Leaderboard

What are your personal pet peeves, when it comes to general web / online / digital stuff? I'm gonna start with one of my most-hated "features": lazy loading technique taken too far (as in "load images / assets only when the user scrolls down") The general idea behind it was certainly fine and noble: Load + show additional content only if the user decides to scroll further down a page, otherwise hide that content. Avoid having to wait for content loading that I might - or might not - be interested in. Problem is, that some site-owners not just use, but abuse that method. Case in point: Flickr. It's almost impossible to reach the footer links / menu, because the page keeps loading new pics forever. You may see the footer for a split second, but it disappears again, when addtl. content loads. What's wrong with using pagination? Is that too "old-school"? Making it hard for visitors to reach a certain navigation section is certainly not "best practise", or user-friendly. Some sites have simply exaggerated it to the point, where I'd rather see a "skip intro" button (and that's saying something!). So, what's your favorite pet peeve, when it comes to websites, apps et al?

Greetings, Thanks dragan for starting this discussion! I've been doing web design since... well, since there was such a thing as "web design," and so I guess I've accumulated some pet peeves. Violating Core Principles Generally speaking, I dislike anything that ignores core principles. It may at times seem that "everything has changed," but when you look at the whole history of the web, it's interesting that certain data/design principles have been constant. Failing the "Concept-First Test" Related to your point about lazy loading: my biggest peeve is applications that start with a visual/design standpoint instead of an information/content/conceptual standpoint. Take a look around, and you can easily find lots of apps that "look cool," but what is the underlying concept? As a test, I like to strip away all visuals and put into simple words what an application does. Then I add in visual/transitional effects, in service of what an application does. It seems to me that too many apps start with the visuals first. I call this the "concept-first test." Overloaded Visuals It feels to me that we are starting to overload our workflow with libraries that simply make superficial visuals work better. I know that Angular.js (for example) can do more than just make things look pretty, but it seems to me that a lot of the applications listed here fail the "concept-first test." I don't mean to pick on Angular.js. It's just an example. (I don't think this direction is sustainable, and we are in for some kind of revolution some point soon.) To make the subject more relevant, I think all this relates to ProcessWire. I have a lot of respect for Ryan and ProcessWire, specifically because Ryan clearly understands the idea of "core principles." CMSs come and go, but the reason Ryan's concept has been around so long and keeps going steady is because Ryan adheres to the principles that govern the ways in which the web works. As the web advances, ProcessWire will continue to be more relevant. I'd definitely like to hear what others here have to say! Thanks, Matthew

This is very dangerous from a security point of view. Never store information coming from the user directly in the session without sanitizing and validating first. Plus, only store necessary informations, for example the e-mail or user-ID. Example: A bad guy can now alter the post data and include 10'000'000'000 new variables which you try to store in a file on the server. Maybe your server crashes Do you use ProcessWire users for the registrated users? I don't understand why you need to store stuff from a post request into the session. I mean, after the registration your system somehow needs to be able to log in users. And if a user is logged in without the "payed" flag, you show the link to PayPal. But in this case you already have a valid session for the user...? The first is true, because "normally" cookies get deleted by clearing the browser. But when leaving a page or closing the tab, the session remains. How long a session is valid, is defined in your php.ini.

You can open the page under /admin/ (in your site tree) and check if there's a Process selected, or add the right one if the field is empty.

@adrian: That looks nice. I think we all struggle with 'how to insert images'. If there's a way to restrict most of the image properties, then it's a good way to go in my opinion.

In case you guys haven't seen it yet: http://ckeditor.com/demo#widgets New content blocks and captioned images etc. Might be a useful addition, although the default box is pretty ugly

Those two projects sound like great ideas to get stuck in to with ProcessWire. I think most people here would recommend using the development version now, and just keep updating as you go along. It's fairly stable right now, and updates are very simple to do (just replace the /wire/ directory, usually). I don't think there's a date scheduled for a 2.4 release - it will be ready when it's ready I started a website using the dev branch in August, and it's almost ready to launch. I've updated the work-in-progress site several times during development and it hasn't been a problem. So I would say go for it

feldtypepage only.have tamplate if u.set ones in feild setting but if u did

I have used a lot of time with this issue. The bigger and more complex sites we build, the more "ad hoc" UA is needed. Defining access on "who can access/edit news" is not relevant when your site has 20 news archives for different groups. Or when you want to create new "workgroup" and choose who can access that. Of course all these can be build custom, but it is lots of work and PW doesn't provide anything as a base for that. I have module baking that solves this very issue. It has basic functionality already made and it for most parts bypass the template based UA totally. This is by intention very simple at this first stage: just view and edit rights and user that belong to groups (one user can belong to many groups). Not sure about the licensing yet, I might release this as a commercial module, but also happy to transform it to a open source project for community (if you guys feel this is promising). Ryan has seen the code and we did hit one core problem: circular reference. My module requires page field that many times references itself. This creates problems in certain situations: actually only when page is saved but no changes is made. It is not ready yet, but the basic functionality should be there. I have build this few times already and always from different angle. I believe this is the most simple one (see the find hook that has been the problem in those previous modules). Anyways, I know at least Teppo and Nik have been interested in this, so here we go: https://github.com/apeisa/UserGroups/

You can combat the by giving the names container a min-height at smaller widths. That way if there are some that only take up one line the container still expands to a consistent height across all items. Yeah. Assuming you are going to go with open sans, just add that declaration to the <body> and remove all the references to Arial.

Thanks for putting together the PR for Ryan. Just for the sake of keeping connected posts linked, this post and the following reply from Ryan are relevant: http://processwire.com/talk/topic/3023-module-ckeditor/?p=42523

I'm using this feature quite a lot. Here's an example of creating a indented list for using with checkboxes, selects. IT creates a new PageArray and modifies the "title" to indent, and here with a (count) used in the context of the categories to see how many products are under that category. $children = $pages->get("/shop/")->children(); $list = new PageArray(); function mylist($children, $ind='', $arr){ $ind .= "– "; foreach($children as $cat) { $countproducts = wire("pages")->find("template=product,categories=$cat")->count(); $cat->title = $ind . $cat->title . " ($countproducts)"; $arr->add($cat); if($cat->numChildren) { mylist($cat->children,$ind, $arr); } } return $arr; } return mylist($children, $ind='',$list); This will produce something like –– Gewürze (0) –– –– Kräuter (27) –– –– Kräutermischungen (7) –– –– Einzelgewürze (0) –– –– –– Anis (2) –– –– –– Blaumohnsaat (1) –– –– –– Bockshornsamen (2) –– –– –– Chili (16) –– –– –– Glutamat (1) –– –– –– Galgantwurzel (3) –– –– –– Ingwer (4) –– –– –– Kardamom (4) –– –– –– Knoblauch (5) –– –– –– Koriander (2) –– –– –– Kurkuma (2)

I originally missed the issue with language switchers not working on the 404 page because all the instances where I'm using language switchers, I was only have it show the language switcher if the page was not the 404 page. The reasoning there is just that in my case, there wasn't much point in being able to show different language versions of a 404 page. If your situation is similar, I recommend simply not showing the language switcher when on a 404 page. Though if you want the switcher to redirect back to the homepage, that seems like another decent approach. But here's what I did: if($page->id != $config->http404PageID){ // render the language switcher }