The Real Trick to Waking Up Early

Devir Kahan at The Geek’s Companion recently wrote a post about The Trick to Waking Up Early.

This post caught my eye since recently I’ve been trying to reform my sleeping habits and start getting up earlier in the day. In doing so I faced the challenge of actually waking up when the alarm clock goes off, then getting up.

Devir suggests:

…when you wake up, roll over, grab whatever screen you choose to use, and use it. Go through emails. Catch up on RSS and Twitter. […] Write. Read. Whatever.

Being a fellow geek, I love this idea. In addition, I can speak from experience that it is effective in waking you up; however it can have a major drawback, depending on the type of person you are.

First, let’s differentiate between waking up and getting up. Waking up is going from sleep to awake. Getting up occurs when you actually crawl out of bed to start getting things done, and this is where the issue is.

On my laptop and phone there are so many things to do that I could spend all day on them without getting anything productive done. Processing email. Reading articles. Reading Twitter. Reading Wikipedia1. Reading Reddit. Playing games. I could go on and on.

I found that I would stay in bed for far longer than planned by wasting time on my phone or laptop, often making me late for classes. While I would wake up, I wouldn’t get up. In addition, by starting the day unproductively it set the pace for the rest of the day and I would have trouble getting things (e.g. homework) accomplished.

This problem certainly won’t apply to everyone, and clearly it doesn’t apply to Devir. It was, however, a real problem for myself and one to be wary of for those trying this technique.

How do I wake up and then get up in the morning, then? A combination of two things:

  1. I have a morning routine, or ritual, that I perform without exception.
  2. Get at least 8 hours of sleep, always.

Let me explain those in a little more depth.

I first learned about the morning ritual from Asian Efficiency — a great site on how to improve productivity. They have an excellent podcast (with transcript) and blog post explaining what a morning ritual is and the kinds of activities you should have in it.

In their podcast, they sum it up quite nicely:

A morning ritual describes what you do from the time that you wake up until the point where you start your day, whether that’s at work or elsewhere. It’s basically your default autopilot way to start every single day.

I would consider this the key for me to wake up and get up in the morning. Having a set of tasks that are easy to accomplish and already laid out is a huge motivator. In turn, it forces me to start my day productively, setting a great pace for the rest of the day and letting my mind fully wake up.

While the morning ritual is the key to me getting up in the morning, getting a proper amount of sleep sure makes everything easier. I realize that getting at least 8 hours of sleep every night without exception is a pipe dream though. Life gets in the way sometimes, and that’s fine. What matters is getting 8 hours of sleep overall. Once again I refer to Asian Efficiency and their post about how to manage sleep for productivity; in particular, the section on Sleep Debt.

The basic concept is that your body requires a certain amount of sleep every night on average. If you fall short one night, you can recoup that through naps or getting extra sleep a following night. I follow this theory religiously, sometimes having to perform a “sleep reset”, as I like to call it, where I get far more hours of sleep than I require (e.g. 12-14 hours) in a single night, then the next night return to my normal schedule.

By being well rested, I find that I am naturally more motivated to get out of bed and then be productive throughout the day.

Ultimately, these are suggestions and I encourage you to try different techniques until you find one that works for you.


  1. In case you haven’t noticed, I really like reading. [return]

More Back Button Confusion

In December I wrote about iOS Back Button Confusion, outlining how having multiple back buttons on the same screen in iOS creates user mistakes or slowdowns as they determine which button they should use. I recently noticed the same issue with my RSS reader of choice, NetNewsWire (which is an outstanding app for reading RSS feeds on the desktop).

In similar fashion to the iOS apps, NetNewsWire has an embedded browser for viewing links. Thus, it also requires having two back buttons: one to exit the browser and one for going backwards in the browser history.

With NetNewsWire, as with iOS, I find myself exiting the embedded browser when I intended to go backwards in the browser history. This causes a small amount of frustration which adds up to a poor user experience over time.

My proposed solution still stands, as the dynamics of the problem have not changed between touch screen and pointer input methods: provide a way for the user to “undo” the act of closing the browser. For a deeper explanation of not only the solution, but the problem, take a look at the previous post.

On Taking Notes

Being a university student, I have taken a lot of notes for my courses. I have not, however, been able to find a method for taking notes that satisfies my desire to use my computer as much as possible, while giving me the flexibility and speed I need when taking notes. Here I have documented the process I have gone through so far in switching to a digital notes system for university.

Let’s first look at the three key reasons why I want to use a computer to take my notes:

  1. I find typing is far faster than hand-writing, especially when having to copy off of a blackboard or PowerPoint slides.
  2. Digital organization is far easier and more expandable than paper organization - I have no qualms about keeping schoolwork from high school that I did on my computer; all paper documents, however, have found their way to the trash.
  3. I’m a geek. I love my computer and love spending time on it; adding yet another place that I can do so productively is always welcome!

Here are my requirements for an ideal note-taking system:

  1. Diagrams - I should be able to easily and accurately reproduce ad hoc diagrams that drawn by the professor (if applicable to the course).
  2. Random characters, formulas and numbers - Being in engineering, I’ve had to write down many formulas and Greek characters that are often very complex (if applicable to the course).
  3. Organized - My notes should be easily and effortlessly organized, allowing quick reference when needed.
  4. Speed - I need to be able to keep up with professors as they move through powerpoint slides or write on the blackboard at near the speed of sound.
  5. Accuracy - While moving at the speed mentioned above, I need to be fairly accurate with my spelling and 100% accurate with any formulas or numbers.
  6. Portability - I should be able to access my notes easily since you never know when a study session with friends might crop up. In the case of digital notes, this access might happen on a device that I do not own (e.g. a university-owned computer)

It can quickly be seen that my reasons for wanting to use a computer deal with requirements 3 and 4 beautifully; better than a traditional system does. How about the rest though?

In the past I had briefly tried Microsoft Word, only to find it was too bloated for my needs; the extra UI associated with making the notes look like a physical notebook detracted from the notes themselves. Also, my notes were not portable being saved in the DOCX format.

I then remembered TextEdit. It has an extraordinarily simple UI and saves plain-text documents for portability. That was all I needed to give it a try. I developed a simple folder structure for organizing the notes, University/Notes/[Course]/[YYYY]-[MM]-[DD].txt, where [YYYY] would be replaced with the year those notes were taken and so forth.

To allow for easy access, I was already storing my “University” folder in my DropBox so naturally my notes were stored there as well. To allow access on the go, I came across PlainText for my iPhone. This beautifully simple app does one thing and does it very well: text editing on iOS while syncing with DropBox.

The first day I started taking notes with TextEdit, I was surprised at just how fast and accurately I was able to type. This isn’t myself gloating, rather it was the effect of systemwide auto-correct implemented in Mac OS X Lion. Having its roots in iOS, it is far more aggressive than what can be found in Microsoft Word or other word processors. This left me with little cleanup to do after a notes session.

Before classes started I knew diagrams would present a major issue - how do you draw with a keyboard? To deal with this, at the beginning of the semester I took notes using TextEdit for as long as I could before encountering an important diagram, at which point I switched to using a traditional notebook for that course. If there weren’t any diagrams, then TextEdit remained.

The other major issue with TextEdit would have been random characters and formulas. As far as I know, there is no easy way to find and enter large numbers of random characters on computers today (though they do have them in the characters menu).1 I’m far enough into my degree that I no longer have any math or physics courses, however in 1st and 2nd year this alone would have killed any dreams of utilizing a digital system.

So in summary, my system with TextEdit handily takes care of 4 requirements: speed, accuracy, organization and portability.

I used this system for two of my five courses last semester (the others had diagrams and/or characters and formulas) and it worked very well. I took more detailed notes while improving my typing speed and accuracy due to the extra practice.

This semester, I am utilizing Markdown, Byword and Marked to help with formatting my notes (and therefore improving upon my organization) while keeping them extremely portable. On my iPhone, I’m currently using Nocs for viewing and editing Markdown since it is free, however if I find I’m doing this often enough (both for notes and blog posts) I will purchase Elements which I find more visually appealing.

I have yet to find the perfect digital note-taking system that can seamlessly handle diagrams and random characters, though my roommate is currently trying a bunch of apps for his iPad that look promising when using a stylus. I truly believe that tablet computing will be the future of note-taking at school since a keyboard is too limiting, though I cannot justify the cost of an iPad to try it myself.


  1. Latex could be better here, though I do not use it enough to justify learning the syntax for all of the different characters. Also, I find the raw syntax unpleasant to read making it not very portable. [return]

Misconceptions About iOS Multitasking

Fraser Speirs posted a very necessary explanation about how iOS background applications work in an attempt to stop incorrect information from being spread.

It is well worth the read and I agree with almost all of his points, except the following statement that he makes a couple of times in the post:

The user never has to manage background tasks on iOS.

While this should be true, I, along with many of my friends and family with iOS devices, have not found this to be the case. If a user let’s the recently used app list grow uncontrolled, it noticeably slows down the device.

This is because an app in the recently used app list may or may not be consuming memory or CPU time (following the rules outlined by Speirs), but if an app is not in that list, guaranteed it is not consuming memory or CPU time. This means removing an app from the list is a method for forcefully freeing memory (or killing the app) and therefore speeding up the device.

As for why iOS noticeably slows down when it needs more memory, but seemingly doesn’t free up what it needs automatically, I don’t know what the reasoning behind that is. Perhaps the algorithm is not aggressive enough on freeing memory, trying too hard to leave apps alone?

The best example of this problem that I’ve seen was when I was in an Apple store and briefly tried out the demo 3Gs (running iOS 5) to find that it was barely usable; when I tried to type out a search on the spotlight screen, the keyboard “hung” with a key depressed for about 5 seconds and all animations were extremely jarring.

Surprised by how slow and unresponsive the phone was, I went into the recently used app list and cleared it out. Removing the first few apps was slow (taps felt unresponsive, animations jarring), but then everything sped up and felt back to normal. Going back to the Spotlight screen and typing again was a bit slower than on my iPhone 4, but very usable, and the animations were quite smooth now.

In addition, usually over once per day I’ve noticed the animations getting slightly choppy and apps becoming less responsive than normal on my iPhone 4. Clearing my recently used app list has always returned the speed and responsiveness that I’m used to with iOS.

I still recommend users clear out their recently used app list whenever they notice their device becoming sluggish as I know from experience that it works to speed up the device.

iOS Back Button Confusion

Since I purchased my iPhone in September, I’ve had a consistent issue that is prevalent in many iOS apps. While small, it is a source of frustration nonetheless. My issue: back button confusion with embedded, in-app browsers.

Let me explain what I mean.

When a link is opened, many apps utilize an embedded web browser to display the web page instead of exiting to Safari (e.g. tapping a link in a tweet in the Twitter app). Upon opening, a traditional “app” back button is placed in the top left (let’s call this the “app back button”) which exits the browser to the screen where the link was tapped. The browser also gets it’s own back button for going backwards in the web browsing history (let’s call this the “browser back button”) placed at the bottom of the screen. This results in two back buttons on the same screen.

While this design is functional, having multiple back buttons on the same screen means I have to choose which one to use: do I want to go back in the app or browser history. While this is a simple decision that makes sense after conscious thought, it doesn’t feel natural. Since I wasn’t in a browser to begin with (remember, I tapped a link in an app), the website feels like an extension of that app and therefore I instinctually reach for the app back button, regardless of whether I wanted to go back in the app or the browser history. If I wanted to go back in the browser history, then I’ve made a mistake. This results in the browsing history being lost with no way to easily undo it; the only way to get back to the last web page is following all of the links again which is both annoying and takes time.

I’ve thought of two possible solutions to this problem (though I know there are far more out there) which I believe would fit nicely with the existing UI paradigms in iOS, and solve the problem explained above.

  1. A single back button
  2. The ability to undo exiting the browser

A Single Back Button

This solution would remove the confusion by having the app back button as the only back button; the browser back button wouldn’t exist. This button would go back through your web history until the first web page, then exit the browser back to the app. There is a major flaw, though; repeatedly tapping the back button could take too long if you’ve built up a history larger than a few pages in that browsing session (not unusual, especially for power users).

The Ability to Undo

This option would give the user a method to undo their action of exiting the browser, returning them to where they were before they tapped the wrong button.

To accomplish this, I propose that tapping the original link in the app immediately after exiting the browser would return the user to the same web page, with the browsing history restored. If, however, you tap another link or change screens in the app, that browsing session would be lost.

An example scenario for this solution would consist of the following steps:

  1. Tap a link to an article in a tweet
  2. Embedded browser opens the link in-app
  3. Tap a link in that article to read supporting information
  4. Want to return to the original article to continue reading, but accidentally tap app back button
  5. Returned to the tweet and realizing the mistake, quickly tap again on the link to the article
  6. Embedded browser opens to the supporting information, not the article, with the browsing history restored
  7. Tap on the back button for browser history to return to the original article and continue reading

I believe this behaviour would feel very natural to users and is similar in concept to fast app switching since state is being restored.

Apple Forgot About Fitts' Law

I have this quirk when I’m reading something on my computer (website, PDF document, etc) that I want to keep my mouse pointer visible so I easily know where it is when I want it, but I don’t want it in the content area itself obscuring what I’m trying to read. The margin of the content is the perfect place for this and that is generally where I leave my mouse when reading.

The other day I opened 5 PDF documents in the same Preview window to read, putting that window into Lion’s handy full-screen mode.

When I went to place my pointer in the left margin of this preview window, however, it wound up hitting the left edge of the screen causing a nifty navigation pane to slide out. I thought to myself “Cool!” and took a mental note for later when I had to switch between the PDF’s.

I was then putting my pointer back in the left margin, being cautious not to hit the edge, when the pane appeared again. This had me very confused and after a few seconds of experimenting I found out that a 40 pixel column along the left edge of the screen (give or take a few pixels) is the trigger area for the pane — not the edge of the screen itself.

This design goes directly against Fitts’ Law. If you don’t know what Fitts’ Law is, I recommend taking a look at the Wikipedia entry (Success and Implications section) for a quick summary. I want to draw attention to one point in particular from there:

Edges and corners of the computer monitor (e.g., the location of the Start button in Microsoft Windows and the menus and Dock”) of Mac OS X) are particularly easy to acquire with a mouse, touchpad or trackball because the pointer remains at the screen edge regardless of how much further the mouse is moved, thus can be considered as having infinite width.

This means making the trigger extend inside the edge of the window is unnecessary due to the infinite depth this pane comes out of. This is due to the fact that once I know the left side of the screen activates this pane, I am not going to carefully place my pointer there, I’m going to “throw” it left, letting it hit the edge (as I do now). This is a similar action to a person showing a hidden Task Bar (Windows) or Dock (Mac OS).

The only reason I can think of for Apple extending the trigger area into the content is to increase discoverability — a person might be more likely to move their pointer near the edge of the screen than actually hit it. Also, unlike the Windows Task Bar and Mac OS Dock which are visible by default, there is no indication that this pane even exists until you activate it. However, this design comes with a major consequence beyond the annoyance for my quirky mouse-placement ways: if there is ever content under that trigger area, it is inaccessible for annotation or selection (despite the margin being prime real estate for annotations)1.

I have high respect for Apple’s design and execution in their software and hardware, but I believe something was overlooked here, or I am not seeing a use-case that this design is the solution for.


  1. Note that the content can still be accessed/annotated by coming out of full-screen, however I don’t believe Apple’s intention with full-screen apps was to have an app’s functionality restricted to support it. [return]

China Drawn Pixel by Pixel

Design Dare has found something truly incredible:

In China, because of the stifling censorship, they decided to recreate the satellite maps by hand-drawing everything.

It is an incredible accomplishment and well worth taking a look at.

TED Talk: My Stroke of Insight

Lately I’ve been in search of interesting, insightful videos to watch. I want videos which expand my knowledge of the world and the universe, as well as push my intellectual thought.

TED (Technology, Entertainment, Design) has some of the best videos for doing exactly that. If you have never heard of TED, in their own words:

The annual TED conferences, in Long Beach/Palm Springs and Edinburgh, bring together the world’s most fascinating thinkers and doers, who are challenged to give the talk of their lives (in 18 minutes or less).

They then provide videos of those talks on their website, free for anyone to watch.

While all of the talks are incredible, there are some true gems. These are the talks that move you as a person and can change how you look at the world. The tough part can be finding them since there are so many talks to browse through.

In searching for a video to watch recently, I stumbled across a TED talks top 10 list where I watched what was previously their number one talk. I cannot describe in words the emotion and insight this talk emanates with an incredible power. It is presented by Jill, a brain scientist, who experienced a brain hemorrhage and she describes what happened during the event. The talk is: Jill Bolte Taylor’s stroke of insight.

I cannot recommend this talk enough so please, when you have the time (about 20 minutes), watch it. I believe it will at the very least be interesting, if not change your outlook on life.

Sort iPod Artists by Display Name

The Problem

I had always found it annoying how the majority of the songs in my iTunes library, when sorted by artist, were sorted by the artist’s last name instead of their display name (e.g. Faith Hill is sorted as Hill, Faith). This is because when I viewed the list of artists on my iPod, they are shown in the First Last format resulting in the first characters of the artist being in random order.

For example, these three artists could be next to each other in the list:

Charlie Daniels Band Kelly Clarkson Billy Currington

When there’s only three it’s easy to see that I’m looking at the C’s. Get that in a larger list (especially when scrolling fast), however, and it becomes impossible to track where you are without stopping since the first characters are changing with no order. This forces you to stop scrolling and analyze the list to determine where you are at.

To make matters worse, not all artists were sorted the same way. For example, Faith Hill might be found under ‘H’ whereas Tim McGraw might be found under ’T’. This is simply poor tagging, but I do not have the time nor will to put into correctly tagging all of my artists for last name sort.

The Fix

I already knew what needed to happen: I needed the artists sorted by whatever value was in the “Artist” field for a song (this is what’s displayed). I also knew there existed the “Artist Sort” field for a song which specifies how a song’s artist should be looked at for sorting.

It turns out that the “Artist Sort” simply takes precedence for iTunes when sorting by artist over the “Artist” field. In short, here’s the basic algorithm for iTune’s artist sorting:

  1. Is “Sort Artist” field empty?
  2. If no, sort by that value. Otherwise move to 3.
  3. Sort by “Artist” value.

So, delete the “Sort Artist” value and iTunes will sort by whatever the displayed value is. Do this for all songs and the entire problem is solved. Simple.

Note: This change is PERMANENT except by going through and manually setting the Sort Artist field for the changed songs.

If you don’t know how to do that, simply:

  1. Select the songs you want to change (I selected my whole library for consistency).
  2. Right click on any selected song.
  3. Click Get Info.
  4. Click the Sorting tab.
  5. Make sure the Sort Artist field is blank.
  6. Make sure the check box next to the Sort Artist field is checked.
  7. Click OK.

iTunes might take a minute or two before the changes are applied if you selected a lot of songs. Then you’re finished! Don’t forget to sync your iPod to see the changes there of course.

I know this seems exceedingly obvious, but I figure if I had the issue for as long as I did without finding a fix (mind you I wasn’t looking terribly hard for one either), there’s a good chance someone else out there is having the same one.

Anyways, I hope this helped to make life a little easier when using your iPod as it definitely did for myself!

Changes

It has definitely been a while since I have posted, but for good reason I believe. I have made massive changes to my website with user-facing design changes, as well as back-end structural changes. Between finishing those changes and working it has kept me pretty occupied, but I’m definitely happy with how everything turned out!

With this post I hope to describe the changes that took place and the reasoning behind them.

Design Changes

So, why did I redesign the look of the site? A couple of key reasons:

  1. It felt very outdated.
  2. It wasn’t a good design for a blog.

Let’s explore those in more detail.

It Felt Outdated

While the internet in the 90’s was filled with animated GIFs, thick borders and a myriad of colors, this is no longer the case. Simplicity is the rule rather than the exception with today’s web and my website previously didn’t follow that rule near as much as I would have liked.

First there was the mix of colors. I had a dark grey background for everything but the content area, a thick lighter grey border, then stark white for the content area. At the top of the site were two images separated by my header text (my name and slogan). These images contained a mixture of green, brown and blue primarily.

The thick border and dark background I mentioned above, beyond the extra color, was an issue all its own. I found the content area felt compressed; almost claustrophobic and that issue came into play on my small 13” laptop monitor, don’t mind larger 20”+ monitors. On these monitors, the content was a skinny column of white amongst a sea of grey, with the black characters getting lost.

By removing the border and change in background color, I feel that the content now fills the page. Not literally of course! For readability, the text is contained to a column width similar to that of before, but without the visual divider. Also, the line under the header stretching across the width of the browser window gives the illusion of the content filling the page. I believe this works well on larger monitors by helping keep the content from getting lost, as well as small ones such as my laptop for general styling.

I eliminated the images in the header because although they look nice, attention was drawn away from the content below by taking up valuable screen real estate and their colors.

These changes simplified how much there is to look at and gave the website that “clean” design with only thin, simple lines to separate the content where necessary. Anything that doesn’t need to exist, no longer does.

Wasn’t a Good Design for a Blog

When I originally designed the flow and layout of this site (back in grade 11!), it was focused around my resume, portfolio and learning about me. Now it contains a blog. With the ever updating content, it made sense to have the blog as the main focus of the site.

Just as a reminder and some information for those who didn’t see the prior design: previously I had a separate home page which contained summaries of the latest 5 blog posts, then the blog page which would show the latest post. All of the other pages haven’t changed.

What was wrong with how I had the blog laid out? Firstly, the home page felt out of place and lost. Many blogs have a home page which contains summaries of the latest few posts (see ThisIsMyNext, All Things Digital and Engadget for just a few examples), so I figured I would do the same. What I had overlooked was the blogs with summary pages post far more often then I ever will. These summary pages work since people need to be able to see the what are the latest posts over the course of a day. Since I’m updating relatively infrequently, it just didn’t make sense; adding a layer of distraction from the content below.

To fix this, I made the blog page the home page and eliminated the summary page completely. This brings the viewer straight to the content and from there each post has a link to the previous one (in the future I will be adding additional navigational aids such as an archive and click-able categories).

The second issue was I didn’t have a sidebar. Yes, I consider that an oversight in the previous design. The reason? People on the web want to get in, read what they want, then get out. The sidebar provides an easy way for me to have a short description of myself (see right) as well as host other content that I don’t want to be the focus, but still consider necessary to the user experience. The alternative for this content before was a completely separate page and link in the main navigation.

Infrastructure Changes

Although the design decisions make a difference in the short term for visitors, how the site is managed and organized on the server can make a profound difference in the long term. That’s why so soon after implementing my blog system, I redesigned the structure that the blog was built off of.

The biggest decision and change that I made was the switch from my blog posts being dynamic pages to static. Let’s first look at the difference between a dynamic and a static web page.

Static vs. Dynamic

These two terms refer to how website content is generated in the web browser for the viewer to see.

On a basic level, static web pages will show the exact same thing to the user every time they go to the page. This would be the large majority of pages on the web.

Dynamic web pages on the other hand, manipulate or customize the content in some way before the user sees it. An example of this is Google’s search where one page (google.com/) shows the results for any search query (with some additional information stored in the URL). Often times a dynamic page requires accessing a database for information. Any moderately large website uses dynamic web pages to allow for easy content management; you don’t have to hire a web developer just to change the wording of something on a page, but can enter the changes into a web form.

How They Affect Blogs

With blogs, either method can be used to present the blog posts. In the initial version of my blog, I used the dynamic method which meant that I had one page which pulled the information for a blog post from the database. This, however, presented a few issues:

  1. Nasty URL: With this system, the blog post being viewed had to be passed to the blog page through the use of a post ID which corresponds to the entry in the database behind the scenes. What results is a URL of the form colincwilliams.com/Blog?post=60. This provides no information to the viewer and just doesn’t look good.
  2. Speed: With the dynamic system, the database is accessed every time a post is viewed. Database access is a slow operation and therefore increases load time for the page. For low to moderate traffic, the slow down is negligible and perfectly manageable. In high traffic situations (i.e. when extremely large numbers of people are accessing blog posts at the same time), this speed drop can become devastating; sometimes causing the website to seem unresponsive.

While I doubt I will ever face enough traffic for the second point to become an issue (although it never hurts to future proof; one link from a site like Daring Fireball will bring a blog to its knees if it’s not set up properly), I took great issue with the first. Once I was linking to my own posts and typing in the URL, I realized how bad that URL style is in this context. For that reason, as well as for the immense learning experience of doing something new, I decided to switch to static pages for my blog.

You can see the change in the URL with this, or any of my posts. It now portrays the year and month in which I posted, as well as has the full title of the blog post right in the URL. This is just another area of polish that cleans up the look of the entire website.

There is an issue with the static system, however. When I decide to change the HTML structure of the website to change the visual design, I’m going to have to regenerate all of the blog post files with the changes. This could be somewhat difficult, or at least a large operation, in the future after I have made many posts. I do believe that the improvement in user experience and future proofing for traffic increases outweighs that single complication though.

Overall my folder structure on the server is clean and efficient, yet I believe I am providing the best end user experience with these structural changes.

Conclusion

I hope you like the new design of the website. Please feel free to let me know what you think so I can keep making the website better!