How about thinking about this.
Garland has had its time, and it starts to really hurt my eye.
Sure going for a recolorable theme is a given necessity.

I opt for Pixture, as I see it is very popular, and also because it looks "Joomly" (this may be a reason for people to NOT use it, I know ;) )
In is actual edition it has layout tables, but I'm working on this flaw with someone. Sure I'm not totally out of hunger for personal fame, but more importantly this would get away from the "Drupal looks like Garland" opinion.

Maybe this is already discussed and I'm only missing it?

Maybe collect candidates?

#40 Central_Administration.jpg55.15 KBCalebD


eigentor’s picture

Title: New core theme » New core theme for D7
eigentor’s picture

Title: New core theme for D7 » New default theme for D7
bensnyder’s picture

Amen to a theme change... garland hurts my eyes too. I think the simple look puts a huge damper on people's first impressions.... And first impressions are everything to non-developers.

eigentor’s picture

After some discussion in the channel, here is a plan how this might be done:

  1. Collect a lot of candidates of themes to go into core
  2. pick the 10 best
  3. Vote on the remaining and by this, pick 3 (Garland has to stay in)
  4. Of the three new ones, pick 1 for default

Candidates should offer:

  • physical existence (no still to-be-built themes)
  • modern, dynamic design
  • maybe have proven themselves by already being popular
  • enough space for sidebar blocks (but it might be two sidebars on the right like in e.g. Deco, or there might be a clever concept to do it otherwise)
  • have settings for flexible and fixed width
  • being recolorable with color menu

It is clear that filling all these features may mean quite some work for
the themes, but it ultimately applies only for those who get in, e.g. Themes
that are not recolorable need an estimate how hard it would be to change this.

btopro’s picture

I'm not sure on the status of including the themeapi in D7 (or something like it) but I think something lacking in current packaged theme's that I've seen is the ability to modify the top header image. They all allow for "change the icon" but changing the icon in garland doesn't make it NOT feel like garland. Changing the color and then having a matching header will at least start to make it move away from the every drupal site is the same feeling. It also can help make those that don't spend a lot of time themeing make their's look at least somewhat unique.

My unit was looking into making something like this in-house because it would help with rapid prototyping of themes (for courses we offer but still I'm sure it's a common issue when developing)


eigentor’s picture

Wonderful idea. Webchick got a completely new lookt to her site just by changing the header (sure with a very special image in that ;) ).

You could talk to Ximo, he is doing the Icon Module as a SoC Project, so maybe no need for double efforts.

kika’s picture

I'd back up a bit and go to a higher level.

Instead picking up new Drupal stock theme, we'd need to gather some issues beforehand (similar what Gabor did for WYSIWYG-in-core debate

1. What are the the drawbacks of Garland theme?
"Eye-hurting" seems to be too vague -- please be more specific. Some good reasons are already provided above (faulty tabs, issues as admin theme) and also several theme-specific points are made in recent Drupal usability studies. Let's collect these.

2. What requeirements the new theme needs to fill?
There are possibly several categorys: aestetics, functionality, ability to customize, themeability, recent CSS developments (CSS reset, CSS frameworks etc.

A BoF session in Hungary perhaps?

rszrama’s picture

I love pixture and use it personally and modified on but I would vote against it being a core theme. Its HTML is a weird blend of tables/divs... difficult to customize, esp. that black bar under the header. : P

It's nice, but not that nice, if you get my drift.

eigentor’s picture

As for pixture: we are having the code fixed right now, so this should not be an issue. Different are the comments from elv in usability group, which weigh heavier for me in it being nice but average and not setting Drupal apart from other CMSes.

But let's not discuss this in depth. Rather let's collect more canditates, 'cause I know there are more. I also propose Litejazz and Aberdeen,
Newsflash is also nice. Personally I also like ported themes like Glossy Blue and Blix, but as a project one should probably avoid non-generic themes.

It may also be thought about having Zen as one of the installed themes (not the default one, because it lacks visual appeal in its core version) because it is so popular.

Collecting reasons and analyze appears a very good idea to me. I was deliberately keeping this issue straight, because of recommendations from catch and addi. The creation and selection of Garland must have been an incredibly painful process with advaned bike shed syndrom. So I wanted to keep this as pragmatic as possible.

As for Garland: I loved it when I first saw it. But: imagine you got a favourite dish. You love it. But then you go to any pub and any restaurant. And the only thing they serve is that dish. And often with strange additions - they throw a banana or a meatball in...

That's how I feel about soooo many people using garland for their drupal site, and maybe in a rush of uncontrolled creativity change the colors or throw a logo in, Shudder. All this would not be so bad if it weren't thousands.

So it is burned out, worn out, shot dead, overdone, misused, done. ;)

Also: you make a new version of Drupal. Keeping in mind that especially for non-technical people the interface IS the product. So new theme = new product. Marketing Strategy.

btopro’s picture

Priority: Normal » Critical

Responding to several ideas.... Definitely need to compile requirements for a new theme as well as justification for why to go about replacing the current one.

Icon project looks interesting but seems confined to admin GUIs / page headings (at least from my early interpretation). Good start but I'm thinking bigger

Zen has traditionally *not* wanted to be a core theme so that they can maintain it separately / independent of module releases (to my understanding). It also is "clean" but that's something most of us probably appreciate, not joe shmoe user and (more importantly) joe shmoe user's boss / manager, the one who is paying the bills and making the financial call on moving to Drupal or not. As a result I *heavily* agree with your last point about Marketing Strategy. Dealing with Artists all day I've come to realize all too well how critical it is to them that the interface be spotless because as you said, to non-technical people the interface IS the product.

That being said maybe Drupal (out of the box) needs something really eye catching and sexy instead of cool, developer friendly, flexible out of the box, and "color-change-able". WE care about those kinds of things but we're (unfortunately) not the people that are the hard sell. Joomla is interesting to a lot of people (I think) because of the (imho) stupid little things like icons, theme, etc. We need to target the user w/ this default theme or maybe even provide them with several right up front and allow them to choose during installation. While it's not what I necessarily want, I see faaaar too many Garland / Minelli's out there to think that it isn't just coincidence that the first thing a lot of users experience is the only.

dahacouk’s picture

I do really like the simplicity of Pixture. However - and I keep on banging on about this issue - the left hand side bar doesn't stretch to fit text. It really should...

Set your minimum font size to 30 and compare:

Pixture is a mess.

HTML Dog degrades beautifully. In fact HTML Dog doesn't even degrade at all. That's what we should be aiming for in the new default theme. Notice how the left side bar expands with the text! And not a table in the house! And totally fluid.

Cheers Daniel

Beauty through design. Design through beauty.

nbz’s picture

We need a good range of themes:

1. basic uncomplicated framework theme - not default. Something like framework, but I guess one of the existing themes would be just as good (but with a better default choice of colours)?

2. Something simple but flashy for blog like sites - default. This is where Garland currently lies and so does pixture. the two have many similarities - but IMO there needs to more care for recolouring - neither are easy to get good looking custom colours for (this could be an issue with colouring in general, or just me and my lack of colour coordination.)

A bonus of pixture over garland is that having more than a couple/three primary links does not look crowded in Pixture.

3. Potentially something with loads of regions for more complicated sites. It does not have to be flashy, but have good white space usage.

Since this is mroe than abotu Garland, maybe the component should be "theme"?

eigentor’s picture

We should not concentrate too much on discussing the pros and cons of pixture IMHO.
More Theme suggestions? Maybe there are Themes far better than that or that can be made far better ...

dmitrig01’s picture

I don't like the idea of "has to be in existence."
I have teamed up with a designer who could potentially design a new core theme. However, he's a bit busy ATM and seeing that it's likely the code thaw will last quite a bit longer, I'd say it's ok if they're created sometime in the next few months

eigentor’s picture

In Existence extended: We could put that point of time a bit into the future.

But, say, there will be code freeze in December - core themes should be chosen then, I guess. And say we need two months for the process of deciding which should go in and refining them - So say October 15 would be a deadline for themes to be there.

Still would be a question how strictly code freeze affects Themes.

But basically at the start of a decision process - whenever this may be - any candidate for core Themes has to be there and working IMHO.

Personally I'd like to see themes designed especially for that purpose, would even like to hold a competition for that. Some experienced persons on #drupal advised against it ;), and maybe it would indeed complicate matters. Anyway - results need to be there at a moment still to be defined to get this really done and avoid the trap of endless discussions.

webchick’s picture

Garland was committed October 29, 2006. The code freeze for Drupal 5 was September 1, 2006. I think it's safe to assume that themes do not count as code freeze material, although we'd need it done by the first beta, for sure.

Pasqualle’s picture

I would like to see a new admin theme for D7..

eigentor’s picture

Dedicated admin theme is sure good - but is another issue:

catch’s picture

It'd be great to have an extra theme in core (or more accurately at least one old one, specifically pushbutton, replaced). That's a different argument to what the default theme should be. Getting a theme core-worthy is a lot of work in itself, changing the default requires also retheming the installer, update.php, maintenance - which are intimately tied to Garland, alongside the inevitable long argument about which should be the default.

Also, core now supports css only themes, it would be great to work on the default page.tpl.php, which iirc was only changed from a table based layout at the last minute during D6, and get a CSS only theme in based on that.

alexanderpas’s picture

+1 for css only default layout!

ica’s picture

Good one Kika! Konigi sums up,
the ways to go for elegant and modular theme framework directions
imo- boilerplate/blueprint seems like most promising option

Also as a conceptual thinking, what would be great is that something like Tripoli framework's selector and layout examples

> turned into a core to a 'theme module' -or current theme settings improved further- to choose plain .css files (as themes), and call $variables, blocks and views on the go (as setter of 'what' goes where and 'when') -and scrap the theme.tpl.php files alltogether.. this seems a truly 'modular' approach.

one obstacle, -afaik- contemplate.module and views.module has overlapping functionality for control of 'what' goes where and 'when' which can turned to advantage looking into all with a fresh approach somehow I presume.

Just a conceptual idea, not sure about module feasibility though as i am not a php coder.

elvis2’s picture

I agree with #8, ditch the tables and stick to css.

After hearing/reading much about the UI tests in the universities, I definitely agree that some admin standard theme is needed. If no theme, then an extra tab on node related pages that allow you to view the content as a non-admin user.

If a new admin theme is aimed for, something iconic would be helpful. I wrote a post on few options ( Personally, I think the "vertical tabs" module has some "core" potential and should be converted to an API for other modules to hook into.

Whatever is decided, a well crafted backend UI would definitely help Drupal adoption :)

BioALIEN’s picture

I don't see the need to keep bluemarine if it is to remain. However, I am all for remixing it and reviving it as a CSS only. Having said that, I think core is sorely missing a definitive framework theme. There are many around such as zen, simple, basic, framework, and the list goes on. If we champion one of these (or convert one of the existing ones in core) to become the ultimate base theme, then this can only be a good thing.

This is what I'd like to see in core, more than a new shiny theme to replace garland. I would rather designers take this base theme further by building on it than to recolour a complex standard theme such a garland.

mustafau’s picture

Bluemarine was converted to tableless layout 3 months ago. See #200685: GHOP #92: Convert bluemarine to a tableless layout.

Robin Monks’s picture

I think a major problem is no matter what new theme goes it, it will look tired after the first 100 sites you see using it ;)

I think a much better solution would be to make sure the next default theme would have at least 50 different color schemes, and then during wizard install have the user choose one, with the default being *random*. This should keep the generic drupal theme eyesore from happening :)


dvessel’s picture

Here are a few things I wouldn't mind seeing:

1.) Flexible and light weight. Similar to Zen when it was considered for core. It's too complicated now.

2.) A library of classes within the style sheet for easy positioning of the layout with the markup to match. Think Blue Print but I would want to stay away from it being solely fixed width.

3.) A combination of fixed width and fluid depending on the classes it uses. Possibly implement theme api so it can be chosen through the admin UI.

4.) Remove the dependancy of Garland and Minnelli for the install, updates and maintenance pages. It's currently hardcoded in an odd way. It seems there are a lot of people who modify Garland directly and that can end up interfering with installs and updates. Have this new theme fill that role.

5.) Have a few sample sub-themes based on the new theme and have it so it's clear that themers can copy it and use it as their own with clear documentation.

6.) Color module support only on a sub-theme if we decided it's really needed. Having it on a base theme would complicate it for everything based on it.

I don't think it has to be the prettiest theme. Garland is still fine IMO but it is difficult to work with. So, it should be the opposite, a basic backbone to let themers do their jobs more efficiently.

There's a faux absolute positioning technique that looks promising. I've played with it briefly. If that can be made into a library of grids, it'd be a godsend as it can be applied in a fluid or fixed layout.

And thanks for the link kika, that roundup of frameworks looks great. It would be a matter of adapting or including them if licensing permits. We need a solid foundation. Everything else is icing on the cake.

Robin Monks’s picture

#26 : +1 !!!

drawk’s picture

Why not select 3 or 4 good quality themes and offer a selection at the time of installation?

Myself I'm generally going to install zen or foundation as the first order of business and build up from there, or start from scratch, but I'm sensitive to the fact that Garland was a big step up from Bluemarine in terms of 'first impact' and that the initial visual experience really does colour most end-users' immediate sense of the platform. Lets pick a few good ones, give the user the choice on installation (with a short description of the themes and the screenshot) via a radio button, and point them to a location where they can scavenge for other themes should none of the core-packaged ones fit their needs.

Let's kill this idea of a default theme, because we will all never agree and we all have very different needs. Much easier to agree on a limited set that meet some defined criteria to be included in the core package.

chx’s picture

+100 on offering a selection of themes on installation. That would help the "all new Drupal sites are theme X". I would be extremely careful with dropping Garland. It takes an ungodly amount of work to make a core worthy theme. I did trust Steef and Steven to create something like that. Who can we trust with such an important task now? They paid attention to such details I did not know you needed to pay attention to...

BioALIEN’s picture

Garland is here to stay. But we do need a framework theme to be included in core for designers interested in making their own theme.

dvessel at #26 expanded on the point I was making in #23 about having the ultimate base theme. Remember the example module that used to ship with Drupal to get developers startd? We need the designer equivalent to that :)

RobLoach’s picture

In my eyes, the contenders are RootCandy (slick administration interface) and BluePrint (slick CSS framework). Switching the default theme completely might be a bad idea right now though because Minelli and Garland have the Drupal recognition. When people see Garland, they think Drupal. Maybe as an alternative, we could add a really slick theme to core, and then switch to it in Drupal 8?

Gurpartap Singh’s picture

Component: Garland theme » theme system
eigentor’s picture

very good points from dvessel.
Also chx remark should be considered: giving the user say three themes to choose from on install would threefould the optical dinner...

Given the fact that an admin theme in core (being different from the front end theme) is another plan that to me would be a good one (though this is surely another issue) - this would reduce the effort to look into admin-section issues for the front end theme greatly.

They would be what they should: front end, facing the world.

lut4rp’s picture

theme choose option - ftw!

catch’s picture

Theme selection is a good idea - opened a separate issue for that #305324: Add theme selection to installer

lilou’s picture

Jeff Burnz’s picture


dropcube’s picture


dvessel’s picture

This is off-topic but it would be nice if we had a system of "tasks" that could be used everywhere, not just the installer. The install should be as simple as possible then it can be handed off to a list of tasks for further configuration, selecting a theme being one of them. The necessary steps for installation and the optional steps could be styled differently letting the end user know that certain steps are optional.

If we did add a theme selection screen for installation, it would look like another mandatory step. Breaking it apart would keep the required steps to a minimum letting the user choose to further configure their site if they decide to do so.

Then later on, other modules can offer their own tasks when they are installed to get people started on the proper setup. Maybe a default block with a list of links and instructions with a "clear" link so it can be flushed quickly. Sort of a built-in to-do list. :p Everything is so spread, it would provide another tool for guidance.

CalebD’s picture

55.15 KB

+1 for dvesel's idea, though it belongs in its own feature request.

For example, when you install SharePoint, on the front page of the "Central Administration" is a list of tasks to complete that are all optional, but recommended for all installations. See attached (shamelessly ripped from some site via Google Search).

RobLoach’s picture

Again, RootCandy does that sort of thing. It gives you a central interface with a bunch of tasks you can do....

Worth considering. Garland screams Drupal though, which is good.

dropcube’s picture

@dvessel: I agree. To achieve things like this, we should organize and improve the installer. I have created a related issue: #316916: Create install directory with multiple files and modularized code

Xano’s picture

In response to nbz in #12:

1. basic uncomplicated framework theme - not default. Something like framework, but I guess one of the existing themes would be just as good (but with a better default choice of colours)?

I don't know about this one. Making a base/framework theme colourable sounds like overkill to me. It's not a base theme for nothing.

2. Something simple but flashy for blog like sites - default. This is where Garland currently lies and so does pixture. the two have many similarities - but IMO there needs to more care for recolouring - neither are easy to get good looking custom colours for (this could be an issue with colouring in general, or just me and my lack of colour coordination.)

I think it's your (and mine, for that matter) lack of colour coordination. Or perhaps color.module, but certainly not the theme itself.

A bonus of pixture over garland is that having more than a couple/three primary links does not look crowded in Pixture.

My problem with Pixture is that it looks big; big bars, blocks that have a very distinct background. All this makes the theme look busy. Too busy for me.

3. Potentially something with loads of regions for more complicated sites. It does not have to be flashy, but have good white space usage.

Good idea, but not too much. I think a three column layout, with a header and a footer and an extra region before the actual content is good enough.

But, enough bullshit. We're running out of time, since it's October already. We need to agree on some requirements soon and let people with good knowledge of the available themes take a look at which themes match these requirements. My short list of requirements:

  1. Recolourable.
  2. Three-column layout that degrades gracefully if one or more of those columns aren't used. It should be possible to put the page content in any column not just the center one.
  3. CSS, CSS and CSS.
  4. No fancy pictures of trees and birds and photographs. (which rules out a lot of contrib themes).
eigentor’s picture

To give this issue a chance to ever succeed, I propose to take a different path. Today It struck me - somebody should be payed to do this. Like with d.o. redesign otherwise it will never be done. But how?

O.K., I thought I had a brilliant idea - run a theme contest. Talked to Amazon and found out the idea might be brilliant, but it is not the least new. He said dozens of people have approached the association in this matter. And to them it is not a priority. As sad as I find this, it is also understandable, cause money is limited.

So we gotta do fundraising ourselves. So here my proposal:

1. Run a theming contest for new Drupal Core themes.
2. Award first, second and third price.
3. A budget of 5000 Euro should be a solid starting point
4. Advertise and plan this initiative very well, there is no hurry
5. The new default theme may be elected of this, the winner obliges himself to do all the work that is needed to make it default-ready

The proposal should be out till the end of the year.

Anyone interested in this, please contact me. People experienced in fundraising much preferred ;) Also designers and official Drupal people for building a Jury are welcome. This will need some energy, but I am convinced, If we really pull this off, will be of enormous benefit for Drupal as a project.

A side effect - lots of great GPL themes for Drupal!

JohnAlbin’s picture

Thomas, you're not the first to think of a theming contest. Unfortunately, no one has pulled it off yet. Although, I'm not sure how hard anyone tried.

At this late date, we need some skunkworks projects, I think.

webchick’s picture

I was talking the other night to beeradb about my vision on how this could happen. This is based on a discussion I had awhile back with stephthegeek, John Albin, yoroy, and some other theming folks in #drupal. Their consensus on what is needed is basically dvessel's post earlier in the thread: many themes to choose from, easy to extend base theme with clean starting markup.

Here's how I think we could get there:

1. A group of Known Knowledgable Themers gets together and evaluates the pros/cons of all the various frameworks and chooses "the one" to go with. This is similar to how the selection of jQuery as Drupal's JS library went. The ideal place for this to be coordinated would be the Theme Developers Working Group. Whoever heads this up, timebox the discussion at X weeks to give a firm deadline for when consensus must be reached (but leave enough time for people to comment).

2. Once "The One" is chosen, we work on making that into core (which means replacing all of Drupal core's default output to match it). It should be emphasized that this is non-trivial. The chosen theme must be well-coded, well-documented, conform to coding and web standards, etc. etc. Therefore, I am completely less than jazzed about the idea of trying to do this more than approximately once. I'm very -1 to the idea of "Let's pick the X prettiest looking themes in contrib and just jam them into core." -- there is far, far more to a good solid theme than how pretty it is.

3. Now that we have our flexible, easy to extend base theme in core, it's time to start a theming contest. This contest revolves around creating sub-themes for The One True Core Base Theme. Pick the best X of these and go to town. These are /much/ easier to review and get into core en masse, because they will probably be changing CSS only, or if they do override markup it'll be very short snippets.


$ cvs rm themes/pushbutton
$ cvs rm themes/bluemarine
$ cvs rm themes/chameleon
$ cvs commit -m "So long, ugly table-based themes!"

5. Profit! ;)

At the same time, I would love to see efforts to make things like swappable image headers, a better color module, etc. etc. to take the customization efforts started in Garland and ramp them up a few levels. Those could be used in conjunction with a solid base theme to offer more flexibility to people who don't know CSS, and could be worked on in parallel.

This approach does the "design by committee" part at the level that it makes sense: the architecture. When talking architecture, you *want* the most knowledgeable people on the subject to each throw in an opinion of how things ought to be. Where design by committee falls apart is in doing actual visual design. And for that, you do a theme contest. Let the designers thrive in their creativity using the solid framework that core provides.

So. Good idea? Bad idea? Ridiculously awful idea? Code freeze is still undetermined, but I wouldn't underestimate how long this is going to take, and it should've ideally been started months back. We'll need a big push. But, admit it! Doesn't the thought of no more Pushbutton fill your heart with glee? ;) Who's going to step up to drive this?

merlinofchaos’s picture

Holy crap webchick, can I +1 that a whole lot?

merlinofchaos’s picture

Let me comment. I want something sleeker than zen as a base theme. And by sleek, I think I mean we don't really want a base theme, because we have templates in core, I don't think we need a whole lot in a base theme, but the real work is the core markup and the core .css so that the 'base theme' we use is, in fact, just Drupal.

So the basic idea here is right, except that what we're looking for, IMO, is something where we have a 'base theme' that is nothing more than a couple of stylesheets to put everything in the right place and Drupal basically works. It looks simple and that's fine. We try to simplify all the /misc/*.css files, maybe moving some of the css into our new base theme (menu css, anyone?) and work from there.

JohnAlbin’s picture

I agree with Earl here. Zen tries to do a lot more than what we need in Drupal core. I've got a development theme that is just a single file: core/

name = Core
description = Uses all of Drupal core’s default template files.
core = 6.x
engine = phptemplate

If you use that, its pretty easy to see what we're starting with. :-)

IMO, we should only need to lightly modify the existing tpl files. Maybe modify the core CSS more heavily; semantic classes? And, if we put a page layout method in Drupal core's stylesheets (debatable), we should make sure any required tpl changes don't hamstring other themes and its CSS is easy to rip out (just like the rest of the CSS in core).

We should definitely get started very soon with the discussions. I, like stephthegeek, talked to beeradb today in IRC. We should strike while the iron is hot!

alexanderpas’s picture

Holy crap webchick, can I +1 that a whole lot?

me too! me too! +1 from me too!

oh and bytheway i think that the core theme should be called... Plain HTML.
we can still use CSS for those one/two/three column layout with heaader and footer, but that's it!

maybe, if we go this way, we could even make liquid/static an (optional) theme option instead of a subtheme ;)
and if we're doing that anyway, we could open some more options to theme developpers.

sort of a zen road, but very different.

Drupal aims to provide a slim, powerful core that can be readily extended through custom modules.

for themes:

Drupal aims to provide a slim, powerful theme that can be readily extended through custom themes.

Jeff Burnz’s picture

I wrote a review of each starter theme just recently, you can view it here, which should help get some up to speed quickly on what's available. You can also test drive them here.

I'm a big -1 to another pretty/recolorable theme in core, esp not Pixture, being the maintainer I know its pitfalls, even Pixture Reloaded (the CSS version) is nowhere near slick enough (looking).

I'd like to see a sleek yet flexible framework that's easy to get started with. The subtheme concept has a lot of merit and this is the approach I have taken with my more recent theming efforts.

alexanderpas’s picture

I'm a big -1 to another pretty/recolorable theme in core

I'm not completely, as long as it is not the core theme, just a "contributed" (default) subtheme of the core theme, that just happens to come with core.

Garland screams drupal, but seriously, it doesn't read drupalese.

Zen is silent, but is able to read some drupalese

Let's make core better!

dvessel’s picture

Webchick, that sounds great!

Merlinofchaos, if we can fit positioning classes into the templates we already have then that'd be awesome but I think it should be secondary to having a layout framework that's easy to use. If no one objects, swap any default templates to fit whatever we need now if necessary. I understand that there aren't many CSS based themes so it shouldn't do much harm I hope. I'm only aware of my own which never got a proper release. (sorry webchick)

IMO, a layout framework is the most important thing here. This is not directly related but adding classes programatically can fit into this too to calculate widths automatically. I've already done this for a job (with blueprint) which worked out very well.

dvessel’s picture

2. Once "The One" is chosen, we work on making that into core (which means replacing all of Drupal core's default output to match it)...

Ah, well.. Webchick already mentions this. +999 :)

quicksketch’s picture

It seems to me we can get the most immediate benefit by making the Zen base theme our default HTML/CSS (this would mean adding the layout.css file as part of core) then we could immediately use all the zen sub-themes as candidates for Drupal 7 core themes. I think after several years of refinement, we're not going to get a better starting point than Zen for our default HTML/CSS. Of course I agree with John Albin and merlinofchaos that we should trim the excess from Zen though (the block-hover links, the theme settings, possibly the specialized tabs.css, etc).

Jeff Burnz’s picture

Zen is a great theme, granted, but the layout options are limited (out of the box), if we could extend it with an easy-to-use set of layout classes that would be great - which is relatively easy to achieve. Personally I wouldn't want to see this as a theme setting (i.e not global), just some classes I can use to set the layout for that particular page template. By layout classes I mean something along the lines of YUI templates, such as how the Genesis theme does it. Yes, no - not what we're after?

Personally I dont like some of the HTML in Zen, the positioning and source order of the nav elements is, for me, a stumbling block. Maybe I just havent used it enough to become highly familiar with extensive customisation techniques, but I see a lot of users having problems with this. For me its not a huge problem, but for many others that appears to be the case.

-1 for a blueprint-like set of grids based classes, adds complexity for simple things like collapsible regions (think about when you want a lot of regions or something more complex than left/centre/right), are difficult for many to get-their-head-around, havent yet seen one that support liquid layouts well and imho would be rather useless without some sort of grid builder/template generator aka the ATCK theme concept.

Sub unit classes like YUI nesting grids can be extremely useful, however I no longer bother and turn to Mini Panels for this sort of thing. I like just having a lot of "big" regions I can drop Mini Panels into. This gets me thinking about who this effort is pitched at - do we give this enough grunt to be taken seriously by advanced designers or strip it right back to bare bones, after all, designers tend to have their own ideas, perhaps it needs the flexibility to support both, which tends to indicate a kiss approach.

So yeah, +1 for a Zen like approach, some of the guff stripped out but more flexibility (read layout options) that are easy to use out of the box. In my opinion this is *all* about layout options - ease of use & flexibility top of mind.

Is a skunkwork group forming, if so where abouts - in the group Webchick pointed towards?

ica’s picture

Shifting a paradigm

- Every individual or community builds paradigms as they are in live and do things in and keep doing of a period of time ~ and there are pros and cons of it, important thing is to adapt and change when is necessary
- It is also true for open source projects
sometimes its useful to think starting with forgetting everything you have and done

a fresh thinking
- what we have to we define as 'content' needs to be composed certain com-positions into a screen?
- they have to appear and disappear when it necessary

- type of nodes with specific fields -which is grouped in different bunches depending of their content types
- blocks - definition of a block has a bit loose meaning often used depending of context as;
~ content type areas outside 'main content' as columns
~ boxes that has functionality i.e admin links, login box, chatbox etc.
~ small 'content areas' i.e weather, blogroll etc.
~ content teasers i.e any newspaper/magazine main pages has or the content blocks you create with Drupal Panels or the Views

all above are chunks of bricks to be assign to structured yet flexible areas and appear and disappear according to
~ content relevancy of the A content itself
~ content relevancy of the other related content
~ relevancy of time
~ relevancy or preferences of/to the user

flexible and areas a challenging/difficult two concepts to match

it is restricting to have a structure of 'fixed' right 'side boxes' and main content ~ yet there has to be main- left - right top -bottom of content types in any case

so the solution to treat a screen as canvas made of grids or bricks of areas defined on instances not of restricted and 'fixed left-right- top bottom- yet they will be places on left-middle-right-bottom on amounts and relevancies i mentioned above.

Nearest example to this is the Yahoo Grids -and the builder

which can give a tool for start - but what is missing from it the instances instances of time, -content type, -relevancy to one content type to another and -admin preference of timing and area and user preference of timing and area

instead instances are usually/sometimes provided in a dispersed fashion by different modules and there is no core approach to it - the best example of approach is the date.module and the views.module combination in terms of Drupal


A fresh fundamental approach to questions of WHEN and WHERE and WHAT tied up tightly in the core rather than 'modules'
and a 'fresh' paradigm based on
WHERE? is provided by a Grid Builder (.ie YUI Grid Builder) as part of the core standard theming
WHEN? is provided to find a solution of instances dealt in the core tied up with the theme builder
WHAT? is already defined by core modules (i.e blog, forum, the CCK types and mini block contents ie weather, blogroll etc. but not tied up with core mechanism to the questions of WHERE (theme) - WHEN? (instances)

dvessel’s picture

I haven't looked into YUI lately but if it hasn't changed much since the year past, I want no part of it. It felt bloated and the class naming gives it an unneeded learning curve.

I'm currently looking at the options. Something flexible and easy to grasp would be nice.

ica, what you mention is way too much as our time is limited.

RobLoach’s picture

YUI Grids is pretty awesome... ... Watch the introduction video to learn more about it.

dvessel’s picture

Just to get this important point out of the way.. It's licensed under BSD which is supposed to be compatible but doesn't host anything that's not GPL. Or am I wrong?

RobLoach’s picture

The BSD is even looser then the GPL. Is there anyway we could talk with them and get written approval to ship it or something?

-1 for a blueprint-like set of grids based classes, adds complexity for simple things like collapsible regions (think about when you want a lot of regions or something more complex than left/centre/right), are difficult for many to get-their-head-around, havent yet seen one that support liquid layouts well and imho would be rather useless without some sort of grid builder/template generator aka the ATCK theme concept.

If you don't understand it, it doesn't mean its a bad option. Blueprint is very clean and we already know it works magic with Drupal.

ica’s picture

@ dvessel
you are right shifting a paradigm takes a time, but also it may save time and energy that you spend it for something that may not last long -yet it still last and do the job for a while within the time frame one have

I agree that class names like [zoom-xlg, shim, layout a-b-c, gr a] is confusing
but class names YUI use for the rest is pretty standard such as [container, section]
besides I gave the YUI grid builder as an example not the way it deals with classes
I also think on YUI grid builder even Header and Footer is not necessary but every 'area unit' made up 'grids' columns and rows and classes named more meaningfully and accordingly
to achieve what i tried to described on my previous post is can be achieved with a 'theme/layout/cotent relations builder tool simlar to the YUI builder panel
+with additions on the similar panel and ASSIGNER options of
+ Instances (Question of WHEN)
+ Content type chooser (Question of WHAT)
we have in this in Drupal Panels and Views modules partially but not as thematic relation

Examples are explain the my concept with some nearest examples not to be taken literally or codewise

the rest what we can call is a 'theme' just plain CSS files makes use of the areas/grids defined by the tool -as a fashionist/tailor artistry ~one comes as a default core CSS's 'theme' a clear simple one preferable.

JohnAlbin’s picture

ica, your points are interesting, but you are talking about more of a theme-engine/block-system/content-rendering re-write.

Which is waaaay out of scope of this issue: “New default theme for D7” i.e. use the existing theme system and make a new theme.

RobLoach’s picture

I think we should go down the road that webchick brought up in #48, the first step being to get a bunch of designers together and potentially pick a CSS library to bring into Drupal core.

Michelle’s picture

Chiming in as a non themer here, the absolutely hardest thing for me when making a theme is the layout. Decorating it with colors and borders and such is cake compared to getting the layout to not break horribly when you throw a different browser at it. If you designer folks could come up with a base theme in core that had solid yet flexible layout for us non designers to paint on top of, that would be awesome. I think that right there would do a lot to dispell the myth that Drupal is hard to theme. Drupal isn't hard to theme. CSS cross browser layout is what's hard.


Jeff Burnz’s picture

Michelle, thats one of the good reasons to use a CSS library, all those nasty cross browser issues have been worked out for you, it can be a paint-by-numbers affair.

JohnAlbin’s picture

Ok. I've taken Angie’s initial plan and expanded it a bit over at

And MIchelle has turned on comments on that page to make discussing it easier!


joshk’s picture

Reviving this thread!

So: the default theme is Zen. WIN!

The (ahem) "legacy" themes like bluemarine, marvin and pushbutton are being quietly retired. WIN!

Now, to make Drupal 7 fly off the shelves (or down the pipes), we need to stock it with at least one more good theme. How do we get there?

Well, we need ot let people know it's possible. I suggest we start posting some information about how a theme can be nominated for Core. Following on what's worked in the past, I have some ideas:

  • The theme needs to be a theme, not an idea or a design draft. This is a theming challenge, not a design contest. Go ahead and make something and check it into contrib if its to be considered (or, improve/lobby for an already contribbed theme)
  • Have clear and high standards. Throw out some ambitious goals (e.g. including color.module support, g960 css architecture, etc) and be sure the code is reviewed as well as the design. Better to have a few quality options than a lot of mediocrity.
  • There should be a Dries-annointed "theme decider" to make the call here. We can flesh out some guidelines, but this is a subjective art, and so doing it by committee would be ineffective.

Big question: can we push past the 9/1 code-freeze deadline? Since this is ui and doesn't affect any of the api...

Other questions: how to get the word out; how to make it clear we're not asking for "spec" work in the traditional sense; what public criteria to use; who to be "the decider"...

joshk’s picture

Title: New default theme for D7 » New theme to include for D7

updating thread title to avoid confusion

joshk’s picture

Just so it's public, I think a big possible win here is to have some Drupal shops step up. This avoids much of the "spec design" exploitation issue, and it also follows a good established pattern (Acquia Slate).

wretched sinner - saved by grace’s picture

+1 to joshk's comment - I think this would be a fantastic opportunity for one/some Drupal theming shops to get some exposure on their work

nbz’s picture

Exposure? I would assume that any accepted themes would not have a link back to the design shop in the footer or anywhere?

I would expect the bigger motivations would be the bragging rights or personal satisfaction of creating a well liked theme.

BioALIEN’s picture

Full ack with #72!

Jeff Burnz’s picture

@joshk - Zen is not going into core, the new starter theme in core is Stark.

I think its a big ask for a rock solid colorable theme to come out of contrib and be of high enough quality to satisfy Drupal core, they just take such a long time to build and need so much testing its not funny. We're talking hundreds of hours of work here. Maybe someone would step up and do it for free, much more likely a design/theme shop would be appointed and it gets paid for.

Isn't this thread dead?

frank0987’s picture

Status: Active » Needs work
Jeff Burnz’s picture

Status: Needs work » Closed (fixed)

Closing, this thread is dead, if you want to post an issue about a new theme in D7 start a new issue, not 60 something comments in on an old thread.

Frankly the time limit is closing in, it would have to be a do or die effort from a pro theme shop like TNT.

frank0987’s picture

Category: task » feature
timmillwood’s picture

Status: Closed (fixed) » Needs work

We are planning to put together a Drupal 7 core theme at Mark Boulton Design.

Our current thinking is a blog/news style design.

We'll be working on it over the next two months and I will open a separate issue to note the development.

webchick’s picture

Version: 7.x-dev » 8.x-dev

Guess we'll try this again in D8. :\

On the plus side, the absolutely spartan selection of available core themes will cause people to go looking for contrib themes that much sooner. ;)

webchick’s picture

Title: New theme to include for D7 » New theme to include for core

Removing version-specific references.

eigentor’s picture

webchick: we are not yet done. The cowboy is dead when his last breath has escaped from him...

timmillwood’s picture

What is the deadline to get a theme into D7?

The day before release day?

The Mark Boulton Design team are still planning on building one. We are planning on building a theme that can be used by Newspapers, bloggers and content producers. The design for these audiences would be pretty similar, and would give a great opening in Drupal to these audiences.

One thought is to very loosely base the theme on our redesign of (not a drupal site, but a redesign Mark Boulton Design worked on).

catch’s picture

@timmillwood, UX freeze was December 1st. Whether a theme-specific exception would be made is up to Dries and webchick, but that's likely to be less and less likely as we get closer to alpha/beta releases, which is the next step.

Michelle’s picture

A theme is a self contained thing, not changing any APIs or really affecting any code outside of itself. Why does it need to be frozen? Day before release in #82 is extreme, of course, since it needs testing but why not leave it an option for a while?


Jeff Burnz’s picture

Planning to build one and actually building one are worlds apart, any proposed theme should have been in contrib for months by now and be deep into heavy testing. With the new usability and accessibility focus of D7 the theme requires a hell of a lot of testing and user feedback.

I think the deadline is gone on this one, deferring to D8 (with a view to replace Garland?) seems like a more pragmatic approach.

catch’s picture

Priority: Critical » Major

Downgrading all D8 criticals to major per

Jeff Burnz’s picture

Version: 8.x-dev » 7.x-dev
Status: Needs work » Fixed

Actually I think we can close this as fixed - we have Bartik in core now so that is that. Not really seeing how this needs to carry on in D8. If we have new theme for D8 thats another issue (88 comments in not a great place to start that conversation).

Status: Fixed » Closed (fixed)

Automatically closed -- issue fixed for 2 weeks with no activity.