FiveYellowMice's Blog

Intrigued by Randomness: Mahjong, Crappy Machine Learning, and Gatcha Games

Tue, 26 Apr 2022 06:28:00 +0000

It's not possible to display full formatting in an RSS reader, different RSS readers may also display drastically different layouts. Therefore it's recommanded to read the article by clicking the web URL</a>.

Humans are pretty drawn to randomness. That’s probably a well-known fact, but I’ve just come to understand on how true it is, and I want to talk about it with regards to some of my personal experiences, as presented in the title.

Mahjong is a complex game. I’ve seen elders gather in community space and play it, but never had any idea why the tiles held by winning players make them win. There are more interesting games to play though, so I didn’t try to learn it. That is, until the online game Mahjong Soul came out, which is just Japanese Mahjong with random players online. But the game is obviously targeted to my generation, with cutesy artstyle and player avatars being anime girls (they even added anime guys later). So lots of my friends started playing it, and pretty enthusiastic, too.[1]</a>

A game in Mahjong Soul</figcaption></figure>
I tried to dabble in it, but the new player guide is filled with undecipherable jargon, unknown phrases explained by even more unknown phrases, 20+ different winning conditions, all intertwined. The rules overwhelmed me to no end, so I gave up. Until this year, when many of friends are still playing it often, they instigated me to play again. I said I can’t understand the rules, threw them bags of questions like "what does phrase X even mean?", in attempt to show them my inability to understand the rules. But somehow, this time, they actually had the patience to answer all of my petty questions. Over an hour later, being the one asking questions, I inevitably started to grasp the rules of Mahjong, or at least, quote one friend, "about 20% of Mahjong rules".
Wait, what, there’s more? That took such a long time and so much patience out of everyone, and I only learned about 20%? I exclaimed, "That just shows how complex Mahjong’s rules are!"
They suggested me to "just dive in, understanding will come naturally". But my experience disagrees. All I can do is play robotically or arbitrarily. There is no time in each turn to think everything through, most of the time I have no idea what I am doing. There is little feedback otherthan winning. If I take some action, it is very difficult to know whether that action is beneficial or harmful. When I lose, even if every action taken in that game are good decisions, I wouldn’t know. If I do win, what does that mean? Which decisions have lead to the win?
Comparied to other tabletop games like chess, Mahjong is much more complex. It requires such a high commitment, one needs to sit down and study it, just to be able to start playing. Whereas in chess, learning the rules is quick and feedbacks are easy to come by during plays. A piece is taken, that likely comes from a previous wrong move; I put the opponent in a difficult position, that probably means I did something right. To me, chess is a superior game than Mahjong.
But I missed a critical factor: Chess has zero randomness [2]</a>, it’s a game of pure skill (or in my case, at least blind luck). Mahjong is over halfly determined by chance, the starting hand is random, each tile is drawn from a shuffled deck. Skills do matter, but each player is still at the mercy of Fortuna. Better players are merely "more likely" to win, which only show up in a large number of games. In an isolated set of several games, worse players still win sometimes. This means, as some other friends pointed out, Mahjong is a great party game, the fun comes from the randomness. Indeed, if all that happens is weak players being crushed by strong players, it wouldn’t be fun for a group of people who have different skill levels. That happens with chess, and all other games that emphisize skills.
The luck factor also brings thrills to experienced players. Something unpredicted can always happen, anyone may have drawn a tile they need at any time, or not having drawn a useful tile for a long time. There can always be something new. I suppose that is why the elders in community space play it from day to night, seemingly tireless.

A couple of years ago (I can’t believe it has been that long), I wrote a chatbot, HoroBot</a>. It runs in a couple of Telegram and IRC groups. It’s core functionality is to send random emojis whenever discussion is happending in the group. It is just for fun, seeing how merely sending random emojis can make a bot seem to blend in a group chat. This was a success, some didn’t even notice that it’s a bot hanging out amongus, for a long while. They joked that it has passed the Turing test. Later, I added more functionalities to it, one of which is detecting whether a chat participant is "of the bot’s own kind". Because Horo</a> is a wolf, this feature is called the "wolf detector".
The way the wold detector works is based on a Bayesian classifier. Don’t worry, you don’t need to know anything technical. And neither did I. All I knew was that it’s a tool that lets a computer program to categorize sentences automatically. First, train the program with a set of sentences and what category each sentence is known to belong, then, with enough training, the program should know what a sentence in each category "looks like", so that when I give it another sentence, it will tell me its category. It’s a form of machine learning. Such things is often used for detecting spam messages, or whether user reviews to products is positive or negative. I wasn’t using it for anything that serious though, I wanted HoroBot to tell whether a chat participant is "of HoroBot’s own kind" or not.

Graphical explanation of this type of machine learning. Source: Yan, Ma et. al.</a> CC-BY</figcaption></figure>
For my application, there are 2 categories, "is" and "is not" (of HoroBot’s own kind). I trained the program with past chat log of the ArchLinuxCN</a> group at the time. Messages sent by anyone with name containing "horo" is considered to be in the "is" category, otherwise "is not". It was obviously pointless, but the whole thing was just for fun. After training, whenever HoroBot sees another message, it will give a category of that message, and the categorization given to the last 100 messages sent by one user determines their score, i.e. "probability of being the bot’s own kind".

Someone (top) using the wolf detector. HoroBot (bottom) replies with "(redacted) has a 56% change of being of my kind."</figcaption></figure>
The results are, as expected, pretty useless. What does even mean to be "of HoroBot’s own kind"? And because there is no way of interpreting this score, there is no way of telling how accurate it is. It appears that most of the results are around 50%, which gives the suspicion that if I put everyone’s scores up, it’s going to be a normal distribution, i.e. the category of each message is pretty much random. I have not tested that suspicion, though. Actually, it’s potentially worse than random guesses, the "original horos" (people who have names containing "horo" in the chat logs I used for training) tried to see their score, and they were consistently lower than others!
I will confess now, that I do have some idea why my Bayesian classifier was so crappy. A Bayesian classifer needs a sentence to be a list of words, rather than a list of characters. I need a way of segmenting the sentences into words. In Chinese, there are no spaces between words, so segmenting is not trivial. I wasn’t bothered to to anything elabourate, so I just decided arbitrarily: every 2 characters is a word! That is clearly a terrible way of doing segmenting. It has probably made every sentence meaningless.
Incredibly, despite all that, this is the most popular feature of HoroBot! To this day, several years later, when I have not updated the bot in a long time, participants of ArchLinuxCN chat group still seem to see their wolf detection scores often. It puzzles me: It’s just a pointless, practically random number! Why have they not gotten tired of it yet! But well, I guess, I made something that people liked, it should be something to be pround of.
This is purely baseless speculation, but I guess one reason of the wolf detector being popular is it’s unpredictability. The result is not random, the same message always yields the same categorization, but over different messages, it appears random. Yet we know the score is somehow related to the messages a person has sent, while that relation is unknown. Such randomness intrigues us, as we are curious to see "Does sending this message make my score go up or down?" Psychologists say that humans loves finding patterns in random events. Here we do have patterns, how unsensible it may be, it’s somewhere within the algorithm, and knowing this makes us wonder all the more strongly.

I’d like to think of myself as a relatively rational person. Playing gatcha games is a thing to be looked down upon in my book. Here, gatcha games refer to video games that has subpar gameplay, and the main attraction is collecting "cards" which are obtained randomly upon paying microtransactions. The cards are usually depicted as lovable characters.
The profits of gatcha game makers is surely betted on our attraction towards randomness. Each time one pays for the microtransaction, they can be intrigued by the anticipation of their new card (or lack thereof, most of time in this type of games, getting a collectible card is a rare event, usually they are less exciting stuff). The problem is, being attracted to that means I become more likely to spend more money on the totally useless cards, and devote time to virtual characters who have absolutely no care for me (other than their creators wanting to extort for more money). This is different to being drawn towards playing Mahjong with friends, which at least practices friendship; or watching over a wolf detector score, which doesn’t cost much.
I would argue that playing gatcha games is a lot like drinking alcohol. Doing it a lot is definetely bad. Doing it a little is probably acceptable, but unhealthy nonetheless. To be fair, it is really difficult to resist paying for the card draws, so I resorted to just stay away from gatcha games. That resulted me in missing out on some games with gatcha mechanics but do have some content, unfortunately.

One became so devoted in Mahjong, he even went ahead and made an actual AI for it</a>. It was just published recently. ↩</a> </li>
Other than maybe using some random means of determining who gets to move first. ↩</a> </li> </ol></section>

Unless otherwise stated, the original contents on this website are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License</a>.While reposting, please state that the article comes from FiveYellowMice's Blog</a>, and include a link to the original article</a>.

致大学招生办：我好菜啊

Mon, 22 Nov 2021 16:56:00 +0000

在 RSS 中无法显示完整的格式，不同的 RSS 阅读器所显示的内容也可能会有很大差距。为了能够获取到完整的内容，最好还是点开网页链接</a>看。

虽然难以接受，但自己已经远远脱离可以自称“中二少年”的年纪，即将被社会毒打了。不过听说有一种办法也许可以减轻这种毒打，同时还可以延缓它的开始——读硕士。讨厌的是，想要申请许多学校的硕士，需要一个黄鼠尤其不擅长的东西：写自荐信。
这自荐信没有框架，没有标准格式。写出来会不会被录取，完全凭靠招生办审阅者的感觉。他们说：“自荐信要表达出你独特的地方，表现为什么你是这个课程的最佳人选而其他人不是。”前半句很好满足，因为黄鼠确实觉得自己很独特——哪个人不是呢？黄鼠有着独特的特征、独特的经历、独特的能力，就像每个人一样。可是自荐信不是征婚启事，也不是可以随口说任何内容的博文，里面写的必须得是和要申请的硕士课程相关的内容，对吧？像“喜欢二次元纸片人”这样的信息，应该是不合适的，除非申请的是美术系。而在这个限制之中，黄鼠就没有那么独特了。会做黄鼠会做的事情的人在时间上数不清，其中也有不少人大概做得比黄鼠更好。至于“表现为什么我是最佳人员而其他人不是”，就完全没有头绪了。黄鼠自己都不觉得自己的相关能力有多独特，怎么能够断言别人做得不如黄鼠呢？比黄鼠条件更好的人大有人在。自己都不相信的事情，要怎样说服招生办呢？
黄鼠高中的成绩很差，因此没有足够的分数在大学进入自己心仪的计算机科学专业，而是不得不学了一个算是相关的专业，网络工程。 4 年即将过去，黄鼠感觉在学校中似乎并没有学到很多（值得花费 4 年的）东西，时间似乎都花在了来回的交通，关心各种行政事物，还有满足各种繁杂考核标准上。倒是在自己的时间里进行的学习，似乎反而还不如中学时的多。有趣的课程是存在的，可是再有趣的东西，被掩埋在时间压力、行政事物、标准考核下，再被批判的目光盯着，并且这个批判目光可以非常轻易地使黄鼠白交学费、白花时间、延迟毕业、打乱一切计划——的时候，也变得枯燥乏味了。于是黄鼠采用了这样的心态：不指望从学校学到什么，主要的能力靠自学，从学校只是为了得到文凭。但是大概是因为自己比高中时变废了，并且也因为在课业中已经敲电脑足够累了，黄鼠在课外进行的自我提升反而减少了。
长辈说，大学并不只是为了学习和得到文凭，还有一个很重要的收获是“人际关系”。可是……黄鼠完全没有在那方面做任何努力，也不会突然无厘头地和其他人搭话。并且因为某个来自武汉的全世界流行病，许多时间都是呆在家里上网课，杜绝了除上课和完成小组作业之外的人际交流。倒是反而与几乎都是在地球另一面的网友们相处的非常好。
现在黄鼠就要从网络工程本科毕业了，终于也许可以有机会在硕士去搞计算机科学了。这两个应该向是相关的专业，想要过去应该是蛮直接的吧？可是光是最低要求就使黄鼠与许多学校的计算机科学硕士无缘：有些是说必须要有计算机科学的学士学位，有些是在后面加注了“或同等学位”。黄鼠怎么知道自己的“网络工程”是不是算是同等？有些地方说这个要求是故意写的这么模糊的，要看每个人的具体情况，这个黄鼠喜欢；但有些地方写了很明确的怎样算是“与计算机科学学士同等”，要包含某些特定的课程，像是“基础算法”、“离散数学”、“数据结构”、“面向对象编程”什么的——都是黄鼠没有在学校学过的东西。就在这时，黄鼠才意识到了自己的本科与真正的计算机科学本科是有多大的差距。编程黄鼠当然是会，但是并没有任何官方机构可以为黄鼠证明；而其它 3 个黄鼠是完全没有系统地学习过。即使是在没有列出这些明确要求的地方，黄鼠想他们大概也是会默认申请者都有这些知识的吧？也许是，也许不是，黄鼠不知道这会不会为自己在审阅者眼中减分。
黄鼠觉得自己还是有资格被称作“极客”的。但是仔细想一想，自己真的有做出什么出众的事情吗？听说贡献开源项目是一件极客们经常做的事情，可黄鼠没有为任何大型开源软件贡献过代码。看一眼自己的 GitHub ，黄鼠有贡献过的最大的开源项目有两个：一个是给软件的文档加了一句话</a>，另一个是 Dress</a>⁠[1]</a> 。听说在技术社区中活跃也是一件极客们经常做的事情，那黄鼠在 ArchLinux 群水群算吗？黄鼠没有打过包、贡献过翻译、管理过什么东西。黄鼠唯独能称作是极客的地方，无非就是写了一些个人项目，一些没什么挑战性、没有什么创新、几乎都是只有自己用、都是玩具质量的软件，换句话说，都是在“造轮子”。朋友建议说，即使是个人项目也可以写它创造了什么，解决了什么挑战。可是黄鼠全部在做的，都只是在学习怎样调用别人事先已经写好的框架，做出玩具的样子。造轮子能够有什么挑战和创新呢？黄鼠能够把自己的轮子们列举出来，并且说明一个个都是做什么的，但是那样听起来无比寒酸。放在企业里面，黄鼠能够做到的事情大概与 3 个月速成培训班出来的人无异，是像民工那样堆砖砌瓦的工作。堆砖砌瓦当然很重要，但是黄鼠不想要做那样的事情一辈子，要是想的话，完全不需要去读硕士。
黄鼠造过什么轮子？有许多 Telegram bot （聊天机器人），大量的 bot 。有会在群里随机发表情的 bot ，有生成图片的 bot ……尽管写它们的过程对于黄鼠来说都很有趣，并且有网友喜欢，但显然不能够被称作是任何“技术壮举”。黄鼠做过一个 LED 时钟</a> ，但是那只是包含了一些基础又粗糙的木工，一些用几个小时就可以上手的电路板绘制，以及一些面向初学者的嵌入式软件开发。黄鼠在做完时钟之后，被网友推荐加入了一个“嵌入式”的群，然而群友们在说的话，大部分时候黄鼠都听不懂。他们都比大概与黄鼠年龄相仿，讨论的内容却高几个层次。黄鼠想，自己要申请研究生的话，竞争对手们大概有许多都是那样的人。
黄鼠有许多失败的个人项目，像是现学现卖的过时视觉小说游戏引擎教程</a>，像是愿景过于宏伟的博客评论系统</a>。黄鼠能够把他们称作自己的“经验”吗？它们的失败，是因为自己失去了兴趣与失败的时间管理。可是“对事情容易失去兴趣”和“不擅长时间管理”应该都是作为研究生申请人不被看好的特征吧？网上说“如果没有相关技术经验的话，无关的经验也可以展现自己的软能力，像是时间管理和团队协作”，可是刚刚说的事情只能够证明黄鼠没有时间管理的能力。而从团队协作而言，个人项目只有自己一个人，有和其他人合作的只有学校的小组作业，而那无一例外都不断在证明黄鼠团队协作能力的缺乏。
计算机科学硕士在黄鼠看到的大学中通常有几个子方向，像是机器学习、人机交互、图形、安全、算法、数据库、编程语言、系统和网络，哪一个是黄鼠想要做的？从初中就决定要做这方面的职业的黄鼠应该有很长的时间来想，可是直到现在自己真的要选择了的时候，才有认真去想。唔……自己好像哪个也许都可以做？那么，哪一个好找工作？不对，黄鼠怎么能这样功利，如果因为功利选择了自己不喜欢的方向，痛苦就会充满未来了。所以黄鼠究竟要在自荐信中说自己要做哪个子方向呢？按照自己在个人项目中写了许多网页前端来想，黄鼠应该会喜欢人机交互，可是其它好像也还行，黄鼠也不知道自己在那个领域能够做什么——黄鼠只是有自己造过轮子，完全没有系统的学习过，不知道真的学习起来会是什么样子，更不知道做研究是如何了——万一黄鼠不喜欢怎么办？导师选定了应该很难后悔了吧？与之相对的，黄鼠在本科所学习的网络工程，最相关的子方向是系统和网络，黄鼠觉得自己的成绩还行，所以应该算是已经尝过了它的味道，还有文凭证明，所以也许做系统和网络会更好？黄鼠当然可以把两个都写上去，可是两个都不能够称作是十分地有基础：一个只有自己造轮子的经验，一个只有上了几门课的经验。
对了！为什么一定要做研究呢？黄鼠只是听朋友说“上课这样久会库走，做点项目更有趣”，觉得很同意，所以想去申请研究型硕士。可是黄鼠从来也没有做过和“研究”有半点关系的事情，怎么知道“研究”不会令黄鼠痛苦万分呢？授课型硕士更加简单，可是那样就感觉自己放弃了很好到的机会——黄鼠连这样的决定都无法做到，审阅者一看：这个人怎么这么没有点子，大概不管是研究和上课都不太行。黄鼠可以写两份申请，但是研究和授课的自荐信侧重点不同，写一份已经如此焦头烂额，写两份……有时间再说吧。
一所大学的“自荐信编写指南”视频</a>中，提到说自荐信要包含“你的方向如何与导师相符”。黄鼠连自己的子方向都不确定，具体要做怎样的研究那是更加没点子了。要和潜在导师讨论的话，总不能说“我完全不知道自己想要做什么研究，等过一年再看着办吧”这样的话，一定会降低自己的亲和力的。在同样的视频中，也提到说如果没有研究点子的话，展现自己有做过相关阅读也可以，读过相关的研究，就可以说“我对这个领域有充分的了解，我有信心在这里做研究”了。黄鼠没有做过相关阅读，要现场抱佛脚读一些论文吗？在现在这个距离报名截止日期十分近的时候，黄鼠觉得要静下心来画上几天时间读论文（黄鼠从来没有读过除了摘要以外的论文！），再对这个领域产生充分了解，再对自己产生信心，是一件十分困难的事。实际上，黄鼠有浏览过学校导师们的过往论文标题，可对它们的感觉都大体是“我难以想象自己可以做与这相似的事情”，要么是因为觉得很平淡，要么是觉得太高端了完全不懂，要么是两个因素的集合——这听起来好像不太妙……吧？黄鼠完全没有头绪自己要做什么。
指南中说，自荐信要展现自己的独特。黄鼠想，自荐信一定要写得真诚：如果虚伪的话，不仅大概写不出来，而且如果按照虚伪的说明被录取的话，大概面对的将会是痛苦的生活了。可是黄鼠的独特优势是什么？黄鼠要真诚地说什么？越是构想，越是觉得自己真诚地应该对招生办说：我好菜啊。你们看，我会这些许多人都会的技能，可能这个技能组合不多见。你们觉得我行吗？
可是黄鼠显然是不能这么说的。申请人自己都不能说服自己适合这个学习和研究，显然是不能够指望阅读上千份申请人的自荐信的招生办被说服的。
黄鼠十分认同，自荐信真的是一个非常好的筛选学生的方式。它强迫每个人必须深刻地审视自己，了解自己，才能够写出来它。对于黄鼠来说真的十分有效，即使没有被任何地方录取，有过这样的自我审视应该也是难得的吧。
然后黄鼠成功地被拖延症带领着写完了这篇博文，自荐信正文却还是只有几行 _(:3)<)_ 。
“直接把这篇博文当作自荐信发过去吧。”黄鼠有这样的冲动，但是那样大概就是自暴自弃了。

一位朋友说，如果自己是跨性别者什么的，也可以写进去。可是……抛开黄鼠是不是不谈，即使是，写进去也不能够是自己变得更被审阅者看好吧？总不见得大学招生会有 LGBT 优待什么的。倒不如说可能会使人觉得申请人在强行博取同情。 ↩</a> </li></ol></section>

除特殊说明以外，本网站原创内容采用知识共享署名-相同方式共享 4.0 国际许可协议</a>进行许可。转载时请注明来源 FiveYellowMice 的博客</a> ，以及原文链接</a>。

More Than a Yuri Story: Thoughts on Heart of the Woods

Mon, 05 Jul 2021 04:09:00 +0000

It's not possible to display full formatting in an RSS reader, different RSS readers may also display drastically different layouts. Therefore it's recommanded to read the article by clicking the web URL</a>.

Characters in the title screen: Abigail (left), Maddie (right)</figcaption> </figure>
Heart of the Woods looks like a nice interesting love story about 2 girls, one human, one ghost. It’s homo, so it’s just what I needed to escape from the real world. Although it appears to involve love across the living and the dead, which is a concept that’s been done to literal death. But hey, the game is well-received, it’s on sale, the artstyle is new to me, and most importantly, it doesn’t appear to be too long. So I guessed it’s going to be a nice fit for the spare time I currently have.
Boy was I wrong. It does not use the old plot device of "one grows old and die while the other stay for eternity" or "the border between life and death is unbreakable so we can’t be together" that human-ghost relationship stories tend to have, which I imagined to also be a leading aspect of this one, too. Heart of the Woods has so much more than that.
Note: </h6>
There is no spoiler beyond the first chapter.</aside>

The opening scene, Tara (left) and Maddie (right) lie on the train seats</figcaption></figure>
The premise of the story involves a YouTuber that talks about paranormal events, Tara, and the assistant, Maddie. The two travels to a supposedly-haunted town to investigate for their show, because a local of the town as well as a viewer of the show, Morgan, invited them. The ghost appears much later in the story, and is named Abigail (feels pretty unimaginative for a ghost name, to be honest).
At this point, even though none of the interesting things have surfaced. I was quickly drawn in by the sheer beautiful writing. Maddie’s opening monologue very effectively introduced a realistic situation: Maddie has been a long-time friend to Tara, and an assistant to her paranormal events show, but Maddie was going to quit and chase different life goals after this trip to the haunted town. Her guilt, her internal struggles and worry for Tara and all the complex emotions was very expressively shown. I felt I could be in Maddie’s shoes. Through her interactions with Tara, I could feel the awkwardness between them. Tara is salty, Maddie is guilty, but they also know they should keep their cool, but at the same time it’s also really hard to get over their emotions, however they still have a final episode to make… This kept going for a long time, it felt real.
It felt funny that I came to this game because of the official pair of Maddie and Abigail, but before that main course came, I was already fascinated by the relationship of Maddie and Tara. I was already worried about them. As Morgan teased them with paranormal phenomena, Tara was hyped and ready to believe, while Maddie was skeptical. Maddie thought Morgan was maniac for claiming crazy things but can’t prove them, and for her infame among townspeople. I actually shared Maddie’s opinion for a while, even though as a reader, I know for a fact that there’s a ghost and so Morgan is at least partially right.
Throughout the former half of the story, the situation keeps getting more complex, and the stakes keeps getting higher. Even without half of the official pair (Abigail) appearing, the 3 characters already seem so interesting. I was just expecting a sweet yuri story, but as I read it, it becomes full-on suspenseful. The suspense peaked at the end of the first chapter, but it did not just decline after that. Things keeps changing, and there is little to no filler texts, no boring segments about "daily lives".
I really liked the romantic parts of the story as well, they complement with the suspenseful parts perfectly. With such good writing, the characters are fleshed out, and their thoughts feels grounded. Their relationships feels human. It’s lovely. With no spoilers, allow me to say that there are lots of hugs, cuddles, and kisses. I felt them, embarrassingly. Even though I’m usually not a fan of erotic scenes in visual novels, I would probably sit through them if this game had any. (There is actually an R18 patch, I didn’t know about it.)
An interesting thing I liked about the character settings is about Tara. The first thing I noticed about her is the voice acting. Her voice sounds unexpectedly deep, almost like… the voice a transgender person could make. I thought "it can’t be", but I looked at the credits, and apparently Studio Élan did find a transgender voice actor</a>, and Tara did mention in one line of dialogue that she is trans. The impressive thing is that nobody in the game ever cared about Tara’s voice, even ones that are unfriendly. Tara being trans is not part of the plot and only briefly mentioned. I think this is a massive "trans rights!" shout of how they are just regular people being themselves and we don’t make dramas on their gender. I’m behaving worse than the characters in the game because I’m making a drama about it in this paragraph.

A guest art by Roarke</a> taken from the game’s guest art gallery. I died as I see this.</figcaption></figure>
Let me hope that this article has made you give some interest to the visual novel. It is sold on itch.io</a> and Steam</a>. Ultimately, I think my experience with Heart of the Woods proves the point in my previous article</a>, how good a story feels comes from subverting expectations. This novel has subverted mine a lot, and hopefully I have not spoiled it too much to let it also subvert yours.

Unless otherwise stated, the original contents on this website are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License</a>.While reposting, please state that the article comes from FiveYellowMice's Blog</a>, and include a link to the original article</a>.

Subverting Expectations

Tue, 13 Apr 2021 22:05:00 +0000

It's not possible to display full formatting in an RSS reader, different RSS readers may also display drastically different layouts. Therefore it's recommanded to read the article by clicking the web URL</a>.


I think a huge factor in how good a piece of artwork feels to me is how much it subverts my expectations. The "artwork" in question can be literature, illustration, music or perhaps more other types.
If a work sets up a high expectation from me, by either making really good marketing materials or its author(s) is/include someone who has produced good works before, it will need to reach or overachieve for it to not be "disappointing". It can still be objectively really good, but for me, because it is comparatively worse than the expectation, it may feel like as if it’s actually bad.
Likewise, if I don’t have much expectation for a work. For example, a low-budget work from someone I have not heard of, or works that does not appear to be serious (like a comedy or a meme). As long as it turns out to be better than that low expectation, I will feel like it’s amazing. Even if objectively, it’s so-so at best.
I want to give some examples. The movie Your Name directed by Shinkai Makoto. It has a highly-reputed director, stunning marketing posters. And before I watched it, countless praises had been seen from my peers. The resultant expectation was floating to the skies. After I watched it, I felt it was "not bad", but was it worth the hype? That was hard to answer. The plot progression felt like there’s too many coincidences, the main characters' experiences seem to just be a road laid down by the writer (which to be fair, it obviously is) rather than a believable adventure. So right after I finished the movie, I gave it 3.5/5 stars. I felt this was exactly what "overrated" means. But let us think rationally, is Your Name a good movie? I suppose most people would agree, and I agreed after I thought rationally.
On a similar vein, also an anime movie, Sword Art Online: Ordinal Scale, but this one faces the opposite scenario. The Sword Art Online series is not known for it’s depth. In fact, its story is considered really shallow and only good for "thrills" by critics. To be honest, I like Sword Art Online, I mean, I like plain thrills without much depth. So I watched the movie, and it offered me exactly what I expected: thrills without much depth. It actually overachieved a little because the movie has flashier visual effects than what’s in the TV series which I was used to. I somewhat felt better after watching this than after watching Your Name, but I would not put the 2 works on the same level.
There is a work I surprisingly liked because of this weird reason. That’s One Room season 1. The 12-episode 5-minute-each anime series that shows each of 3 girls (they don’t appear together, each one has 4 episodes dedicated to one) talking to the camera as if she is spending time with the viewer. The makers call it "first-person anime". It felt like a novel concept, but as you may guess from its description, it’s going to be an awkward anime to watch. I obviously didn’t expect much from stereotypical anime girls talking to a silent camera. I just watched for the novelty since even if it was terrible, I only lose 5 minutes each week. The first 2 characters (8 episodes) was as I expected, kind of sweet but still boring, but it only lasts 5 minutes so I didn’t mind wasting those 5 minutes 4 more times. Each character has a song sung by the voice actor and used as the ending song, the songs are ok.
Spoiler Alert: </h6>
The following 1 paragraph contains spoiler for One Room season 1 episode 9-12. Although you may not think spoilers for this work is worth worrying about, the way I wrote has indicated that those episodes are unexpected. They’re probably worth a watch, they are only 5 minutes each anyways.

违背期望

Tue, 13 Apr 2021 22:05:00 +0000

在 RSS 中无法显示完整的格式，不同的 RSS 阅读器所显示的内容也可能会有很大差距。为了能够获取到完整的内容，最好还是点开网页链接</a>看。


我觉得影响自己对一篇艺术作品印象好坏的很大的一个因素，是它与我的期望中有多不同。这个艺术作品可以是文学的，绘画的，或者音乐的。
如果一部作品使我对它产生了很高的期望，不管是通过宣传材料还是它的作者（们）（之一）是一位曾经创作过优秀作品的人，那它都需要至少达到、或者超越这个期望值以使得它不显得“令人失望”。这部作品可以客观上很优秀，但是对于我来说，因为它比期望值要差，就有可能感觉好像它真的就是差一样。
同样的，如果我没有对一部作品有很高的期望，比如可能是一部来自从来没听说过的作者的预算比较低的作品，或者是看起来就不是很认真的作品（像是喜剧或迷因）。这时只要它最后比这个低的期望指要做得好，那我就会感觉它好像很厉害，尽管实际上它可能最多只能称作是“普通水平”。
还是来给一些例子吧。被新海诚导演的电影《你的名字》，有一位声望很高的导演、惊艳的宣传海报，并且在我观看之前还看到了非常多的来自其他人的夸奖。其结果就是对它的期望漂浮到了天上。然后在我观看完毕后，感觉到的却只是“不错”，这部电影真的值得那么多赞扬吗？这个问题不好回答。剧情的发展感觉有太多的巧合了，主角们的经历感觉更像是编剧为它们铺好的道路（当然，实际上也确实是这样），而不像是使人信服的冒险。所以在我刚刚看完电影的时候，只给了它 3.5/5 星的评价。我感觉这就是“被高估了”所代表的意思。但仔细想一想，《你的名字》是个精彩的电影吗？大部分人应该都会认同，我也在仔细想了想之后认同了。
在相似的领域，也是一部动画电影，《刀剑神域：序列之争》是相反的情况。《刀剑神域》这个系列可不以有深度的故事而闻名。实际上，很多评论家都觉得它很肤浅，只适合看看爽快。诚实说，我喜欢看《刀剑神域》，应该说，我觉得看看没有深度的爽快很开心。于是我去看了电影，然后电影也提供了刚好我所期望的——没有深度的爽快。实际上可能还稍稍超过了一点，因为电影的视觉效果要比我所习惯了的动画的视觉效果要更加华丽。甚至感觉在看完这部电影后比看完《你的名字》后还要感觉更好，但认真地说，两部作品可不会被我放在同一个级别。
有一部神奇的作品，恰好因为这个神奇的原因使我喜欢，它是《One Room》第一季。是一部共有 12 集，每集 5 分钟的动画系列。它展现的是 3 位少女（不是在一起出现的，而是每位各有属于自己的 4 集）对着镜头说话，像是在和观者在一起渡过时间一样。动画的创作者们称它为“第一人称动画”。这个点子听起来很神奇，不过从上面的描述应该可以看出，应该会是一部看起来会令人感到尴尬的作品。我显然没有对它抱有很高的期望，既然只是动漫美少女对着沉默的镜头单方面说话。只是它反正只有 5 分钟，为了这个神奇的点子，即使它就是很烂，每周浪费了 5 分钟也可以接受。前 2 位角色（前 8 集）的故事同期望中一样，有一点趣味但实际上还是比较无聊，不过没有关系，反正每集只有 5 分钟，再继续浪费 4 次这样的 5 分钟也没有关系嘛。每一位角色有着属于自己的角色歌，被作为片尾曲使用，这些歌曲还可以。
剧透提醒: </h6>
以下 1 个段落包含对 One Room 第一季第 9 到 12 集的剧透。尽管你可能不觉得这是一个值得担心剧透的作品，但我写文章的方式应该已经表示了这最后 4 集是出乎意料的。也许可以值得看一看，反正每集只有 5 分钟嘛。

Making a Custom Smart LED Clock

Sat, 20 Mar 2021 09:10:00 +0000

It's not possible to display full formatting in an RSS reader, different RSS readers may also display drastically different layouts. Therefore it's recommanded to read the article by clicking the web URL</a>.

The finished product displaying a time of Tuesday, 16 March 2021 13:37:07</figcaption> </figure>
After around 2 years, my first (presentable) DIY project is finally complete! It’s a device with a 16x16 LED matrix display, which is primarily used to display time. It’s also WiFi-connected, syncs time via NTP, can have 3 alarms, and potentially do more interesting stuff with its internet capability.
So, this post will document the process of me making this thing. I didn’t have many mishaps, the process is relatively smooth, and the reason it took so long is mostly me procrastinating.
For a lack of name ideas, I used a random pet name generator online, and it gave me the name "Baby Tiger". This is probably not the best name for a digital clock (it sounds like a very generic name for a cat), but I have no better ideas. Name suggestions are welcome.</section>
The Electronic Components</h2>
I guess it’s because of watching YouTubers like GreatScott!!</a> and DIY Perks</a> piqued my interest and I wanted to make something of my own. And a clock with LEDs sounded like a fairly conservative thing to make that would not present much trouble, since similar hardware and design patterns seems to be done by others a lot. So with this thought, and the knowledge from YouTube videos that a kind of microcontroller⁠[1]</a> called the Arduino would be suitable for controlling LEDs, I bought an Arduino Nano knockoff⁠[2]</a>, breadboard⁠[3]</a>, jumper wires and RGB LEDs.

The first RGB LED I got</figcaption></figure>
This RGB LED is a package of 3 LEDs of colors red, green and blue. I need to power each color individually in order to display different colors. I made it to work, but there were big catches:

I can’t adjust the brightness of each color components directly, so I can only have the primary and secondary colors, plus white and black (off). I had to use PWM⁠[4]</a> in order to make fancier colors.</li>
The weirder problem was this RGB LED is unlike just 3 LEDs in parallel, the red, green and blues LEDs can’t be on simultaneously, which means I had to do PWM faster and the software needs to be a little more complicated.</li>
Each LED consumes 3 pins for the 3 color components. I was making a clock, so there eventually needs to be a large quantity of LEDs connected before displaying anything resembling time. There aren’t that many pins on an Arduino Nano. I learned about shift registers that can "expand" the number of controllable pins, but then I had to do PWM with software which felt very complicated. Shift registers will also make outputs slower, combined with the issue above, I was afraid that the speed of an Arduino Nano wasn’t fast enough.</li> </ol></div>
I complained about this online and prepared to buy some shift registers to try nevertheless, then someone online suggested me to take a look at a more sophisticated RGB LED called the WS2818B. I looked it up, and it seems like a very popular choice of RGB LED. With a WS2818B, it will only need a single data pin to receive the color data, and will retain the color without the controller doing PWM. It also appears that those things can also be chained so I can have theoretically as much LEDs as I want, using just a single pin to transmit color. These features compared to the original LEDs, sounded like driving a car compared to crawling. I was carried away, and thought of expanding the project into something fancier.
I won’t make any regular old digital clock, I wanted make it to have a matrix of LEDs that functions as a screen, so that it can display more interesting things than just time. Coincidentally, I saw a product listing</a> of 16x16 WS2818Bs pre-made into a matrix, which sounds like exactly what I want, saves me the trouble of laying out the total of 256 LEDs and soldering them.

The LED matrix, taken from the product listing web page</figcaption></figure>
As I researched, the device won’t function with just a microcontroller and 256 LEDs. It will also need these less important, but still necessary things to be a decent clock:

A power supply. Both the Arduino Nano and the WS2818B matrix wants 5V power. According to an article I read online</a>, a single WS2818B LED can draw 50mA of current, which sounds small. But I have 256 of them, so it multiplies to 12.8A. I couldn’t find a mains power supply that converts to 5V and is rated for more than 12.8A. What I was able to find were a mains to 12V power supply that can carry that much power, plus a converter that converts those 12V into 5V. The 12V-to-5V converter can only carry 10A though, but according to the article mentioned above, 50mA is just a maximum, so as long I don’t run it at full brightness and every pixel being white, it should be fine.</li>
5.5mm DC power barrel jack. This allows the 12V power supply to connect to the input of the 12V-to-5V converter.</li>
DS1307 RTC module. This module keeps the clock running in the background, even when power is disconnected, so I would not need to set the time every time power is reconnected. I suppose it may also a bit more accurate than counting the time with the Arduino’s internal CPU clock.</li>
More wires and breadboard.</li> </ul></div>
At the same time, I came across something called an ESP8266. They say it’s similar to an Arduino, but it can connect to WiFi. I then excitedly decided to use that as the central microcontroller and ditched the now-dumb-looking Arduino Nano, and started daydreaming about the potentials of having internet access: I could synchronize time from the internet, and possibly displaying messages and notifications, even stream videos! (Yes, I know, playing a video on a 16x16 screen is extremely impractical. It could probably play Bad Apple</a> though, which sounded pretty cool to me.)
</div>
Now the project has become more ambitious, I started to think of even more features that I could add to this clock.
Displaying current temperature? Yes, let’s add a temperature sensor.
Having the ability to sound alarms? Yes, let’s add a buzzer.
Buttons too small and ugly? Let’s use the 4 Cherry MX key switches⁠[5]</a> I happened to have lying around instead of regular small tactile buttons, and use some custom key caps which I’ll worry about later.
So here are the aforementioned things I ended up having inside the clock, plus some other details:
(This photo was taken after the whole thing has been built, it obviously didn’t look like this at that stage. They were stuff plugged onto breadboards, I have a photo of that later in the article, but that photo wasn’t suitable for explaining what the components are.)

① 12V-to-5V DC converter
② 5.5mm DC barrel jack
③ Arduino Nano
④ ESP8266
⑤ DS1307 RTC module
⑥ BME280 temperature sensor
⑦ Buzzer
⑧ 5V-3.3V bidirectional logic level shifter, this allows the components operating at 3.3V (ESP8266 and BME280) to communicate with the components operating at 5V (others) without blowing up
⑨ The board that holds the 4 keys of the device, a.k.a. key board, not to be confused with a computer keyboard
⑩ Where the key board connects to the main board
⑪ Where the 16x16 LED display connects to the main board, the display is on the other side (front)
⑫ Unused connections of the display glued to the box so they don’t get in the way
⑬ Unused connections of a reset button, I decided later that this is unnecessary for normal use
⑭ Unused hole where the reset button was supposed to be
Why do I still have an Arduino Nano in the picture above? Didn’t I say I ditched it? Well, it turns out that controlling 256 LEDs with 1 pin requires a lot of work from the microcontroller. It requires very precisely timed signals so the microcontroller can’t do anything else during signal transmission, which is bad if I want to have WiFi connected because it requires some WiFi stuff to be constantly done, I also want time for my own code to run. There are tricks</a> to use dedicated hardware to transmit these signals, but they either consume too much memory, or occupies the pin I need for other purpose. So I put the Arduino Nano back, and programmed it so that when the ESP8266 wants to update the display, it transfers the display data into the Arduino Nano, the Arduino Nano then does the time-consuming job of transmitting precisely-timed signals, while the ESP8266 carries on doing more important things. Was is a waste to put a whole Arduino just for this singular purpose? I suppose so, but it was the easiest way of solving my problem.</section>
The Software</h2>
After I have got those components on the breadboard and connected (it didn’t look like the picture above, it was just on the breadboard), I started programming the software.
First, I decided the interactions I could have with the clock. It will have 4 buttons: , , , . When it’s powered on, it will display the current time (I’ll call it "clock mode"), pressing will turn off the display, and pressing will bring up the main menu, where the various additional functionalities (like alarms and settings) can be selected. and changes the selection, confirms the selection, and returns to an upper-level menu or the default clock mode.

Flowchart of basic interactions of Baby Tiger</figcaption></figure>
While making the clock mode, I realized that there isn’t enough pixels to display everything about the current time. A 0-9 digit takes at least 3x5 pixels, each 2 digits needs to have 1 pixel between them so they don’t fuse together. It ends up being a little challenging. I have to use different colors for different parts of the time for them to look distinct. I made a separate screen to show the year number and month name when is held down. Pressing will on the other hand switch the time display into temperature/humidity and temperature/pressure obtained through the BME280 temperature sensor.

Examples of the different screens in clock mode</figcaption></figure>
At this stage, I was pretty happy about what I have accomplished, so I posted</a> its photo on social media.

What it looked like after completing the clock mode</figcaption></figure>
Afterwards I made the main menu (and by afterwards I mean 1.5 years later, yes I procrastinated for that long). It shows the name of the menu item at the bottom half of the screen. Right now they are Alarms, Games and Settings. Since the screen is so small, only one option is shown at a time, and the text scrolls from right to left.

Main menu showing the "Settings" option</figcaption></figure>
The settings menu consists of these options:

Time: Setting the current time manually. There aren’t much to be noted, except writing the UI interactions were kind of tedious. Anyways, it works.</li>
Timezone: Because I use NTP to synchronize time with the internet, and NTP only cares about UTC, I have to take care of the local time offset locally, which is changed from here.</li>
NTP: I originally planned to make an On/Off switch for NTP, but didn’t ended up bothering, so right now it just shows the last synchronization time, and if there is an error synchronizing.</li>
WiFi: Setting the SSID and password of the WiFi network to connect to. Since the clock does not have a keyboard to type those things, I have to do it the around-about way: Let the ESP8266 make a WiFi access point for a phone to connect to, present a web page for inputting the SSID and password, receive them from the phone through HTTP, and finally try to connect to the network. Writing code for this process was the most tedious of them all (I wrote an HTTP server from scratch!). It eventually worked, kinda, although I don’t dare looking at the spaghetti code I wrote ever again.</li> </ol></div>
In the alarms menu, I made 3 alarm slots. The enabled alarms are green, while the disabled ones are red. Since I don’t have enough vertical pixels to display 3 rows of timestamps, the last alarm will have its last bit cut off, and everything will get shifted up when the last alarm is selected. Going into an alarm setting (by pressing ) will let me change the hour, minute, on/off, whether it repeats on each day of of a week, and which music to play when the alarm triggers. It ended up looking pretty good, if I do say so myself.

Alarm setting on Baby Tiger. There are 3 alarms, the 1st and 3rd alarms are disabled, the 2nd alarm is enabled and set to 7:15, repeats every Friday, music set to Katyusha.</figcaption></figure>
I originally planned to put some games on there, but they aren’t exactly necessary, and seemed like a lot of work, so in the end I didn’t write any. The "Games" on the main menu is just the demo music player that plays Senbonzakura.
I mentioned 2 musics that’s on Baby Tiger, Senbonzakura and Katyusha. They aren’t really musics though, as I could only play square waves from a ESP8266. It can’t play regular music file, nor does it play MIDI. These 2 tracks are just notes I typed into the code manually. (I really should write a script to convert from MIDI, if I really want many different music to select from. Doing this by hand is pretty time-consuming.)
And then I felt like the software was done enough. So the next step is to figure out how put them in a good-looking enclosure. But before that, I got to move the electronic component out of the wobbly breadboard, onto a more sturdy circuit board.</section>
The Circuit Boards and Keys</h2>
So the plan was to have 2 separate circuit boards connected by a ribbon cable. One was going to be the main board, holding most of the stuff; the other was going to be the key board, holding the 4 key switches, as well as the capacitors and resistors for debouncing.
Let’s talk about debouncing⁠[6]</a>. There are 2 ways of debouncing, hardware and software. I chose hardware because the article I read to learn about the topic said "It’s a hardware problem, so it’s best to be solved with hardware." and I agreed that hardware feels more elegant. The problem is, I don’t have an oscilloscope, so I just had to guess how long and how often the buttons bounce, and in turn had to guess the capacitance and resistance values in the debouncing circuit. The guesswork was a weird experience. The idea of hardware debouncing is to let the capacitor charge or discharge slowly when the button is pressed or released, so even if it bounces a little, the voltage would not change too much as to let the microcontroller think as extra button presses. That means the bigger the capacitance, the bigger the resistance, the slower the voltage changes, so the less likely for the bounces to have an effect, with the trade-off of the button press feeling more laggy, right? Well maybe my understanding is wrong, because changing to bigger capacitor apparently make the bouncing worse than having a small capacitor. Changing resistors could also make it worse – or better, it feels completely random, and the ESP8266 is so weird that some pins seems to work the best with different valued resistors than other pins. I still couldn’t understand why that is, maybe the answer will be more apparent if I had an oscilloscope. Eventually though, I figured out a circuit that worked well enough.

Schematic of the key board</figcaption></figure>
I didn’t plan to draw and order a custom-made PCB, I thought just soldering stuff onto a perf board⁠[7]</a> should work well enough for my purpose. I was wrong.
I started out making the key board since it’s the simpler of two. Then I realized what appears to be simple on a YouTube video may not be so simple if I do it myself. Those YouTubers had experiences and I didn’t. The process of realizing the circuit diagram above was slow and painful. I couldn’t keep the components in place before soldering them, it ended up resulting in none of the key switches are seated nicely, they tilted in all different directions. It’s really unideal, but what’s been done is done, I can’t do it again unless I buy new key switches. It was also difficult to make the wires connect to where I want them to connect to while not touching where I don’t want them to have contact with. After I finished making the key board, I concluded: There is no way I can do this manually for the much more complicated main board.

Schematic of the main board</figcaption></figure>
Defeated, I sat back to the front of the computer, and decided to draw a PCB with KiCad. I didn’t know how to draw PCBs, but luckily it didn’t take too much time to learn with the help of guides found online. And I was glad to not go down the manual perf board soldering path, because the PCB I ended up drawing looked like this: They are all necessary connections.

PCB design of the main board, red lines are traces on the front layer, green lines are the traces on the back layer</figcaption></figure>
I sent the PCB design to a company that makes custom PCBs for individuals. The board itself was not very expensive, costing US$ 4, but the shipping though, they cost $14 and was estimated to take about a month. The DHL express shipping costed $21 but is estimated to take less than a week. I chose the latter. The slow shipping was already that expensive, so it felt justifiable to buy the 1.5 times as expensive but 4 times as fast shipping method.
It took less than a week to arrive, as expected from DHL. I then soldered all the remaining components, completed with significantly less sweat than when I made the key board.
Now comes the moment of truth: If I drew the PCB wrong, then things will not work and I’ll have to spend $25 again. Or worse, I’ll blow up components. Luckily nothing seemed wrong as I plugged in power – a successful milestone.

The photo I took when posting on social media</a></figcaption></figure></section>
The Enclosure</h2>
It was the time for the non-electronic part: making an enclosure for everything to sit nicely and look good in. In my imagination, I figured that the shape of the enclosure should be a rectangular box, with a shallow depth, like a picture frame. The front has the screen, covered with translucent acrylic, and behind the screen are all the control circuits. I figured that wood is probably suitable as its main building material. It’s heavy enough so the device won’t tip over, and it also sounds like a very commonly used material.
I had no experience in woodworking. I was a bit overwhelmed by all the information I needed to get started without spending too much money. There were questions like what suitable types of wood are, how to make right angled joints that aren’t easily broken, what types of glues I need, how to cut square holes, how to make a detachable backplate. A lot of the problems weren’t hard, they seemed to be pretty easy with powerful enough tools. But I didn’t want to spend that much money on powerful tools. They were really expensive. OK, they were not that expensive if I was going to use them all the time, but I couldn’t think of when the next time I use them will be, and investing hundreds of dollars into tools I could only think of using once is definitely not worth it.
To demonstrate my point above, let’s talk about how I wanted to cut 20x20mm square holes onto a piece of wood in order for the keys to poke out. What options did I have? The tools I had were an electric dill (which I bought from Ikea a long time ago) and a hacksaw. An electric drill can only dig round holes, how do I make square ones? The answer I got from the internet was to use a "router". Well, I did have a router to use, it even had 1 Gigabit Ethernet, but it’s apparently not the "router" I need to cut square holes. The router for cutting square holes was something I didn’t have, and costed quite a lot. And apparently, a router was not enough! To assist the router into cutting straight lines, I would also need a "template", and templates seemed to be quite a deep rabbit hole.
I guessed I could first drill a round hole that the diameter is the same as the side length of the square, and then use a saw to cut the corners. Like the left halt of the image below. But I could not find affordable drills bits that are 20mm in diameter, and I could not find saws that would fit into a 20mm round hole even if there were 20mm drill bits.

Circular holes are drilled first, then the material in shaded areas are removed by other means, creating square holes</figcaption></figure>
So sawing were out of question, what else? Maybe using a file instead of a saw? It’s probably going to be more time-and-strength-consuming though, but I was able to find small enough files for not-too-high prices. As for drilling the round holes, I guessed using the largest drill bit my drill could hold (8mm), and drilling multiple of them should suffice. As seen of the right half of the image above.
After painstakingly researching every potential problem I though of (they were not enough), many WikiHow-reading and YouTube-watching. I settled on a design that balanced ease-of-making and looked passable.

The mockup 3D model. From left to right: front, back, back with backplate off</figcaption></figure>
I originally planned to use wood for the square frame (yellow in the 3D model), the display support (orange in the 3D model, put there for the display to be glued on) and the back plate (also yellow in the 3D model). I didn’t expect regular wood to be so expensive though! So I turned to the alternatives, which are still at least made of wood fibres. Plywood, which are made by gluing thin pieces of wood together, and somewhat cheaper. And MDF, which are made by compressing wood fibres (I feel like they are similar to paper, just thicker and stronger. MDF also disintegrates with water, just like paper does.), and quite a bit cheaper. I contemplated on which to use, watched a few videos explaining the differences. It turns out plywood seems to be better because the dust is less fine and thus less harmful for the lung, and does not disintegrate with water. But I still turned to MDF since it’s quite a bit cheaper. It shouldn’t matter much for my project, I guessed.
So I went shopping for tools and materials. I bought:

A piece of 12mm thick MDF, very large, much more than I needed, but they don’t sell smaller ones.</li>
2 pieces of 3mm thick MDF, for some reason they had small piece for 3mm, and the size was just enough for what I needed.</li>
A large file.</li>
A set of small files, they only had small files in a set and not individually.</li>
A handsaw.</li>
Wood glue.</li>
4 clamps.</li>
A square stick of cheap and soft wood for support while gluing and drilling.</li>
8mm drill bit.</li> </ul></div>
Then I started cutting wood into the pieces shown below. The reason I used this shape with protrusions and slots for the side pieces was to increase the surface area of the joints. The 4 side pieces joins end-to-end, and the protrusions slots into the slots. I am not sure if this was worth the effort since this thing doesn’t need to be that strong, but everyone online seems to emphasize how important it is to increase the surface area of joints. And no one on YouTube seems to make a right angled joint by just sticking one end of a piece onto the side of another (a.k.a. a butt joint).

Wood pieces I cut. Black shapes are 12mm thick, orange shapes are 3mm thick. Units are in mm.</figcaption></figure>
I tried to make the slots on the side pieces by sawing off as much material as the saw can reach and filing down the remaining. It turned out the large file I bought wasn’t suited for the job. I filed for a long time but little material were visibly removed. The small files were obviously going to be worse, since they are even finer. Now making the slots were already this hard with the files I have, making the square holes in the later steps were definitely going to be harder. So I went back to the hardware store, returned the small file set, and bought a set of rougher files. Filing was still slow with the rougher files, but at least my work could have visible progress.
Making the square holes was as painful as I imagined. Drilling the 8mm-diameter circular holes were easy with an electric drill. Removing the rest of the material in the square hole area was a long and tiring process. I filed and filed. Working from afternoon to evening to early night. Had my neighbour complaining about noise. The "square" holes were irregularly-shaped. But I was really tired, so I decided it was good enough and stopped for the moment. It later turned out the holes were not large enough to fit the keys, so I had to enlarge them a little after gluing.

The "square" holes as seen on the finished product. Don’t mind the key caps, I’ll talk about them later.</figcaption></figure>
After cutting the 4 side pieces, I glued them together to make the frame. Since the protrusions and slots on the ends of the pieces were not that perfect either, the wood glue did not really make great contacts. But they still held up, which was good enough. I was really scared of the frame turning out to be not square (slanted into a parallelogram), but luckily it was square enough, after waiting 24 hours which was how long the label of the wood glue bottle said I needed to wait.
I also cut the display support and backplate, and glued the former with hot glue.
Since I originally planned the box to be made of wood, I didn’t plan to paint them. Wood grain patterns seemed pretty nice. But now I’m using MDF, I had to paint it so it is not MDF-brown-coloured. The front panel was supposed to be a black translucent acrylic in order for the screen to appear black where pixels were not lit. So I guessed I might as well paint the whole thing black. But also because it was MDF, which disintegrates with water, many paints cannot be used on it, including acrylic paint which was the one of the only paint types I could get in small quantities (I guess most paint containers were designed for people who paint walls and stuff. Buying 2kg buckets of paint just to paint this small thing seemed super not worth it for me.). I went to research online again, and concluded that I should use spray paint cans.
I went to the hardware store once again to buy the spray paint cans. The spray paint cans’ label said I need a primer as well, so I bought a can of flat black paint and a can of primer. Back home, I laid down paper of the floor of my balcony and started spraying. It smelled pretty bad even as I did it on the balcony. It was also a long process, since instructions online as well as the label said I need to wait 15 minutes between each coat. So after painting a few coats of primer and more coats of black paint, it became dark once again. Fortunately I finished before the sky was as dark as I stopped being able to see clearly.
Remember the hot glue I used to secure the display support? One corner came loose after the spray painting. So I added wood glue as it should be stronger than hot glue.
At this stage, I still did not have a way to put the back plate on. It needs to be removable as I need to access the electronics inside. I first thought of using screws, but people said screws in wood becomes loose after unscrewing. A friend suggested using magnets, which I guessed was a good solution.
I bought 5mm diameter neodymium disc magnets online. 5mm is a size that can fit into the sides of my 12mm thick side panels, and also a size I have a drill bit for. I drilled a hole on the middle of each of the 4 sides and placed magnets in them. I also taped magnets of opposite poles onto the corresponding locations on the back plate. There were more problems:

My hand were not still during drilling. So the holes were not in the exact centers of each sides. They drifted a little, enough drift for me to having to adjust the locations of magnets on the back plate. Since the drifts were irregular, I had to lay the frame onto the back plate, poke the magnets into each of the 4 holes, then carefully lift the frame, leaving the back plate with the magnets on the correct location.</li>
I put the poles wrong! Same poles of magnets repel each other which was the opposite of what I want, so I had to take them off, and repeat the step above.</li>
1 magnet pair on each side turned out to be a lot weaker than I expected (Aren’t neodymium magnets supposed to be insanely strong?). I tried drilling the holes deeper and adding more magnets. Eventually I put 4 magnets on each sides of the frame and the back plate, a total of 32 magnets.</li>
Hot glue could not make a strong bound for gluing neodymium magnets to MDF. I had to buy yet another type of glue: the 2-part epoxy. It came off once more, and had to try again with thicker layers of the glue.</li>
One magnet on one side of the back plate did not get glued well, and got broken off and stuck inside the corresponding hole on the frame. It seems to have stuck pretty tightly, so I didn’t bother getting it out and gluing it again. It became a feature.</li> </ol></div>

Finished back plate, this side faces the inside so I didn’t bother painting, and the hole is for plugging power</figcaption></figure>
Before putting the front panel on, I glued the display onto the display support, also with epoxy glue.
I wanted the acrylic front panel to be translucent, not semi-transparent, in order to diffuse the light from inside so I don’t see the pixels when they are off. The 200x200mm acrylic I bought from the beginning was translucent when the protective films were on, but I knew it would just be a semi-transparent black when the films were peeled off. Leaving a protective film on the finished product is definitely ugly, so I needed another light diffusing layer. I thought tracing paper would be fitting, but they were actually quite expensive and the stores I went to had them sold out. I found a piece of frosted plastic in alternative, which I think actually worked better than if I had used tracing paper. I glued one side of the acrylic to the cut frosted plastic with epoxy glue. And then with the plastic facing inside, I glued them onto the frame. It came off once and I had to glue again.
By the way, 2-part epoxy glue smelled a little gross, the B part smelled like fish for some reason.
With all these stuff done, I could finally put the electronics in. I glued the key board, main board, the 12V-to-5V converter, the DC power barrel jack in this order, with a combination of hot glue and epoxy glue. So far they have stayed in place.

Inside the enclosure after all components have been put in. It’s the same picture from before, ignore the numbers.</figcaption></figure></section>
Completion</h2>
Plugging everything in, now it’s the most important moment of truth. If this somehow ended in smoke, I would be very upset. It shouldn’t have any problems though, as I have tested it in every stage. I was nervous, my hands trembled as I plugged in the power, my heart apparently stopped beating for a second before the display came on. But finally, my worries turned out to be needless, as the product of my hard work blinked each second with accurate time on the screen.
I called it a success.
I posted</a> the completed photos on social media once again, and people congratulated me.

The final results. Left shows the time, right shows the temperature and humidity.</figcaption></figure>
Well, it’s not flawless though. First of all, the whole thing looks totally handmade and imperfect. The surface finish is rough, and marks left by dried glue is all over the place. Some unpainted areas shows from the outside. The back plate does not attach well not only because the back of the frame is not flat at all, but also because the glue around the magnets on the back plate formed a chamfer, preventing them from going all the way in. Unlit pixels still shows though the front panel a little, though I guess it’s acceptable. And not to mention the ugly-shaped corner joints and holes for the keys. Also there’s an unoccupied hole on the left side that was supposed to house the reset button I ditched.
Second of all, the temperature measurement is a lot higher than the actual room temperature. The reason is probably the other components are dissipating a significant amount of heat so the inside becomes hot. I didn’t foresee this, I thought this thing wasn’t going to emit enough heat to affect the temperature measurement, I was totally wrong. The amount discrepancy is unacceptably high (like, about 8 ℃). Also because all the other environmental measures (humidity and pressure) are based on the temperature measurement, they are all unusable. I took off the back plate so it’s a bit better, but, like, there’s no back plate then.
It’s also not really finished. I used Cherry MX key switches, so I got to use key caps designed for a computer. But my buttons are , , and . On a computer keyboard, there are left and right arrows, but obviously no cross and circle. I could use X and O letter keys for and , but they are differently shaped from each other and the arrow keys. I saw online that epoxy resin⁠[8]</a> is something people often use to make custom key caps, so I guess I could do that as well. But that’s another deep rabbit hole I’ll go into later in time. Meanwhile, I just randomly took 4 same-shaped key caps from a pile. They’ll eventually get replaced.
The software also doesn’t have much functionality. The excitement of having WiFi when I switched to ESP8266 did not do a lot. The only thing that currently uses WiFi is syncing time via NTP. I guess I might want to expand the software more later in time, adding countdown timers, stop watches, displaying notifications, and maybe actually playing Bad Apple.
Throughout the making of this project, it made me realize that DIY is probably not going to be a money-saving options oftentimes. The amount of money I spent on tools and materials is something I had lost count on, even though I tried to be as cheapskate as possible. Not to mention the amount of effort and time. I guess if the project is simple, and you have most of the tools, DIYing may be cheaper, I dunno. I wanted a custom digital clock that is not sold anywhere though, so I’d like to think my project was worth it for me.
That’s it. Finally I finished writing this long article which took 2 days to write. I have published all the code and design files</a>, in case you want to know any technical details, or use them as references.

A microcontroller is a micro electronic component that can be programmed to control circuits. It’s a very simple kind of computer. ↩</a> </li>
An Arduino Nano is basically the same as the most popular Arduino Uno, but smaller and cheaper. A Uno’s size is way too large in my opinion for something with this little processing power. ↩</a> </li>
A breadboard is a mainly-plastic board with holes all over them. One can insert electronics and wires onto it and have them work without soldering. ↩</a> </li>
PWM means turning on and off the LED so quickly that it appears to the human eye as if it’s continuously on but dimmer. Adjusting how much time it’s on verses it’s off changes how bright or dim it appears. ↩</a> </li>
A key switch is what’s underneath the key cap on a mechanical computer keyboard. Although it’s primarily used for keyboards, in terms of how it interacts with electronics, it’s the same as a regular button. I had the 4 key switches left over from repairing my computer keyboard. ↩</a> </li>
Buttons are mechanical construct, in the real world, the mechanical contacts will swing back and forth a bit when they are pushed together or released away. So pressing a button by hand once will be seen as the electrical contact has been closed and opened multiple times thus being thought as multiple key presses. Debouncing is the act of removing the effect of these unwanted openings and closings. ↩</a> </li>
A perf board is something like a circuit board, but with no internal connections and have a matrix holes filling the entire surface (So it doesn’t need to be custom-made). One would insert components on one side and solder and connect them on the other side. Search for images should give you a good idea on what it is. ↩</a> </li>
Epoxy resin is not the same as epoxy glue, although they are similar. Epoxy glue usually sets faster (tens of minutes), have a color tint, and sold in small volumes. Epoxy resin usually sets slower (about a day), are clear, and sold in larger bottles. Epoxy resin seems like a pretty popular material to make artistic objects, like custom key caps. ↩</a> </li> </ol></section>

Unless otherwise stated, the original contents on this website are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License</a>.While reposting, please state that the article comes from FiveYellowMice's Blog</a>, and include a link to the original article</a>.

明天再和朋友相见

Sat, 15 Aug 2020 19:08:00 +0000

在 RSS 中无法显示完整的格式，不同的 RSS 阅读器所显示的内容也可能会有很大差距。为了能够获取到完整的内容，最好还是点开网页链接</a>看。

读着视觉小说 [1]</a>，是三位高中生一起在社团活动后的傍晚互相道别的场景。其中一人向另外两人开心地说道：“那么明天再和两位相见喔！”然后跑跳着离开了。
刚刚还在吵吵闹闹的场景，在一人离开之后一下子就安静了下来。看着屏幕上笼罩着夜色的背景，正在带入着主角思考的黄鼠，一瞬间感到了一丝孤独。尽管刚刚还在很开心，但不可避免地，一天马上就要结束了，大家都需要各自分道回家了。有趣的活动到现在已经停止，而朋友们也会是在接下来的一段时间里见不到的了。
但黄鼠在下一刻就意识到这是不需要过度感伤的，因为我们还有明天，我们在明天、十几个小时之后就会马上再见了。刚刚也是这样说的，“明天再和两位相见”，不是吗？这不是长时间的道别，所以黄鼠的孤独感转变成了期待——好期待明天的到来呀。只要熬过今晚的几个小时，再睡一觉，在明天就又可以很快和朋友相见，产生更多有趣的经历了。
不过很遗憾的是，这一切都只是在小说里发生的，黄鼠的上述的感受也只是针对那个虚构的故事的。而在现实中，除去有期待少有的特别活动（像是旅行计划）的到来外，似乎已经忘记自己上一次这样期待第二天的日常的到来是什么时候的事情了。
黄鼠印象中的校园日常，几乎就是充满了缺乏活力的轮回。每天早上起来，急急忙忙度过早晨，就开始了不一定很无聊，但大都很累人的课。结束了之后，回到家才可以放松一会，但也不能放松很长时间，因为马上父母就要回来监督自己完成作业了。每一天的自己，整天几乎全部的期待，都放在了放学回家后到父母回家前的那段“自由时间”里，似乎每天都是为那一小段时间而活。在这段时间结束后，下一次能够留给自己的时间就只有完成作业后（或放弃了完成作业后）躺在床上的时间了。而再往后，就又得挨过一整天的学校才能再次享用这份安逸了。
黄鼠不是在学校中没有能够愉快玩耍的朋友，但与朋友的交往，却很少能够像这次在小说中感到的那样，抵消上学的劳累，令黄鼠充满对明天的期待。
现实中的黄鼠，劳累于上学而期待在家的时间，而这段视觉小说中的小小对话，却把两者完全反转过来，开始期待与在学校的朋友们相见，而对在家的时间反而感到孤独了。
不过这只是视觉小说，黄鼠不是主角。而站在读者的角度的话，黄鼠只能对这样的日常表示羡慕了。能够有对每天怀有这样愉快的期待，真好呀。
黄鼠回不到像高中那样的校园生活了，因为在现实中只能向前看。在未来，即使能够找到工作而不是饿死或啃老，所面对的也大概只能是比校园更加乏味且劳累的社畜生活——实习的经历让黄鼠意识到了这一点。每天回家后几乎就像被抽干了体力一样，除了生活必要的事情（像是做饭），唯一想做的事情就是躺在床上耍废，看一些没有营养的 YouTube 视频，刷一些更缺少营养的社交媒体。游戏——甚至是相对低投入的手机游戏——都很少有心思去玩。写兴趣项目和学习更多知识的动力什么的，就更加被鸽子叼走了。在这样的日常中，每天值得期待的东西，就比学生日常更加稀少了。
黄鼠就经常这样想，做社畜的日常好辛苦呀好累呀好可怕呀。所以在读视觉小说时产生了“对明天日常的期待”的心情时，真是被自己惊讶到了呢。原来自己也可以对日常感到开心和期待。但可惜只是针对在小说中存在的虚构日常。
黄鼠看到这么一小段对话，就写了这么一长篇文章，诚实说，也是令自己觉得蛮惊讶的。不过在写完之后，黄鼠并没有什么深的感悟。要说整片文章有一个什么中心结论的话，那就是——二次元真好呀。

尽管这并不重要，但如果在好奇的话，黄鼠在读的视觉小说是《魔女的夜宴》</a>。但请不要把这当作是黄鼠的推荐。故事的主要内容并没有和黄鼠在这篇文章中的感悟有很大联系。 ↩</a> </li></ol></section>

除特殊说明以外，本网站原创内容采用知识共享署名-相同方式共享 4.0 国际许可协议</a>进行许可。转载时请注明来源 FiveYellowMice 的博客</a> ，以及原文链接</a>。

Something Incredible Between Them

Sat, 11 Jul 2020 22:09:00 +0000

It's not possible to display full formatting in an RSS reader, different RSS readers may also display drastically different layouts. Therefore it's recommanded to read the article by clicking the web URL</a>.

Mathew feels troubled today. It’s not because of the mock final exam that has just passed. Although indeed he didn’t do well, it doesn’t bother him too much. He believes that studying is easy, relatively speaking. All he have got to do is to read those textbooks, remember these bullet points, do things that have all been successfully done by billions of people. Even he has trouble understanding, he has literally all the resources to help him. As long as he studies seriously, his grades are bound to rise to a respectable level. That’s what he believes, at least.
What troubles him, is the tiny screen glowing in his hand. On the top of the screen is the logo of a popular social media. On a more prominent part of the screen is a name that can drive Mathew’s heart up and down in milliseconds, Remi. The post on screen is not sent by Remi however, it’s another person that occupies a large part of Mathew’s mind, Karter. The post reads: I just witnessed something incredible of Remi.
Someone asks “What’s incredible?” and judging from the response Karter has given, it’s apparently very secretive.
“It’s always been like that.” Mathew mumbles. “Saying some prompt as if wanting to tell us more of a story, but just as the prompt provoked others’ curiosity, they tell us no more.”
Well, that’s what couples do, isn’t that right? They share things on social media not for others to see, but for themselves to have the pleasure of showing off and being showed off. Mathew doesn’t think it’s a badly selfish thing. In fact, he likes it when people announce their love, and enjoys reading romantic stories, both real and fictional. In fact, he craves for it. It doesn’t only make him happy for those in love, but also gives him a space for illusions as if he is part of it. This is usually the case with people that Mathew is unfamiliar with, and especially the case when the couple in question includes the one he admires.
Indeed, “admires” is the most Mathew dares to say about his feeling towards Remi.
Mathew met Remi when he was a first-year high school student. Remi was in the second year. In a school-organised activity, where 3-5 students from both first-year and second-year collaborated to complete a collage on a topic freely chosen. The groups were randomly arranged, although students had a chance to rearrange if they were unhappy about their groups. Mathew, a couple of unimportant people, and Remi were in the same group, they were quite happy about the arrangement, so it stuck. Mathew quickly took a liking on Remi. Remi is interesting, artistic, caring, and knows quite a lot. Remi didn’t seem to think Mathew as a nuisance either, so they exchanged contacts and chatted regularly. The collage didn’t end up being a hit, but it was good enough that both their parents praised them for it.
Later on, Matthew loved hanging out with a bunch of second-year students after each class, including Remi. The first-years and second-years were in the same building, so it wasn’t a rare sight in this high school. To others, they just looked every other students hanging out together, but to Mathew, the times they spend together were incredible. The more Mathew knew about Remi, the greater he thinks about Remi. And Mathew thought, this was going to last forever.
But obviously it couldn’t. A year passed by, Remi had gone up to the third year. The third-years were in a different building, “In order to better assist their studies.” The time Remi could spend with Mathew had drastically decreased, their interaction was then mostly restricted to social media and weekends. It was also around the start of the academic year that Karter started to appear in Remi’s social media posts.
Mathew didn’t care that he didn’t know Karter personally. “It’s just someone Remi thinks is cool.” Mathew thought. He knew, by reputation, that Karter was a prominent member in the school soccer team, but then resigned for some reason. “Whatever, it doesn’t bother me at all.” Mathew told himself.
But then Karter was mentioned more and more by Remi. They posted more and more photos of them together. They also put each other’s username in their social media profile. “Wow, they must be really good friends.” Mathew thought, “Karter must be a genuinely interesting person. I kinda want to know more about.” Mathew wasn’t jealous or anything, at least he knew he shouldn’t be.
One day, Mathew got a 2-ticket bundle at a local movie theatre, it could only be used when 2 people went at the same time. So Mathew invited Remi.
“Is it just us two?”
“Yes, it’s a 2-ticket bundle.”
“I guess… but Karter probably wouldn’t be happy about it.”
The rest of the conversation didn’t matter. The last sentence Remi said stuck in Mathew’s head. “Why would Karter be unhappy about Remi and I going to a movie?” Mathew pondered, “What kind of people would be unhappy? Is it… what I imagined it is?”
There wasn’t a sure answer, but later observations only confirmed Mathew’s guess. Remi and Karter, are in a relationship. They didn’t announce it when it started, but they are, now.
It is when Mathew made this guess that he also realized how jealous he was over having someone that holds an exclusive lock on Remi. There are now things that Remi must keep Mathew away from. There are now things that Remi cannot do for Mathew.
It was time that Mathew needs to accept the new reality, one that has Remi occupied by someone Mathew wasn’t close with.
This reality has been largely accepted by Mathew today. He knows he can only watch Remi and Karter from a friendly distance. He has also attempted to get close with Karter, but they don’t much in common. Mathew knows that there will always be a large part of Remi he can’t acquire forcibly, so the best course of action, is instead watch for things that Karter and Remi shared about their life publicly, consuming that, relieves his insatiable hunger for more Remi.
But Mathew also knows that he can’t do this forever. In matter of months, Remi and Karter, both as third-years, are graduating. Whether they’re going to the same college is unkown to Mathew. Obviously he could just ask Remi. But he is too scared to ask, as he is with many more questions to Remi. The same goes for many things Mathew wants to say to Karter and Remi, but he will never say them, as they are, at most, irrespective and irresponsible. Mathew thinks it’s best for him to befriend Karter, and be friends with both Remi and Karter for as long as they can keep in contact.
Yet, Mathew is feeling troubled by something incredible between them that he can’t know about.

Unless otherwise stated, the original contents on this website are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License</a>.While reposting, please state that the article comes from FiveYellowMice's Blog</a>, and include a link to the original article</a>.

狗粮！唔嗯……想要……还要更多！咳！

Mon, 25 May 2020 20:30:00 +0000

在 RSS 中无法显示完整的格式，不同的 RSS 阅读器所显示的内容也可能会有很大差距。为了能够获取到完整的内容，最好还是点开网页链接</a>看。


恋爱真是件美妙的事情呢。尽管要享受它，还必须得付出巨大的代价才行。得承担起两人份的责任，得照顾好对方的心情，得掌握好所有人的看法和态度，还得知道全方位的处理冲突的办法。但很幸运地，对于像黄鼠这样，不想付出代价又想吸收恋爱带来的甜美的人，我们还有替代方案——吃狗粮。
我们可以寻找虚构的故事。除了少数的例外，故事们一定都是被设计成读后会让人开心的。作家们用了全力，只为了将人物写得无比可爱，把交互写得超级幸福。即使有冲突，那也一定是会得到解决的，而在解决之时，他们也一定不会放弃机会，再让读者释放一次快乐的荷尔蒙。
黄鼠可以放心地、全力地喜欢虚构故事中的 CP 们，因为不用担心它们会令我失望——如果这是两个黄鼠不会喜欢的角色的话，那么从故事开始就会知道了。角色们的可爱是被刻意设计如此，因此是完全不用担心会消逝的。
而在现实中，我们也可以不劳而获地享受别人的恋爱带来的幸福感。现实中的情侣会秀恩爱，会撒狗粮——这是人之常情，得到了那样稀有的机会、付出了那样多的努力，黄鼠也一定会忍不住使劲炫耀的。
黄鼠通常来说，都认为自己是一个擅长接受狗粮的人。尽管有少数时候会有些许嫉妒，但那不重要，因为那是和自己无关的事情。面对别人的感情、别人的幸福，自己只要在周围开心的祝福就好了。狗粮发放者本人的的幸福，才是故事的主角。而自己能够免费蹭到这份幸福，是值得感谢的事情。
黄鼠看 YouTube ，在 YouTuber 们的镜头前，几乎每对人都是那样地美丽。他们互相扶持着，一同行进着，一起将美好展示给全世界的观众。在镜头之外也许仍然会有冲突，但那跟黄鼠无关，也无法获知，因此黄鼠不去思考它。
黄鼠看着 Veritasium 的 Derek 讲述自己的人生故事</a> ，那本身就是一篇很棒的、启发人的视频，而在其中 Derek 感谢到自己的妻子对自己的长期支持的时候，在提及两人相遇的时候，在视频中播放他们在从前求婚时在室外用手机录下的模糊录像的时候，那份狗粮是突然的，但是是美味的。幸福的感觉遍布了黄鼠全身，唯一能组织出来的语言，是“好棒啊，好想祝福”；唯一拥有的想法，就只有祝福他们在这样多年的时间后，依然能够持续这份幸福。
黄鼠看着 Aki</a> 与 Joey</a> 的频道。看着这两位做着相似种类的视频的 YouTuber ，从乎不相干，到公布了跨越大洋的感情，再到现在生活在了一起，除了祝福他们以外，难道还有什么其他的事情可以思考吗？而在如今，两人已经频繁地在视频中互相出镜，互相提及一起做的事情，这样的生活成为了日常，这样的日常是多么令人向往呀。而黄鼠，除了感谢他们一直带来的有趣视频以外，就只有希望这美妙的日常可以继续下去了。
决不能忘记提及的是 Rachel and Jun</a> 的频道。从这样的频道名称看来，还能指望更多的什么呢？几乎每一篇的视频，都是优质的狗粮。美妙的两人，被三只可爱的猫咪包围着。对着样的美妙生活，那是连嫉妒都难以做到的。
放下 YouTube ，黄鼠看了看自己所认识的人们，那也是一个大量的狗粮来源。黄鼠看着大家的幸福，自己也会被幸福感所感染。黄鼠自诩是一个思想开放的人，会接受各种各样关系，会理解他人各种各样的决定，会明白这是与自己无关，自己无权评判的事情。于是黄鼠享受着从他人借来的快乐，开心地祝福着他们。
至少，这是黄鼠认为自己应该做的事情，在所有时候。
但事实上，黄鼠并没有像自己想象中那样强的承受能力，也并没有能够完全保持开放的思想。黄鼠有时会私下评判别人的身份和行为，也会拿着自己个人的衡量标杆，去评判与自己完全无关、自己完全无权干涉的事情。黄鼠无法总是纯粹地享受狗粮而不添加自己的主见。
黄鼠会偷偷地给人打上“难以喜欢上”的标签，尽管这可能仅仅是依据于粗浅的认知。黄鼠会觉得某个人可能太高冷、太凶、说话不好听、做事太绝情。可是人家是独立、有权做自己的人，是不需要黄鼠的欣赏的。事实上，黄鼠的存在有可能令对方厌烦也说不定。
尽管是人家自己的事情，自己完全无权评判的事情，但是可笑地，黄鼠会觉得自己无法接受现实中一些人的配对。就像对虚构作品中自己不喜欢的 CP 一样，想“诶，为什么那两个人会在一起”。可是对虚构作品这样做是没关系的，因为他们是虚构的人，可是在现实中，每个人都是有自己独立的思想，有着自己的事才对。与谁在一起是别人自己思考做出的决定，这样子否定别人的人生决定，是太大的不尊重了。
黄鼠知道这是自己不该拥有的想法，所以想要尽量避免它。因为这样，黄鼠十分害怕获知自己在意的人的恋人。万一那是一个黄鼠“难以喜欢上”的人该怎么办？那样的话，黄鼠就会悄悄地否定它们了。尽管这一切都不是黄鼠有权去评判的事情。
黄鼠是充满控制欲的，不愿意面对不在自己控制中的事情。如果熟悉的人宣布了新开始的感情，却又不愿意让黄鼠了解他的恋人，黄鼠会觉得气愤。尽管黄鼠知道，这是人家自己的决定，人家自己私人的事情，不愿意分享明明是十分正常的。可是不平衡的内心，仍然会无法接受自己熟悉的人正在隐藏其十分重要的一面，并且将大量的经历与时间花费在这个隐藏的地方，黄鼠难以获知的事实。但显然，这样的想法是不能说出来的，只好让黄鼠自己活该地承受。
黄鼠更不愿意接受的，是来自特定的人的狗粮。
那其中包含着黄鼠对对方不了解的气愤。自己明明声称那样喜欢一个人，却不知道人家居然有认识来自其它地方的其他人。不能够完全了解一个人的信息来源，令黄鼠气得不断跺脚：“他们到底怎么认识的啊！为什么黄鼠不知道啊！”
黄鼠没有“为什么不是我”这样的想法，因为黄鼠明白，之所以不是自己，一定是有合理的原因的。黄鼠因此想要了解对方，希望能够使用和平的方式，来让自己内心的控制欲平衡一些。可是对方大概完全不会有想要被黄鼠了解的兴趣，更不用提黄鼠的动机了。
但这样的事情，能有什么经决方案呢？在黄鼠能够改变自己的想法之前，大概每次吃到那颗特定的狗粮后，只能够继续被这些禁忌的想法所围绕了。“为什么这个人能与这个人关系这样好”、“为什么我不能成为其中的一分子”，这些不该被提出，也不该被回答的问题，希望能够停止吧。黄鼠想要能够全心全意地吃狗粮，这是足够的，也是黄鼠唯一有权做的。

Image by Mat Coulton</a> from Pixabay</a>
除特殊说明以外，本网站原创内容采用知识共享署名-相同方式共享 4.0 国际许可协议</a>进行许可。转载时请注明来源 FiveYellowMice 的博客</a> ，以及原文链接</a>。

天亮啦！动物森友会

Thu, 09 Apr 2020 21:11:00 +0000

在 RSS 中无法显示完整的格式，不同的 RSS 阅读器所显示的内容也可能会有很大差距。为了能够获取到完整的内容，最好还是点开网页链接</a>看。

一次在游戏中的合照，除最左村民外，从左至又依次是：芙兰，黄鼠，奎妮，萌狼， Cocoa</figcaption> </figure>
“动物之森？那样无聊的游戏，我还是不要浪费更多时间和金钱了吧”，第一次看到许多人为即将推出的新游戏《集合啦！动物森友会》兴奋时，黄鼠这样想到。
但这风潮似乎远比想象中更大，越来越多的玩家开始考虑是否要入手这部游戏，而其中的一些，已经开始他们的无人岛生活了。一位同样看到这场风潮，正在捉拿不定的朋友询问黄鼠会不会购买，不过黄鼠同样给出了否定的回答。
而黄鼠觉得这部游戏会无聊，也不仅是凭借粗浅从预告片看到的感觉呢。在蛮久之前，黄鼠曾使用 NDS 模拟器玩过一点《欢迎来到动物之森》，就觉得好无聊，没有玩很久就放弃了。所以这次在同样的系列发布新作的时候，黄鼠也继续认为新作会同样地无聊。
但是越来越多的人开始跳上前往无人岛的飞机了。在发售之后的一天下午，忍不住黄鼠一咬牙，决定也要冲动消费了！临时地赶在最近的百货商店关门之前，从货架上拎走了一盒《集合啦！动物森友会》。
在那之后，黄鼠的行为就变得和王镜泽一样了。
整一天的时间，几乎有做的唯一事情，就是在玩动物之森。本身就因为学校停课、整天呆在家里而变得混乱的作息，被动物之森加速成为几乎完全的黑白颠倒了。每天的睡觉时间成为了凌晨 7:00 ，起床时间成为了下午 16:30 （±3 小时）。在睡觉之前，黄鼠会望一眼窗外的天空，感叹道：“呀，天亮了呢。”

动物之森让黄鼠的作息更快坏掉、让黄鼠无法及时学习、让黄鼠整理房间的计划更加延后——这些游戏带来的坏影响，是实实在在的，不过在这里，黄鼠想要讨论的不是这些话题。
黄鼠想要讨论的，是关于人的话题。
无需质疑，动物之森使得许多平常几乎不会有交集的朋友得到了非常丰富的交流机会。平时完全没有理由说上一句话，在交流关于游戏的时候却有了许多的内容。即使只是简单的日常业务性对话，像是询问大头菜价格和交换水果，在这样的来往间，创造的羁绊大概也是存在的。
黄鼠得以了解许多可爱的人——动物之森是一个能够通过设计和搭配表达个性的游戏，每个人都拥有着的漂亮的室外景观和室内装饰，还有各种各样的可爱衣服可以收集。这是难得的体验。
可是在享受着玩家之间互相游玩的乐趣之后，黄鼠总是无法停止想到游戏所带来的排他性。在 Switch 持有者之间、在动物之森持有者之间，这部游戏让大家联系地更紧密，可是更多的没有游戏或没有主机的人，却被悄悄地排除在外了。
本身就不是联系很紧密的人群，现在被进一步分割成为了“动森玩家”和“非动森玩家”。而在意动森玩家的非动森玩家，在几乎所有人都在讨论一个自己不懂的事物的时候，大概会有很大的被排挤的感受吧。这样的感受也是黄鼠在许多人都在讨论技术话题的时候，不停地感受到的。对于本身就已经充满焦虑的人而言，这是最不想要的东西之一了。
黄鼠在意着非动森玩家的成员，但作为一个需要 Switch 主机才可以玩到的游戏，想要同时玩到实在是太大的投入了。黄鼠曾构思了一个可以让网络上的人远程玩 Switch 的计划，利用微控制器模拟手柄键入远程玩家的输入，同时利用采集卡或摄像头将画面实时传输给对方。那大概是一个可行的计划，但一定会很复杂吧，而且网络延迟也许会使游戏体验差到无法接受。似乎没有更多人有兴趣的计划，黄鼠也没有动力开始做了。
黄鼠在玩着动物之森。在接近着其他玩家的同时，却不知觉地也在同时疏远更多人呢。这样一个难得的交流的理由，却无法在更多在意的人身上使用。这样的事实，令人担心呢。

除特殊说明以外，本网站原创内容采用知识共享署名-相同方式共享 4.0 国际许可协议</a>进行许可。转载时请注明来源 FiveYellowMice 的博客</a> ，以及原文链接</a>。