embeddedt's blog

Why it's OK to be a 'sugar cookie'

Sun, 11 Jan 2026 14:00:00 GMT

About a month or so ago, I listened to Admiral William McRaven’s commencement speech given at the University of Texas in 2014. The full 20-minute speech can be found on YouTube and is definitely worth the time to listen to (or if you prefer reading, the transcript is readily available on the university’s website).

McRaven shares 10 lessons from his time in Navy training that can be applied to many circumstances in life. Each of those lessons could be worth a post on its own, but in this one, I want to reflect on what it means to be a “sugar cookie”, in his words, and more broadly, why working too hard to avoid failure (and/or maximize success) ends up being counterproductive.

The Admiral starts this part of the speech by talking about the uniform inspection drill he and his fellow students had to partake in during SEAL training. The instructors would ruthlessly examine all aspects of the students’ uniforms, looking for even the slightest imperfection.

It turned out that no amount of polishing the uniform was enough to satisfy them. For failing the inspection, students would be instructed to wet their clothing and then roll around in the sand, creating a coated appearance that was deemed a “sugar cookie”. Several students were unable to put up with this continuous humiliation or cope with their inability to create the ideal uniform. McRaven comments that “those students didn’t make it through training.”

At first, we as listeners are tempted to sympathize with the students. After all, how immaculate does a uniform really need to be in such harsh conditions? Upon further reflection, however, we may start to realize that the story described here is not so different from how life works.

There are countless situations in life in which no matter how hard we try, it will be impossible to get the “perfect” result, whether defined as avoiding an undesirable situation, achieving some standardized grade, or just meeting personal expectations. Oftentimes there are factors outside of our control, but sometimes, we just didn’t know better to avoid the problem, or chose to ignore warning signs.

Regardless of the cause, the outcome is the same: we find ourselves stuck in a situation we didn’t ask for, possibly publicly or privately humiliated. The most natural response in such circumstances is to become bitter and negative, blame the environment around us, and/or look for any justification we can find to absolve ourselves of responsibility for the problem. While natural, this tends to just make things worse.

A personal example: some time ago I took a combinatorics course. It had a rather distinct grading scheme from your typical postsecondary math course. Instead of assignments, tests, and an exam, it had only one graded submission: the final portfolio & its associated interview, which would be due around the last week of classes. The lectures had no graded attendance, and were thus, in theory, completely optional. The only other opportunity to submit work was an optional weekly submission of material we might want to include in our portfolio, but the feedback had no marks associated with it.

Most students described combinatorics as one of the easiest math courses in the program. Without the pressure of their final mark being based on only a few hours of proctored assessment, they were free to explore and engage the mathematical content at their own pace. Indeed, the average in the course had been relatively high for many years. As a high-achieving student, the course naturally attracted my attention as a way to have some fun while still working towards my program requirements.

Unfortunately, what I failed to realize was how much I had come to rely on the structure of a typical university math course. My time allocation skills, which had worked very well up to this point, failed me as I subconsciously deprioritized combinatorics (which had no pressing deadlines) in favor of the multiple other courses I was taking, as well as my responsibilities for the 2 courses I was also teaching as a teaching assistant. By the midterm review checkpoint, my portfolio was quite polished, but far too short. I had not put enough effort into the course exercises to have content to fill it with.

It is worthwhile to note that at this stage, recovery was theoretically still possible - half the course was yet to come and there were plenty of weekly feedback submissions left. Had I dedicated time each week to work on the portfolio, it would likely have been in great shape by the end. As you can probably guess, I did not do this.

The final interview was approaching rapidly and my portfolio remained extremely incomplete. In the weeks before my interview, things went from bad to worse. I had forgotten that my grading work for one of the courses I taught would take place within the last 2 weeks of the semester. Grading that course was extremely slow & monotonous, and would leave little time and energy to properly absorb the combinatorics content and write my portfolio. I would have to choose one or the other. The choice, however, was obvious: the students needed their marks to prepare for the exam, and I could not let them down especially over a situation that was entirely my fault for creating. I finished the grading as soon as I could, and spent a few spare hours I had left adding some mediocre content to the portfolio.

Later when we got on Zoom for the long-anticipated interview, the professor, sympathetic but forced to be consistent with his prior standards for portfolios, informed me that my work in the class was only good enough for a final mark of 65.

This was a big blow to my self-esteem and I started doubting my abilities a bit. 65 was a very low mark, the worst mark I had ever gotten in any university course. To get it in the supposedly “easy” course was even more humiliating. Most of my friends couldn’t believe it was possible for me to do so poorly in a course.

This was one of those “sugar cookie” situations. The intelligence & opportunities were there, but through a combination of external circumstances and my own poor decisions, I had humiliated myself, and somewhat publicly as well. I had shattered any chance of nudging my GPA back to the admired “near-perfect” range.

But, the reality was that none of this mattered within a couple weeks. I went on to complete the rest of my exams and my marks in those courses were fine. The professor for the course I was teaching was happy with my work, and rehired me the following semester. I would also go on to double my teaching workload from the one prior, successfully.

If I had not fumbled combinatorics in that way, I may not have paid as much attention to my time management and the next semester may have gone very poorly. The outcome in the short term was negative, but overall there were many good lessons that came out of it.

Being a sugar cookie isn’t always a bad thing. Sometimes it’s the fastest way to deeply understand a lesson that may otherwise have taken months or years to slowly learn. The harm comes when you don’t look for the lesson and stay stagnant.

The key to making progress in life is to not be permanently discouraged by hardships. Look for the lesson being taught to you, and as Admiral McRaven concludes bluntly and powerfully, “get over being a sugar cookie and keep moving forward.”

There are no guarantees in this world

Mon, 16 Feb 2026 14:00:00 GMT

I haven’t written a blog post in a few weeks. Partially because I’ve simply had a busy schedule, and partially because of figuring out how to express the next thing I wanted to write about.

My blog hasn’t turned out exactly the way I envisioned when I wrote the first post. I originally thought it’d be mostly focused around technical commentary. That is still something I’d like to write more about! I spend a huge part of my time dealing with tech, both in my day job and in various side projects.

At the same time, I’ve found it very therapeutic to treat my blog as a philosophical outlet, almost a public journal of sorts. I am quite sure that this type of content isn’t rewarded by today’s search and social media algorithms, and also that many people will not find it particularly interesting. But I’m not writing for an algorithm or for the masses. If just one person reads one of my posts and takes away something meaningful from it, then my time wasn’t wasted.

This time I want to talk about one of the biggest sources of stress I’ve personally faced, and that I think we all face: the uncertainty of the future.

I am a very logically minded person, and I tend to like things being predictable. Not repetitive, but possible to anticipate and prepare for.

In today’s world, we work so hard to analyze situations and craft well-thought-out plans, whether it be for our finances, social life, things we want to build, or anything else that might come to mind. But the truth is, we don’t have any guarantees in this world, and anyone who claims otherwise is lying. If you don’t believe me, let me give some practical examples. Can you be sure you’ll wake up tomorrow morning? Absolutely not; you could die of a heart attack. Your home could burn down. You could be in a terrible car, plane, or train accident. You could be in the wrong place at the wrong time and get caught in an act of mass violence.

Perhaps you find death to be an extreme example. Consider socializing. Most of us would take the ability to meet up with friends for granted, but anyone who lived through early 2020 and the height of the COVID pandemic knows otherwise.

Programming is another good example. As programmers we try very hard to eliminate bugs from code and make it reliable. We commonly think of algorithms as “guaranteed” to work correctly. The reality, however, is that all code is written under some set of assumptions. We in the field like to call them “invariants”. As soon as any invariant is violated, the stability of the whole program comes into question. Even the most hardened code is vulnerable to issues outside its control like memory errors or disk corruption.

Yet another example is in business. We have all seen the numerous products that offer “lifetime” access, in which you are “guaranteed” to be able to access the product indefinitely. Most products nowadays are no longer feasible to download and save as an individual, especially if they rely on or are cloud services. The company might go under one day and that access is then gone forever. (I believe many lifetime access clauses include fine print for this reason.)

The last example is very “meta”: I thought this blog would be very technically focused, and that’s not the direction it ended up going in. Since I write about things I feel inspired to write about, and not a specific topic, I really have no hard guarantee on what my next post will be about. Occasionally I write a few at once and post them over a few weeks, but beyond that it really comes down to whatever situation is inspiring me in that moment.

We don’t have any guarantees, and so the best thing we can do is take life one day at a time. This is not to say that we ignore the future. It’s wise to be prudent and consider how our actions today might affect our situation tomorrow. But it’s also wise to remember it’s a privilege to wake up each morning.

Understanding that nothing in life can be taken for granted helps with being grateful for what we do have. Being grateful for what we have is something I’ve heard many times, but I found it hard to get into that habit mentally. It’s hard to take inventory of one’s life and pick out the positives when there also seem to be many negatives. What has been working is to narrow down the scope a bit more.

This past month, I’ve been trying to set a habit of finding one thing to really be grateful for each day. It was a little bit hard at first, but doable (even on a bad day, it’s usually rare to not find at least one really good moment). The neat part was that each day, this got a little easier. At this point let’s just say it’s been hard to limit the count to just one. ;)

Every day is precious, and there are no second chances or do-overs. We might get a chance to retry some things from any given day, but we can never go back and live that day again. Moreover, we don’t know that there’ll be another day. We cannot control the past or the future: the only moment we control is right now. So dive in and make it count.

How modding a voxel game taught me to slide head-first towards obstacles

Sun, 25 Jan 2026 14:00:00 GMT

If you want to change the world, sometimes you have to slide down the obstacle head first.

This intriguing statement is the concluding remark in one section of Admiral William McRaven’s 2014 University of Texas commencement speech, which I have written more about in a previous post.

The context of his remark was, of course, the rigorous and challenging training that McRaven and other Navy SEALs were put through at the start of their career. This part of the training was an obstacle course with a thin rope stretching from a tall tower down to a short one. The intended way to cross was by pulling oneself hand over hand from one side to the other.

McRaven describes how one particularly brave and creative student set out to beat the record time. The student devised the highly risky strategy of sliding on top of the rope instead of incrementally pulling himself along underneath it. While McRaven doesn’t provide many more details in the speech, it’s not too hard to envision the many ways in which this can go wrong. To start with, the student was almost certainly much wider than the rope, and extremely good balance would be essential to avoid rolling off one side. According to McRaven, one of the towers was 30 feet tall - a massive drop.

Next, excess friction between the student’s body and the rope could tear their clothing and scrape them badly, as well as reduce their ability to steer while sliding. Lack of friction, on the other hand, might make a controlled & safe stop at the other tower impossible.

With all these problems in mind, it’s not hard to see how the record for this obstacle course stood untouched for so long. Yet the student was seemingly determined to do better. Despite the immense risks involved, he chose to slide head-first down the rope anyway, and McRaven notes that “by the end of the course he had broken the record.”

In life, most problems have multiple solutions. Oftentimes, there are a few conventional, low-risk, perhaps even “boring” ways to tackle the problem. These are the approaches that most people giving advice will suggest, and it’s because they typically always work.

I by no means intend to suggest that being different just for the sake of it is good. On the contrary, I myself tend to prefer “boring” solutions to problems when possible. They’re predictable, and there’s lots of information to be found about how they work and don’t work. If you’re a developer like I am, a great article that talks about this is Dan McKinley’s Choose Boring Technology.

There are still times when thinking outside the box is worthwhile or even necessary, though. Perhaps you’re trying to get a new product idea off the ground, time to market matters, and “boring” takes too long to set up. Or, maybe you know the “boring” solution is the right one but you don’t have the budget for it. Or maybe you have already been trying the boring solution, you have hit its limits, and you’ve determined that no existing solution adequately addresses those limitations. In all of these situations, innovation is certainly worth a try.

I accidentally learned this lesson in late 2022/early 2023, when I was developing the first versions of ModernFix, a performance mod I wrote to address several problems that plagued large-scale modded Minecraft instances. One of the performance issues at the time¹ was with listing & reading resources embedded inside mod files. Minecraft mods are distributed as .jar files, which are essentially ZIPs with a special manifest as the first entry. After some research & experimentation, I found that caching seemed like a good solution to tackle this issue.

Yet the reaction in online spaces by some prominent & experienced community members to this idea was… less than stellar…

To be clear - I harbor no ill will towards anyone in these screenshots! We all get along well nowadays and are all part of the current NeoForge team (as of the time of writing).

ModernFix was my first real foray into making a “serious” mod for a modern version of Minecraft - all my previous mods were either built for older versions declining in popularity, or very simple mods that were respins of abandoned concepts. It’s also important to note that this segment of the Minecraft community also had a very negative outlook on newcomers writing performance-oriented mods after some prior bad experiences. With this context in mind, it would be entirely reasonable to drop the idea of caching and not pursue it any further. Despite that, I had a suspicion that they were wrong, and the only way to prove or disprove that suspicion would be to press on and try the idea out.

So I did. I committed the initial version of resource pack caching along with some other significant patches (1, 2) and released a test file relatively quietly in spaces I was more familiar with. It quickly became evident that I was right, and I had stumbled on some critical optimizations.

I continued to maintain an open mind towards optimization ideas people had previously dismissed. One of ModernFix’s most powerful optimizations to this day is the “dynamic resources” system, which dynamically loads & unloads block/item models as they are needed, rather than preloading them all at game startup and keeping them all in system memory forever. In large modpacks this has frequently made the game open 30+ seconds faster on weaker machines and saved at least 1GB of RAM².

Today, ModernFix sits at just over 163 million downloads, and has become an integral component of almost every large modpack for Minecraft. But the main takeaway from this post is not intended to be “how to make a successful Minecraft mod”. There is always some risk, luck, and/or providence involved whenever straying off the beaten path. I happened to be working on the right ideas at a time when not many other people were looking into them. Now that I am “old” in terms of years spent modding and the underlying technology in Minecraft is changing rapidly, I try to be more humble, knowing that it is likely not long till this story repeats itself with some new mod developer.

With experience comes the wisdom to decide when playing it safe is the right call, versus when the risk is low enough to be worth experimenting. In the case of Minecraft, experimenting was relatively low risk, as players can just remove a mod if it causes problems for them (provided they can easily identify which one to remove). In many other real-world scenarios (including corporate software development) the risk is much higher, and it would be prudent to conduct careful research and properly assess the risk before trying something novel.

Regardless, the core lesson remains clear. If you want to make a difference in some part of your life, or other people’s lives, you have to be willing to get over your fears and find a path through the barriers that you find yourself presented with at the time. It may not always be easy, but the good things in life don’t often come the easy way.

Recently I’ve seen evidence suggesting this performance problem is sometimes still present in much more recent Minecraft versions, though I believe several design changes in both Minecraft and modloaders have made it significantly less severe than it was in the past. ↩
Numbers are back-of-napkin and based on my own (flawed) memory; conduct your own testing to verify before citing them as scientific figures. ↩

How to (actually) start changing your life

Sat, 17 Jan 2026 14:00:00 GMT

(Apologies for the clickbait title, but the opportunity seemed too good to pass up.)

We humans are creatures of habit, and habits are hard to break. Our natural instinct is to gravitate to things that make us comfortable. The problem is: sometimes things that are comfortable are also harming us. It might be easy to see this harm or we might be totally unaware of it. The harm could also be indirect; perhaps our habit isn’t bad in and of itself but it takes our time and energy away from more productive or fulfilling things.

There’s an endless volume of advice on the internet about how to break habits/break out of our comfort zone, as well as a few common sayings. One somewhat humorous one that comes to mind is “eat the frog” - referring to an imaginary scenario where you have several things you need to eat, and one of them is a very unappealing frog. The idea is to get the undesirable situation over with first before tackling the rest of the day.

My issue with most of the advice is it largely restates the problem, while not providing much in the way of actionable steps one can take to make it easier. Mantras like “if it doesn’t challenge you, it won’t change you” are somewhat helpful midway through the challenge, but don’t really provide the motivation needed to get started.

I want to share some advice that did help. As I mentioned in a previous post, I was made aware of Admiral William McRaven’s speech at the University of Texas in 2014. If you have not seen the speech, I highly encourage you to check it out.

McRaven is only a minute or two into the speech when he summarizes his first key point: “if you want to change the world, start off by making your bed.” This sounds simple but there’s a very important lesson behind it.

For a bit of context, McRaven and his fellow students were going through simulated rigorous military experiences in the Navy SEALs training program. At the start of each day, they were expected to perfectly make their beds - a seemingly simple task which McRaven even remarks “seemed a little ridiculous at the time.”

“What’s so important about this?” you might ask. Making their beds each morning forced the students into a routine where they always accomplished something. It might not be anything glamorous, but it was still productive. Moreover, it meant that even if the rest of the day went poorly, they could expect to always come back to a prepared bed at night.

I believe that the “make your bed” example can be applied more broadly. It helps to push yourself out of your comfort zone in a very small way consistently. Don’t let people ridicule you for taking baby steps. Just be consistent and be content with the progress you’re making. Once you adopt this mindset, a change of habits becomes a lot easier to implement because you no longer have to do everything at once.

There is another relevant point in McRaven’s speech: as part of SEAL training there were intense physical activities. The students who didn’t perform well enough would be selected to participate in what they called a “circus”; it was actually several more hours of intense physical activity.

Nobody wanted to be part of a circus, because it meant that you were the underdog, that you would be more tired at the end of the day, and that you would probably end up in another circus the next day. That said, no matter how hard they tried, everyone eventually found themselves forced to partake in a circus.

Unsurprisingly, though, the circuses weren’t all bad. Students who regularly participated in them became noticeably stronger over time. McRaven concludes this part of the speech by advising the class to “[not] be afraid of the circuses.”

As I mentioned at the start, one really hard thing about changing habits is the discomfort of trying something new. Like the circuses in SEAL training, change is hard and it can make us feel disoriented and discouraged at the beginning. It’s important to build a positive mental space to have the motivation to keep at it (more on this in an upcoming post).

Still, once we can get over that initial stumbling block, things become a lot easier very quickly. Seeing the benefits to a new habit becomes a little bit of extra motivation to keep going, or maybe to try changing another habit.

It is very helpful to pair these types of changes together: for example, when trying to break a bad habit, do it alongside the introduction of another enjoyable (and ideally not harmful) habit. If you balance them correctly, it will be surprisingly straightforward to quit the old. However, it is critical to choose the new habit correctly, otherwise it may be even harder or more painful to break than the old one.

While nobody can expect to change their life overnight (absent the presence of a miracle), it’s already a great step forward to start by changing your mindset about it.

Why I am starting a blog in 2026

Sat, 03 Jan 2026 14:00:00 GMT

As we start a new year, I have decided to finally create and publish my personal website, as well as attempt to maintain a blog (what you are reading now). In this first post, I want to talk about why I am choosing to start a blog when blogs as a concept are declining in popularity.

There is no shortage of content on the Internet. On the contrary, with the proliferation of social media platforms and LLMs, I feel it is relatively uncontroversial to suggest that the Internet might have too much content. In recent months I have seen a growing sentiment that search engines have gotten worse in an effort to steer users towards AI-based alternatives. Perhaps, instead, the Internet has simply gotten harder to search.

It has certainly become harder to access information as a user. Blogs & forums have historically been treasure troves of information, especially when the content was posted by a subject matter expert. Nowadays, more and more expertise is being locked up in walled gardens, such as Discord servers or even traditional social media sites like Twitter/Facebook/Instagram. Content on these sites generally receives little to no indexing by search engines, and may not be possible for humans to access without creating an account.

A related trend I have seen is that users invest far less effort in learning or researching how to solve their problem. For several years I provided active support in my Discord community for Minecraft modding. It is important to note that mods primarily appeal to a younger audience, and so the user behavior patterns I saw there may not translate to the Internet at large. Still, I noticed a trend of newer users lacking initiative to look for existing solutions to their problem or at least collect as much useful information as possible for the person helping them. It is tempting (and easy) to blame them, the environment that neglected to teach them this skill, or various other factors. But again, perhaps the problem is much simpler: can they even readily find the info we expect them to?

Discord forums offer a search feature, but it is rather simplistic, and relies on user posts to have clear and distinct summaries. The classic paradigm of searching for information breaks down in an era of unprecedented complexity in the tools and world we use & live in today.

I would like to propose that perhaps blogs, for all their faults (lack of peer review, ephemerality) do a good job of solving these problems. Blogs are rather easy to index and search. They are typically public sites that require no authentication to access. Each post tends to focus on a single topic, and today’s LLMs offer a potential path to SEO for authors without time to write explicit keywords and summaries. I may be wrong in this proposal, but at the very least, the information I choose to publish here will not be forgotten so easily as in a random Discord server. :)