----- Original Message -----

Does anyone know the rationale behind the moving and shrinking of the fedora logo to this position?

The original impetus for moving it came out of multimonitor testing sessions we did last fall:

https://bugzilla.gnome.org/show_bug.cgi?id=685852

On Mon 18 Mar 2013 01:30:36 PM EDT, Ryan Lerch wrote:

On Mon 18 Mar 2013 01:07:28 PM EDT, Allan Day wrote:

...

Hi Ryan,

I agree that this doesn't look very good. There's an ongoing discussion in the upstream bugtracker [1].

Allan

[1] https://bugzilla.gnome.org/show_bug.cgi?id=694912

So from what I can gather from the GNOME bugzilla that you linked to, it was moved to the top panel to save space for when the userlist has many entries.

However, it was also moved to an area (the top panel), that is usually populated with elements a user interacts with. In fact when i first saw the fedora logo in the top left corner, i clicked it several times, thinking it was a user interface element.

cheers, ryanlerch

Allan,

Just to clarify, the new suggested solution (after the previous solution of moving the fedora logo from above the user list to the top bar) from is to have just a string with "fedora" in it up in the top left corner[1]

No fedora logo, No fedora logotype, just the string "Fedora" in the default gnome font?

cheers, ryanlerch

[1] - https://bugzilla.gnome.org/show_bug.cgi?id=694912#c16

On Mon, Mar 18, 2013 at 3:43 PM, Allan Day allanpday@gmail.com wrote:

That's right. It's looking like we will be targeting this for 3.8.1 though, so there's a bit of time for discussion.

Please, let's discuss this -- because as a former Fedora Project Leader, I can definitively say that the Fedora community is very sensitive to having their branding removed or changed in a significant way.

-- Jared Smith

On Mon 18 Mar 2013 05:54:42 PM EDT, Clyde E. Kunkel wrote:

I am not a programmer, but, that said, why can't you put a decent sized fedora logo above the users section. If the number of users exceeds a certain number, then create a window with a slider bar on the right. This just does not seem like rocket science. :-)

Isn't GNOME shell primarily designed as a single-user experience? How frequently are we expecting there to be a case where there are enough users in the list for a logo placed as suggested by Ray in comment 13 (https://bugzilla.gnome.org/show_bug.cgi?id=694912#c13)

I see a 'solution' looking for a problem here. The problem isn't the Fedora logo looks bad when given exactly 9 pixels of vertical space (what logo would?) The problem is... I'm really not sure what. Can you help me understand? I see this multi-monitor use related bug has been cited:

https://bugzilla.gnome.org/show_bug.cgi?id=685852

It says "The distribution logo is displayed in the middle above the list. This visually clashes with the layout of the rest of the login screen. It would be better to place the logo in the top-left hand corner of the screen, in the top bar. "

There is no screenshot, so I am unsure of how it 'visually clashes' with the rest of the login screen, and I'm not sure what it has to do with multi-monitor, if anything. A GNOME wiki link is provided, which I clicked on (https://live.gnome.org/Boston2012/Multimonitor) and found the exact same text:

"The distribution logo is displayed in the middle above the list. This visually clashes with the layout of the rest of the login screen. It would be better to place the logo in the top-left hand corner of the screen, in the top bar. #685852"

Please help me understand what exactly the problem is here? From the thread on devel-list about improving the boot experience, my understanding of the problem was:

1) The logos are not consistent between GDM and Plymouth. And they aren't, 100% agreed. Plymouth uses the logomark only, GDM uses the logotype+logomark. Their balance and general shape and size are inconsistent, and there's no reason they need to be. We could use the same version of the logo in both places.

2) The logos do not transition smoothly from Plymouth to GDM. Again, agreed 100%. If we used the same consistent version of the logo and used a consistent placement, one would fade into the other in the Plymouth-to-GDM transition, and this would be improved.

I am at a total loss as to how making the logo 9 pixels tall and shoving it in the upper left corner in both Plymouth and GDM is necessary to solve the above two problems. Why not have a 45x45 px version of the logomark only halfway between the clock and the top of the users list (again, as suggested by Ray) and have the same placement in plymouth? Then you would get consistency and a smooth transition.

Removing the logo completely and replacing it with a string is completely unacceptable from a Fedora point of view, and I'm very surprised this is the suggested solution if the problem is #1 and #2 above. If there is some other problem-to-solve that I am missing here that necessitates complete removal of the logo from the operating system, please let's talk about it so we can work on a solution together.

~m

As an end user of Fedora, I prefer having the logo centered above the login section. Having said that, I still have not understood why the login section is fairly large. There is a very plain dull grey background and this relatively huge login screen. I never found it to be visually appealing, and the Fedora logo was the only thing that made sense. On Mar 18, 2013 6:23 PM, "Máirín Duffy" duffy@fedoraproject.org wrote:

On 03/18/2013 06:13 PM, Máirín Duffy wrote:

...

Isn't GNOME shell primarily designed as a single-user experience? How frequently are we expecting there to be a case where there are enough users in the list for a logo placed as suggested by Ray in comment 13 (https://bugzilla.gnome.org/show_bug.cgi?id=694912#c13)

Sigh, unfinished thought. How frequently are we expecting there to be so many users that the placement suggested in the linked comment would be an issue?

~m

desktop mailing list desktop@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/desktop

On Mon, Mar 18, 2013 at 8:03 PM, Matthias Clasen mclasen@redhat.com wrote:

Jared, thats what we are doing here, discussing the best way to do justice to Fedoras branding while preserving the integrity of the login screen design.

Yes, and I'm glad we're having the discussion -- because that's how we get more ideas and help build consensus.

https://bugzilla.gnome.org/show_bug.cgi?id=694912 has a number of ideas already. We can't get them into 3.8.0 at this point, because we are entering hard code freeze with the .92 release tonight. It seems quite possible to get 3.8.1 into F19, considering that is planned for April 15.

I've read the bug, and can see that there are a number of ideas already. And I think I understand the timelines as well.

Lets just keep it constructive.

I'm sorry if my previous message came off sounding offensive or unconstructive -- that wasn't my intention.

-- Jared Smith

On Tue 19 Mar 2013 07:28:27 AM EDT, Allan Day wrote:

A bit of background:

The change during the 3.8 cycle was based on a couple of factors. First, the logo interfered with the layout of the login screen: it's a prominent visual presence that creates another anchor point which conflicts with the other elements on the screen (ie. it is horizontally centered, which clashes with the anchor points in the user list). Second, the logo was felt to be a distracting presence. We've made an effort to make sure that the most important elements are the most visually prominent, and we want the primary interaction points to be the ones that jump out at you. The logo was a strong visual presence placed above the user list: this drew the eye to it, making it the first thing you saw, and distracted you from the parts of the screen that are actually useful to the user (ie. the user list). Third and finally, having the logo in its previous position limited the size to which the user list could grow when there are a large number of users.

My preference is to focus the user interface on providing the best user experience possible. That means prioritising the things that people need to use, reducing distraction and making the UI look great. The addition of a logo diminishes the user experience along each of these dimensions.

The proposal to replace the logo with a simple string in the top-left hand corner is intended to mitigate the negative impact of including a logo while retaining a visual reference to the distributor. However, the usability issue that Ryan brought up is a valid concern about this proposal.

It has been suggested that not including the logo somehow weakens distributions' ability to brand their products. My view is that this is not the case. Branding is not the practice of slapping logos onto products. Instead, it is the attempt to instill and promote positive associations with the brand. The best way to do that, in my opinion, is to make the user experience as good as it can possibly be. If you diminish the user experience through the addition of a logo, then you actually harm the brand: you make the product worse, and in the process you make it less likely that people will think good things about your brand.

I suppose we are at a complete impasse then; to me it is completely unacceptable to completely debrand the operating system. You do realize there are usability implications with that - namely, people don't even know what they are running in order to obtain help with it or even identify the type of system they are using.

You have basically posed here that there is no way you will accept a logo on the login screen. How is a compromise possible then?

~m

Máirín Duffy duffy@fedoraproject.org wrote:

I suppose we are at a complete impasse then; to me it is completely unacceptable to completely debrand the operating system. You do realize there are usability implications with that - namely, people don't even know what they are running in order to obtain help with it or even identify the type of system they are using.

You have basically posed here that there is no way you will accept a logo on the login screen. How is a compromise possible then?

I was trying to explain my position and my concerns. I'm happy to explore the options. However, I would be much more comfortable doing that in an upstream setting.

Allan

On 03/19/2013 07:28 AM, Allan Day wrote:

Máirín Duffy duffy@fedoraproject.org wrote:

...

On Mon 18 Mar 2013 05:54:42 PM EDT, Clyde E. Kunkel wrote:

...

I am not a programmer, but, that said, why can't you put a decent sized fedora logo above the users section. If the number of users exceeds a certain number, then create a window with a slider bar on the right. This just does not seem like rocket science. :-)

Isn't GNOME shell primarily designed as a single-user experience?

I think it's a good idea to special case the single user case. We still want to provide a good multi-user experience though.

...

How frequently are we expecting there to be a case where there are enough users in the list for a logo placed as suggested by Ray in comment 13 (https://bugzilla.gnome.org/show_bug.cgi?id=694912#c13)

Even if it is a corner case, it's something we should handle, I think.

...

I see a 'solution' looking for a problem here. The problem isn't the Fedora logo looks bad when given exactly 9 pixels of vertical space (what logo would?) The problem is... I'm really not sure what. Can you help me understand? I see this multi-monitor use related bug has been cited:

https://bugzilla.gnome.org/show_bug.cgi?id=685852

It says "The distribution logo is displayed in the middle above the list. This visually clashes with the layout of the rest of the login screen. It would be better to place the logo in the top-left hand corner of the screen, in the top bar. "

There is no screenshot, so I am unsure of how it 'visually clashes' with the rest of the login screen, and I'm not sure what it has to do with multi-monitor, if anything. A GNOME wiki link is provided, which I clicked on (https://live.gnome.org/Boston2012/Multimonitor) and found the exact same text:

"The distribution logo is displayed in the middle above the list. This visually clashes with the layout of the rest of the login screen. It would be better to place the logo in the top-left hand corner of the screen, in the top bar. #685852"

Please help me understand what exactly the problem is here?

A bit of background:

The change during the 3.8 cycle was based on a couple of factors. First, the logo interfered with the layout of the login screen: it's a prominent visual presence that creates another anchor point which conflicts with the other elements on the screen (ie. it is horizontally centered, which clashes with the anchor points in the user list).

I agree with this. The alignment of the usernames to the left while the full logo (with both logo and logotype) was centre aligned did not really work.

However were any other alignment / layout options considered here? if we reduced the design to just the logo, you lose some of the interference between the anchor points of the logo and the userlist layout:

http://ryanlerch.fedorapeople.org/login/justthelogo.png

Or could the usernames be presented in a out horizontal layout? (although, this one makes the login screen very close to the OSX designs)

http://ryanlerch.fedorapeople.org/login/horizontal.png

Second, the logo was felt to be a distracting presence. We've made an effort to make sure that the most important elements are the most visually prominent, and we want the primary interaction points to be the ones that jump out at you. The logo was a strong visual presence placed above the user list: this drew the eye to it, making it the first thing you saw, and distracted you from the parts of the screen that are actually useful to the user (ie. the user list).

Were any tests done to prove the hypothesis that the placement of the old fedora logo was distracting users from the interactive parts of the screen? If that was the issue that was trying to be solved, was any thought given to putting the logo towards the bottom of the screen?

http://ryanlerch.fedorapeople.org/login/logobottom.png

My preference is to focus the user interface on providing the best user experience possible. That means prioritising the things that people need to use, reducing distraction and making the UI look great. The addition of a logo diminishes the user experience along each of these dimensions.

The proposal to replace the logo with a simple string in the top-left hand corner is intended to mitigate the negative impact of including a logo while retaining a visual reference to the distributor. However, the usability issue that Ryan brought up is a valid concern about this proposal.

It has been suggested that not including the logo somehow weakens distributions' ability to brand their products. My view is that this is not the case. Branding is not the practice of slapping logos onto products. Instead, it is the attempt to instill and promote positive associations with the brand. The best way to do that, in my opinion, is to make the user experience as good as it can possibly be.

True, you want to promote positive associations with a brand. And if the logo is too prominent, or there are too many logos, you end up diluting the brand. If all text is bold, nothing stands out.

http://ryanlerch.fedorapeople.org/login/logotoomany.png

However if the mark of the brand is *nowhere* you are merely promoting positive associations with a generic grey background.

http://ryanlerch.fedorapeople.org/login/nologo.png

I am not suggesting that we slap logo(s) over everything, simply include a clear definition to the user of brand at the point they start using the system.

regards, ryanlerch

Why don't we explore the options right here in the affected community?

Fedora is not the only consumer of GNOME.

am not a member of the GNOME community anymore.

What does that mean? You are even listed at http://www.gnome.org/foundation/membership/

Cheers, Debarshi

On 03/19/2013 10:25 AM, Debarshi Ray wrote:

...

Why don't we explore the options right here in the affected community?

Fedora is not the only consumer of GNOME.

It's a pretty important one, don't you think?

...

am not a member of the GNOME community anymore.

What does that mean? You are even listed at http://www.gnome.org/foundation/membership/

I delisted my blog from planet gnome well over a year ago and withdrew my foundation membership - I should *not* be on that list.

Either way, my point is let's not spend a week talking about how we need to talk about this and instead talk about it on neutral ground everyone is comfortable with.

~m

On Tue, 2013-03-19 at 10:59 -0400, Ryan Lerch wrote:

...

Genuine question: why do you feel the login screen is a good place to promote any kind of positive brand association?

This is the entry point to the Fedora user experience. Having the logo on the sign in screen is conceptually similar to having a sign on the door.

If you want to follow this analogy, it would be more as if the door of my house had a permanent logo/sign of the realtor agency I bought it from...

...

All I usually want to do from that screen is to get out of it as quickly and seamlessly as possible. Under this perspective, having it plainly unbranded and focused exclusively on its task — get me to my session and my work — makes sense to me.

Cosimo

From a usability perspective, I fail to see how a subtle logo here makes it harder to use. Having some user testing the hypothesis that a logo reduces productivity of the login screen would be a good place to start before removing whole elements.

I didn't say it makes it *harder* to use. To my taste, it just feels redundant/distracting and maybe a bit cheesy though.

Finally, a question. If we don't have the sign the doorway to Fedora, then where should the name of the brand be presented? I think it would be a step backwards to have the OS brand mark polluting the main interface of the desktop that a user stares at for 8 hours a day.

Good question; having it always visible in the desktop is a huge step backwards of course (and we agree it's a non starter). My personal opinion is it's debatable you should present it at all, since the user knows what it downloaded and installed himself.

Who is supposed to look at that doorway and from where? In other words, I don't think I can have an answer to that question unless it's very clear who's the intended target of that branding.

Cheers, Cosimo

On Tue, 2013-03-19 at 12:26 -0400, Máirín Duffy wrote:

On Tue 19 Mar 2013 12:22:51 PM EDT, Cosimo Cecchi wrote:

...

Good question; having it always visible in the desktop is a huge step backwards of course (and we agree it's a non starter). My personal opinion is it's debatable you should present it at all, since the user knows what it downloaded and installed himself.

Let's be fair - in a multi-situation, it is necessarily that only one of the users involved installed the system. In a computer lab situation or in an employee preload situation (maybe the latter wouldn't be Fedora, but RHEL) the users did not install it themselves.

That's true. But in a preloaded/locked-down situation where the user doesn't have full, or any, control on the software installed on the machine the requirements might be different: - the logo would likely not be the stock distributor's one, but one identifying the organization providing that leased/temporary service to the user. I can think of a number of reasons related to billing/support/accountability as of why this makes a lot more sense than in the "personal use" case. - it's still unclear to me the benefit to the user and/or the distributor in showing a logo in that scenario. For example, if the machine has a corporate lockdown, that fact alone might discourage users trying to install the same OS on another personal machine. This is in my opinion related to the question I was asking in the second part of my above previous message.

Cheers, Cosimo

Cosimo Cecchi (ccecchi@redhat.com) said:

On Tue, 2013-03-19 at 12:26 -0400, Máirín Duffy wrote:

...

On Tue 19 Mar 2013 12:22:51 PM EDT, Cosimo Cecchi wrote:

...

Good question; having it always visible in the desktop is a huge step backwards of course (and we agree it's a non starter). My personal opinion is it's debatable you should present it at all, since the user knows what it downloaded and installed himself.

Let's be fair - in a multi-situation, it is necessarily that only one of the users involved installed the system. In a computer lab situation or in an employee preload situation (maybe the latter wouldn't be Fedora, but RHEL) the users did not install it themselves.

That's true. But in a preloaded/locked-down situation where the user doesn't have full, or any, control on the software installed on the machine the requirements might be different:

• the logo would likely not be the stock distributor's one, but one

identifying the organization providing that leased/temporary service to the user. I can think of a number of reasons related to billing/support/accountability as of why this makes a lot more sense than in the "personal use" case.

... which, since many distributors of GNOME need to account for this case, seems reasonable enough reason to have that space reserved in both the login screen and any boot-time display (e.g. plymouth) already for a logo, and for Fedora to put its logo there.

Bill

On Tue, 2013-03-19 at 14:38 -0400, Bill Nottingham wrote:

...

• the logo would likely not be the stock distributor's one, but one

identifying the organization providing that leased/temporary service to the user. I can think of a number of reasons related to billing/support/accountability as of why this makes a lot more sense than in the "personal use" case.

... which, since many distributors of GNOME need to account for this case, seems reasonable enough reason to have that space reserved in both the login screen and any boot-time display (e.g. plymouth) already for a logo, and for Fedora to put its logo there.

It sounds like we're going to have a logo in Fedora in any case, but just for the sake of making my point clear, I really don't see why the former (having that possibility available for specific use cases of some users) would necessarily imply the latter (put a Fedora logo there for all users).

Cosimo

It is since we started shipping Gnome 3.0 in Fedora 15 that I feel the Fedora branding was reduced to almost zero. You can, with a good confidence, instantly tell if a user is running Windows, OSX or Ubuntu when he attaches his PC to a projector. Not so with Fedora and Gnome.

With GNOME3, GNOME's visual identity is clearer than ever before. I am not sure about Fedora, but one can instantly recognize if a computer is running GNOME3.

Cheers, Debarshi

Hi,

----- Original Message -----

...

With GNOME3, GNOME's visual identity is clearer than ever before. I am not sure about Fedora, but one can instantly recognize if a computer is running GNOME3.

So the question is this:

Is the user installing Fedora or are they installing Gnome?

I think it is Fedora.

I don't think there's any question that the user is installing Fedora. I seriously doubt anyone disagrees with that.

The points of contention seems more to be whether the logo needs to be prominently placed or not, and if so where on the screen.

--Ray

On Wed, Mar 20, 2013 at 1:30 PM, Máirín Duffy duffy@fedoraproject.org wrote:

On 03/20/2013 12:07 PM, seth vidal wrote:

...

So the question is this:

Is the user installing Fedora or are they installing Gnome?

I think it is Fedora.

I think Fedora ships many desktops and we should get some precedence in the login screen that the user sees no matter which desktop they are using.

Here's another important & related question: When something goes wrong, who can help?

In the case of my family the answer is usually "me", as the person who installed Fedora on all these laptops. :) I also believe the System Monitor has a panel with more information specifically to help with situations like that, where it includes the distribution name, version, kernel and important hardware details.

If a user encounters an SELinux error, or a kernel oops, or their laptop stops suspending, or their network card is no longer recognized - is the GNOME community where they can go to get help? Is the GNOME community equipped to help them with those things? Should they file those bugs in GNOME bugzilla?

The OS is not GNOME. The OS is Fedora. Branding the OS as if it was GNOME isn't going to change the reality of the situation.

I don't think that's what's being discussed here. GNOME is not trying to remove Fedora logo in favor of a GNOME logo.

The goal of the redesigned login screen is, as far as I understand it, to provide a better user experience. Designing and going over every detail of the user experience is what the GNOME project does across all of its components, to the benefit of Fedora and all other distributors and users of GNOME. That includes the login screen, the activities overview, the settings panel, etc. By shipping GNOME and thousands of other packages as their creators originally intended is how Fedora accomplishes its goal of leading the advancement of open source and free software imo.

-- Evandro

Cosimo Cecchi (ccecchi@redhat.com) said:

On Tue, 2013-03-19 at 14:38 -0400, Bill Nottingham wrote:

...

...

• the logo would likely not be the stock distributor's one, but one

identifying the organization providing that leased/temporary service to the user. I can think of a number of reasons related to billing/support/accountability as of why this makes a lot more sense than in the "personal use" case.

... which, since many distributors of GNOME need to account for this case, seems reasonable enough reason to have that space reserved in both the login screen and any boot-time display (e.g. plymouth) already for a logo, and for Fedora to put its logo there.

It sounds like we're going to have a logo in Fedora in any case, but just for the sake of making my point clear, I really don't see why the former (having that possibility available for specific use cases of some users) would necessarily imply the latter (put a Fedora logo there for all users).

If nothing else, having it be the default makes it a good test case to ensure that it works, and also to give those that want to customize it later an entry point.

With respect to having a logo, the suggestion posted by Ryan of having it center-bottom would seem to be the cleanest one - there it would seem less likely to me to distract from the center-focus of the login screen itself, but would be notable/sized enough that a reasonable logo could be used instead of the top-left titlebar approach.

Bill

On Wed, 20 Mar 2013 12:30:01 -0400 Máirín Duffy duffy@fedoraproject.org wrote:

On 03/20/2013 12:07 PM, seth vidal wrote:

...

So the question is this:

Is the user installing Fedora or are they installing Gnome?

I think it is Fedora.

I think Fedora ships many desktops and we should get some precedence in the login screen that the user sees no matter which desktop they are using.

Here's another important & related question: When something goes wrong, who can help?

If a user encounters an SELinux error, or a kernel oops, or their laptop stops suspending, or their network card is no longer recognized - is the GNOME community where they can go to get help? Is the GNOME community equipped to help them with those things? Should they file those bugs in GNOME bugzilla?

The OS is not GNOME. The OS is Fedora. Branding the OS as if it was GNOME isn't going to change the reality of the situation.

Here's my 2c take on it. GDM isn't the only Display manager. If the gnome project doesn't want to do what we want in terms of a logo on the DM then let's move to lightdm or one of the others and theme it however we want.

GDM isn't the only game in town and afaict we wouldn't even need to package something new, looks like lightdm and lxdm both exist and are maintained.

We don't want to require patches to lots of programs, of course, but in this case it is not about patching, afaict, it's about adding a different pkg by default and adding some theming to it.

Seems like a simpler win and would involve more designing of a theme and less arguing over various people's impression of cleanliness.

-sv

Hi,

Here's my 2c take on it. GDM isn't the only Display manager. If the gnome project doesn't want to do what we want in terms of a logo on the DM then let's move to lightdm or one of the others and theme it however we want.

Well hang on now. There's a huge overlap between the fedora desktop people and gnome people. Like I said earlier, we all where multiple hats.

This clearly isn't a "fedora desktop" versus "gnome" fight here. It's just a technical disagreement between specific individuals. And that's fine. We don't all have to agree one every detail as long as we agree on the general direction.

And we can resolve technical disagreements through discussion. We don't need to pack up our toys and go home or whatever.

--Ray

On Wed, 2013-03-20 at 13:21 -0400, seth vidal wrote:

Here's my 2c take on it. GDM isn't the only Display manager. If the gnome project doesn't want to do what we want in terms of a logo on the DM then let's move to lightdm or one of the others and theme it however we want.

GDM isn't the only game in town and afaict we wouldn't even need to package something new, looks like lightdm and lxdm both exist and are maintained.

We don't want to require patches to lots of programs, of course, but in this case it is not about patching, afaict, it's about adding a different pkg by default and adding some theming to it.

Seems like a simpler win and would involve more designing of a theme and less arguing over various people's impression of cleanliness.

I would normally be ashamed and frustrated to the idea alone of replying to a message like this, but I will make myself an exception, because I feel it's important for me to say it.

I think this message of yours was rude, inflammatory and condescending. It's technically misinformed. It misses the entire point of this discussion. It moves the argument to a space where it can only die, or burst in an unproductive flame that will have no positive outcome other than making people more frustrated and less willing to engage in a serious conversation on this list in the future.

Communication and collaboration can be hard, and I think we can only succeed if we carefully weight what we're saying and respectfully consider the nature of people involved in a given discussion. I personally believe project leaders especially should realize this, and I am often saddened to observe that this is not the case in Fedora.

Regards, Cosimo

On Wed, Mar 20, 2013 at 7:29 PM, seth vidal skvidal@fedoraproject.org wrote:

On Wed, 20 Mar 2013 14:24:03 -0400 Cosimo Cecchi ccecchi@redhat.com wrote:

...

I would normally be ashamed and frustrated to the idea alone of replying to a message like this, but I will make myself an exception, because I feel it's important for me to say it.

I think this message of yours was rude, inflammatory and condescending. It's technically misinformed. It misses the entire point of this discussion. It moves the argument to a space where it can only die, or burst in an unproductive flame that will have no positive outcome other than making people more frustrated and less willing to engage in a serious conversation on this list in the future.

Communication and collaboration can be hard, and I think we can only succeed if we carefully weight what we're saying and respectfully consider the nature of people involved in a given discussion. I personally believe project leaders especially should realize this, and I am often saddened to observe that this is not the case in Fedora.

Cosimo,

I completely disagree - This is a fedora desktop list - I recommended a simple and technical solution to our disagreement. There was nothing inflammatory about it whatsoever. I recommended using another, available, program that steps around the issue and is immediately available to fedora.

If this were a gnome mailing list I would agree that it would be inappropriate but this is not a gnome mailing list - this is the fedora desktop list.

If you could address the technical issues of my suggestion, I'd be happy to hear.

OK I'll bite there are technical and usability advantages of using gdm when gnome is used as the default desktop. For one consistency you have the same look and feel on the login screen as in your desktop, nice transition effect from going from greeter -> desktop, faster login due to usage of the same libraries, support for accessibility in the login screen keeping the accessibility options on if the user needed them for login and many others ...

It is not as simple as replacing it with something else as you thing ... sure it isn't impossible but it causes lots of regression. I'd argue coming up to an agreement regarding the logo issue is way easier so lets focus on that.

On Wed, Mar 20, 2013 at 7:37 PM, seth vidal skvidal@fedoraproject.org wrote:

So when Cosimo said it was 'technically misinformed' what, specifically, was he talking about?

That's a question Cosimo can answer better then me so I defer ..

Transition effects doesn't sound like a road block.

Not a roadblock but it contributes to a nice professional modern looking UX.

We don't seem to have this concern with gdm->xfce or people using lightdm and then install gnome afterwards, do we?

No but that does not change the fact that it is a nice thing to have (in addition to the other bits I mentioned).

I completely disagree - This is a fedora desktop list - I recommended a simple and technical solution to our disagreement. There was nothing inflammatory about it whatsoever. I recommended using another, available, program that steps around the issue and is immediately available to fedora.

If this were a gnome mailing list I would agree that it would be inappropriate but this is not a gnome mailing list - this is the fedora desktop list.

1. The default Fedora desktop is GNOME.

2. You seem to be trying to draw a sharp line between Fedora and GNOME users or contributors, merely based on the fact that you run a different desktop.

Cheers, Debarshi

In Fedora 18, the Details page only shows a GNOME logo

See "Base system" and https://bugzilla.gnome.org/show_bug.cgi?id=695691

Cheers, Debarshi

On 03/20/2013 03:26 PM, Colin Walters wrote:

So if the question is "what entity is responsible for the gdm package", the answer really is ultimately neither GNOME, nor Fedora, nor Ray, nor any other entity.

You're being pedantic. There is liability and then there is our community responsibility to work with the end-user on any bugs they encounter. Fedora bears the latter, and will continue to do so, via bugzilla. Obviously, this is without legal guarantees of any sort, but to say "No entity is responsible" isn't really accurate.

Any packager on any component who doesn't make any attempt to address bugs in their Fedora component is likely to be removed from that package.

~tom

== Fedora Project

Hi,

----- Original Message -----

There is liability and then there is our community responsibility to work with the end-user on any bugs they encounter. Fedora bears the latter, and will continue to do so, via bugzilla. Obviously, this is without legal guarantees of any sort, but to say "No entity is responsible" isn't really accurate.

This is an interesting topic, and perhaps one that should be brought to the more general developer list, but i'll chime in here.

I see bugzilla has a vehicle for users in the community to contribute to the community by providing insight into how the os is working. It's a very important service those users provide. It helps us tract where the serious problems are, and it helps us see the impact of bugs.

I don't see bugzilla as obligation for the package owner. The ratio of packages to package owners, and bugs to package owners is really too big for that to be true. I'm sure you've heard before and probably even said before the mantra "bugzilla isn't a support forum". I see it as more of a way for a user to give something back, than it is a way for a user to get something.

Furthermore, most bugs should go upstream. In gnome at least, almost every commit has to have an upstream report associated with it, anyway, so the bugs should really start there. The exception are bugs that block a release or have a big enough impact that they warrant an asynchronous update.

Any packager on any component who doesn't make any attempt to address bugs in their Fedora component is likely to be removed from that package.

That's 100% the wrong way to look at it. If there isn't enough man power to deal with something, you don't take the man power that's there away. You augment it. But I don't think the goal post should be every bug triaged, investigated, and fixed. it's not realistic, and it's never been that way.

--Ray

On Mar 20, 2013 12:49 PM, "Tom Callaway" tcallawa@redhat.com wrote:

On 03/20/2013 03:26 PM, Colin Walters wrote:

...

So if the question is "what entity is responsible for the gdm package", the answer really is ultimately neither GNOME, nor Fedora, nor Ray, nor any other entity.

You're being pedantic. There is liability and then there is our community responsibility to work with the end-user on any bugs they encounter. Fedora bears the latter, and will continue to do so, via bugzilla. Obviously, this is without legal guarantees of any sort, but to say "No entity is responsible" isn't really accurate.

Any packager on any component who doesn't make any attempt to address bugs in their Fedora component is likely to be removed from that package.

~tom

In a judge's opinion, (myself a political science undergrad), the responsibility for software bugs in our product probably lies with Red Hat, since that is our commercial aside, which, on the open side, as Tom concluded, would likely result in removal "from that" and probably other packages. Responsibility for a software bug, unlike Bill Gates who society let sell a horrible, bug-ridden system to them, has to rest somewhere.

Disclaimer: the attitudes and opinions of the author are those of the author alone and not reflective of Red Hat or its Fedora open source communities

== Fedora Project -- desktop mailing list desktop@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/desktop

On Wed, 2013-03-20 at 13:38 -0700, Adam Williamson wrote:

On Wed, 2013-03-20 at 15:12 -0400, Bill Nottingham wrote:

...

• some people feel that there is a concerted effort in upstream GNOME to deemphasize and/or remove downstream branding. (The existence of a concept called GNOME OS probably doesn't help work against that perception.)

This seems a pretty inarguable trend to me: there is much more of a vertically integrated GNOME stack in Fedora than there ever used to be. We are replacing the generic Fedora firstboot with gnome-initial-setup, for GNOME installs. GNOME has been systematically attempting to replace system-config-* with environment specific changes.

I don't necessarily think this is a *bad* Thing, but it is definitely a Thing. There is a definite trend towards our default live image and package set becoming a lot more GNOME-y and a lot less Fedora-y.

Oh, I forgot: GTK+ is getting tied closer and closer to GNOME, and there is a strong blurring of the lines between GTK+ apps and GNOME apps. GDM used to be a GTK+-based, fairly desktop-agnostic login manager: it is now a special instance of GNOME Shell, strongly integrated with GNOME. It can load other desktops, but this is almost a coincidence at this point.

Again, I don't necessarily think these are bad things, but they're definitely things, and possibly ones Fedora as a project has not thought about systematically at the level of 'what does this mean for Fedora, and do we have to adjust anything else to account for these changes?'

Hi again,

----- Original Message -----

...

I don't necessarily think this is a *bad* Thing, but it is definitely a Thing. There is a definite trend towards our default live image and package set becoming a lot more GNOME-y and a lot less Fedora-y.

I don't see why more gnome integration implies being less fedora-y. A more integrated experience is good for Fedora and Fedora's image. We're way better than we used to be and it's because we've made things more seamless.

I just wanted to add that I do think Fedora should have a strong identity.

How best to do that, I think is a valid question and we should try to figure out.

--Ray

Hi,

----- Original Message -----

Well, it's less Fedora-y in that we used to have this kind of conception where there were desktop environments, controlled by the desktop team.

Fedora desktop team has only ever controlled the desktop spin, which uses GNOME. Other SIGs have controlled other desktops since there have been spins.

Then the login manager, system config tools, and probably some other stuff I'm not thinking of were controlled more or less by the distribution.

what do you mean by "the distribution" ? We all work on the distribution right?

GDM did stuff like setting language and keyboard layout, and wasn't really considered a part of the GNOME stuff (I don't think).

GDM has always been a part of the gnome ecosystem. that's what the G in GDM stands for.

It was expected that you could just swap out DMs (like skvidal thinks is still the case) and everything else should just deal with it.

So now, gnome-shell does sort of require GDM to have full functionality, and GDM does require gnome-shell to have full functionality. That's definitely an example of integration that didn't used to exist that does now. I don't see why that integration makes Fedora less Fedora-y though.

The system-config-* tools were part of Fedora and used across all desktops - it didn't matter what desktop you ran, you used system-config-keyboard and system-config-display and so on.

It's true, we've tried to get rid of the need for the system-config-tools. In some cases we've made them completely unnecessary (thanks ajax, thanks dcbw, etc). I think make things like those more integrated doesn't make Fedora less Fedora-y though. I think it just makes Fedora more cohesive.

I think other desktops still think of things somewhat in those terms, but GNOME definitely doesn't: GNOME wants its own stack, almost top to bottom.

Not sure about GNOME, but I can speak for myself as a gnomie and long time fedora desktop member. I want users to have a top-to-bottom integrated experience. But I think if you asked anyone working on fedora if in general they want the user to have the opposite of a smooth integrated experience, they would say of course not. So it's really a matter of the specifics of how to get there, I guess.

--Ray

On Wed, 20 Mar 2013 14:02:58 -0700 Adam Williamson awilliam@redhat.com wrote:

Well, it's less Fedora-y in that we used to have this kind of conception where there were desktop environments, controlled by the desktop team. Then the login manager, system config tools, and probably some other stuff I'm not thinking of were controlled more or less by the distribution. GDM did stuff like setting language and keyboard layout, and wasn't really considered a part of the GNOME stuff (I don't think). It was expected that you could just swap out DMs (like skvidal thinks is still the case) and everything else should just deal with it.

Adam, That is exactly what I was thinking. Back in the mists of time my department at duke was one of the larger deployments of gnome (ximian gnome at the time). We had 100ish desktops running red hat linux and defaulting their logins to ximian's gnome. The gdm they shipped worked pretty well then at one point things went wrong and in order to allow for logins using nfs-automount homedirs we had to switch to kdm. We rethemed the kde login so it had the duke physics logos on it (which were cool pictures of lasers and masses of atoms!) and then swapped out gdm for kdm. No other issues crept up. The bits were all behaving and kdm did what I needed to do, so that was a problem solved.

I came at this discussion fully from that perspective and from the perspective of wanting to find a resolution that:

1. would end up with LESS need for discussion of aesthetics - since that seems to me to get murky, quickly.

2. where we didn't need to shove the maintainer of gdm in an awkward position trying to stand astride what fedora wants and what gnome wants - in terms of logo placement.

3. that allowed the fedora design team's decisions to be followed - since branding in this discussion does seem to fall under their purview.

Additionally, I want to point out that I agree with Notting that it's not without precedent that we replace some piece that is considered a component by gnome upstream b/c the distro wants something else - firefox being a prime example.

Finally, this was not meant to be inflammatory or as a 'nuclear option' - I didn't believe replacing a piece like the DM to be as tightly coupled as you seem to be saying it is. I'm a bit concerned as to what happens to our users installing from one of the other desktop env spins and then adding gnome or kde later. If I can dig out a spare laptop I may give that a whirl to see what goes off the rails.

-sv

On Wed, 2013-03-20 at 23:24 +0100, drago01 wrote:

On Wed, Mar 20, 2013 at 10:34 PM, Adam Williamson awilliam@redhat.com wrote:

...

On Wed, 2013-03-20 at 17:19 -0400, Ray Strode wrote:

...

Hi,

----- Original Message -----

...

Well, it's less Fedora-y in that we used to have this kind of conception where there were desktop environments, controlled by the desktop team.

Fedora desktop team has only ever controlled the desktop spin, which uses GNOME. Other SIGs have controlled other desktops since there have been spins.

...

Then the login manager, system config tools, and probably some other stuff I'm not thinking of were controlled more or less by the distribution.

what do you mean by "the distribution" ? We all work on the distribution right?

Well sure, but we kinda felt a project wide responsibility that those tools should work, because they were Our Tools For Making Stuff Work.

Now no-one really works on a lot of the s-c* tools, partly because 'GNOME replaced them'.

Well it seems that nobody cares enough to do that work. So either it is not that important for the other desktop environments (user prefer to do do the configuration using other tools / text editors) or there is simply no one that cares enough *and* has the ability to step and do the work.

Well, sure, but I don't place much faith in this exact formulation of this common argument. 'Caring' isn't something that just magically Happens or Doesn't Happen. It's not like anyone was maintaining s-c-* before because it was a bundle of fun, they were maintaining them because they were the Fedora system configuration tools. And, bluntly, RH was paying most of them. But now RH's paid resources and most of the 'we care because they're our most prominent configuration tools' resources get re-directed into working on the GNOME tools, to the detriment of the desktop-agnostic tools.

Configuration tools that are part of and integrated into the desktop do offer a better and consistent user experience,

They offer a better and more consistent user experience *for users who use that desktop and only that desktop*. They offer a less consistent user experience for users who use multiple desktops, and they offer nothing at all for users who don't use that desktop.

On Thu, 2013-03-21 at 00:17 +0100, drago01 wrote:

...

...

Well it seems that nobody cares enough to do that work. So either it is not that important for the other desktop environments (user prefer to do do the configuration using other tools / text editors) or there is simply no one that cares enough *and* has the ability to step and do the work.

Well, sure, but I don't place much faith in this exact formulation of this common argument. 'Caring' isn't something that just magically Happens or Doesn't Happen. It's not like anyone was maintaining s-c-* before because it was a bundle of fun, they were maintaining them because they were the Fedora system configuration tools. And, bluntly, RH was paying most of them. But now RH's paid resources and most of the 'we care because they're our most prominent configuration tools' resources get re-directed into working on the GNOME tools, to the detriment of the desktop-agnostic tools.

So RH cared and payed for the maintenance now RH no longer cares .... (you get the pattern)

That's not what I said at all. To put it cynically, my evaluation is that RH kinda cares a bit (let's not get too carried away with how much RH cares about the desktop at all...) that the most obvious bits of the default desktop pretty much work. So when the default desktop used desktop-agnostic tools, RH paid for those to be maintained. Now the default desktop uses its own tools, RH pays for those to be maintained instead. The default desktop is still happy: the desktop team has never seen this as a problem because, for the desktop team, it is not a problem. RH is happy; insofar as RH cares at all, which is only moderately, it cares that GNOME works. For Fedora as a whole, I suspect it's a problem that we have not been very good at recognizing and formulating.

...

...

Configuration tools that are part of and integrated into the desktop do offer a better and consistent user experience,

They offer a better and more consistent user experience *for users who use that desktop and only that desktop*. They offer a less consistent user experience for users who use multiple desktops, and they offer nothing at all for users who don't use that desktop.

Yeah so the goal would be to have such tools integrated in every desktop (upstream!) ... yes we are not their yet and probably should add something as a stop gap solution but the end goal for each desktop should be to provide a consistent and integrated UX and not a mixture of random tools that do not fit at all.

So two points, there: one, be careful of telling other projects what their goals are. You don't like it when KK sends his sixth mail of the week saying how GNOME should become exactly like KDE, and I doubt fluxbox or LXDE users really think 'the end goal' of those projects should be to write a bunch of configuration utilities.

Two, on a purely practical level, even if that were the end goal for everyone, as you say, we are not there yet, or even close. Practically speaking, I think we (Fedora) would be in a much better place if those boring, old, unsexy, non-desktop specific, 'random' tools were properly maintained and respected within the distro.

On Wed, 2013-03-20 at 15:53 -0400, Tom Callaway wrote:

Any packager on any component who doesn't make any attempt to address bugs in their Fedora component is likely to be removed from that package.

Please do. After triaging the bugs and seeing that most of them are already fixed upstream where the bugzilla has triagers, per module components, inline, searchable, backtraces and images are opened in the browser.

Hi,

Like I said, I think there are areas where we need to look at things from a Fedora-as-a-whole perspective and say, well, look, is the distro covered here? Not just 'does it work on our default desktop' but does it work on a distro-wide level?

I thinks it's far more important to focus our efforts on making the desktop spin good than expending effort bringing feature parity for the other spins. Getting the default experience good is a huge effort all on its own and we aren't really measuring up as well as we could be if we focused more.

That's not to say we should discourage specific individuals/subgroups who have a vested interest in other desktop environments from working on them. A spin by KDE guys for KDE guys is fine and good thing.

I just think that the lion's share of fedora's effort should go toward making the default spin as good as possible, since that's what we're telling our users to use. We have a (big!) userbase depending on the defaults we put out, and we should spend time on making it a good experience for them. I think we have orders of magnitude more of those users than all FAS members, for instance.

Well, they probably wouldn't all agree on which top-to-bottom integrated experience they want to have ;)

No question. Even more so since GNOME 3.0, where we have made some non-conventional UI changes.

And actually I think there are people who _don't_ want that. The guys who run nine terminals in a 3x3 square on fluxbox or whatever probably don't want a top-to-bottom-integrated experience, but they might want to configure their keyboard once in a while.

There are ways to configure the keyboard with one of the nine terminals, without using gnome-control-center or using system-config-keyboard. I certainly don't mind if a user chooses to go custom, but we have so far to go on the stock experience, I really think we need to prioritize our efforts there (of course, others may disagree, and will work on what they want).

--Ray

On Wed, 2013-03-20 at 15:12 -0400, Bill Nottingham wrote:

Still, I'd like to think we can bring it back down, and just concentrate on the simplest way to fix any logo issues in GDM.

Yeah I agree; this whole thing got seriously derailed. The last comment here https://bugzilla.gnome.org/show_bug.cgi?id=694912#c24 implies both a proposed fix and an owner.

On Thu, 2013-03-21 at 14:47 -0400, Tom Callaway wrote:

On 03/20/2013 07:55 PM, Bastien Nocera wrote:

...

Please do. After triaging the bugs and seeing that most of them are already fixed upstream where the bugzilla has triagers, per module components, inline, searchable, backtraces and images are opened in the browser.

FWIW, triaging bugs counts as addressing bugs, in my opinion.

I was asking you to strip me of my maintainerships after triaging the bugs. The reasons I listed make it impossible for me to do any work with the Red Hat bugzilla.

While I don't want to prolong the login screen thread unnecessarily, now that *hopefully* it's concluded I wanted to reply to this briefly.

On Wed, 2013-03-20 at 14:02 -0700, Adam Williamson wrote:

It was expected that you could just swap out DMs (like skvidal thinks is still the case) and everything else should just deal with it. The system-config-* tools were part of Fedora and used across all desktops - it didn't matter what desktop you ran, you used system-config-keyboard and system-config-display and so on.

It's not just about the client though - several of these tools were designed for more "enterprise" use cases, like system-config-httpd. At the moment as far as I'm aware the current direction is that gnome-control-center will be focused more on things that matter to client ("desktop") systems.

I think other desktops still think of things somewhat in those terms, but GNOME definitely doesn't: GNOME wants its own stack, almost top to bottom.

It's difficult to speak of a diverse collection of individuals as collectively "wanting" something, but I think even looking just at those who are the strongest proponents of vertical integration, this is an inaccurate characterization.

But there are real world problems that go straight down the vertical stack, and are easily re-broken if one of the components gets out of alignment (i.e. is swapped with a different package). The init system shutdown versus suspend is a perfect example (covered here http://lwn.net/Articles/520892/ ).

So no one wants a stack just to have a stack - it's about solving problems.

On 03/21/2013 03:08 PM, Bastien Nocera wrote:

On Thu, 2013-03-21 at 14:47 -0400, Tom Callaway wrote:

...

On 03/20/2013 07:55 PM, Bastien Nocera wrote:

...

Please do. After triaging the bugs and seeing that most of them are already fixed upstream where the bugzilla has triagers, per module components, inline, searchable, backtraces and images are opened in the browser.

FWIW, triaging bugs counts as addressing bugs, in my opinion.

I was asking you to strip me of my maintainerships after triaging the bugs. The reasons I listed make it impossible for me to do any work with the Red Hat bugzilla.

If you're serious about this, you should consider orphaning your Fedora packages.

~tom

== Fedora Project

I would be interested in becoming a triager, but I am not sure what to do or how to get started.

On Fri, Mar 22, 2013 at 11:43 AM, Bill Nottingham notting@redhat.comwrote:

Tom Callaway (tcallawa@redhat.com) said:

...

On 03/21/2013 03:09 PM, Ray Strode wrote:

...

It's not a good use of time compared to addressing customer issues,

doing development, and fixing bugs.

...

In the context of Fedora, user == customer. They are putting their issues into bugzilla.redhat.com.

I do not believe that you (or Bastien) or anyone is ignoring user issues with their packages. Nor am I saying that triaging every bug is always possible. That said, if you prefer to work these issues upstream, it is a necessary evil to do some triaging.

Honestly, this screams for a need for dedicated Fedora triagers, but that's tricky to do unless we have some really excited volunteers show up in the SIGs.

Bill

desktop mailing list desktop@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/desktop

On 03/22/2013 11:53 AM, Lynn Dixon wrote:

I would be interested in becoming a triager, but I am not sure what to do or how to get started.

I am also interested in starting to triage some of the Fedora desktop bugs

Generally, there is the BugZapper SIG[1] for general bug triage.

However, to focus on triaging Desktop(GNOME) bugs in fedora specifically, I have created a page with the vision to set up a plan under the Desktop pages[2] to start a triage regiment.

cheers, ryanlerch

[1] - https://fedoraproject.org/wiki/BugZappers [2] - https://fedoraproject.org/wiki/Desktop/BugTriaging

On Fri, Mar 22, 2013 at 11:43 AM, Bill Nottingham <notting@redhat.com mailto:notting@redhat.com> wrote:

Tom Callaway (tcallawa@redhat.com <mailto:tcallawa@redhat.com>) said:
> On 03/21/2013 03:09 PM, Ray Strode wrote:
> > It's not a good use of time compared to addressing customer
issues, doing development, and fixing bugs.
>
> In the context of Fedora, user == customer. They are putting their
> issues into bugzilla.redhat.com <http://bugzilla.redhat.com>.
>
> I do not believe that you (or Bastien) or anyone is ignoring
user issues
> with their packages. Nor am I saying that triaging every bug is
always
> possible. That said, if you prefer to work these issues
upstream, it is
> a necessary evil to do some triaging.

Honestly, this screams for a need for dedicated Fedora triagers,
but that's
tricky to do unless we have some really excited volunteers show up
in the
SIGs.

Bill
--
desktop mailing list
desktop@lists.fedoraproject.org
<mailto:desktop@lists.fedoraproject.org>
https://admin.fedoraproject.org/mailman/listinfo/desktop

Hey Lynn and Ryan,

----- Original Message -----

I would be interested in becoming a triager, but I am not sure what to do or how to get started.

Thanks for volunteering. Ideally the process is something like:

1) read the report carefully, and ask for clarification on any details that are vague or unclear 2) verify that it's filed against the correct component based on the description of the bug 3) search in bugzilla for any duplicates of the problem and close DUPLICATE if possible. 4) try to reproduce the problem, if possible, and then if successful, list specific step-by-step reproduction instructions for others to reproduce (unless the reporter already did this, and following their instructions worked). 5) ensure the fedora version listed on the report is the latest version of fedora the problem still happens in. 6) if the issue isn't distro specific, search upstream bugzilla for an existing report that describes the same bug 7) If found, mention the bug number as an external bug link on the page 8) If found, post a comment in the upstream bug report linking to the downstream report. 8) If not found, either ask the reporter to file upstream or file the report upstream yourself. 9) If the issue is not something we specifically want to track for a release, close the bug UPSTREAM. 10) if the issue is something we want to track for a release move the bug to ASSIGNED and it will get closed when the update is pushed out, or when the fix is applied. 11) if the issue is something we want to track for an upcoming release and it's a relatively high impact bug, add something to the CommonBugs wiki page.

Obviously depending on various constraints (time, hardware to reproduce, motivation, etc) not all of those steps will always get followed.

But in an ideal world, bug fixing isn't "worked on" downstream. The bug fixing needs to happen upstream so non-Fedora, upstream developers are available to do peer-review and chime in.

As far as what bugs to focus on? I'd say gnome-shell and gnome-settings-daemon are good places to start.

--Ray

On Mar 22, 2013 10:40 AM, "Ray Strode" rstrode@redhat.com wrote:

Hey Lynn and Ryan,

----- Original Message -----

...

I would be interested in becoming a triager, but I am not sure what to do or how to get started.

Thanks for volunteering. Ideally the process is something like:

1. read the report carefully, and ask for clarification on any details

that are vague or unclear

2. verify that it's filed against the correct component based on the

description of the bug

3. search in bugzilla for any duplicates of the problem and close

DUPLICATE if possible.

4. try to reproduce the problem, if possible, and then if successful,

list specific step-by-step reproduction instructions for others to reproduce (unless the reporter already did this, and following their instructions worked).

5. ensure the fedora version listed on the report is the latest version

of fedora the problem still happens in.

6. if the issue isn't distro specific, search upstream bugzilla for an

existing report that describes the same bug

7. If found, mention the bug number as an external bug link on the page
8. If found, post a comment in the upstream bug report linking to the

downstream report.

8. If not found, either ask the reporter to file upstream or file the

report upstream yourself.

9. If the issue is not something we specifically want to track for a

release, close the bug UPSTREAM.

10. if the issue is something we want to track for a release move the bug

to ASSIGNED and it will get closed when the update is pushed out, or when the fix is applied.

11. if the issue is something we want to track for an upcoming release

and it's a relatively high impact bug, add something to the CommonBugs wiki page.

Obviously depending on various constraints (time, hardware to reproduce,

motivation, etc) not all of those steps will always get followed.

But in an ideal world, bug fixing isn't "worked on" downstream. The bug

fixing needs to happen upstream so non-Fedora, upstream developers are available to do peer-review and chime in.

As far as what bugs to focus on? I'd say gnome-shell and

gnome-settings-daemon are good places to start.

--Ray

desktop mailing list desktop@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/desktop

Sorry I didn't trim the quotes above; I could try to help out for the time being, before classes start up. This seems like it could get to be a lot of work.

Richard

On Fri, 2013-03-22 at 17:12 +0000, "Jóhann B. Guðmundsson" wrote:

On 03/22/2013 04:07 PM, Ryan Lerch wrote:

...

Generally, there is the BugZapper SIG[1] for general bug triage.

No there is not.

That wiki page only exist because me or Adam have not found the time to remove it.

I don't necessarily want to remove it until we've sucked the juice from it, but we should certainly edit it in some way to make clear that BZ as an independent group doesn't really exist any more.

On Fri, 2013-03-22 at 12:42 -0700, Joe Wulf wrote:

Adam,

Wondering, as an organizational issue, if bugzappers has died out because it is just too painful a 'job' to keep up with. Sure people would come with high energy in the beginning, but their enthusiasm wanes when the rewards don't keep up with the level of effort to accomplish it diligently and ever-so-successfully. So many, I am sure, have other interests, and life in general can just get in the way of hobbyists. Your suggestion about having a person who's JOB it is to accomplish triaging, might really be the right course of action. They'll be paid to stick around and invest themselves in making Fedora better.

As Bill said, RH has paid people to triage in the past (mcepl comes to mind), and we may want to do so again. So far as the volunteer side goes, your theory may well be correct; it's actually quite hard to tell. One of the issues with triaging is that it's less of a communal thing than you might initially imagine: you tend to wind up with a group of people doing entirely separate work. If I'm triaging X bugs, and you're triaging Firefox bugs, and Anna, Bob and Charlotte are all triaging other components, we don't actually have much to talk to each other about. So you don't get that catalyst of discussion that other groups get, people just wind up ploughing their solitary fields. We often had weekly meetings where we'd talk about the white elephant project of producing some stats on triage (another thing that never quite got done) and then...just leave, because there wasn't really anything else to say.

If you managed to hit some kind of critical mass where you could have five people working on some components, at least each group of five people could talk to each other. But we never got there.

On 03/22/2013 07:12 PM, Adam Williamson wrote:

On Fri, 2013-03-22 at 17:12 +0000, "Jóhann B. Guðmundsson" wrote:

...

On 03/22/2013 04:07 PM, Ryan Lerch wrote:

...

Generally, there is the BugZapper SIG[1] for general bug triage.

No there is not.

That wiki page only exist because me or Adam have not found the time to remove it.

I don't necessarily want to remove it until we've sucked the juice from it, but we should certainly edit it in some way to make clear that BZ as an independent group doesn't really exist any more.

I'm not sure what juice you are referring to It had been tried and revived three times before the current failure I think all the juice in it has been tried and tested.

It's time to merge the reporter and triagers into one process ( as had been planned in the long run ) which in my theory at least should improve reporting since the same individuals will be standing on both sides.

There will be other problems to solve once we have merged these two processes but we will cross that bridge if and when we get there.

JBG

On Fri, 2013-03-22 at 21:39 +0000, "Jóhann B. Guðmundsson" wrote:

On 03/22/2013 07:12 PM, Adam Williamson wrote:

...

On Fri, 2013-03-22 at 17:12 +0000, "Jóhann B. Guðmundsson" wrote:

...

On 03/22/2013 04:07 PM, Ryan Lerch wrote:

...

Generally, there is the BugZapper SIG[1] for general bug triage.

No there is not.

That wiki page only exist because me or Adam have not found the time to remove it.

I don't necessarily want to remove it until we've sucked the juice from it, but we should certainly edit it in some way to make clear that BZ as an independent group doesn't really exist any more.

I'm not sure what juice you are referring to It had been tried and revived three times before the current failure I think all the juice in it has been tried and tested.

Just the actual content relating to triaging that we have on the existing BZ pages, I don't want to lose that. Of course it's always in the page history, but eh.

On 03/22/2013 11:18 PM, Adam Williamson wrote:

On Fri, 2013-03-22 at 17:06 -0400, Lynn Dixon wrote:

...

Like I mentioned in the previous thread, I am more than happy to help out in any way. I have always wanted to get involved in Fedora, and I am currently an Ambassador. I would like to do more "hands-on" work in the distro itself, but I have no programming experience. Bug triage seems like a cool thing to do for someone like me.

Would it be possible to start an ember on BugZappers? What can I do to help?

Right now, as a practical consideration, you can still read all the instructional stuff in the BugZappers wiki space, pick a component, and put a lot of it into practice. It doesn't really depend on anything other than Bugzilla being there. So far as getting triaging going as an organized project again (a sub-project of QA this time), there was a thread started on the test@ list I think a few weeks back by someone else who is interested; unfortunately I wasn't able to contribute a constructive reply yet as I've been busy with F19 stuff :( But that makes two of you who are interested in trying to drive triage in some way. The thread I'm thinking of is https://lists.fedoraproject.org/pipermail/test/2013-February/114008.html

• William got in touch with me directly before I encouraged him to post

to the list, and he's definitely interested in getting triage going again, so I'm sure he'd be happy to hear from someone else who's interested. Thanks!

it's better the individuals interested in triaging just contact the maintainer(s) of the relevant components they are interested in triaging for directly, get their permission and start working with them rather then trying yet again to start another official QA effort which will just die in $next-release JBG

On Mon, 2013-03-25 at 10:24 -0400, Máirín Duffy wrote:

On 03/22/2013 05:47 PM, "Jóhann B. Guðmundsson" wrote:

...

And as has been proven in the original thread the Gnome maintainers dont deal with bugs et all filed in RH bugzilla and consider it more of a burden having to do so.

Why do we continue to have components under Fedora in our bugzilla that the maintainers of said components don't watch?

Can we remove them from bugzilla so we don't give a false impression?

Please, no; a) the whole 'we don't watch downstream' thing seems somewhat overplayed to me (for at least some components), but more importantly b), we can't handle blocker status in an external BZ. We need to use RH Bugzilla for tracking blocker bugs.

On 03/25/2013 11:53 AM, "Jóhann B. Guðmundsson" wrote:

The components have to stay the number of maintainers need to be increased behind relevant component(s) and or those maintainers that are unwilling or otherwise unable to due their due diligence in their distribution maintainership for one reason or another be replaced with once that are.

I believe they *are* doing their due diligence, it's just that they're upstream. I've fallen into the trap of reporting certain things downstream (even outside of GNOME) and finding out later that it's better to report upstream for certain components.

We can't expect our users to know which components in Fedora's bugzilla are *real* and which you're better off not bothering with.

Maybe if we can auto-detect a maintainer in our downstream is the same as a maintainer upstream, we pre-fill the bug template for their component with 'don't use this, file at bugzilla.upstream.org'? And then if the downstream maintainer ever changes, reactivating the component?

~m

On 03/25/2013 03:58 PM, Máirín Duffy wrote:

On 03/25/2013 11:53 AM, "Jóhann B. Guðmundsson" wrote:

...

The components have to stay the number of maintainers need to be increased behind relevant component(s) and or those maintainers that are unwilling or otherwise unable to due their due diligence in their distribution maintainership for one reason or another be replaced with once that are.

I believe they *are* doing their due diligence, it's just that they're upstream. I've fallen into the trap of reporting certain things downstream (even outside of GNOME) and finding out later that it's better to report upstream for certain components.

We can't expect our users to know which components in Fedora's bugzilla are *real* and which you're better off not bothering with.

Maybe if we can auto-detect a maintainer in our downstream is the same as a maintainer upstream, we pre-fill the bug template for their component with 'don't use this, file at bugzilla.upstream.org'? And then if the downstream maintainer ever changes, reactivating the component?

As I mentioned years ago that this can be solved technically as in the problem for reporters having to have gazillion upstream bugzilla accounts or upstream maintainers having gazillion downstream bugzilla accounts ( same problem for both parties involved, you just need bi-directional communication between two or more bugzilla instances then all the downstream reports could be handle by triagers who would forward it upstream via flag in bugzilla and upstream do the same ) there even is a an RFE somewhere deep in the closet of the RH Bugzilla maintainers and I think James actually had one of them look into it with mozilla developers but I think that guy has quit RH and we cant do nothing about hacking bugzilla since it's not our own instance but we share it with RHEL which arguably we should not be doing in the first place since it goes against our foundation of "freedom" .

JBG