themes."
>> Without a restart is nice ;-) - naturally - And? Is this all?
>
> Easy install/update, easy to create and tweak, very very sandboxed (no
> chrome access, no trust decision required).

Yes, as well correct ... and you had for example forgotten to mention
that many many have a very very bad usability .. and they can not
replace the other themes, when someone want more ;-)

>
>> Personas can do some things better and themes can do other things much
>> better.
>> This will be never changed. Personas can have many new user and themes

>> have already much more � i still don't understand why themes should be


>> hidden.
>
> To turn it around, why should current themes get primacy over another
> type of customization like tab addons, or toolbars?
>

Themes do not need or earn a primacy.
A fair treatment is enough. For example have themes not even anymore an
own name and must share it now with these Personas background images.

By the way,
what is now the official name of these themes?
For example: Old rusty heavyweight XPI extension themes, or what? ;-)


> -- Mike
>

Aronnax

Aronnax

unread,
Nov 23, 2009, 8:01:19 AM11/23/09
to
Hi,

We have for example alredy some theme-extentions
Brand Thunder https://addons.mozilla.org/en-US/firefox/user/37347

Even my Mac OS X only themes have higher download numbers ;-)
They are hidden somewhere in Appearance Add-ons + Toolbars Add-ons
and this is probably one of the reasons.
When you want to kill themes, when treat all the other themes in the
future like these theme-extentions ;-)

On the other hand have we more and more themes, which includes
additionally an extension.
For example: https://addons.mozilla.org/en-US/firefox/addon/10674
I think many themer would be happy with possibilities, but not when they
are then hidden in nowhere of the Appearance selection ;-)

Cheers

>
> -- Mike

Pardal Freudenthal

unread,
Dec 17, 2009, 1:43:51 PM12/17/09
to
> > Nope. The idea of "XPI themes" was shorthand for taking the limitations
> > off what a "theme" can do and have them simply be a type of extension,
> > with all of the same privs for authors. Rather than simply styling the
> > elements (alternate CSS+images) we would enable theme authors to do more
> > radical changes, and even provide alternate widgets (i.e. separate
> > back/forward buttons, or even different UIs). It's really more "stop
> > treating extensions and themes as separate things in the UI" as a concept.

Ohh. It sounds wonderful, but NO THANKS.
As Aronnax! has pointed, we have already such possibilities in those
two ways: the "extension-themes" (and that example "ruins" another
themes with their "wonderful skinned" toolbars; and looking for the
number of "themes" they produce (82?), one can imagine the propo$e$
from them...) and the bundles (that also need some guidelines to avoid
missing up other themes). So, we have already the possibilities (with
the bundle) you are talking about.

The great advantage from themes against extensions is the way they
register the skin providers and how they are selected. The skin
providers brought from themes substitute the originals. This is a
great feature, since if the user has installed a crap theme, he/she
must only to select another theme and will have no annoying side
effects.

Styles brought from extensions that are made only looking for the
default theme will still living there even if the user choose a third-
party theme. This could be simply avoided, we have enough
documentation and guidelines for it. I can say I spend 70% from my
work on my themes repairing things that extensions break that should
never get broke...

Again, we have already a great theme system, that really makes the
difference comparing to the boring models from the others. Why kill
it?

And what about the link to the real themes on personas home page? It
still not existing there.

Cheers

Mike Beltzner

unread,
Dec 17, 2009, 2:50:47 PM12/17/09
to Pardal Freudenthal, dev-apps...@lists.mozilla.org
----- "Pardal Freudenthal" <par...@gmx.de> wrote:
> Again, we have already a great theme system, that really makes the
> difference comparing to the boring models from the others. Why kill
> it?

Nobody has ever stated that we are "killing" anything. Also, the uptake and ease of development, application and deployment of Personas fix real problems with the existing theme system. This is win-win.

I'm sorry you feel differently, but don't think there's much point to discussing this further.

cheers,
mike

Aronnax

unread,
Dec 17, 2009, 3:39:45 PM12/17/09
to

>
> ... deployment of Personas fix real problems with the existing theme system. This is win-win.

Currently have some Personas user as well some difficulties to see this
win-win ;-)
https://addons.mozilla.org/en-US/firefox/reviews/display/10900

Or they have maybe as well another opinion about a real win-win ;-)

Cheers

CatThief

unread,
Dec 19, 2009, 8:57:57 PM12/19/09
to
Pardal Freudenthal wrote the following on 12-17-2009 1:43 PM:

> Again, we have already a great theme system, that really makes the
> difference comparing to the boring models from the others. Why kill
> it?

I'm with you, Pardal.

--
Regards,
CatThief

To reply privately, please PM me at MozillaZine...
http://forums.mozillazine.org/ucp.php?i=pm&mode=compose&u=25774

Alfred Kayser

unread,
Jan 4, 2010, 10:05:36 AM1/4/10
to Mike Beltzner, Pardal Freudenthal, dev-apps...@lists.mozilla.org
I am sorry, but I don't see this as a 'win-win'.
A win-win situation is where both Personas and 'xpi' themes can live
together.
But currently, FF3.6 with and without Personas 1.5 extension:
1. doesn't allow a theme together with a personas. It even switches back
from a selected theme to the default theme.
2. doesn't point to the addons/theme site.

Especially 1 is an active method of disabling/removing existing theme
usage: or 'killing' it.
Killing is never a 'win-win'...

I don't understand why the mozilla organization thinks that
'heavy-weight' themes are such a trouble.
They are 'sandboxed' and cannot do the harm that normal extensions do.
Allowing (or even forcing) the existing themes to become an XPI
extension, just paves the way for new issues...

Introducing Personas as a type of theme is fine (even with the extra
competition),
but locking out the old type of themes in the way as described above,
is considered 'unfair' in terms of competition (see the IE case in the
European open market issue).

Shawn Wilsher

unread,
Jan 4, 2010, 10:59:59 AM1/4/10
to dev-apps...@lists.mozilla.org
On 1/4/2010 7:05 AM, Alfred Kayser wrote:
> I don't understand why the mozilla organization thinks that
> 'heavy-weight' themes are such a trouble.
> They are 'sandboxed' and cannot do the harm that normal extensions do.
This is hardly true. It's trivial for a theme to do extension-like
things. It wasn't the original intent, sure, but it is certainly possible.

Cheers,

Shawn

Pardal Freudenthal

unread,
Jan 4, 2010, 11:38:07 AM1/4/10
to

Ok! I agree with you that the idea to "kill" the theme system is quite
absurd.
Such a decision could only be done as a long time target, with a lot
of discussions. Killing the theme system now, would mean that this
decision was made a long time ago.
So, which kind of people would encourage a lot of developers to work
without any payment sacrificing their free time (and believe me,
making themes demands a looooot of work and time), having in mind to
"kill" this work after a while? What would these people say to these
developers now? Some of these developers are working on this for more
than 5-6 years now!
No, this is impossible.

So, I suppose the document from Mike Connor:
https://wiki.mozilla.org/User:Mconnor/PersonasUplift#Deprecation_of_old-style_themes
contains a typo. "Deprecation of old-style themes" must be read as
"*Depreciation* of old-style themes".
This is exactly what is happening now. Themes are going to be
"depreciated", pushed to the status of second-class citizens.
Maybe having seen the issue in this way, it serves as a basis for
further discussions?

Themers have already a lot of issues to deal with. like changes on XUL
that kill backwards compatibility, lack of classes and ID's that would
be helpful for other UI approaches than that from default theme, style
rules living on content, hard codded images, inline styles, poor
written stylesheets brought from extensions and so on... With this
"depreciation", the chances for solving these issues are minimal.
(Maybe are exactly these issues the reason for "depreciating" third
party themes?)

I don't think that theme developers deserve "depreciation". They
deserve just a little respect...

Boris Zbarsky

unread,
Jan 4, 2010, 11:46:14 AM1/4/10
to
On 1/4/10 10:05 AM, Alfred Kayser wrote:
> They are 'sandboxed' and cannot do the harm that normal extensions do.

Would that it were so. It's not. The false perception that this is the
case is a huge problem from a security standpoint. The obvious solution
to that is to make reality match expectations, by either changing
reality (a lot of work in core code, breaks some nontrivial fraction of
themes, makes themes less useful than they are now) or changing
expectations (e.g. eliminating the distinction between XPI themes and
XPI extensions)...

-Boris

Mike Connor

unread,
Jan 4, 2010, 12:31:38 PM1/4/10
to Boris Zbarsky, dev-apps...@lists.mozilla.org

We've changed the install flow so it should be the same for XPI installs
in general. I don't know if there's still an education problem, but I
think the reality is that XPI themes shouldn't be sandboxed at all.

-- Mike

Alfred Kayser

unread,
Jan 4, 2010, 4:53:25 PM1/4/10
to Pardal Freudenthal
On 4-1-2010 17:38, Pardal Freudenthal wrote:
> I don't think that theme developers deserve "depreciation". They
> deserve just a little respect...
Indeed, so if we are indeed talking about depreciation here,
it would have been nicer to introduce this to the 'old' theme authors
in a much more nicer way. At least by telling why this happing,
the timing of this, and by supporting these authors in converting and/or
migrating and/or bailing out of this...


Philip Chee

unread,
Jan 5, 2010, 12:09:34 AM1/5/10
to

Also please note that "Heavy Weight" XPI themes are *not*
deprecated/depreciated in SeaMonkey. I very much hope that when you
disable these themes in Firefox you don't remove the underlying
toolkit/gecko functionality that other applications expect.

A bit of consideration for other consumers will be gratefully appreciated.

Phil

--
Philip Chee <phi...@aleytys.pc.my>, <phili...@gmail.com>
http://flashblock.mozdev.org/ http://xsidebar.mozdev.org
Guard us from the she-wolf and the wolf, and guard us from the thief,
oh Night, and so be good for us to pass.

Mike Beltzner

unread,
Jan 5, 2010, 12:39:56 AM1/5/10
to Philip Chee, dev-apps...@lists.mozilla.org
On 2010-01-05, at 12:09 AM, Philip Chee wrote:

> Also please note that "Heavy Weight" XPI themes are *not*
> deprecated/depreciated in SeaMonkey. I very much hope that when you
> disable these themes in Firefox you don't remove the underlying
> toolkit/gecko functionality that other applications expect.

I don't know how many times I have to say this, but this will definitely be my last contribution to this thread which shows little evidence of people actually listening, but lots of people desperate to speak "truth to power."

There are presently no plans to delete, dilute, deprecate, disable, destroy or any other d-verb "heavy weight" themes. There are also no plans to invest deeply in developing that technology, but (and this is important) there weren't every such plans in the works in the years that I've been involved with the project, really.

There are plans to move the primary UI mechanisms for theming towards "Personas" style lightweight themes which can be designed, developed, and delivered more easily to users. That's all that's happened. We think that they provide a better user experience for the majority of users, and so we've added them in 3.6, and are trying to figure out how to enhance them (add styling to buttons) so that they displace (not *replace*) Themes in the primary customization UI in a future release. Beyond that we've made no firm decisions.

We may, at some point, decide that we need to stop supporting classic themes. That decision would be made based on the costs vs. benefits, but we're not there yet.

> A bit of consideration for other consumers will be gratefully appreciated.

I don't mean to pick on Philip, but as someone who has spent literally years doing my absolute best to gather input, weigh costs and benefits, and guide the project in ways that are most beneficial for not only the Firefox product but the overall community, I find the implication in this comment to be hurtful.

I think everyone needs to be much more careful in the way we discuss these sorts of decisions. These are not arguments that need to be "won" or "fought." They are decisions that need to be debated on their merits and with respect.

cheers,
mike

Aronnax

unread,
Jan 5, 2010, 6:31:10 AM1/5/10
to
Hi,
thanks

We have still the question, if a theme should work (again) together with
a personas in future versions of Firefox.


Cheers

Pardal Freudenthal

unread,
Jan 6, 2010, 11:22:43 AM1/6/10
to
Please, we are not discussing about what was planned or not, what you
said or not. I don't care about words, neither d-verbs nor l-verbs. I
believe on attitude. The facts are clear enough and we provide a lot
of examples from what we are concerned about.

To make things easier I will try to list the complains:

1. WE DON'T WANT Personas being treated as "themes". They are not!
Personas are "skins". Don't suppose our users are not "smart" enough
to understand the difference between them. Of course they are smart!
They use Firefox ;-)

2. WE DON'T WANT a "Get Themes" link pointing to the "Get Personas"
page. This is quite offensive (not only for us, theme developers, but
also for our users). For more than two months ago we talked about this
on Bugzilla and the talk was about "primarily" Personas page but with
a link to the themes section. Until now nothing happened... And this
is also not really something that would satisfy us....

3. WE DON'T WANT Personas listed as Themes on Add-ons Manager. This is
confusing and will bloat the UI. 1. and 2. explain why.

4. WE WANT that the new feature, to change skins (personas) at real
time, also works with third party themes. (Could you imagine if
Microsoft would allow the "feature" Glass only for Microsoft
Products?)


For kind of feedback on the above points, look at the few reviews to
Personas from people using the 3.6 betas. Also this comment on MZ is
very enlightening: http://forums.mozillazine.org/viewtopic.php?p=8385025#p8385025


5. WE DON'T WANT Themes working as Extensions. The benefits of the
actual system, concerning chrome registration, how themes substitute
the skin providers for the main packages and how the
general.skins.selectedSkin key works are enormous.

6. WE WANT that themes will be listed like they used to at AMO, not at
the end of the list. (Interesting that everything there is listed on
alphabetically order but Personas...)

So, please treat these complains seriously and with the respect they
deserve. Please explain us why these decisions, that obviously
displease the whole theme community, were taken.

Mike Beltzner

unread,
Jan 6, 2010, 11:48:28 AM1/6/10
to Pardal Freudenthal, dev-apps...@lists.mozilla.org
I believe I have actually spoken to all of these concerns previously in the thread. The core answer, though, is that while we are *not* removing the functional capabilities that enable "heavy" themes, we are definitely migrating the user interface towards promoting "lightweight" themes, starting with Firefox 3.6.

I would like to move towards a point where lightweight themes can work with heavy themes, allowing users to skin and theme in completely flexible ways. So far no solution has presented itself, but I think that it would definitely be functionality that would be advantageous to users and developers, so it should be investigated.

I am sorry that this will leave you unsatisfied and upset, but it is a Firefox product decision based on our belief that lightweight themes are closer to what the majority of Firefox users are looking for in terms of visual customization of the browser. This is supported by metrics, user feedback, and comparing experiences with other browsers.

cheers,
mike

> _______________________________________________
> dev-apps-firefox mailing list
> dev-apps...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-apps-firefox

Aronnax

unread,
Jan 6, 2010, 12:37:31 PM1/6/10
to

> I would like to move towards a point where lightweight themes can work with heavy themes, allowing users to skin and theme in completely flexible ways. So far no solution has presented itself, but I think that it would definitely be functionality that would be advantageous to users and developers, so it should be investigated.
>

D�o G. has already build the lightweight themes system inside 3.6 in a
kind of way that every theme could use even now.
(By the way, he asked therefor the themer, what we want and need and
build it then - very obliging ;-) )

We have only the problem, that the 3.6 Add-on manager not approve it and
has some missing features to choose a themes and a Personas together.
All we have to do is to find therefor an easy to use solution and add
these changes to the Add-ons Manager.

Themes on the other hand must only add some code changes and everthing
is fine again.

Cheers

Mike Shaver

unread,
Jan 6, 2010, 1:25:26 PM1/6/10
to Pardal Freudenthal, dev-apps...@lists.mozilla.org
On Wed, Jan 6, 2010 at 11:22 AM, Pardal Freudenthal <par...@gmx.de> wrote:
> 1. [YOU, PARDAL] DON'T WANT Personas being treated as "themes". They are not!

I don't understand: are you saying that if there were an XPI,
chrome-override theme that just customized the background the way a
Persona does, you wouldn't consider it a theme? Or you wouldn't want
it listed in AMO under Themes? Unless Personas are somehow *more*
powerful than Themes, which I very much doubt, they are truly just
another type of theme, and the user shouldn't care about whether it's
packaged as an XPI, has a chrome.manifest, etc.

They *should* care about whether they can be applied without a
restart, are quick to download and update, etc. The tradeoff for the
flexibility that "full-power" themes provide (especially since they
can run script) is that it is very hard to reliably change between
them without reinitializing the browser, and that they require more
care for users when installing. Personas provides a way for theme
authors to constrain themselves to a "sweet spot" of theme capability
-- simple visual customization of the dominant main-window chrome
areas -- "in exchange" for which the theme can participate in a more
streamlined installation experience. This improved user experience
is, in our not-totally-uninformed opinion, very valuable in having
*more* people experience the power of customizing their browser. Some
of them will want deeper customization, and once they have a hint that
it's possible, I think they're much more likely to seek out
"full-power themes" that fulfill their wishes.

I love themes, and the shockingly creative things that people do with
them. There are a pile of us running and tweeting about the Strata
family of themes and add-ons right now, for example, and some that are
considering an operating-system switch to do so! This isn't about
disliking themes, and certainly not about disliking their creators.
(I daresay that the least respectful and co-operative tone
demonstrated in this thread has been from indignant full-power theme
authors, unfortunately.)

> 4. [YOU, PARDAL] WANT that the new feature, to change skins (personas) at real


> time, also works with third party themes. (Could you imagine if
> Microsoft would allow the "feature" Glass only for Microsoft
> Products?)

My understanding of this issue is that, indeed, any theme can support
personas if it makes appropriate code changes to accommodate the
persona customizations, and that "we" (Dao, really) went to some
lengths to make sure that was possible. This is almost exactly
analogous to how applications have to add support for Aero Glass,
which we have in fact done for Firefox. (There may be some rough
edges still to sort out, but enumerating those in bugs and helping
find resolutions to them is much more likely to get to an outcome you
want than nailing your demands to the newsgroup with liberal use of
ALL CAPS. This is again quite analogous to us finding issues with
different types of OS integration!)

I can understand that there is an emotional reaction here, and it's a
reasonable one: it's now possible for people to create a Firefox theme
(albeit a minimally-powerful one) with much less effort than before.
Even I could do it! For some people, it will feel like the prestige
attached to the label "Firefox theme creator" has been diminished, but
I don't think that's an appropriate view. It's *always* been the case
that some themes were more complex, more fully polished, and indeed
more widely-used than others. Personas as a tool for a subset of
theme capabilities doesn't change that, it just improves the user
experience where we are effectively able to do so, and ideally
introduces more people to both the use *and* creation of Firefox
themes. It may be that more people choose to use Personas than
full-power themes, for any number of individual reasons that are
difficult to predict crisply. But that means that those users are
getting a browser they like better than without it, since it is indeed
an active choice to use a theme of any "power level", and that seems
like what everyone in this thread is after.

A similar, though not analogically perfect, case from my own work: I
originally dreamed up and got people to work on FUEL, as a way to make
extension authoring easier and more future-proof. It didn't really
pan out, though it's used by some people. Jetpack is a much more
promising direction for reaching that goal. Do I feel a bit of sting
that we're going to deprecate FUEL in favour of Jetpack? A little: I
"was there first" with API improvements for extension developers, and
very similar goals. But not a lot, when I think about it: I did FUEL
because I wanted to make it easier for developers to make extensions,
and that goal is being attacked much better than it was before. You
made a theme because you wanted users to have another attractive
options for how their browser looks -- we want to give them a LOT of
such options, so it seems like we are allies here, not predator and
prey.

(If you made the theme because you wanted the exclusive cachet of
being able to do the difficult work of manipulating XUL and cascades
and packaging and overlays, and you don't think people should be able
to claim that they are theming the browser without going uphill both
ways in the snow, then we are not aligned on outcomes, and that we
helped each other was pretty much coincidental in the first place, I
fear. It wouldn't really be possible in good conscience to impede
improvements to user experience to preserve the exclusivity of the
developer circle, after all. I'm not at all saying that you, Pardal,
have that motivation, just recognizing that some authors might, and
that *those* people are not going to come out of this process with
what they want.)

Mike

Mandy

unread,
Jan 6, 2010, 2:06:33 PM1/6/10
to
On Jan 6, 6:25 pm, Mike Shaver <mike.sha...@gmail.com> wrote:

> > 4. [YOU, PARDAL] WANT that the new feature, to change skins (personas) at real
> > time, also works with third party themes. (Could you imagine if
> > Microsoft would allow the "feature" Glass only for Microsoft
> > Products?)
>
> My understanding of this issue is that, indeed, any theme can support
> personas if it makes appropriate code changes to accommodate the
> persona customizations, and that "we" (Dao, really) went to some
> lengths to make sure that was possible.  This is almost exactly
> analogous to how applications have to add support for Aero Glass,
> which we have in fact done for Firefox.  (There may be some rough
> edges still to sort out, but enumerating those in bugs and helping
> find resolutions to them is much more likely to get to an outcome you
> want than nailing your demands to the newsgroup with liberal use of
> ALL CAPS.  This is again quite analogous to us finding issues with
> different types of OS integration!)

Even I know that Firefox 3.6 only allows Personas to be used with only
the default theme. This is because the Persona chosen appears in the
Addons Manager and as only one theme can be in use at any one time,
then as soon as you select that Persona a restart is required to
disable any third party theme that was being used. After restart the
Persona appears, using the default theme.

How could you write such a long post here without knowing that?

Robert Strong

unread,
Jan 6, 2010, 2:21:47 PM1/6/10
to dev-apps...@lists.mozilla.org
He does know that and is just stating that additional work would need to
be done (by Dao, really) on the Firefox side and on the theme side (by
Theme authors) to make it so other themes can also be used with Personas.

Robert

Philip Chee

unread,
Jan 6, 2010, 2:28:44 PM1/6/10
to
On Wed, 6 Jan 2010 13:25:26 -0500, Mike Shaver wrote:

> I can understand that there is an emotional reaction here, and it's a

I believe that you are misreading the situation here. As far as I can
tell there are at least two issues.

1. You (collective you) are treating Personas vs traditional themes as a
zero sum game. In order to promote LWTs you feel the need to
de-emphasise HWTs. I'm sure that most if not all themers aren't actually
objecting to LWTs. What they seem to be objecting to is that (due to
current limitations of technology to be sure) a LWT and HWT cannot be
allowed to be active at the same time.

2. Secondly there is the deliberate obscuration of the difference
between LWTs and HWTs in both the Add-ons manager and on AMO. While it
is true that both are "themes" in the greater scheme of things it could
be also argued that the reductio ad absurdium is to remove the various
extension/theme/plugin panes in the addons manager and on AMO and put
everything in one window or category on the grounds that they are all
just "addons" after all.

I believe that there is significant public benefit to having separate
categories for LWTs and HWTs, or at least some obvious signifier to
differentiate these two. For example, distinctively different icons in
both the addons manager and on AMO.

I think that part of the paranoia currently floating around is due to
the current implementation on AMO where every other category is sorted
alphabetically except in the themes category where personas /appear/ to
be listed out of order. And as we all know perception can sometimes be
more powerful than reality. *And* there is no obvious way to search only
for HWTs or only for LWTs.

As an aside, I believe that there may be actually two distinct (though
overlapping) groups of users, one that prefers LWTs and another that are
interested only in HWTs.

There is also another issue with the combined theme category on AMO. I
think it is instructive to look at some past events on AMO. Some of you
might remember the decision to combine themes together with extensions
that were tagged with Appearance into one mixed category. That was I'm
sure everyone now agrees an extremely suboptimal decision causing a
great deal of confusion not just in the themer community but also for
the end users looking for some specific theme. I seem to remember a
similar problem caused by lumping dictionaries with extensions in the
"Language" category, but I could be mis-remembering.

May I suggest that in line with Mozilla's public aim of promoting choice
on the web that you strongly consider providing some way of clearly
distinguishing LWTs from HWTs for those users who desire such? For
example it could be one of the options in the advanced search option on
AMO to allow users to search only in LWTs or only in HWTs. For the
majority of users who don't care for the distinction, they should then
see none.

And then for the Addons Manager there could be an optional way of
visually separating LWTs from HWTs, perhaps as a toggleable option or a
checkbox. This should also go some way of assuaging complaints from
themers that the get more themes link now points only to LWTs instead of
giving the end user a choice of LWTs, HWTs or both. (I could be wrong I
don't actually use Firefox except to test my extensions in).

Also I might add that in the themer forums there seems to be a movement
to add CSS to their themes to hide the "Get more themes" link as long it
points exclusively to Personas. By all means do promote LWTs, but please
don't remove the ability of the user to find HWTs (or make the user jump
through various hoops in order to do so).

Aronnax

unread,
Jan 6, 2010, 2:39:47 PM1/6/10
to

> I can understand that there is an emotional reaction here, and it's a
> reasonable one: it's now possible for people to create a Firefox theme
> (albeit a minimally-powerful one) with much less effort than before.
> Even I could do it! For some people, it will feel like the prestige
> attached to the label "Firefox theme creator" has been diminished, but
> I don't think that's an appropriate view.

Hi,
i think not one themer has therefore an emotional reaction ;-)
Personas is now a core Firefox feature and it is the first time that a
theme can not use or support every Firefox feature.
Themes are now second class, compared with the default theme, and when
you know that there are only excuses, why this happens, have some themer
a problem with it.
But they have now the trouble, when user ask why his (her) theme works
not anymore with Personas and they maybe think it is the fault from a
buggy theme.

And for a themer is Personas on the other hand completely boring (not
powerful enough).
When many new Personas themer other have now as well fun with an easier
to use theme system is this only as well nice.

Cheers

Mike Shaver

unread,
Jan 6, 2010, 2:50:27 PM1/6/10
to Philip Chee, dev-apps...@lists.mozilla.org
On Wed, Jan 6, 2010 at 2:28 PM, Philip Chee <phili...@gmail.com> wrote:
> 1. You (collective you) are treating Personas vs traditional themes as a
> zero sum game. In order to promote LWTs you feel the need to
> de-emphasise HWTs. I'm sure that most if not all themers aren't actually
> objecting to LWTs. What they seem to be objecting to is that (due to
> current limitations of technology to be sure) a LWT and HWT cannot be
> allowed to be active at the same time.

Why is that more objectionable than two "HWT"s not being active at the
same time? Two Persona-based themes can't be active at the same time
either.

Now, some structural-theme developers may wish to allow further
customization of their visuals based on the Personas infrastructure
and gallery. For now, they will have to do that work in their theme
(or perhaps an accompanying extension, though I don't think that's
strictly a necessary separation) manually. It may be that some common
model emerges that we can support. I would welcome your input on what
that might look like, if there are specific things you want to see to
make it easier for your theme to use the Persona infrastructure to
allow additional customization.

> 2. Secondly there is the deliberate obscuration of the difference
> between LWTs and HWTs in both the Add-ons manager and on AMO. While it
> is true that both are "themes" in the greater scheme of things it could
> be also argued that the reductio ad absurdium is to remove the various
> extension/theme/plugin panes in the addons manager and on AMO and put
> everything in one window or category on the grounds that they are all
> just "addons" after all.

What is the relevant difference for the user between those, in Firefox
3.6? Only one of them (LWT or HWT) can be active at a given time.
Changing between and HWT and another HWT, or an HWT and an LWT
requires a restart, while LWT -> LWT doesn't, but I don't think that
warrants a separation.

You are welcome to whatever absurdums you want, but they're not useful
in making product decisions. We have evidence that users do know what
"themes" are, from other software experience as well as other Firefox
users and articles and such. SeaMonkey may choose to go another path
with this, and that would be a *perfectly* valid product decision for
you based on your knowledge of your users.

> I believe that there is significant public benefit to having separate
> categories for LWTs and HWTs, or at least some obvious signifier to
> differentiate these two. For example, distinctively different icons in
> both the addons manager and on AMO.

> I think that part of the paranoia currently floating around is due to
> the current implementation on AMO where every other category is sorted
> alphabetically except in the themes category where personas /appear/ to
> be listed out of order. And as we all know perception can sometimes be
> more powerful than reality. *And* there is no obvious way to search only
> for HWTs or only for LWTs.

> As an aside, I believe that there may be actually two distinct (though
> overlapping) groups of users, one that prefers LWTs and another that are
> interested only in HWTs.

Why do you believe that a meaningful number of users even know there's
a difference, let alone have an interest in a theme based on its
underlying technology? Our research indicates exactly the opposite,
to my knowledge.

> May I suggest that in line with Mozilla's public aim of promoting choice
> on the web that you strongly consider providing some way of clearly
> distinguishing LWTs from HWTs

How does that promote choice on the web? Why would a user explicitly
want a HWT, per se? What about its heaviness is part of what they're
looking for?

> for those users who desire such?

I haven't seen user demand for it, though I can imagine users wanting
at some point to limit their search to add-ons (Persona-based themes
or Jetpack-based extensions) that don't are safer (less-privileged)
and don't require restarts. If you have reason to believe that there
is Firefox user demand for that sort of search criterion, I recommend
you file a bug against AMO and take it from there.

> And then for the Addons Manager there could be an optional way of
> visually separating LWTs from HWTs, perhaps as a toggleable option or a
> checkbox.

Maybe -- want to write an extension and see if people find it useful?
Is this what SeaMonkey will do?

> This should also go some way of assuaging complaints from
> themers that the get more themes link now points only to LWTs instead of
> giving the end user a choice of LWTs, HWTs or both.

The first ones we present are those which can be instantly installed,
without restart. I think that's a pretty important user experience to
promote, now that it's possible.

> (I could be wrong I

> don't actually use Firefox).

Well, don't let that stop you from telling us how to build it! :-P

> Also I might add that in the themer forums there seems to be a movement
> to add CSS to their themes to hide the "Get more themes" link as long it
> points exclusively to Personas.

It points first to the Persona-based themes, with a link to more, just
as in the past it has pointed to the recommended themes list with a
link to more of them. What is the objection, exactly? It has always
promoted a way for users to get another theme, and I don't know why
themers object to users changing to an LWT more than to an HWT -- can
you elaborate on why, once the user has chosen to use a different
theme from an author's, the author should care what technology or
tools were used to build the user's new theme?

Mike

Benjamin Smedberg

unread,
Jan 6, 2010, 2:51:23 PM1/6/10
to
On 1/6/10 2:28 PM, Philip Chee wrote:

> 2. Secondly there is the deliberate obscuration of the difference
> between LWTs and HWTs in both the Add-ons manager and on AMO. While it
> is true that both are "themes" in the greater scheme of things it could
> be also argued that the reductio ad absurdium is to remove the various
> extension/theme/plugin panes in the addons manager and on AMO and put
> everything in one window or category on the grounds that they are all
> just "addons" after all.

This is what Firefox Mobile does, and it works well for the small screen
real-estate there.

But from a user perspective, a lightweight theme and a heavyweight theme
*are* the same thing: it's a way to modify the appearance of the browser.
The technical details of how it's done are irrelevant to most users, and
therefore the UI decision to obscure the differences between them is the
correct decision.

--BDS

Pardal Freudenthal

unread,
Jan 6, 2010, 4:20:57 PM1/6/10
to
On Jan 6, 8:51 pm, Benjamin Smedberg <benja...@smedbergs.us> wrote:

>
> But from a user perspective, a lightweight theme and a heavyweight theme
> *are* the same thing: it's a way to modify the appearance of the browser.
> The technical details of how it's done are irrelevant to most users, and
> therefore the UI decision to obscure the differences between them is the
> correct decision.
>
> --BDS

If each Persona *is* a theme, then please list them on the themes
section from AMO.

Asa Dotzler

unread,
Jan 6, 2010, 7:39:27 PM1/6/10
to

What is the compelling explanation from heavy-weight theme authors here
why a heavy-weight theme is not just another extension.

If heavy-weight themes are just like extensions in all the important
ways, (require a restart, have more serious security implications, and
can make major changes to the browser up to and including making it
unusable) then what is so horrible about moving all of your work over to
the extensions panel and letting the theme panel be exclusively used for
light-weight themes which don't require a restart, don't have serious
security implications and cannot make the browser completely unusable?

What is the objection to heavy-weight themes just being merged with the
extensions world?

- A

Robert Strong

unread,
Jan 6, 2010, 7:44:11 PM1/6/10
to dev-apps...@lists.mozilla.org
Because as far as the user is concerned it is a theme and not an
extension and the user doesn't care about these differences.

Robert

Philip Chee

unread,
Jan 6, 2010, 8:54:26 PM1/6/10
to
On Wed, 06 Jan 2010 16:39:27 -0800, Asa Dotzler wrote:

> What is the compelling explanation from heavy-weight theme authors here
> why a heavy-weight theme is not just another extension.

I hope you aren't serious about this question!

> If heavy-weight themes are just like extensions in all the important
> ways, (require a restart, have more serious security implications, and
> can make major changes to the browser up to and including making it
> unusable) then what is so horrible about moving all of your work over to
> the extensions panel and letting the theme panel be exclusively used for
> light-weight themes which don't require a restart, don't have serious
> security implications and cannot make the browser completely unusable?
>
> What is the objection to heavy-weight themes just being merged with the
> extensions world?

You do know that something like this was tried on AMO before right? And
it was universally unpopular with the *end users*. What makes you think
it would be any more popular the second time around?

Aronnax

unread,
Jan 6, 2010, 9:22:32 PM1/6/10
to

> What is the compelling explanation from heavy-weight theme authors here
> why a heavy-weight theme is not just another extension.

A theme change the style and an extention add features.

>
> If heavy-weight themes are just like extensions in all the important
> ways, (require a restart, have more serious security implications,

Do you have an example that a theme ever caused a serious security
implications? And themes have many limitations, because of security
questions.
A Personas image can theoretical import a virus - as well a serious
security implications, or?

and
> can make major changes to the browser up to and including making it
> unusable) then what is so horrible about moving all of your work over to
> the extensions panel and letting the theme panel be exclusively used for
> light-weight themes

Themes change the style - Personas change the style.
It makes in a horrible kind of way no sense to treat similar stuff
different ;-)

which don't require a restart, don't have serious
> security implications and cannot make the browser completely unusable?

You have not tried many Personas, or?
There are many, which makes the browser completely unusable ;-) (hard to
read text for example - even serious usability problems)

>
> What is the objection to heavy-weight themes just being merged with the
> extensions world?

And what change it?
A new extention feature works with all the other installed themes and
extentions. But a theme should not affect another theme, or you could
install only one.
It must be then anyway again a new and different kind of extention.

The same stuff only with new name is not really an improvement ;-)


Cheers

Mook

unread,
Jan 7, 2010, 1:41:54 AM1/7/10
to
On 2010-01-06 10:25 AM, Mike Shaver wrote:

> The tradeoff for the
> flexibility that "full-power" themes provide (especially since they
> can run script)

I'm curious where this is possible - are there any examples of this? As
in, actual, working ones, and not theoretical things? I'm mostly asking
because (really old, and possibly no longer true) documentation and
experience said normal scripts don't load from skin packages; knowing
when I'm wrong is always a fun thing for me :)

--
Mook

Michael Lefevre

unread,
Jan 7, 2010, 5:11:34 AM1/7/10
to
On 07/01/2010 00:39, Asa Dotzler wrote:
>
> What is the compelling explanation from heavy-weight theme authors here
> why a heavy-weight theme is not just another extension.
>
> If heavy-weight themes are just like extensions in all the important
> ways, (require a restart, have more serious security implications, and
> can make major changes to the browser up to and including making it
> unusable)

The theme authors didn't and don't actually want most of that, as far as
I know. The additional security implications and the restart are the
result of Mozilla's design decisions (and your failure to fix bugs that
would eliminate them).

> What is the objection to heavy-weight themes just being merged with the
> extensions world?

Themes are for changing the appearance of the browser, extensions are
for adding extra functionality. If you think it's a good idea to design
Firefox's UI based on how the code works, then knock yourself out...

Michael

Dao

unread,
Jan 7, 2010, 7:22:16 AM1/7/10
to

Themes can attach xbl bindings, from where they can do anything. Since
xbl bindings can be detached, I don't think that's the central reason
for why the restart is still required.

Robert Kaiser

unread,
Jan 7, 2010, 8:02:27 AM1/7/10
to
Asa Dotzler schrieb:

> What is the compelling explanation from heavy-weight theme authors here
> why a heavy-weight theme is not just another extension.

You can install any given number of extensions together and don't need
to switch between them, while you need that for themes.

Robert Kaiser

Aronnax

unread,
Jan 7, 2010, 9:18:31 AM1/7/10
to
> what is so horrible about moving all of your work over to
> the extensions panel and letting the theme panel be exclusively used for
> light-weight themes which don't require a restart, don't have serious
> security implications and cannot make the browser completely unusable?

By the way,
to prevent a similar discussion some months later.

With the same arguments can you remove the current extensions out of the
extention panel and replace it then with the new Jetpack stuff.

We should also create already now a new panel for example with the name:
Rusty but still lovely extention and themes stew.
This then for 3.7 and later with Firefox 4.0 is this stew naturally no
anymore lovely and we delete it then again.

And every developer who has a problem with it can Mozilla explain:
Nothing can be useful enough that restart is justifiable and the
competition has as well only limited theme and extensions systems.

Cheers


Pardal Freudenthal

unread,
Jan 7, 2010, 9:26:01 AM1/7/10
to
On Jan 7, 1:22 pm, Dao <d...@design-noir.de> wrote:
> Themes can attach xbl bindings, from where they can do anything.
Dão, XBL Bindidngs allowed for themes are very, very limited. No
javascript allowed.

Dao

unread,
Jan 7, 2010, 11:46:42 AM1/7/10
to
On 07.01.2010 15:26, Pardal Freudenthal wrote:
> On Jan 7, 1:22 pm, Dao <d...@design-noir.de> wrote:
>> Themes can attach xbl bindings, from where they can do anything.
> D�o, XBL Bindidngs allowed for themes are very, very limited. No
> javascript allowed.

That's news to me. In that case, I second Mook's question. Is there
another way for themes to do scripting, an official one or a hack?

Benjamin Smedberg

unread,
Jan 7, 2010, 11:56:24 AM1/7/10
to

A hack, yes. I'd prefer not to go into details here, but suffice it to say
that a combination of data:, javascript:, themes and XBL can cause themes to
execute arbitrary script, and so we decided that rather than trying to patch
all the possible holes we would just give up on the pretense that themes
can't execute script.

--BDS

Mike Shaver

unread,
Jan 7, 2010, 12:07:41 PM1/7/10
to Benjamin Smedberg, dev-apps...@lists.mozilla.org
On Thu, Jan 7, 2010 at 11:56 AM, Benjamin Smedberg
<benj...@smedbergs.us> wrote:
> A hack, yes. I'd prefer not to go into details here, but suffice it to say
> that a combination of data:, javascript:, themes and XBL can cause themes to
> execute arbitrary script, and so we decided that rather than trying to patch
> all the possible holes we would just give up on the pretense that themes
> can't execute script.

You refer here to skin/ code, right -- did we disable themes packaging
script and overlays that way as well? Can't remember now how far we
got along that whack-a-mole.

Mike

Boris Zbarsky

unread,
Jan 7, 2010, 12:09:04 PM1/7/10
to
On 1/7/10 11:56 AM, Benjamin Smedberg wrote:
> A hack, yes. I'd prefer not to go into details here, but suffice it to
> say that a combination of data:, javascript:, themes and XBL can cause
> themes to execute arbitrary script, and so we decided that rather than
> trying to patch all the possible holes we would just give up on the
> pretense that themes can't execute script.

Just to be clear, this was done after it was discovered that trying to
block all methods of executing script would also break various other
things that themes were doing....

-Boris

Benjamin Smedberg

unread,
Jan 7, 2010, 12:09:19 PM1/7/10
to
On 1/7/10 12:07 PM, Mike Shaver wrote:

> You refer here to skin/ code, right -- did we disable themes packaging
> script and overlays that way as well? Can't remember now how far we
> got along that whack-a-mole.

Yes, themes are only allowed to package chrome://foo/skin/, never */locale/
or */content/.

--BDS

Mike Shaver

unread,
Jan 7, 2010, 12:11:56 PM1/7/10
to Benjamin Smedberg, dev-apps...@lists.mozilla.org

And their components/ directory isn't traversed? OK.

Mike

Pardal Freudenthal

unread,
Jan 7, 2010, 12:33:50 PM1/7/10
to
On Dec 17 2009, 7:43 pm, Pardal Freudenthal <par...@gmx.de> wrote:

> The great advantage from themes against extensions is the way they
> register the skin providers and how they are selected. The skin
> providers brought from themes substitute the originals. This is a
> great feature, since if the user has installed a crap theme, he/she
> must only to select another theme and will have no annoying side
> effects.
>

More: http://www.tudobom.de/articles/yatt/

and: http://forums.mozillazine.org/viewtopic.php?p=7469305#p7469305

Philip Chee

unread,
Jan 7, 2010, 12:38:50 PM1/7/10