Recently Browsing 0 members
No registered users viewing this page.
I'm trying to understand how I could have users log in as "members", provide them with a customized experience, but still serve them cached pages?
For example let's say I have 3 roles
Guest Member SuperUser And let's say I have a NavBar type component that has a menu. Plus it has "log in!" if guest, or "logged in as Mike Smith"
I would want guests to see a cached site with guest-only page access. Members should also see cached pages, but their menu may have access to members-only pages or fields or information. Furthermore they might have a "logged in as Mike Smith" element. SuperUsers can stay uncached. How does one go about creating that cache so that visitors aren't constantly re-creating the menu? Or that when "Sally Baker" logs in, she doesn't see a homepage that is cached for "Mike Smith" with all his elements?
Thank you in advance for any replies.
I like to showcase my new website acniti on the forum here.
Building and managing a website is a hobby, over the years, making websites got more complicated and more technologies, knowledge and wisdom are required. I started building my first website around 1997. It started out with a static site built with FrontPage, a WYSIWYG HTML editor. A few years later it was time for the first content management system, I looked at Joomla but settled for MediaWiki. I run those websites for 2 years on the MediaWiki platform and then moved on to WordPress. WordPress was good, it did a good job but over time, it became more complicated to make something out of the box, if it's not a blog, it becomes complicated and to have a feature rich website requires a lot of plugins. Little by little it became less fun and more and more hassle juggling the various plugins.
In 2014 I became interested in learning PHP programming, I wanted to do this already for many years, but never had enough time to bite the bullet and work my way through the basics. At the end of the courses I though and now what have I learned, how to put this into action? To built modern website with PHP only is difficult, it also requires knowledge of html, MySQL, CSS, java-script etc. I started looking for a framework experimented a little with CakePHP and then came across Processwire via a CMS Critic blog post.
I developed the acniti website on a Linux Ubuntu 16, with PHP 7 and MySQL as the development server. For the IDE I use PhpStorm, before using Storm I have used and tried some other IDE's such as Zend, Eclipse, Netbeans, Aptana but none of them I liked, some were feature poor, Zend and Eclipse were slow and use a lot of memory. PhpStorm not free but definitely worth the investment. I make use of the free tier Git repository of AWS called CodeCommit, I still use GIT Cola to commit the changes, I could also use PhpStorm for this but I never took the time to change my workflow.
For project management I am a big fan of Redmine, Redmine is a web-based open-source project management and issue tracking tool. I use this also for my other work so it easily integrates with the website building flow as well. It's easy for maintaining lists of features you want to carry out per version, it supports a wiki which is easy for making notes and keeping a log of the activities. I use it everyday and it runs on Ruby. For images and graphics I switch back to Windows for some Photoshop.
The acniti website runs on the latest stable Processwire version at the time of writing 3.0.62, the website has 4 languages including an Asian language. The Japanese language URL's are implemented with their 3 alphabets kanji, hiragana, katakana i.e. https://www.acniti.com/ja/インレットフィルタ. Some images on the site have text and image language tags help to select the correct language, the Processwire blog post from 30 June was helpful to get this running.
The main site has a bootstrap theme, for the mobile version of the site the google AMP specification is implemented. This was really fun to do but challenging at times as the AMP specification is still a little limited. To visit the AMP pages type /amp/ behind any URL like https://www.acniti.com/amp/ for the homepage. The Google webmaster portal is really easy to troubleshoot and check for the correct AMP implementation. Finally structured data according to schema.org is added to the site via the JSON-LD markup.
The commercial modules ProCache and Formbuilder are installed. The ProCache module is really amazing and makes the website lightning fast. Besides the commercial modules around eleven open-source modules are used, Database Backups, Tracy Debugger, Wire Mail SMTP, Protected Mode, Batcher, Upgrades, PublishAsHidden, URL (Multi-language), Twitter Feed Markup, Email Obfuscation (EMO), Login History, Selector test.
During development the Processwire forum is really helpful and checked often. The forum is good for two reasons, most of the questions, I had during development of the site, are already on the site. Secondly the only 6 questions I posted over the last 2 years, are quickly and accurately answered. The downside I didn't become a very active member on the forum but see that as a compliment.
An open issue on the acniti site is the AMP contact form with Formbuilder, the restricted usage of java-script for the AMP specification requires some more in-depth study.
For the hosting services the acniti site uses Amazon EC2, I use AWS already many years to manage my cloud office so it was easy to decide to use it for the web hosting as well. The site is running on a micro instance of EC2 and with the ProCache module CloudFront is serving webpages worldwide fast. Updates from the development server are sent to CodeCommit and from there to the production server. From a site management point of view it would be nice to use AWS RDS to manage the MySQL databases, but from a cost perspective I decide not to do that for now. Via a cron I have set up automatic MySQL backups and these are via another cron job uploaded to AWS S3. To make sure the server is safe, a cron job runs daily snapshots of the server, this is getting initiated via AWS Lambda. Lambda also removes older snapshots because during creation a delete tag is attached for sevens days after their creation. It's important to make a separate MySQL backup as with snapshots the database often gets corrupted and its easier to restore a database backup than to fix a corrupted database. Another nice feature to use AWS Lambda for is a simple HTTP service health checker, which reports to you by email or sms when the website is down.
Making use of all these Amazon services cost me probably somewhere between 10 - 15 $ a month, I have to estimate a little since I am running a lot more things on AWS than only the website.
The site is running on a Comodo SSL certificate but next year I will change to the free LetsEncrypt, as it is easier to add and will automatically renew after 90 days. The Comodo certificate requires manually copy pasting and editing the certificates in place.
The content for the site I write in the Redmine wiki, most of the content I write requires research and it takes about two weeks before I publish the content to the Processwire site. For writing content I use the google spell checker with the grammar checker, After the Deadline. To ensure catchy headlines they are optimized with the Headline Analyzer from CoSchedule
Now the site is running, it needs promotion. The robots.txt files shows the search engines the way as does the sitemap.xml both of these I have made in a template file. Most of the blog articles I promote are republished on social networks like, LinkedIn, Tumblr, Google+, Twitter, and some branch specific networks as the Waternetwork and Environmental XPRT. To check, the search engines index the site well, Google webmaster and Bing webmaster check for any problems with the site. For statics on the same server there is an instance installed of Piwik. Piwik is a leading open alternative to Google Analytics that gives full control over data. The Piwik setup works very well and gives a good overview of the site usage both on the desktop via the site or via a mobile app. As a part of a test I have installed the open-source SEO-panel on the same server to manage keywords and to further improve the scores in the search engine, a nice feature is that you can also track your competitors. I am still new to SEO panel and have to learn more how to use the tool effectively.
I want to showcase my website made in processwire. https://chrysemys.nl/
Goal of the website
There are quit some websites with information about turtles. Most of them are very limited in information. This website main goal is to be very complete in information. Information about turtles, there natural behaviour, detailed descriptions. Next is also associations ( (bi-)directional links) with caresheets, books, food, illnesses, etc...
Second goal is to learn PHP as I started an education for application engineering 4 months ago.
It is build on the latest version (3.0.61) and makes use of the latest template strategy. I use the pw-append, pw-replace classes in combination with _main.php and several template files.
Most of the rendering is done from custom functions in _functions.php.
In the back-end the pages are styled with the options for visibility and group in logical blocks. This is a very nice feature with processwire and makes the pages easier to fill.
Next to the core modules I have used the following extra modules:
AIOM+ - obviously
Connect Page Fields - This module is very important. The whole site is considered a relational database. The base is the enity turtle. This entity has bidirectional associations with books, caresheets, locations (voor google maps), regions (to group locations), the different classifications of the animal taxonomy, food, etc. To make these kind of associations it greatly simplified the output on the other templates.
This module was a mandatory feature for being able to create a site as I wanted to. Otherwise it would it would be very inefficient to keep everything up to date. Now I can add a new book and connect it to one ore more turtles. After the save you can see the book also on the turtle page. This goes for all the bidrectional links.
This page is about a species - https://chrysemys.nl/schildpadden/emys-orbicularis/
This species has several sub species (basicly the same turtle but live in another country/region and has some different charasteristics). I have decided to make no distinction between species and sub species (on a template level) and use this module to connect them to each other. One reason was that this way the url would not be too long.
This turtle lives in several different habitats. These are also bidirectional associations. The same goed for food, books, food to mention some.
The biological taxonomy is also build this way as entities that have no parent/child relation. They are connected to each other with page reference (in this case).
Both methods (page reference or parent/child) have their pro's and cons, in the end this setup seemed to work best for me.
Map Marker - This is used for the map marker field for the locations (verspreiding). Everyt turtle can live in one ore more locations and the will show up on the google map. Also I have the possibility to add location specific information about the climat on the pages (missing on almost all other websites).
Map Markup (Google Maps) - For the output of the google map. The js script for this is only outputted on the pages where it is used (template strategy).
Markup Sitemap XML - obviously
Social Share Buttons - Added this as it is recommended for a better SEO score.
Get Video Thumbnails - The blog story can have a youtube link. This module automatically grabs an image and places it as the image for the blog. No more extra steps needed.
Video embed for YouTube/Vimeo - I slight changed the output for this module so it always is placed under the body.
Datamaps.js - Also I had some very important help from @adrian with help of using datamaps.js. This page is a datamaps script that read all turtles that live in the US and shows them in the correct state. https://chrysemys.nl/kaart-verenigde-staten/
A php function renders the output for the script and fills the variables. These are added to te script on rendering the page.
The website is not finished and probably never will be as new information and functionality will be added all the time. But it is finished enough to serve as a showcase.
There are several things I want to add in the future:
A find selector for overview pages (select by size, region, family, etc). Worked on this but, could not get it working.
Datamaps.js for all regions (also with a selector to select another region).
Optimize the queries to make the site faster.
Several front-end improvements (marging/padding/font-size, etc...)