20260828
- work
- explored a town alone (it was fun)
- long sleep + long nap
- I worked so hard the last two weeks.
- journaling, diet log
20260827
I was so busy getting things done and didn’t have enough time to do
journaling today.
20260826
- work (hard)
- LeetCode
- 230. Kth Smallest Element in a BST
- traverse to get a sorted list
- use heap to handle kth smallest
- journaling, diet log
## Daily Report — Aug 26, 2026
- **Weight:** 74.7 kg
- **Calories:** ~2,300 kcal
- **Meals:** Fruits 300, Protein shake 200, Strawberry Matcha Boba 400, McDouble 300, Nuggets 200, Fries 200, Milk soup 200, Nasu don 500
- **Hunger:** 4/10
- **Stress:** 5/10
- **Steps:** 21,339
- **Exercise:** 2h 1m
- Outdoor run: 15.01 km — 1h 22m (5:29/km)
- Yoga: 19m
- Yoga: 18m
- **Active Calories (COROS):** 1,707 kcal
- **Sleep:** 6h 10m
- **HRV:** 55 ms — Above normal (baseline 48 ms; normal 44–52)
- **COROS Stress:** 32
**Summary:** Very high-activity day with a 15 km run. 2,300 kcal is reasonable today; prioritize recovery and sleep rather than trying to compensate tomorrow.
20260825
- work (hard)
- LeetCode
- 1137. N-th Tribonacci Number
- 509. Fibonacci Number
- journaling and diet log
I’ll try to resume working on Grind 75
20260824
- work (very hard)
git fetch --unshallow
- tried Claude Code
- hackathon
- MongoDB University
- Networking Security: Atlas
- TLS (Transport Layer Security)
- SSL (Secure Sockets Layer)
- CA (Certificate Authority)
- I keep forgetting these things.
- AuthN & AuthZ (I wasn’t aware of the abbreviations)
- VPC (Virtual Private Cloud)
- VPC means shared network by default; peering means two separate
networks agreeing to interconnect.
I will try to do LeetCode tomorrow…
## Daily Report — Aug 24, 2026
- **Weight:** 75.6 kg
- **Calories:** ~2,200 kcal
- **Meals:** Protein bar 200, Potato salad 500, Unreal 100, Nasu don 900, Fries 200, Snack 100, Fruits 100, Ice cream 100
- **Hunger:** 3/10
- **Stress:** 4/10
- **Steps:** 13,579
- **Exercise:** 1h 28m
- **Active Calories (COROS):** 933 kcal
- **Sleep:** 5h 52m
- **HRV:** 37 ms — Low (baseline 48 ms; normal 44–52)
- **COROS Stress:** 31
**Note:** Visited SF to meet clients and felt tired.
**Summary:** High activity and calories stayed on target, but short sleep and low HRV fit with feeling tired. Prioritize recovery.
20260823
- MongoDB University
- Building an App with Code Agents with MongoDB
- Data Resilience: Atlas
- WORM (Write Once, Read Many)
- マンガで分かる心療内科 34
- LeetCode
- 509. Fibonacci Number
- memoization, tabulation,
@functools.cache
- journaling
## Daily Report — Aug 23, 2026
- **Weight:** 75.1 kg
- **Calories:** ~2,100 kcal
- **Meals:** Protein shake 100, Ramen 600, Fruits 400, Chips 100, Rice 200, Eggplant & Chicken 400, Soup 100, Fries 200
- **Hunger:** 2/10
- **Stress:** 4/10
- **Steps:** 12,937
- **Exercise:** 1h 54m
- **Active Calories (COROS):** 1,272 kcal
- **Sleep:** 37 min (COROS; likely unreliable due to unusual sleep schedule)
**Summary:** Very active day with low hunger and calories comfortably within the 1,900–2,200 target. Weight bounced slightly from 74.8 → 75.1 kg; nothing to adjust.
20260822
- rested to recover from a very tiring week at work
- MongoDB University
- MongoDB Aggregation Pipeline
- Encryption at Rest
- encryption data while it is stored, as opposed to while it is
being transmitted.
- LeetCode
- 1137. N-th Tribonacci Number
- diet log
20260821
- work (very hard)
- LeetCode
- diet log
- MongoDB University
- drawing
20260820
- work (very hard)
- LeetCode
- MongoDB University
- diet log (since ChatGPT now has access, I’m thinking about setting
up a pipeline to put daily summary in this diary)
## Daily Report — Aug 20, 2026
- **Weight:** 75.1 kg
- **Calories:** ~2,200 kcal
- **Meals:** Protein bar 200, Buldak 300, Breads 400, Sausage 100, Fruits 200, Coffee 100, Greek yogurt 100, Rice 300, Soup 200, Fishes & Chickens 300
- **Hunger:** 4/10
- **Stress:** 5/10
- **Steps:** 8,079
- **Exercise:** 1h 19m
- **Active Calories (COROS):** 859 kcal
- **Sleep:** 9h 7m
- **HRV:** 49 ms (Normal; baseline 48 ms)
**Summary:** Weight trend remains encouraging. Calories were at the upper end of the 1,900–2,200 target, with manageable hunger, good sleep, and high activity. No changes needed.
20260819
- work (very hard)
- LeetCode
- MongoDB University
20260818
- dental appointment
- work (very hard)
- LeetCode
- 70. Climbing Stairs
- 146. LRU Cache
20260817
- connected ChatGPT with COROS MCP Server, so now I don’t need to
manually log my diet
- work (hard)
- LeetCode
- 146. LRU Cache
- CPython’s
lru_cache and OrderedDict are
basically the same design. Hash Table + Circular Doubly Linked List
- created a pull request
- 70. Climbing Stairs
- MongoDB University
- B+ trees are one of the most common data structures used to
implement database indexes, especially in relational databases. internal
nodes = routing by ranges → leaf nodes = actual index entries
- In PostgreSQL, the same B-tree index can often be used for both
searching and sorting. That’s possible because a B-tree index is already
ordered by its keys. PostgreSQL can walk the index forward for ASC or
backward for DESC.
20260816
- 75.7 kg, arm curls, side raises, chest press, lat pull down, 20
minutes indoor bike, diet log
- MongoDB University
- earned 3 badges
- In my experience, aggregation pipelines get too long to comfortably
maintain
- LeetCode
- 237. Delete Node in a Linked List
- A trick that does not work for a tail node
- 146. LRU Cache
- work (some code reviews)
# First attempt
# We are not given a head node, so I thought about ways to avoid rewiring
class Solution:
def deleteNode(self, node: ListNode) -> None:
while node.next is not None and node.next.next is not None:
node.val = node.next.val
node = node.next
node.val = node.next.val
node.next = None
# This seems to be the optimal / most concise one.
# One I understand the solution, it just makes sense.
class Solution:
def deleteNode(self, node: ListNode) -> None:
node.val = node.next.val
node.next = node.next.next
20260815
- 75.4 kg, 11km run, 15000 steps in total, diet log
- work (some code reviews)
- Garlic High
- MongoDB University
- Hierarchical Navigable Small Worlds (HNSW)
- Vector Search Index
- quantization: None
- only turn it on if you’re seeing real index-memory or latency
pressure
- similarity: cosine
- cosine because you care about semantic orientation across
heterogeneous embedding models, without relying on unit-norm vectors the
way dotProduct does.
- earned three badges
- LeetCode
- 203. Remove Linked List Elements
- 621. Task Scheduler
# Definition for singly-linked list.
# class ListNode:
# def __init__(self, val=0, next=None):
# self.val = val
# self.next = next
from typing import Optional
class Solution:
def removeElements(self, head: Optional[ListNode], val: int) -> Optional[ListNode]:
dummy = ListNode(next=head)
prev = dummy
node = dummy.next
while node is not None:
if node.val != val:
prev = node
node = node.next
continue
prev.next = node.next
node = node.next
return dummy.next
Iterable -> Iterator
20260814
- 75.4 kg, 7km run, arm curls, push ups, 10000 steps, diet log
- work (hard)
- played Bomberman with my wife
20260813
- 75.7 kg, 14000 steps, diet log
- attended a MongoDB event
- https://mongodb-developer.github.io/Data-Modeling-Worksheets/web/viewer
- As long as data won’t frequently change, we can embed in a
document.
- The lecture on embed vs reference was interesting and helpful.
- I now understand how to model data in MongoDB, but I still don’t
understand advantagages of MongoDB over RDBMS.
- The employees seemed to have the same view as me; it really depends
on data and objectives. Auto-scaling is a win, but in general, SQL and
MongoDB are evolving into (kind of) similar things.
- earned 3 skill badges
- a lot of swag
- work
20260812
- drawing
- 75.8 kg, 12000 steps, diet log
- work (hard)
- LeetCode
- team dinner
20260811
- 76.3 kg, arm curls, push ups, diet log
- LeetCode
- 283. Move Zeros
- 621. Task Scheduler
- work
- drawing
20260810
- 76.3 kg, 5k run, 15000 steps in total, diet log
- work
- LeetCode
- 621. Task Scheduler
- 146. LRU Cache
A JavaBean is a normal Java class that follows standard
getter/setter naming so frameworks can read and write its properties by
name.
20260809
- 76.8 kg, squats, 4000 steps, diet log
- LeetCode
- 283. Move Zeros
- 27. Remove Element
- 2460. Apply Operations to an Array
- 21. Task Scheduler
- I gave up. I didn’t fully understand the solutions but created a
pull request to ask for review from my mentors.
- drawing
20260808
- 76.9 kg, 10 km run, arm curls, push ups, diet log
- Super Bomberman
- Hunter x Hunter
20260807
- work
- 76.2 kg, 11000 steps, diet log
- played Bomberman with my wife
20260806
- work (very hard)
- 76.3 kg, squats, chest press, lat pull down, diet log
- LeetCode
- 621. Task Scheduler
- 283. Move Zeroes
20260805
- work (hard)
- LeetCode
- spent some time with my wife
- squats, 30-minute indoor bike, 11000 steps, diet log, 77.1 kg
- I may be able to share my ChatGPT log later
class Solution:
def moveZeroes(self, nums: List[int]) -> None:
"""
Do not return anything, modify nums in-place instead.
"""
non_zero_end = -1
for checking in range(len(nums)):
if nums[checking] != 0:
non_zero_end += 1
nums[checking], nums[non_zero_end] = nums[non_zero_end], nums[checking]
class Solution:
def moveZeroes(self, nums: list[int]) -> None:
"""
Do not return anything, modify nums in-place instead.
"""
zero_count = 0
while 0 in nums:
nums.remove(0)
zero_count += 1
for _ in range(zero_count):
nums.append(0)
20260804
- arm curls, squats, push ups, 6 km run, 13000 steps in total, diet
log
- work (hard)
- LeetCode
20260803
- work
- created a small PR for my friend’s project
- LeetCode
- 27. Remove Element
- 621. Task Scheduler
- squats, 9000 steps, diet log
20260802
- 限りなく透明に近いブルー
- 12 km jog, 18000 steps in total, diet log
20260801
Augest already? Time flies by when my life is falling apart.
- My wife surprised me with a new backpack. It looks really cool.
- LeetCode
- squats, chest press, lat pull down, 10 minutes indoor bike
(intense), 15000 steps in total, diet log
- work (a bit of)
- 限りなく透明に近いブルー
- drawing
index 202607 202609