Module Adminbar

[ryan]

I'm not at a place where I can look in the code right now, so I'm going to go back and do that later. But initially I'm thinking that Apache took over this request: /ot/processwire/&modal=1 ... because there is no "?" in the URL So it's not a valid URL, at least not one that PW will get. Do you know if PW or AdminBar generated that URL? I'm guessing PW since AdminBar works in 2.0, but just wanted to check if you knew. The error message you posted was from Apache rather than PW, because PW excludes itself from getting URLs it doesn't understand via it's htaccess file.

Also in 2.1, $pages->find() and $pages->children(), etc. exclude pages that you don't have access as an option in each template's settings. You can reverse that behavior in a selector by adding: "check_access=0" to the selector. I don't think that's the case here, but just wanted to mention this option.

[apeisa]

Ok, this helped a lot.

&modal=1 is added by AdminBar (line 280):

<?php
/**
 * Hook to take place right before a redirect occurs
 *
 * We intercept the redirect URL and modify it to add 'modal=1' to the query string
 *
 */
public function sessionRedirect($event) {
	if(!$this->page || $this->page->template != 'admin') return;
	if(!$this->input->get->modal) return;
	$url = $event->arguments(0);
	if(preg_match('/[?&]modal=/', $url)) return;
	$url .= (count($this->input->get) ? '&' : '?') . "modal=1";
	$event->arguments(0, $url);
}

If I comment $url .=... line out, then it redirects to admin page, but there is "you are logged in" notification in place. Same happens if I go directly to /processwire/adminbar/ab-sitemap/.

But if I go to setup -> templates and set that "adminbar" template manages it's access and set it to go 404 page when there is no access, then I go to 404 page. So this is definitely some kind of access problem. Maybe in page create?

<?php
// create /adminbar/ page under admin
	$abPage = new Page();
	$abPage->template = $abTemplate;
	$abPage->parent = $this->pages->get($this->config->adminRootPageID);
	$abPage->name = $abName;
	$abPage->status = 1031; // hidden & locked
	$abPage->save();

	// after that create /adminbar/sitemap/
	$abSitemapPage = new Page();
	$abSitemapPage->template = $abTemplate;
	$abSitemapPage->parent = $abPage;
	$abSitemapPage->name = 'ab-sitemap';
	$abSitemapPage->status = 1031;
	return $abSitemapPage->save();

There is no view button on adminbar pages when browsing page tree, but this must be because there is no actual template file (we set that at runtime):

<?php
// We change the filepath of adminbar template file
public function beforePageRender($event) {
	$page = $event->object;
	if($page->template == 'adminbar') {
		$page->template->filename = $this->config->paths->AdminBar . "templates/sitemap.php";
	}
}

And I get this working when I copy /site/modules/AdminBar/templates/sitemap.php to /site/templates/sitemap.php and after that edit template "adminbar" and "advanced" tab, and set it to use sitemap.php template.

So this is pretty minor issue, but not sure if I understand where the real problem lies?

[ryan]

I'm thinking the problem might be in wire/core/Template.php. It caches the state of whether the template filename exists. But it wasn't resetting it when the value of filename was changed. PW will only display a 404 for a page that exists to the superuser when there is no template to render the page.  AdminBar appears to be setting the filename, but PW still thinks it doesn't exist because the previous "non existing" value was cached. I'm thinking this is it anyway. If you grab the latest /wire/core/Template.php from the PW 2.1 source, does this make a difference? I will also test locally later today, but wanted to get this posted before a call, hoping this was it.

Thanks,

Ryan

[ryan]

If the above doesn't fix it, this might: Get rid of the beforePageRender hook that sets the template filename, and move it into your init() function:

$template = $this->templates->get("adminbar"); 
if($template) $template->filename = $this->config->paths->AdminBar . "templates/sitemap.php";

I think this would work in either version of PW, though have not tested. The reason I think this might fix it is because this would set the template's filename much earlier than it currently does (well before PW would have any opportunities to check whether the sitemap page is viewable). So theoretically, this should prevent it from deciding that the page isn't viewable due to lack of a template file.

[apeisa]

Thanks Ryan. New template.php didn't fix it, but moving that template change to init fixed it. I will commit this sooner than later to Github (will test on 2.0 first).

[almonk]

I did a fork of AdminBar and slightly remixed the UI aesthetic to better suit my tastes. See attached for screenshot

https://github.com/almonk/AdminBar

[apeisa]

Almonk: Nice work, I like it!

I have been using Adminbar pretty much last few weeks and I have few improvements in mind. One of them is "smaller UI" - no sliding "more info" tab etc. Probably less buttons (like you have). Also current modal wastes space (which is not problem on wide screen, but is very annoying on smaller resolution).

I think I will replace "sitemap" button with link to admin site (that opens sitemap there on the right page). So there would be three buttons: edit, add child, go to admin (to sitemap view).

Any thoughts people? I am having very little time besides client projects, so it will probably take few weeks before I can continue develepment.

[Adam Kiss]

Ha! It was supposed to go into the admin before you changed it!

The smaller @Almonks' UI looks very nice

[apeisa]

Ha! It was supposed to go into the admin before you changed it!

Heh, I know That was before we could link to an single page on pagelist though. After using adminbar myself I realised that another sitemap is just confusing. Also I'm not sure how important it is to see additional info about current page through adminbar? Things like who edited last time, who can access this page etc? I start to lean towards less is more philosophy here...

[Adam Kiss]

If I recall correctly, I did that expand on click just because I wanted to show more with less – in original UI, you had just one pencil, which you had to click to get to editing buttons and more information. I put the buttons out and put a bit deeper the more information parts.

I think that this more information could be useful, but under certain circumstances out of scope of this post.

[apeisa]

If I recall correctly, I did that expand on click just because I wanted to show more with less – in original UI, you had just one pencil, which you had to click to get to editing buttons and more information. I put the buttons out and put a bit deeper the more information parts.

I think that this more information could be useful, but under certain circumstances out of scope of this post.

Actually you recall wrong: http://processwire.com/talk/index.php/topic,56.msg235.html#msg235 (huh - that version does look ugly.. ). Original had two buttons: open panel + quick edit.

Most common use case is by far "quick edit". After that I think what is needed is quick link to admin view - and linking to pagelist would be best option there (now that we can link to certain page on list), since that would kill the need for sitemap. I have to test different variations here... I don't think that current "toggle view" is bad solution - but even I haven't used it at all and always had trouble to find my way back to admin site (I started to use sitemap -> go to administration link).

[apeisa]

Ok, I have been doing site building and some content management lately (+ building data import module), so no new progress with AdminBar in long time. But I have some user feedback from this topic, clients and from myself after using this in real situation.

I am not happy with current ui. It goes over the page and I don't think that showing/hiding it is good solution, neither is the current "hide" button. Also like I talked before, I need to consider functionality again (at least sitemap is not needed). No one uses "toggle" or "advanced view" so there should be only one view. I don't like how current modal waste space. So there is lot's of little details that just doesn't seem right.

I tried to mockup how this thing should work, and I am pretty happy with the results. I dropped the adminbar to the bottom. This gives it a lot more space and more conventional shape (horizontal). Also top of the sites are often graphically busy (logos, headers etc) so top is also nice in this regard.

The other "big change" is that modal is no longer modal. It will be content area that slides from bottom to almost top of the page. So the underlying page is visible, but we are not wasting any space on sides and just a little bit from top.

And one thing I would like to try is optional quick edit. So in templates you could say for some areas that they are quick editable. Maybe something like this:

<h1 class="quickedit"><?php echo $page->title; ?></h1>

This allows quick edit functionality. Button for this shows only when there is something that is quick editable. This is something that I am thinking of, not something I will build for sure, so please tell me what you think?

I attached current mockups, but need to say few words about these: I'm not sure about the icons (probably will settle with something more simple and single color). Also I probably will remove or change location of logout. Link to admin will be used a lot, so it needs to be easy click and without possibility to miss and hit logout accidentally. And if someone wants to help me with UI graphics (adam, almonk, anyone) you are more than welcome.

[ryan]

Great thinking with these updates. It makes a lot of sense what you are saying about the location of the AdminBar. I also like your new approach non-modal overlay–can't wait to try it out. One thought is that some may prefer this same type of horizontal admin bar anchored to the top of the screen rather than the bottom, but I think it will depend on the person (maybe a configuration choice?)

The quick edit sounds great. I'm interested to learn more. If you implement this, I wanted to mention that you can now optionally pass a field name to the editable function(), which may make sense with the quick edit capability:

<?php
if($page->editable()) { /* page is editable, same as before */ }
if($page->editable('field_name')) { /* New: field_name on page is editable */ }

Down the road, PW will support field-level permission granularity. Meaning you can let a user edit some things on a page and not others. It doesn't support that yet (beyond template, parent and sortfield), so I mainly mention this for future-proofing if you want to use it. Currently this editable('field_name') function only applies to template, parent, and sortfield to see if the user has access to modify those properties of the page. It returns true for everything else.

[almonk]

I'm not sure about 'quick edit' functionality - adminbar should strive to make everything quick to edit in my opinion

That said, I do like the idea of somehow integrating edit in place functionality somehow - http://www.aloha-editor.org/ < this is a nice little add on that is being used a lot recently, notably in Locomotive CMS (http://www.locomotivecms.com/)

Other than that, love the new look and UI for adminbar, although, as Ryan said, I'd prefer it at the top

[apeisa]

http://www.aloha-editor.org/ < this is a nice little add on that is being used a lot recently, notably in Locomotive CMS (http://www.locomotivecms.com/)

Exactly the implementation I was looking for! Feels very fast for quick edits (you can test it by logging in to locomotive demo site, and there top right corner -> view website). Not sure how much convenient it is in real life than default adminbar edit (so is it worth to have two different methods for quick editing the content). That really makes a good demo/sales effect though

I will probably switch the bar to the top.

[ryan]

Aloha Editor looks great. I've not had time to investigate the technical details of it yet, but based on what I understand so far, we might need to create a new Inputfield module to support this?

It looks to me like locomotive is using TinyMCE on their admin/back end.

[almonk]

An inputfield sounds like a tidy solution - would love to see this.

[ryan]

I've taken a closer look. This page answered a lot:

http://www.aloha-editor.org/wiki/How_to_install_Aloha_Editor

Implementation on the front end looks easy. Not yet sure what to do on the back end, maybe the same thing. There would be a couple modules involved:

The first AlohaEditor.module would be an autoload module that triggers itself when it sees that you are viewing a page on the front end that you have access to edit (hooking in via Page::render). It would insert AlohaEditor's <script> tags before the closing </head> tag in the markup. It would also look for an ajax post and handle any submitted changes. I can communicate it better in code than text, so here goes:

<?php

class AlohaEditor extends WireData {

    public static function getModuleInfo() {
        return array(
            'title' => 'Aloha Editor',
            'version' => 100,
            'summary' => "Makes fields editable on the front end",
            );
    } 

    public function init() {
        // if it's not editable, don't setup any hooks
        if(!$this->page->editable()) return;

        // attach hook to Page::render
        $this->addHookAfter("Page::render", $this, "pageRender"); 

        // check if changed data posted
        if($this->input->post->aloha_save) {
            $field = $this->input->post->aloha_field;
            $value = $this->input->post->aloha_value; 
            $this->page->setOutputFormatting(false);
            $this->page->set($field, $value); 
            $this->page->save($field); 
            $this->page->setOutputFormatting(true); 
        }
    }

    public function pageRender($event) {
        $out = $event->return;
        $out = str_replace("</head>", "script tags </head>", $out); 
        $event->return = $out;  
    }
}

Next would be a TextformatterAloha module that you select as the "text formatter" for any text-based field types that you want to be editable on the front end. (this is in each field's settings). It would wrap a <div class='aloha-editable'>...</div> around the output of any fields you wanted to be editable (when it sees that you have access to do so). i.e.

<?php

class TextformatterAloha extends Textformatter {
    public static function getModuleInfo() {
        return array(
            'title' => 'Aloha Formatter',
            'version' => 100,
            'summary' => "Select for any fields you want to be editable",
            );
    }

    public function format(&$str) {
        if(!$this->page->editable()) return;
        $str = "<div class='aloha-editable'>$str</div>";        
    }
}

The last part would be creating another JS file to be included with the ones AlohaEditor comes with. That JS file would activate AlohaEditor for any fields that have the aloha-editable class. And it would work with Aloha Editor's callback and getContents() function to send the jQuery ajax post to the current page whenever a change is made (where it would get saved in the AlohaEditor.module's init function above). I need to do more reading at the Aloha site before I could code up an example of that.

This is all just written in the browser, so some of it may not be quite right. If anyone wants to take a stab at this I'll be happy to help. It also seems like this might be a natural fit in AdminBar, though I'm not sure–what do you think Antti? AdminBar's existing method may still be preferable for several reasons, but Aloha may be good for those times when you want to edit content in the context of the site's exact placement and styles. Sites are usually designed for presentation not editing, so it's one of those things that would be useful sometimes and not others, but definitely seems like an interesting option.

[apeisa]

Ryan's solution looks nice and clean. I was first thinking implementation without that textformatter, so person who creates html template can add that on wherever he thinks quick editing is reasonable. But using textformatter here is much cleaner solution!

I am still worried if this is good in overall, compared to modal-type of editing. Quick edit only supports text fields, where processwire sites (at least those that I have build) have lot's of other fields too, mainly page-fields. Of course we can add two different edit modes, but is that really convenient? Other option is to ditch modal editing all together and simply only offer aloha inline editing + real admin.

Anyway, I got working version of the new ui ready yesterday. It will get some more polish, but I find it already more pleasing to use than old one. "Sliding" is nicer than modal in this situation. You can grab this version from: https://github.com/apeisa/AdminBar/tree/AdminBarNew

[apeisa]

Did some comparison between adminbar and locomotive's inline editing. Not so impressed with locomotive's current implementation anymore. Clicking edit button creates full page refresh and again saving (or clicking "done editing") creates another. It doesn't feel fast and it's not fast. I think editing with current way adminbar is done is much quicker (it loads editing page in background and shows it in iframe), and creates page refresh after you hit save. Also things like saving or canceling editing are much clearer in adminbar.

Of course inline editing could be done without any page refreshes at all, and that should be quicker than adminbar's current implementation. I need to think about this a little, but it seems that I won't be getting into that inline editing thing after all. It is nice demo and great for small and simple sites, but it just don't seem to bend very well for flexibility that pw offers. I think it really is more confusing than beneficial for many reasons. Situations where someone should really consider implementing inline editing are when they are building something like drupal gardens is for drupal.

[Adam Kiss]

I don't even think that inline editing has any place in PW system; not with the multiple inputfiled types, not with this simple API. Current solution in Locomotive (Aloha) allows to edit only text fields, and on current page. I, on the other hand, take data from multiple pages (in page tree) almost on every page (on the front-end), so the inline editing would allowed only on little fraction on fields – not to mention I imagine it quite cofusing for end-user, imagine this:

You have some listing of blog posts. Why can't you edit their titles here on the list page? You would have to click the blog post and edit it there. Or, you have images, you couldn't edit their description, because 1., it might be javascript slideshow for instance, and that would get cumbersome, not to mention 2., you can't use aloha editor on image description, because it's not textfield.

In-place editing would have to be much much much more tied into system, and templates would just get plain messy, just to let JS know  what parts are editable with what type of editor, not to mention, that you'd have to have special CSS for most of the items, because just adding 5px border would mostly just fuck up (my) CSS, not display what content you're editing.

So I believe Aloha is currently no-go, simply because making inplace editing for PW with the wide range of possibilities is god damn straight up inhuman task, so... there.

[ryan]

Nice AdminBar updates! I've installed it on the site I'm developing now and I know the client will love this.

I'm not suggesting that inline editing would ever be a replacement for existing editing tools. Nor am I suggesting it should be part of the core (just a 3rd party module option). And I probably wouldn't use it myself. It goes against CM accessibility best practices– content shouldn't have a specific presentation context dependency. But I do think there are some situations where it would be a useful option to provide... like when the needs are really simple and the client is your Grandma (and I can think of other cases, few and far between).

But the main reason I like it is...

Putting on my salesperson hat– If I going in to pitch a proposal for a site and I can show the client that they can update their address, phone, headline, directions, etc., right in the page where it appears, then I've just sold the decision makers. Many times, these people don't understand content management best practices and may not even know exactly what a CMS is (and rarely are they the ones that will actually use it). They understand the concept of print and printing-presses, and they don't like thinking beyond those boundaries. Dynamic content is an abstract concept that's understood only as a buzzword. These are the folks I have to convince. Despite the drawbacks of inline editing, these people will "get it" immediately, and I will be their problem solver. When the next guy comes in pitching the benefits of taxonomies and nodes, then they won't even be listening.

So whether or not the inline editing capability is useful or good is debatable. But I think we can all agree it's at least compelling eye candy that can help to market a product and connect with a broader audience.

[ryan]

I also meant to add that I think this thing would best be developed as a module on it's own. I know that this came up in talking about AdminBar, but AdminBar's utility is clearly far greater than that of any inline editor. Or to word it differently, I would install AdminBar on all sites, but I would only install Aloha as another option in some cases. The two could of course also run together, but it seems like Aloha should probably exist as an independent 3rd party module rather than integrated into anything else (if that was even a consideration). If anyone else decides to develop an Aloha editor module, let me know. Otherwise, I will probably put one together at some point in the near future because it seems like it will be pretty easy to do (building from the code posted earlier).

I mentioned above that AdminBar is something that I would install on all sites (and I'm guessing more than half of others would too). I think it might make sense for AdminBar to be a core module (maybe with 2.1 or 2.2 release?). That way it would be ready to install (1-click) for anyone that installs ProcessWire. Antti, would you have any interest in this, or would you prefer to keep it separate?

[apeisa]

Great discussion. This is really interesting question. One article I remember reading from this topic is here: http://allinthehead.com/retro/357/the-lure-of-on-page-editing (great comments also).

You have some listing of blog posts. Why can't you edit their titles here on the list page? You would have to click the blog post and edit it there.

This is true. But it is also true on current AdminBar implementation and it is also true on real admin site: you cannot edit blog posts when you are editing blog listing page.  Non-tech people would probably find the actual blog post easier by clicking blog post, than by looking for children pages (or somewhere else - you just have to know) in admin site.

Or, you have images, you couldn't edit their description, because 1., it might be javascript slideshow for instance, and that would get cumbersome, not to mention 2., you can't use aloha editor on image description, because it's not textfield.

Yes, special cases like these are where true admin view is much better or even only option.

In-place editing would have to be much much much more tied into system, and templates would just get plain messy, just to let JS know  what parts are editable with what type of editor, not to mention, that you'd have to have special CSS for most of the items, because just adding 5px border would mostly just fuck up (my) CSS, not display what content you're editing.

I'm not so sure about this. Ryan showed example code which means you don't have to do anything special in your templates. Also no need to use 5px border (special background icon on that wrapper div that TextformatterAloha creates should be enough). So you do not need anything special in your CSS neither.

Putting on my salesperson hat– If I going in to pitch a proposal for a site and I can show the client that they can update their address, phone, headline, directions, etc., right in the page where it appears, then I've just sold the decision makers.

This is exactly the same reason why I have been thinking for this. It looks damn good on sales meeting, might even give one or more "wow, that is nice" comments from audience. And depending on the site your "content editing" might be mostly small edits, typo corrections, linking and stuff like that. Then you would love tool like aloha editor. So there is also real life situations where it would be good.

I also meant to add that I think this thing would best be developed as a module on it's own.

I was earlier thinking this in realm of adminbar, and somehow didn't even think that option. I think that it would make great sense to build this separately, but in a way that they work well together. It keeps adminbar codebase clean and simple - and gives it path to grow in more "admin" direction if needed.

If anyone else decides to develop an Aloha editor module, let me know. Otherwise, I will probably put one together at some point in the near future because it seems like it will be pretty easy to do (building from the code posted earlier).

I might put some effort to this on my summer holiday (first week going, three left). It seems simple and rewarding project. Or maybe I do something after we finish the movie tonight.... (edit: decided to close computer and go to bed)

I mentioned above that AdminBar is something that I would install on all sites (and I'm guessing more than half of others would too). I think it might make sense for AdminBar to be a core module (maybe with 2.1 or 2.2 release?). That way it would be ready to install (1-click) for anyone that installs ProcessWire. Antti, would you have any interest in this, or would you prefer to keep it separate?

Of course, that would be an honor actually

[almonk]

I totally agree with Ryan on this.

I don't think Aloha is appropriate for 100% of problems, but it doesn't need to be to be useful. As Ryan says, it's a cool little utility that provides a useful function and I personally would love to see support for it.