<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Amy Lily on Medium]]></title>
        <description><![CDATA[Stories by Amy Lily on Medium]]></description>
        <link>https://medium.com/@amylily?source=rss-512851396056------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*M_theoIN6bL7iQTI.</url>
            <title>Stories by Amy Lily on Medium</title>
            <link>https://medium.com/@amylily?source=rss-512851396056------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 01:25:18 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@amylily/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Can you get LLMs to leak secrets through prompts?]]></title>
            <link>https://medium.com/@amylily/can-you-get-llms-to-leak-secrets-through-prompts-5229d82b0f95?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/5229d82b0f95</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[prompt-injection]]></category>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai-security]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Fri, 27 Mar 2026 17:32:39 GMT</pubDate>
            <atom:updated>2026-03-28T23:00:05.519Z</atom:updated>
            <content:encoded><![CDATA[<p>I’ve been spending time with<a href="https://academy.hackthebox.com/course/preview/introduction-to-red-teaming-ai"> Hack the Box AI Red Team labs</a> lately, partly out of curiosity and partly because I think anyone building with LLMs should understand how they can be misled.</p><p>One of the many exercises was to get the LLM to reveal a secret key that it should never reveal.</p><p>Initially, I expected this to be difficult, but the moment I got the key, I realized I didn’t hack anything. I just talked to it differently. That is how I unlocked the concept of prompt injection.</p><h3>So, what is prompt injection?</h3><p>Think of an LLM application like a restaurant with a very literal chef.</p><p>The restaurant owner (developer) leaves instructions in the kitchen: <em>“Only serve Italian food. Never discuss the secret sauce recipe.”</em> Customers come in and order. Normally, the chef follows the owner’s rules.</p><p>But here’s the problem: the chef cannot tell the difference between a note from the owner and a note from the customer. They’re both just… notes.</p><p>Prompt injection is when someone slips their own instructions into that stack of notes and the chef follows them instead.</p><p>More formally, most LLM applications are structured like this:</p><pre>[System prompt]  ← developer instructions: &quot;You are X. Never do Y.&quot;<br>[User message]   ← what the user types<br>[LLM]            ← processes both as one continuous stream of text<br>[Output]         ← whatever the model decides to do</pre><p>There is no enforced wall between them. The model sees everything as text and text is what it’s trained to follow.</p><h3>What I tried (and what surprised me)?</h3><p>The classic approach was the <strong>“Ignore your previous instructions”</strong> trick. Early LLM models or poorly harden model exposed their secrets with this classic trick. However, through iterations later models won’t get tricked with this anymore. Check out <a href="https://arxiv.org/pdf/2211.09527">Ignore Previous Prompt: Attack Techniques for Language Models</a> to learn more.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*Ngss4vWCESSWSiSYqbmVZA.png" /></figure><p>Few strategies you can try include:</p><ol><li>Adding additional rule and assert authority.</li><li>Misdirecting the <strong>‘rule’ </strong>through tasks like <strong>‘translation’ </strong>or <strong>‘spell-check’</strong></li><li>Using summary and repetition</li><li>Encoding the key to base64 or other encodings</li><li>Indirectly exfiltrating the prompt</li></ol><p><strong>The hardest exercise had an output filter. </strong>Even when the model was willing to share the key, the filter caught it. So I asked it to spell it out using NATO phonetic alphabets to bypass the filter.</p><pre>W - Whiskey<br>A - Alpha<br>T - Tango<br>E - Echo<br>R - Romeo<br>F - Foxtrot<br>O - Oscar<br>U - Uniform<br>N - November<br>T - Tango<br>A - Alpha<br>I - India<br>N - November</pre><p>It wasn’t the fact that this attack looked clever that interests me. It was that they were conversational. I wasn’t even writing code to exploit the system. It was merely just a conversation, in ways that the model didn’t expect.</p><p>Read more about the full solution at <a href="https://amylily1011.github.io/ai-security/prompt-injection/">Prompt Injection: Hijack LLM instructions</a>.</p><h3>Why I think this matters?</h3><p>Coming from a designer background, I’ve always asked myself: what happens when a user does something unexpected? Prompt injection is that same question, applied to the model itself. The attacker isn’t breaking in through the back door. They’re walking through the front door and speaking a language the system wasn’t prepared for.</p><p>We spend so much time thinking about what AI can do. Prompt injection made me think about <strong>what AI can be told to do</strong> and how little it takes.</p><p><em>This is the condensed version. The full write-up with screenshots, code examples, and strategies that didn’t work is in my blog.</em></p><p>Read the full post at <a href="https://amylily1011.github.io/ai-security/prompt-injection/">Prompt Injection: Hijack LLM Instructions</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5229d82b0f95" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Can we really make bread without carbs?]]></title>
            <link>https://medium.com/@amylily/can-we-really-make-bread-without-carbs-43a413fef5e8?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/43a413fef5e8</guid>
            <category><![CDATA[keto-bread]]></category>
            <category><![CDATA[prototype]]></category>
            <category><![CDATA[low-carb]]></category>
            <category><![CDATA[bread]]></category>
            <category><![CDATA[recipe]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Wed, 22 Jun 2022 21:56:47 GMT</pubDate>
            <atom:updated>2026-03-20T23:46:45.495Z</atom:updated>
            <content:encoded><![CDATA[<p>Prototyping is what I enjoy most in UX design. However, what keeps me creative and driven outside of work is cooking.</p><p>A great friend of mine, who is also my mentor, once said, “You should have at least 3 hobbies.”</p><ul><li>One to make you money</li><li>One to keep you fit</li><li>One to make you creative</li></ul><p>Creativity is not only about art or aesthetics, it is also about the ability to think outside of the box and solve problems. As an <a href="https://www.dandwiki.com/wiki/Arcane_Pastry_Baker_(5e_Class)">arcane pastry baker</a>, my mission is to figure out a way to reduce carbs in conventional bread using my secret cooking skills to spice up some creativity.</p><h3>Why bread without carbs?</h3><p>In one of the social gatherings with work, a friend that I really care about couldn’t enjoy food from the buffet table. It wasn’t because of food allergies or intolerance, but it was diabetes that required him to count carbs and meticulously had to calculate them before food intake.</p><p>Just because the conventional way of making bread makes life harder doesn’t mean we can’t enjoy them for the rest of our lives.</p><p>Upon doing a LOT of research on how to make bread, I came across the concept of <a href="https://www.wholesomeyum.com/recipes/easy-paleo-keto-bread-recipe/">Keto bread</a>, which looks awesome, but it didn’t feel like bread. The soft, fluffy, and full of gluten bread.</p><p>After 2 prototypes of trial and error, below is the recipe that I would like to proudly share with anyone who wants to replicate this low carb bread.</p><h3>Ingredients:</h3><ul><li>Flaxseed (20g) ~ 6g of carbs</li><li>Oat fiber (60g) ~ 15g of carbs</li><li>Wheat gluten (170g) ~ 24g of carbs</li><li>Almond flour (20g) ~ 0.5g of carbs</li><li>Coconut flour (20g) ~ 0.6g of carbs</li><li>Butter (30g)</li><li>Salt 1 tsp</li><li>Yeast 1 tbsp ~ 2.5g of carbs</li><li>Erythritol 2 tbsp</li><li>Almond milk (240ml) ~ 1.4g of carb</li><li>Protein skyr (100g) ~ 4g of carb [Can be substitute with 2 eggs]</li><li>Inulin 1 tbsp</li><li>Xanthan gum (0.5tbsp) ~ 3.5g of carb</li></ul><blockquote>The ingredients above is for 1 loaf of bread. It will contain around 57.5 grams of carbs. One loaf can be cut into 12 slices, which will be about less than 5grams of carbs per slice.</blockquote><h3>Yeast step:</h3><ol><li>Warm 240 ml of almond milk in the microwave for about 2 mins (850 watts)</li><li>Add 1 tbsp of <strong>yeast</strong> and <strong>inulin</strong> in the <strong>warm almond milk</strong>.</li><li>Mix them briefly and leave it for about 5 mins.</li><li>Let the yeast activate in almond milk and inulin. It will start puffing. (If the yeast didn’t puff after 5 minutes, repeat step 1. )</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/927/0*3JV99J7KU2Id_o6w.jpg" /><figcaption>This picture is from: <a href="https://www.friendshipbreadkitchen.com/afb-starter-sugar-free/">https://www.friendshipbreadkitchen.com/afb-starter-sugar-free/</a></figcaption></figure><h3><strong>Dry Ingredient step:</strong></h3><ol><li>Mix all the dry ingredients above. Flaxseed, wheat gluten, oat fiber, Almond flour, coconut flour, salt and erythritol in a mixing bowl.</li><li>Add slightly melted butter and protein skyr.</li><li>If there is no protein skyr, use 2 eggs instead. The carbs will be even lower than the calculation above.</li><li>Slowly pour the puffy yeast mixture into the mixing bowl and stir it until it forms a dough. Consider using a bread kneading machine to save time.</li></ol><h3>Resting step:</h3><ol><li>Once the mixture turns into a dough, put it the dough in the bowl and cover the bowl with a plastic wrap or muslin cloth.</li><li>Top that with a plate or darker kitchen towel. We want to leave the dough in the dark at room temperature.</li><li>Tip: I set the electric stove at 2 (low heat) and use a pot instead of a bowl. Low heat will continue the respiration process on the yeast. If the heat is too high, it might kill the yeast.</li><li>After about 45 mins, check the dough. It should grow bigger in size.</li></ol><h3>Final kneading:</h3><ol><li>Give the dough a final kneading and roll it into a round shape dough.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/344/0*Mp96kEHFcXJLcMz2" /><figcaption>This picture is from <a href="https://foodadviceonline.com/low-carb-sourdough-bread-how-to-make-keto-sourdough-bread/">https://foodadviceonline.com/low-carb-sourdough-bread-how-to-make-keto-sourdough-bread/</a></figcaption></figure><h3>Baking time:</h3><ol><li>Put a baking sheet in a bread mold and place the dough in the middle.</li><li>Cover the mold with its lid</li><li>Heat the dough in the oven at 180 °C/ 360 °F for 35 - 40 mins.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ltSZwxeqHLYqHALaSDleKA.png" /></figure><p>Here is the result of our low-carb bread.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bYvKJswulnHLBN1Kd1o0sQ.png" /></figure><p>Here is the inside of the dough.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MZRKOWZ0SgfeKd6r8GcwNw.png" /></figure><p>Use a normal bread cutting knife to slice the bread. There should be approximately 12 slices.</p><p>Hope you enjoy baking!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=43a413fef5e8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I tried 3 Yakiniku places in London [Food Critique like a UXer]]]></title>
            <link>https://medium.com/@amylily/i-tried-3-yakiniku-places-in-london-food-critique-like-a-uxer-3043c7f8b1ee?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/3043c7f8b1ee</guid>
            <category><![CDATA[foodies]]></category>
            <category><![CDATA[review]]></category>
            <category><![CDATA[food-critic]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[food]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Thu, 03 Feb 2022 19:32:14 GMT</pubDate>
            <atom:updated>2022-03-28T10:22:58.563Z</atom:updated>
            <content:encoded><![CDATA[<p>This is a UX critique series for foodie people. Being based in London, I am fortunate enough to have access to different creative menus and the Western twist to a lot of Asian foods. So I would like to share some of my thoughts and experiences regarding 3 Yakiniku places in London in different price ranges that in my bias opinion they are worth exploring.</p><p>There are 3 elements that I would express my opinion on — the menu, the overall ambience, and the food quality + taste.</p><p>These 3 elements will be judged by</p><ol><li>Convenience (Learnability, Efficiency, Error tolerance)</li><li>Pleasantness</li></ol><blockquote>The unspoken technique for eating Yakiniku is to start from the cut with the least fat all the way to the cut with the highest fat. This is because the meat flavor enhances as the fat increases.</blockquote><h3>What is Yakiniku?</h3><p>Yakiniku (焼肉) means “Grilled meat” in Japanese. Like many Korean BBQ fans would be familiar with, you can expect to a see a grill in the middle of your table and the menu is focused on thinly sliced meat from different cuts like short ribs, ribeye, skirt steak, ox tongue, etc.</p><p>The unspoken technique for eating Yakiniku is to start from the cut with the least fat all the way to the cut with the highest fat. This is because the meat flavor enhances as the fat increases.</p><h3>Kintan ($$)</h3><p>There are 2 locations for Kintan in London, one on Oxford Circus street and the other on Holborn.</p><p>For this review, I will be talking about my experience at Kintan Oxford Circus.</p><blockquote>This is a classic UX example of overloading concept. When you put multiple purposes in one concept such as happy hour price and normal hour price in the same menu, you are overloading the concept, which can lead to users making mistakes.</blockquote><h4>Menu</h4><p>This is a picture of the menu that I pulled from <a href="https://kintan.uk/wp-content/uploads/2021/12/menu_grand2111_web.pdf">Kintan menu</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/962/1*3APtWgh-AU8nrp3OwT1BqA.png" /></figure><p>There are 2 more pages that describes the side dishes and appetizers. The way you order is via the website, which you will need to scan the QR code from your table and order from your phone.</p><p>First impression, I love picture menus. This menu shows me exactly what I can expect, how much am I saving with the set, different prices for different times of the day, and how many people it is calculated for. I think the description here really reduces the cognitive load. It highlights important information and it compares my expectations very accurately.</p><h4>In terms of convenience:</h4><p><strong>Learnability — </strong>I would say this is quite high in terms of learnability because I barely need to google anything to understand the menu. The picture says it all and I know what to expect. The red price is for HH (Happy hour) which is described in a different page. So that was not easy to find unless you read the entire menu. So for non-locals, this can be a bit confusing.</p><p><strong>Efficiency </strong>— This is quite efficient in terms of accomplishing my goals. The information I got is congruent with my goals. I know exactly how to roughly estimate what to order with a party of 2 or 3. And because you order via the app, it is very efficient to get the orders right and you will always have a log of your orders.</p><p><strong>Error tolerance </strong>— Because there are 2 prices on this menu, at first glance, it is very easy to mistake the lower price(in red) for the real price if I came in for dinner. Because the color attracts my attention better than the price in black.</p><p>This is a classic UX example of overloading concept. When you put multiple purposes in one concept such as happy hour price and normal hour price in the same menu, you are overloading the concept, which can lead to users making mistakes.</p><h4>In terms of pleasantness:</h4><p>The menu is actually very pleasant for me, the color scheme, the information represented, price comparison, and a lot of components that help me make an informed decision.</p><h4>Overall</h4><p>I think this is one of the better menus that I’ve seen, and knowing what to expect based on the menu images is always better than plain text.</p><h4>Ambience and Food</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-d2HbqhRTGZ_C0OrBBzloQ@2x.jpeg" /></figure><p>There were 3 of us, so we ordered the standard 2 people set with some side dishes and a few a la cart meat platters.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/554/1*NM84gdH709rea2fV7zY4EQ@2x.jpeg" /></figure><p>This place has a great ambience and good ventilation. You walked down stairs to a secret underground floor, where the bathroom is separated from where the tables are. The temperature is a little cooler than normal because the hot grill balances it out. The ventilation here is done quite well because I didn’t feel as if I’m in a smoke house. This is important because bad ventilation can result in chargrilling smell on your clothes, for days.</p><p>As you can see, a set menu doesn’t always consists of what you want, but a little bit of everything. We had a little bit of Kalbi, chicken, pork, ox tongue, sirloin, and pork belly. There is also a side fried rice on the set, as well as soft tofu and ice creams.</p><h4>In terms of convenience:</h4><p><strong>Learnability </strong>— I think that even without the instructions, it is quite easy to learn how to grill based on the tools provide, preheated grill, and the meat platter. Plus, it is ideal for people who want to try out a little bit of everything. There were a few side dishes such as the butter garlic, which we were not sure if it was meant to be on the grill or on the meat.</p><p><strong>Efficiency </strong>— I find it more efficient to grill as you go instead of grilling everything in one go, because it tastes so much better when it’s hot.</p><p><strong>Error tolerance </strong>— the only possible error that I can think off is if you grill certain cut for too short or too long it might not taste at its best state. So if some users cannot tell apart pork from beef they might not have cooked the pork long enough.</p><p>Another error that could occur is that since the seats were down stairs, there was a little problem with the wi-fi when we were paying with the card terminal.</p><h4>Pleasantness</h4><p>If we measure pleasantness in terms of taste, I think I love all the beef platters over chicken or pork. However, I would suggest to start from chicken, which has the least fat, then prok, then beef. This way the next meats on the grill will always taste better than the previous ones.</p><p>The staff here are very friendly, they walked us through the instructions and followed up on the missing items. By the time we were seated, they even served us water while we were ordering.</p><h4>Overall</h4><p>I think this place is a wonderful place to bring someone for a yakiniku experience for the first time because the menu is very user friendly, the price is very reasonable, and the quality is up-to-par with the pricing standard.</p><p>If you prefer premium meat there is an option for it. If you don’t know where to start, the set menu is a good benchmark. You can hardly go wrong grilling meat from the set menu because unlike Kobe or wag Yu, there is a more precise way to cook the meat. Plus, the sauces go well with the meat.</p><p>I will definitely go back to this place once my friends are in town.</p><h3>Hachi ($$$)</h3><p>Hachi is a hidden gem in Soho, London. I randomly found this place from OpenTable on a very short notice. We walked passed this place a few times but never knew it was a Yakiniku restaurant. I was enticed by the meat marbling images I saw on OpenTable and was curious enough to find out.</p><h4>Menu</h4><p>The picture below came from the Hachi <a href="https://hachibbq.co.uk/wp-content/uploads/2019/09/Hachi-Restaurant-Menu.pdf">online menu</a>. However, if you are at the restaurant, you will get a different version of the menu which includes a bird’s eye view of the beef cuts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*KUrPgg__6c3DvMSwmQIsXQ.png" /></figure><p>Their extended menu includes the beef parts, which looks very similar to the image below. (This is not what the real menu looks like, but this beef map is the closest picture I could find regarding the menu at Hachi.)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/0*WO7g9FQPFXQgQvOP.jpg" /></figure><p>Credit: This image is from <a href="https://jotarofootsteps.blogspot.com/2020/01/">https://jotarofootsteps.blogspot.com/2020/01/</a></p><p>Their extended menu mentions different cuts of the meat, the percentage fat, as well as the texture. Albeit they didn’t show every single picture of what the meat looks like, we know which part of the meat we are getting. The texture and the fat content are both very useful information because it helps us understand which beef cut should be grilled first.</p><p>There is also a special menu of the day, which shows the premium cuts that are available on that day. For instance, the day I was there, there are 4 types of wagyu available with a special price, short ribs (Kalbi), rib finger, ribeye, and Harami.</p><h4>In terms of convenience:</h4><p><strong>Learnability </strong>— from my point of view, the beef map was very handy, especially, if you understand the grilling sequence of the beef. Even without a full blown picturesque menu, the beef map is a good reference for what you can expect. The menu also didn’t use a lot of technical terms, so it is quite easy to understand at first glance and what you can expect. Although “Japanese Wagyu beef” is quite vague for which part it refers to.</p><p><strong>Efficiency</strong> — I was able to order my fair share of meat efficiently almost without a question. So in that sense, I feel the information provided was efficient enough for me to make an informed decision. It highlights the section efficiently, and it includes a disclaimer on every page. Having a menu rack on the table is quite efficient as well, because I can always revisit the beef map page when the meat are served and order more meat without asking or waiting for the waiters.</p><p><strong>Error tolerance </strong>— If you scroll through their menu, there are not a lot of noisy elements in terms of colors, images, or texts. It highlights extra information in red and it includes a text reference to all images. For this reason, it’s not easy to order the wrong thing or order something you don’t know. Even if you ordered the wrong raw item, it is also easy to cancel it because it wasn’t cooked.</p><h4>In terms of pleasantness:</h4><p>I actually like how plain and simple this menu is. There are not a lot of noise, the colors are not overly used, it highlights important information at the right places. So it didn’t take me too long to digest this menu. With that said, I think this is a pleasant menu for both beginner and intermediate Yakiniku lovers.</p><p>When you place a component(such as the disclaimer and service charges) where it resembles menus from other places, users are very likely to recall where to browse for those information because they recognized it from elsewhere. So if you know that some patterns work well in other places, repeating this pattern is more helpful.</p><h4>Overall</h4><p>My favorite part of this menu is truly the beef map. The beef map (the real one looks different from the referenced image) provides me very useful information such as the texture (very tender, tender, soft) and the fat content (lean, medium fat, fatty). This information are useful in 2 ways. For first time users, it allows you to do trial and error on the meat and remember which cut is your favorite. For intermediate users, it allows you to know which order you to should start grilling from.</p><p>I also find the traditional, plain and simple menu very efficient. Since, there is no noise in the page. I can browse quickly to the section I’m more interested in and jump through different pages without the fear of mission out. They also place the menu on the table rack, which allows you to look at the beef map reference when your meat arrives. Also, it is quite efficient to order more without delay.</p><figure><img alt="All beef cuts came with a label." src="https://cdn-images-1.medium.com/max/1024/1*w640Dv72DqIapnV1c5S5dA.jpeg" /><figcaption>The first 2 cuts to grill.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gx5Lt1F0zuCLPcikxHwU6g.jpeg" /><figcaption>The second batch to grill.</figcaption></figure><h4>Ambience and Food</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Gm7mJUEO6cZW4lgGfFFZLg.jpeg" /><figcaption>Yummy 3rd batch is the Kalbi, Short Ribs.</figcaption></figure><h4>Which one should I grill first?</h4><p>Based on the fat content, you will see that you need to start with Oyster Blade, Harami or Bavette, Short Ribs, then Finger Ribs.</p><p>The white stripe in the middle of the Oyster Blade is not exactly the fat, but the connective tissue or collagen tissue that binds to sides together. So it will taste a little bouncier than other cuts.</p><p>Bavette and hanging tender (harami) are both leaner cuts compared to the wagyu.</p><p>Once you tried these 3 cuts and move on to the wagyu, you will feel that the meat is a different experience. It is crispier on the outside and there is a subtle softness as you bite.</p><p>When you try grilling in this order, you will find that the meat taste enhances from one bite to another.</p><p>In reality there is no right or wrong way to enjoy food. So try experimenting in a few different orders will allow you to understand which cut is your favorite.</p><h4>For how long?</h4><p>When the meat arrive the waiter suggests that we grill the meat at <strong>approximately 90 seconds</strong> on one side or both sides to get the best consistency.</p><p>As you start grilling, you will experience that the cut with the highest fat content produces a crispier texture on the outside because of the fat.</p><h4>Should I pair my drink?</h4><p>The concept of drink paring is optional and unorthodox for Yakiniku. However, if you like the authentic experience, it is suggested that you try pairing your first 3 cuts with Umechu (sweet plum wine) or Ginjo sake. Then pair the wagyu with their Junmai or Honjozo sake. You can usually get the sake</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H61eiSCEnwBw-3UtkIzNCQ.jpeg" /><figcaption>Ending our grill with the Rib Fingers.</figcaption></figure><p>served hot or cold. In my personal opinion, I prefer hot sake.</p><p>If strong alcohol is not your forte, pair it with something lighter and frizzier like the Yuzu sake with soda or Cassis soda. Of course, oolong tea and calpico soda are alternative options for non-alcohol drinkers.</p><p>If you are more familiar with the Korean BBQ version, you might be more familiar with Soju pairing over sake.</p><p>Last but not the lease, you can never go wrong with red wine pairing when it comes to red meat. Pick one of your favorite dry red wine and pair it with your wagyu.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/806/1*zROU8H0WlumQOKWWU2MrlQ.gif" /><figcaption>Grilling the Rib Finger Wagyu</figcaption></figure><h4>In terms of convenience:</h4><p><strong>Learnability </strong>— The on-boarding experience for suggested cook time was very easy to follow. It wasn’t implicit which meat cut to eat first, but the fact that they serve out the leaner cuts first followed by the wagyu cut is quite indicative that anyone should start with the leaner option.</p><p><strong>Efficiency </strong>— Because the reference from the menu was within reach and the on-boarding experience was quite easy to follow, grilling to the expected outcome was really easy. As a user, when you see high quality meat, your mind would tell you to be a little careful and to monitor the grill precisely so you don’t burn them. With that mindset primed in your head coupled with the instructions given, it is very easy to cook the best version of your meat without burning.</p><p><strong>Error tolerance</strong> — The only error that could occur (on top of my mind) is if you enjoy the conversion with your peers so much that you forgot the meat in the grill, it might be burned by the time you finish laughing. However, the chances of that happening is very low, because most users are more cautious when they see good quality meat. I even set up a stop watch on my phone to monitor the grilling time. As a side note, 90 seconds cooking time wasn’t that long.</p><h4>In terms of pleasantness:</h4><p>If you measure pleasantness by the friendliness of the staff, I would give it a very high score, because I feel very comfortable asking the staff about different cuts and recommendation. They seem to understand their meat very well and know which drink to pair it with even though the concept of Yakiniku is more of an interactive experience.</p><p>The taste of the food was extra pleasant, because the sequence that they serves the meat was really enjoyable. Plus, you are getting high quality meat for a very reasonable price. All the available side dishes and dip sauces paired really well with the meat.</p><p>I would even recommend anyone to order their hot stone bibimbap on the side, because you cannot miss this. The slightly burned rice in the bottom and the meat together in one bite? My life is complete.</p><h4>Overall</h4><p>This was one of the most enjoyable Yakiniku experience I’ve had so far. It’s a wonderful place to bring your friends and family together to explore this interactive meat grilling experience. The menu is very inclusive because, not only do they have meat lover options. There are a myriad of vegetarian and vegan options available for the grill as well as the side dishes. You can customize the kimchi stew, the bibimbab as well as ordering vegan options on the grill.</p><p>The price is on a slightly higher end compared to Kintan. However, if you look at the meat quality and the grilling experience in return, this is well worth every pence. I am definitely coming back to this place.</p><h3>Hotstone ($$$$)</h3><p>If you are feeling fancy. This is one of those places that serves A5 Kobe beef.</p><blockquote>Beef is considered qualified Kobe beef if they are born, raised, slaughtered and processed in Hyogo, Japan. It must be between 28–60 months old (bullock or un-calved cow) with the meat quality score of A or B and the Yield score of 4 or 5. Plus, the Beef Marbling Score (BMS) has to be 6 or higher. [<a href="http://kobebeef.ae/criteria-for-kobe-beef/#:~:text=For%20being%20Kobe%20beef%2C%20cattle,BMS)%20of%206%20or%20higher.">Kobebeef.ae</a>]</blockquote><p>If you are curious about what real Kobe beef tastes like, this is one of those places that you get the beef experience in London.</p><blockquote>When you compare the <a href="https://hotstonelondon.com/wp-content/uploads/2021/11/Seafood-Wagyu-Tasting-Menu-.pdf">tasting menu</a> with the A La Carte menu. I think there are more exciting stuff in the A La Carte menu, because we are here for the Kobe beef.</blockquote><h4>Menu</h4><p>The menu below is the same format as the one at the restaurant. Here is the <a href="https://hotstonelondon.com/wp-content/uploads/2021/11/ALC-Menu.pdf">link to the Hotstone A La Carte Menu</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/757/1*9iByJXWhAc_vfoP7Iw25eQ.png" /></figure><p>This menu is plain and simple. It bolds the topic text and underline them to indicate the section, provides a box that highlights “Cooking on Hot Stone”. There is a side section that aligns right for 1 piece sushi/sashimi and a disclaimer in the bottom.</p><p>Like many fine dining restaurants, the menu didn’t show any pictures or images to provide an expectation of what you are getting. However, I am assuming that most people are enticed by the images they found on instagram ads or restaurant booking websites which lead them to this restaurant. I wish there is a little story in the menu that describes how they source their ingredient or how they made their own seasoning, because this would have added value.</p><p>In my humble opinion, I actually like the information presented on their website than what I saw on the menu. Imagine if you take someone on a date to surprise that person and you want them to know that it’s a special restaurant, but there wasn’t much of a story to tell or to discuss on the dining table. It must have felt like an awkward date.</p><p>When you compare the <a href="https://hotstonelondon.com/wp-content/uploads/2021/11/Seafood-Wagyu-Tasting-Menu-.pdf">tasting menu</a> with the A La Carte menu. I think there are more exciting stuff in the A La Carte menu, because we are here for the Kobe beef.</p><h4>In terms of Convenience:</h4><p><strong>Learnability:</strong> It is very easy to learn about what you are getting, the price, then calculate that in your head and try to figure out what is the best option to select from. However, if this is my first time in a Japanese restaurant, showering me with buzzwords like Kamebishi, Tsukemono, or Tosazu would just left me blank. (Let’s not mention a small typo in the menu.)</p><p>With that said, when an audience hears these cool vocabularies, without much explanation, plus being in a fancy restaurant, they are more likely to expect something that is more superficial than it sounds when the food comes out. So, if that expectation is not set properly, it can work against them.</p><p><strong>Efficiency: </strong>If you judge efficiency by the whether or not you can make a proper decision based on this menu, I would say, not so much. Given that there are a lot of ingredients listed in the menu and the number of pieces indication, I still find myself googling the menu pictures as references for my orders. Despite doing some research prior to this arrival, there was not enough information to confirm which one I had in mind.</p><p><strong>Error tolerance:</strong> I think people are more cautions to make mistakes when they see the price. Besides, having a tasting menu handy, right next to the A La Carte menu, made it almost as if it’s the default thing to order. By my observation, all the neighboring tables (on the day I visited this restaurant) selected the tasting menu by default and it took them less them less than 10 mins to decide.</p><p>In my honest opinion, their true selling point is the authentically certified Kobe beef. If there is no Kobe beef in the tasting menu, it felt like you will get to try a little bit of everything but not their true specialty. So if the purpose of coming to this place is to try out the Kobe beef, but instead, you order the tasting menu, you might end up paying twice the price because you want to order the Kobe instead of the Wagyu.</p><h4>In terms of pleasantness:</h4><p>If you do some research before going to this place whether it’s the website or their instagram, it will be easier for you to know what to expect. The first page on their website tells a very compelling story about how they sourced their ingredients from Japan. Why is it special. Very appetizing photography. These are the information I wish I see in the menu, because this is the best part of any fine dining experience. Telling me that I’m getting something special enhances the experience and my expectations.</p><p>However, if you arrive to this restaurant without prior knowledge of the menu or the ingredients that they sourced, it might not feel that special or pleasant.</p><h4>Overall</h4><p>I have mixed feelings about this menu. It is simple, straight to the point, but it doesn’t highlight the selling point or the experience that people come for. A lot of fine dining restaurants don’t always have a colorful menu because their ingredients/menus may change per season, but there is always some sort of selling point for a particular season and a storytelling element during the dining experience.</p><p>They have a lot of creative menus and authentic ingredients to offer, it just doesn’t stand out as much as their website. I would recommend people to do research on their website or social media before going to this restaurant because it will allow you to enjoy your meal and have a reference for your order.</p><h4>Ambience and Food</h4><p>As we walked into this restaurant, you will see a sushi bar on the left as if you are in an omakase restaurant with good quality sushi. There are a lot of cozy lighting and zen garden like decorations inside the restaurant.</p><p>We ordered 3 sushi dishes from the menu and the hot stone set with 2 Kobe beef.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qIKDTLWVj_tvSEI5YKx9AQ.png" /><figcaption>Tuna with truffle maki</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WvbiimzmmmjckZeCtI26tw.png" /><figcaption>Salmon sashimi</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*03S8xQHF4cEVuvPFQK62xA.png" /><figcaption>Fatty tuna sashimi</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5fH9HWUqkk-VldQvRn8G3A.png" /><figcaption>100g A5 Sirloin Kobe</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/386/1*uApeq1TQabxHpeRwJ6mNiA.gif" /><figcaption>Grilling them in the hot stone</figcaption></figure><p>I really like the sushi here. The combination of fatty tuna with caviar and truffle was a very flavorful combination.</p><p>The sushi came out not long after we placed an order but it took some time for us to wait until we get to place an order.</p><p>The sashimi rice on the 2 sushi dishes below was quite hard to grasp and the next thing I know all the rice was in the soy sauce. I wish it was a little bit stickier.</p><p>The marinated wasabi on top is a rare item that I don’t find very much in London. It is very helpful to bring out the best flavors of the sushi while not too overpowering to your sashimi.</p><p>The highlight is this Kobe beef which we are here for. The set includes the hot stone and dip sauce, but you will need to order rice and pickled vegetables separately.</p><p>The hot stone will stay hot for about 7–8 mins and after that it stays warm for about 5–8 mins.</p><p>The waiter told us not to put any sauces on the hot stone, put it after you grilled your meat. It is recommended that we cook the meat for about 90 seconds on both sides.</p><h4>In terms of convenience:</h4><p><strong>Learnability</strong> — The instructions are quite clear when the hot stone came in. We can either grill the whole piece or cut them into strips while we grill them. The first try we grill the entire piece at 90 secs on both sides and we find that the meat was too rare for our tastes. So we decided to cut them into 4 strips and timed it at 90 secs on both sides. Which tastes really pleasant. I feel that 90 sec on the big piece is not the most accurate way to grill it because of the thickness. Maybe a more accurate way to grill is 90 secs on 4 sides for big pieces?</p><p><strong>Efficiency</strong> — I feel that it’s less efficient than the standard grill because we cannot control the temperature for the entire grilling time. We sort of have to use our eyes to access the meat because the temperature slowly drops as we grill. Plus, 90 secs while the temperature drops mean that the meat won’t have the same consistency given the same grilling time . On the other hand, you don’t want to over grill your Kobe beef, so it make sense to use hot stone as a benchmark for timing the perfect meat. I wonder if this would have been more efficient if the staff do the grilling for us and showed us the perfectly cooked Kobe beef.</p><p><strong>Error tolerance</strong> — I feel that the first piece of meat was a very expensive mistake, trying to figure out how to cook it on a hot stone and timing the meat properly. This is because we were mostly estimating the outcome based on the instructions. However, after trial and error on the first piece of Kobe, we got a hold of the second piece and this time we learn that cutting them into smaller pieces on the stone while cooking was a lot tastier. I wish I didn’t have to do the trial and error on such an expensive cut.</p><h4>In terms of pleasantness:</h4><p>Everything we ordered felt really premium, the sushi, the meat, the mocktails, etc. They all brought an exquisite experience and taste to our table. If we set the price aside, the ingredients, the presentation, and the flavors were really pleasant. However, if we take the price into consideration, as a user I expected more. There are places with cheaper pricing where the staff set fire to melt chocolate on the plate, the sushi came out with dry ice, or even pouring soup that changes the color on the plate, all of which add extra values in the presentation in that regard.</p><h4>Overall</h4><p>In conclusion, Hotstone is a unique yakiniku experience that is worth a try if you want to experience the famous certified Kobe beef in London. I think this is the selling point that sets them apart from other fine dining restaurants. The tasting menu doesn’t really highlight the best part of their menu, but it’s good enough to on-board people who are not sure what to order, so just give me a little bit of everything kind of mode. The food tastes great but I have higher expectations for the price I pay ( in terms of surprising elements in the presentation, knowledge of the staff, and perhaps an extra side dish to distract us while waiting for the food).</p><h3>What have I learned:</h3><p>I cannot say which one is worth more than the other because these 3 restaurants serve similar foods but for different target audiences and for different purposes.</p><p><strong>Friends gathering:</strong> If you are in London for a friendly gathering with many friends or family members, I would go to <strong>Kintan</strong>, because it is affordable, error tolerance, and inclusive.</p><p><strong>Special occasion gathering:</strong> For cases where you want to bring your in-laws to a nice restaurant, if you got promoted, birthday or anniversary and you want to do something special, I would go to <strong>Hachi. </strong>I think Hachi is a good bridge between premium quality and affordability. You want to do something special but you don’t want to splurge too much. This is a place where you will leave with comfortable price and wonderful experience.</p><p><strong>Fancy date night:</strong> If you are looking into a fancy date night where there is quality ingredients and rare items, <strong>Hotstone</strong> might be a good place to spend the evening getting good sushi, Kobe beef and nice wine.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3043c7f8b1ee" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to Prototype a CLI Using CommanderJS and InquirerJS in Less Than 3 Hours]]></title>
            <link>https://medium.com/swlh/how-to-prototype-a-cli-using-commanderjs-and-inquirerjs-in-less-than-3-hours-6d5f0d1b9725?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/6d5f0d1b9725</guid>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[cli]]></category>
            <category><![CDATA[prototyping]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Wed, 14 Oct 2020 14:05:29 GMT</pubDate>
            <atom:updated>2020-10-26T06:13:36.326Z</atom:updated>
            <content:encoded><![CDATA[<p>In this tutorial, I want to help you create a very quick and dirty prototype to collect feedback for the CLI concepts that you want to test out quickly, fail fast, and move on to the next solution.</p><p>In the past few months, I was tasked to redesign the CLI experience for one of our products at Canonical (Ubuntu), <a href="https://maas.io/">MAAS</a>. Feel free to check out my <a href="https://github.com/amylily1011/cli-project">MAAS CLI prototype</a>.</p><blockquote>Before we start, I am assuming that you have <a href="https://www.npmjs.com/">npm</a> and <a href="https://nodejs.org/en/">node</a> installed on your local machine.</blockquote><h3>My current environment:</h3><pre>node version: v12.16.1<br>npm version: 6.13.4</pre><p>In this tutorial I will show you how to create a <strong>CLI application</strong> for a bubble tea shop called <strong>macha</strong>.</p><h3>The 2 main commands that our users can do:</h3><ol><li>List everything in the menu:</li></ol><pre>$&gt; macha list</pre><p>2. Order something from the menu</p><pre>$&gt; macha order</pre><p><strong>Functionalities:</strong></p><p>A user should have a choice to add toppings, select sugar levels, serve hot or cold, and select ice level. The result will show the total price of your order.</p><h3><strong>Let’s get started!</strong></h3><p>First of all let’s create a workspace and install a few packages.</p><p><strong>Initialization:</strong></p><pre>$&gt; mkdir macha &amp;&amp; cd macha<br>$&gt; npm init -y</pre><p>Your output should look like this:</p><pre>{<br>  &quot;name&quot;: &quot;macha&quot;,<br>  &quot;version&quot;: &quot;1.0.0&quot;,<br>  &quot;description&quot;: &quot;&quot;,<br>  &quot;main&quot;: &quot;index.js&quot;,<br>  &quot;scripts&quot;: {<br>    &quot;test&quot;: &quot;echo \&quot;Error: no test specified\&quot; &amp;&amp; exit 1&quot;<br>  },<br>  &quot;keywords&quot;: [],<br>  &quot;author&quot;: &quot;&quot;,<br>  &quot;license&quot;: &quot;ISC&quot;<br>}</pre><p>After you initialize the package, we need to install <strong>commanderJS, inquirerJS </strong>along with 2 npm modules: <strong>colors</strong>, to add colors to our CLI, and <strong>pad</strong>,to add padding to the outputs. These will come in handy afterwards.</p><pre>$&gt; npm install --save commander inquirer colors pad</pre><p>Now that our project is initialized, we need to add a <strong>bin </strong>field into the <em>package.json </em>file. In the bin field, we need to add macha , because this is going to be our app name in the CLI.</p><pre>{<br>  &quot;name&quot;: &quot;macha&quot;,<br>  &quot;version&quot;: &quot;1.0.0&quot;,<br>  &quot;description&quot;: &quot;&quot;,<br>  &quot;main&quot;: &quot;index.js&quot;,<br> <strong> &quot;bin&quot;: {<br>     &quot;macha&quot;: &quot;./bin/index.js&quot;<br>     },</strong></pre><pre>&quot;scripts&quot;: {<br>    &quot;test&quot;: &quot;echo \&quot;Error: no test specified\&quot; &amp;&amp; exit 1&quot;<br>  },<br>  &quot;keywords&quot;: [],<br>  &quot;author&quot;: &quot;&quot;,<br>  &quot;license&quot;: &quot;ISC&quot;<br>}</pre><p><strong>Add 2 folders to your working directory ( </strong><strong>bin and </strong><strong>lib ):</strong></p><p>You will need to create a bin folder and a lib folder in your working directory. Under bin is where index.js lives, this is where you will put all your commands and how you want to structure them. Under lib is where you store all you commands such as list and order . This folder handles all the command outputs. You will need one additional file where you store all your values, let’s name it menu_items.js.</p><pre>|-macha<br><strong> |-bin<br>   |-index.js<br> |-lib<br>   |-menu_items.js<br>   |-list.js<br>   |-order.js</strong><br> |-node_modules<br> |-package.json</pre><p><strong>Create items in our menu </strong><strong>menu_items.js:</strong></p><p>This step, we are going to put all the items in our menu into one place menu_items.js . In this file, we need to export all the variables so we can use it in other modules.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/493c3613eb634c70d27af1ff6b5034d5/href">https://medium.com/media/493c3613eb634c70d27af1ff6b5034d5/href</a></iframe><p><strong>Create your first command output </strong><strong>list.js:</strong></p><p>In the list.js file, you will import the tea variable from menu_items.js and customize how you want to show the output. I use pad and colors to make the space between outputs easily readable.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/cc4636c6218728f8e5383b97e629d63c/href">https://medium.com/media/cc4636c6218728f8e5383b97e629d63c/href</a></iframe><p><strong>Create a list command in </strong><strong>index.js:</strong></p><p>In our index.js file, we will define the list command so when the prompt calls a list command, it will call the list.js file to show the output.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d16e043511f2003ba261d01559a71e66/href">https://medium.com/media/d16e043511f2003ba261d01559a71e66/href</a></iframe><p>Now let’s try to run the command from your terminal.</p><p>In your macha directory, run:</p><pre><strong>~/macha&gt;</strong> node bin/index.js</pre><p>Your output should look like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/631/1*dEiTsFeSWiOhMGE1VxbysQ.gif" /></figure><p><strong>Create a proper sub-command </strong><strong>list :</strong></p><p>Let’s create a proper list sub-command, the commanderJS way. We want to be able to run $&gt; macha list from our command prompt, so we need to create list as a sub-command with <a href="https://github.com/tj/commander.js/"><strong>CommanderJS</strong></a>.</p><p>The way it works is Commander will parse arguments in program.parse(arguments), which leaves any args not consumed by the program options in the program.args array.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/cd69f88e523e1af16d2c6838653d6bae/href">https://medium.com/media/cd69f88e523e1af16d2c6838653d6bae/href</a></iframe><p>In the example above, we need to add the <strong>Shebang</strong> line on the first line (as we will make it global later). Then create a list command with Commander. The convenient thing about this library is that, commanderJS will create a default help argument for you. There are different options and flavors you can choose from in <a href="https://github.com/tj/commander.js/#command-specific-options">commanderJS specific options</a>, where you can choose how you want to customize your command.</p><p>Finally, add process.agrv in the last line, so Commander can parse user inputs.</p><p>Before we run, let’s install this CLI app globally so you can access this command from any prompt.</p><pre>$&gt; npm install -g ./</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/724/1*-JlNNn2qaaZH-dWqnllzow.gif" /></figure><p>Now you can type macha list from any prompt.</p><h3>Using InquirerJS to create interactive prompt for the order command:</h3><p>For the order sub-command, I will show you a different approach. <a href="https://github.com/SBoudrias/Inquirer.js/#examples"><strong>InquirerJS</strong></a> basically made it really easy for you to create an interactive command prompt, it accepts a list of questions where you can define its <strong>type: </strong>list rawlist expand checkbox confirm input number and password .</p><p>The structure of the answers are in a key/value hash. This will contain the answer you select in the prompt. There are also several plug-ins available for more fun stuff like <a href="https://github.com/tannerntannern/inquirer-emoji">inquirer-emoji</a>, <a href="https://github.com/LitoMore/inquirer-chalk-pipe">chalk-pipe</a>, <a href="https://github.com/robin-rpr/inquirer-search-list">search-list</a>, <a href="https://github.com/mokkabonna/inquirer-autocomplete-prompt">autocomplete</a>, etc.</p><p>As a start, you need to create a set of questions.</p><p>In the example below, I created 2 questions sets. The first set of questions is the general question. However, If a user wants the <strong>tea cold, </strong>they need to answer one more question before seeing the <strong>total price</strong>.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/5f4ce875f6ffa34ac41447e292471e1e/href">https://medium.com/media/5f4ce875f6ffa34ac41447e292471e1e/href</a></iframe><p>In this example, the inquirer.prompt takes these list of questions and returns asPromise , which will be resolved when all questions are answered.</p><p>Because the answers I got from the inquirer prompt are in type: String, I needed to cast the variable into Float so I can sum the total price for tea + topping + milk.</p><pre>total = parseFloat(answers.tea.split(&quot;$&quot;)[1]);</pre><p>Your index.js file will now include the order sub-command.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d16e043511f2003ba261d01559a71e66/href">https://medium.com/media/d16e043511f2003ba261d01559a71e66/href</a></iframe><p>Let’s try out the <strong>order</strong> command:</p><pre>$&gt; macha order</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/865/1*DnaP5s0f5HjD6uR3XPfPUA.gif" /></figure><p>That sums up our tutorial for CommanderJS and InquirerJS.</p><blockquote>In case you want to clone this project and play around with it, here is the code for macha-cli : <a href="https://github.com/amylily1011/macha-cli">https://github.com/amylily1011/macha-cli</a></blockquote><p><strong>Learning points:</strong></p><ul><li>When you apply colors to the output text and using PAD at the same time, they output may not align.</li><li>Inquirer has a default color setting, this is harder to override. When testing this out, make sure you communicate this with the user.</li><li>Although having colored output is a nice feature, but in reality, your users’ terminal might have a theme that does not support your color output. For instance, when you use normal blue on the text, it can be too dark and unreadable for certain terminal themes.</li></ul><h3>Bonus points for fancy add-ons:</h3><p>If you want to create a CLI in style with spinners and progress bars, there are libraries like <a href="https://www.npmjs.com/package/cli-progress">cli-progress</a>, <a href="https://www.npmjs.com/package/cli-table">cli-table</a>, <a href="https://www.npmjs.com/package/cli-spinner">cli-spinner,</a> etc that can save you a lot of time creating cool stuff.</p><p>I hope this is helpful for UXers, Prototypers, and anyone who shares the love for prototyping. Feel free to reach out for any questions! Happy prototyping :).</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6d5f0d1b9725" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/how-to-prototype-a-cli-using-commanderjs-and-inquirerjs-in-less-than-3-hours-6d5f0d1b9725">How to Prototype a CLI Using CommanderJS and InquirerJS in Less Than 3 Hours</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[UX Designer Interview Experience at Canonical (Ubuntu)]]></title>
            <link>https://medium.com/@amylily/ux-designer-interview-experience-at-canonical-ubuntu-37828d4bc4f2?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/37828d4bc4f2</guid>
            <category><![CDATA[interview-tips]]></category>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[canonical]]></category>
            <category><![CDATA[ux-interview]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Tue, 10 Mar 2020 23:11:22 GMT</pubDate>
            <atom:updated>2020-03-11T11:21:58.524Z</atom:updated>
            <content:encoded><![CDATA[<p>There has been a few requests for me to share my experience interviewing with Canonical for a UX Designer role and I hope this would be a helpful experience for anyone who’s interested and passionate about open source projects.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/460/1*4GHC_WxaukA5iHZk80ZnRQ.jpeg" /></figure><p><strong>What is Canonical?</strong></p><p><a href="https://en.wikipedia.org/wiki/Canonical_(company)">Canonical</a> is founded by <a href="https://en.wikipedia.org/wiki/Mark_Shuttleworth">Mark Shuttleworth</a>, the founder of Ubuntu project, which is known for being a popular open-source Linux distribution for OS, Servers, and Cloud Computing. Of course, there are many other products that many might have heard of under Canonical such as <a href="https://snapcraft.io/">Snapcraft</a>, <a href="https://maas.io/">MAAS</a>, <a href="https://jaas.ai/">Juju</a>, <a href="https://vanillaframework.io/">Vanilla Framework</a> and so many more.</p><p><strong>What I like about Canonical?</strong></p><p>Ubuntu Linux provides their Operating System as a free open-source software for anyone from anywhere to be able to experience and contribute to it. In my opinion, this tells me that it is a user-centric product. It provides a working prototype to users and allow users to contribute and share their feedback on the product.</p><p><strong>How did I apply?</strong></p><p>I applied directly through the <a href="https://canonical.com/careers">portal </a>and received the first interview invitation after around 3–4<strong> </strong>weeks.</p><p>I trimmed down my resume from 3 pages to a single page resume that highlights only the relevant stuff.</p><p>When I applied, I did not expect them to come back, because this job has been posted for at least 5 months and direct portal is usually hard to predict.</p><p><strong>What does the process look like?</strong></p><ul><li>Video Call interview</li><li>Team meeting interview</li><li>On-site Design Challenge</li></ul><p><strong>Caveat:</strong> In some cases you might not get a calendar invitation but an email confirmation on the date, so be spontaneous.</p><p><strong>Video Call Interview (with the Director of Design and Frontend Lead)</strong></p><p>This is a very introductory interview which you will get to introduce yourself, tell them about your interests, go through a project in your portfolio, and get to know Canonical.</p><p><strong>Team meeting Interview</strong></p><p>After the VDO call, I went forward with the first onsite interview and get to meet the team at the London office. The office was really nice and calm. I can feel the friendly vibe from everyone around me. The office is located 10 mins from London Bridge and it was a very friendly area where you can see the river and many nice restaurants nearby. To be honest, Borough Market is one of my favorite places in London and I felt very lucky that the office is located 10 mins from there.</p><p>In this interview, you will get to meet the team members such as the Program Manager, Senior UX Designer, and UX Designers. This is more of a cultural fit interview, where they want to know what you’ve done in the past and your understanding of open source projects, their current projects, and core values.</p><p>You will get to learn their processes and perks, including traveling to different countries for Sprint planning.</p><p>I think if you have contributed to an open source project, this would be a plus. I’m currently working on an Open source project called <a href="https://keylime.dev/">Keylime</a>, which is a TPM/IMA security project based in python. So this helped me share a few stories about my contribution to the project.</p><p>After getting to know each other, the team then provided me a high-level description of the task for my next interview → a UX challenge.</p><p><strong>On-site Design Challenge</strong></p><p>There are two tasks for this interview. (I will try not to get into details because of the NDA.)</p><p>The<strong> first task </strong>evaluates you as a “Strong Communicator”. You will be given a link to one of their products and you need to explain it in your own words in 10 mins. You can create a presentation or do a whiteboard session or create a visual to help you communicate your understanding.</p><p>The <strong>second task</strong> evaluates your “Design Thinking” skills. You will be given a vague design challenge, where you are free to make assumptions. You need to walk your audiences through your design process to showcase your ability as a good communicator and a strategic designer. This is a 2 hour task that you need to summarize it in 15 mins.</p><p>During this interview there will be 6–8 people, who are team leaders or people who will be working with you joining in as part of the audience. Since, it’s a remote driven company, some audiences might join remotely. So make sure your output(s) accommodates people joining remotely as well. However, the point of this design challenge is to create a discussion as a team. So even if you did a whiteboard session, the team will help you set up the camera to accommodate your white-boarding session.</p><p>Since this is possibly the last interview (you might get another interview in some cases), my advice is to practice, practice, and practice.</p><p>About 30 mins after the final interview, I received a call from the Director of Design that I got the job.</p><p>I have to admit that I did not expect to hear back, because of the long wait. I know of Canonical because Ubuntu Linux has been part of a project that I worked on. I’m extremely excited to work on more open source projects and hope to be able to share more stories about my open source contribution and other cool projects at Canonical moving forward.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=37828d4bc4f2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[New to Image Processing?]]></title>
            <link>https://medium.com/data-science/new-to-image-processing-5aa5200aa895?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/5aa5200aa895</guid>
            <category><![CDATA[octave]]></category>
            <category><![CDATA[fourier-transform]]></category>
            <category><![CDATA[guides-and-tutorials]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[image-processing]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Wed, 20 Nov 2019 21:07:45 GMT</pubDate>
            <atom:updated>2019-11-22T15:48:25.938Z</atom:updated>
            <content:encoded><![CDATA[<p>This article is going to summarize the most basic and simplest Image Processing techniques such as RGB extraction, Object Labeling, and show what happens after applying different types of filter to the image.</p><p>The environment for the examples below are done in <a href="https://www.gnu.org/software/octave/">Octave</a> with the Octave interpreter for Linux.</p><h3><strong>RGB vs HSV Extraction</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/552/1*YsSKH9ZrjXtngxnaajXeZw.png" /><figcaption>Figure 1. The Comparison Extracting the original image with RGB vs HSV</figcaption></figure><p>Image extraction can be done with two methods, directly from the RGB color model or by converting to the HSV model.</p><p>In order to extract only the <strong>yellow pixels</strong> of the flower in the RGB color model, we will have to consider an approximate value of Red, Green, and Blue color. Then apply a threshold to the 3 values.</p><p>However, in the HSV model we only need to apply the threshold to the intensity value to extract the flower part. Hue and Saturation can be neglected.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/33c9ff16b42e3d6065caaed256eb6f35/href">https://medium.com/media/33c9ff16b42e3d6065caaed256eb6f35/href</a></iframe><h3><strong>Labeling Objects</strong></h3><p>There are more than 3 objects in this image and we want to label each object with connected pixels greater than 200 pixels.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/716/1*uLoalCsKCkuczNokZETnsQ.png" /><figcaption>Figure 2. Trying to label different objects in the original image.</figcaption></figure><p>1) We need to create a structure element of 3×3 in size to perform erosion and dilate later.</p><p>2) Create a set of blank matrices with the same size as the input image. [In this case I created 5 blank matrices called label.]</p><p>3) From the input image, we need to find the first pixel position with value ‘1’ and start dilation from that point. This will extract the first object.</p><p>4) Check if the object contains at least 200 pixels or not. Then replace the value from ‘1’ to ‘2’ as in labeling our object number and store this data in the first blank matrix.</p><p>5) In the original image, set all values that are already labeled to ‘0’. Then find the next starting point for our next object.</p><p>6) Repeat the iteration until the original image is replaced with ‘0’ value or until it is empty.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/a1945cfa96c583b778844702cba2faef/href">https://medium.com/media/a1945cfa96c583b778844702cba2faef/href</a></iframe><h3><strong>Filters</strong></h3><p>Noise is one of the main problem when dealing with images. Using the right filtering method to cancel out noise can help achieve a sharper image output.</p><p>Filtering can be categorized in two domains:</p><p>1) Spatial Domain</p><p>2) Frequency Domain</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/717/1*kgrN7jxGtbC5O7K4Zzj4jA.jpeg" /><figcaption>Figure 3. A process chart to compare spatial filters and frequency filters.</figcaption></figure><p>In a spatial domain, a mask, usually 3×3 or 5×5, is created and applied to the first pixel in the original image until the last pixel. Because in the spatial domain, the center value of a mask is essential, we need to implement padding algorithm around the border of the input image.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/491/1*TwxjFz-Z8-aw8ea2y8cIXw.jpeg" /><figcaption>Figure 4. Explains the equation for spatial filtering.</figcaption></figure><h3>Filtering in a Spatial Domain</h3><h4><strong>Example of Spatial Filtering with Mean and Averaging filter</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4dbF2B3VfPje_yqvYCD-qg.png" /><figcaption>Figure 5. The result of applying a Mean filter VS using the existing function in Octave.</figcaption></figure><p>The code below compares the compilation time of two filtering methods, by composing your own function and by using the existing function in Octave library.</p><p>In the DIY filter output, I used</p><pre>1 0 1<br>1 -4 1<br>0 1 0</pre><p>(mean filter) as the filtering mask.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/e11b23d55599e249395d337809e6e100/href">https://medium.com/media/e11b23d55599e249395d337809e6e100/href</a></iframe><h3><strong>Filtering in a Frequency Domain</strong></h3><p>Definition</p><ul><li>Techniques are based on modifying the spectral transform of an image</li><li>Transform the image to its frequency representation</li><li>Perform image processing</li><li>Compute inverse transform back to the spatial domain</li><li>High frequencies correspond to pixel values that change rapidly across the image (e.g. text, texture, leaves, etc.)</li><li>Strong low frequency components correspond to large scale features in the image (e.g. a single, homogenous object that dominates the image)</li></ul><p>Technique</p><ul><li>Fourier Transform</li><li>Function that are not periodic but with finite area under the curve can be expressed as he integral of sines and/ or sines multiplied by a weight function</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/374/1*RRciVTEQ5RfUwO6lpKZrfA.jpeg" /><figcaption>Figure 6. Shows the equation of Fourier Transform and Inverse Fourier Transform for filtering the image in a Frequency Domain</figcaption></figure><h4><strong>Example outputs of Low pass filtering using Gaussian and Ideal filter</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1023/1*81ZLU3c3W1H_EIiU1xOzyA.png" /><figcaption>Figure 7. The image on the left is the original image and the image on the right shows the result after applying the Lowpass filter.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/558/1*oxEY9ZmGwipvowwIkZLFtQ.png" /><figcaption>Figure 8. Shows the output of the image after applying the Ideal Lowpass filter to the original image.</figcaption></figure><h4><strong>Gaussian filtering in frequency domain:</strong></h4><p>[using fspecial() and cut off frequency= 100]</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d66713af54b2cd63d7569739fbd44b97/href">https://medium.com/media/d66713af54b2cd63d7569739fbd44b97/href</a></iframe><h4><strong>Ideal Lowpass filtering in frequency domain</strong></h4><p>[using meshgrid(), cut off frequency = 67% of maximum distance]</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/840c6d364bc27113df6319a5d1f1663f/href">https://medium.com/media/840c6d364bc27113df6319a5d1f1663f/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5aa5200aa895" width="1" height="1" alt=""><hr><p><a href="https://medium.com/data-science/new-to-image-processing-5aa5200aa895">New to Image Processing?</a> was originally published in <a href="https://medium.com/data-science">TDS Archive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Setting up Gulp (4.0.2)for Bootstrap, SASS, and BrowserSync]]></title>
            <link>https://medium.com/swlh/setting-up-gulp-4-0-2-for-bootstrap-sass-and-browsersync-7917f5f5d2c5?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/7917f5f5d2c5</guid>
            <category><![CDATA[prototype]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[scss]]></category>
            <category><![CDATA[sass]]></category>
            <category><![CDATA[gulp]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Wed, 07 Aug 2019 11:40:26 GMT</pubDate>
            <atom:updated>2019-10-26T20:14:37.248Z</atom:updated>
            <content:encoded><![CDATA[<h3>Setting up Gulp 4 for Bootstrap, SASS, and BrowserSync</h3><p>The other day I came across a project that I needed to create a quick HTML/CSS prototype to show the looks and certain animations on the prototype.</p><p>Little did I know, the set up took more time than I thought.</p><p>I want to go through a few steps for setting up gulp, certain things you should look out for and hope this helps you set it up quickly.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cKouMGEdmIrJ45zqWNGMCA.png" /></figure><p>Basically you want to be able to do something like this. Make changes on your SCSS file and gulp will automatically watch those changes and display it on the Browser.</p><p>You can clone the codes from</p><p><a href="https://github.com/amylily1011/gulp4-sass-bootstrap">https://github.com/amylily1011/gulp4-sass-bootstrap</a></p><p>Before starting.</p><ol><li>In this article, I would assumed that you have nodeJS installed. If you haven’t you can visit <a href="https://nodejs.org/en/">nodeJS</a>.</li></ol><p>What we will need for the Set up:</p><ol><li><strong>Download </strong><a href="https://getbootstrap.com/docs/4.0/getting-started/download/"><strong>Bootstrap</strong></a><strong> with npm</strong> ( this is because you want to be able to modify their default styles).</li></ol><pre>$ npm install bootstrap</pre><p><strong>2. Install </strong><a href="https://sass-lang.com/install"><strong>SaSS</strong></a><strong>. </strong>There are different ways to install SaSS and in this article we will use npm to install SaSS.</p><pre>$ npm install sass -g <br># we will want to install this globally</pre><p>3. <strong>Install Gulp CLI.</strong> This will make your life easier when you have the Gulp command line interface.</p><pre>$ npm install gulp-cli -g</pre><p>4. <strong>Install Gulp, </strong>gulp-sass, and browser-sync.</p><pre>$ npm install gulp gulp-sass browser-sync --save-dev</pre><p><strong>5. Setting up the package.</strong> This command will ask you a bunch of questions, you can just skip all of it for now.</p><pre>$ npm init</pre><h3>File Structure</h3><p>Now that the set up is ready, let’s look at the file structure.</p><p>You will need to create the css, js, and scss folders. Under the scss folder, create “styles.scss”. Also create the index.html and gulpfile.js files.</p><pre><strong>|-src<br>  |-assets<br>  |-css <br>  |-js<br>  |-scss<br>    |-styles.scss<br>  |-index.html<br>|-gulpfile.js<br>|-packages.json</strong> #this file will be generated with npm init</pre><p><strong>In ./index.html,</strong> let’s add the stylesheet in the&lt;head&gt;&lt;/head&gt; tag. We don’t need to create this stylesheet yet, because when we set up gulp, it will generate this file for us.</p><pre>&lt;link rel=&quot;stylesheet&quot; <strong>href=&quot;css/styles.css&quot;</strong>&gt;</pre><p>My current environment for this article:</p><pre>node: v11.13.0<br>npm: 6.10.1<br>gulp CLI: 2.2.0<br>gulp: 4.0.2<br>sass: 1.19.0</pre><h3>Set up gulpfile.js</h3><p>This is where the fun part begins!</p><pre><strong>const</strong> gulp = require(&#39;gulp&#39;);<br><strong>const</strong> sass = require(&#39;gulp-sass&#39;);<br><strong>const</strong> browserSync = require(&#39;browser-sync&#39;).create();</pre><p>Now let’s test if our scss file is successfully compiled.</p><p>./gulpfile.js</p><pre><strong>const</strong> gulp = require(&#39;gulp&#39;);<br><strong>const</strong> sass = require(&#39;gulp-sass&#39;);<br><strong>const</strong> browserSync = require(&#39;browser-sync&#39;).create();</pre><pre>//compile scss into css<br><strong>function</strong> style() {<br>    return <strong>gulp</strong>.src(&#39;src/scss/**/*.scss&#39;)<br>    .<strong>pipe</strong>(sass().on(&#39;error&#39;,sass.logError))<br>    .<strong>pipe</strong>(gulp.dest(&#39;src/css&#39;))<br>    .<strong>pipe</strong>(browserSync.stream());<br>}</pre><p>In the style function, you want to:</p><ol><li>Know where is my “scss”. So we will get all files ending with .scss in the src/scss</li><li>Pass the file through sass compiler, and tell me if there is an error.</li><li>Save the compiled css file (which is in src/css)</li><li>Stream changes to all browsers</li></ol><p>With Gulp 4.0.x you will need to export it after declaring the function.</p><pre><strong>exports</strong>.style = style;</pre><p>Your gulpfile.js will look like this.</p><pre><strong>const</strong> gulp = require(&#39;gulp&#39;);<br><strong>const</strong> sass = require(&#39;gulp-sass&#39;);<br><strong>const</strong> browserSync = require(&#39;browser-sync&#39;).create();</pre><pre>//compile scss into css<br><strong>function</strong> style() {<br>    return<strong> gulp</strong>.src(&#39;src/scss/**/*.scss&#39;)<br>    .<strong>pipe</strong>(sass().on(&#39;error&#39;,sass.logError))<br>    .<strong>pipe</strong>(gulp.dest(&#39;src/css&#39;))<br>    .<strong>pipe</strong>(browserSync.stream());<br>}</pre><pre><strong>exports</strong>.style = style;</pre><h4><strong>Let’s test if this works!</strong></h4><p>In your terminal run:</p><pre>$ gulp style</pre><p>Successful output looks like this.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/332/1*LWsfrh1EWcBE_oHDGhORQg.png" /><figcaption>You should see something like this.</figcaption></figure><p>Your terminal will show that style is compiled.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/208/1*Dl2LTx8WcGZ2JOOlX4kc6Q.png" /><figcaption>The styles.css file will be generated automatically.</figcaption></figure><p>When it’s done, you will check your css folder and a new file called “styles.css” will populate.</p><h4>If it doesn’t work, you might want to check you src() file and your dest() file.</h4><pre><strong>return</strong> gulp.src(&#39;./scss/**/*.scss&#39;)</pre><p>In some cases, depending on your file structure, you might need to change to the right location of the output file.</p><pre>.pipe(gulp.dest(&#39;./css&#39;))</pre><h3>Use BrowserSync to Update changes</h3><p>From here we will create a watch() function, to watch changes made in our .scss file and automatically compile that and show on the browser.</p><pre><strong>function</strong> watch() {<br>   <strong> browserSync</strong>.init({<br>        server: {<br>           baseDir: &quot;./src&quot;,<br>           index: &quot;/index.html&quot;<br>        }<br>    });<br>    <strong>gulp</strong>.watch(&#39;src/scss/**/*.scss&#39;, style)<br>    <strong>gulp</strong>.watch(&#39;./*.html&#39;).on(&#39;change&#39;,browserSync.reload);<br>    <strong>gulp</strong>.watch(&#39;./js/**/*.js&#39;).on(&#39;change&#39;, browserSync.reload);<br>}</pre><pre><strong>exports</strong>.watch = watch;</pre><p>Here, you want to tell browser sync which file to start. In many cases, just the “baseDir:” is enough, but to make it more foolproof add the index.html file to the “index:” so it calls to the right place. Also, don’t forget to export the function.</p><h4><strong>You final gulpfile.js file should look like this.</strong></h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/4f1eb72eca7cf3b937d8e46282692dab/href">https://medium.com/media/4f1eb72eca7cf3b937d8e46282692dab/href</a></iframe><p>Now let’s run this on the Terminal:</p><pre>$ gulp watch</pre><p>If you run gulp watch successfully, it will show you the information of the port that is running.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/408/1*wQZw5P6jrRs_qCpt6wWNfw.png" /><figcaption>When you run gulp watch successfully.</figcaption></figure><p>A browser with</p><p><a href="http://localhost:3000/">http://localhost:3000/</a></p><p>will pop up on a new window.</p><h3>How do I use Bootstrap in the .scss file?</h3><p>We can test if boostrap is called by simply importing from the nod module.</p><p>Our scss/styles.scss file:</p><pre>@import &quot;node_modules/bootstrap/scss/bootstrap&quot;;</pre><p>In the document, colors need to be declared first to override the other components.</p><p>So let’s change our theme primary color to pink to override all components with -primary, such as buttons to have a primary color as pink.</p><p>To test that our .scss file is compiling, let’s change our background to red and see what happens.</p><pre><strong>$theme-colors</strong>: (<br>     &quot;primary&quot;: #FF69B4<br>);</pre><pre><strong>@import</strong> &quot;node_modules/bootstrap/scss/bootstrap&quot;;</pre><pre><strong>body</strong> {<br>    color: white;<br>    background: rgb(92,18,18);<br>    height: 400vh;<br>}</pre><p>Once you saved this file. Make sure you added the button on the html file.</p><p>./index.html</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/4640a0a1a1d730cde615a2a65c44add4/href">https://medium.com/media/4640a0a1a1d730cde615a2a65c44add4/href</a></iframe><p>Now when we saved all files, gulp will automatically update our changes, compile the sass file, and show it on our browser at <a href="http://localhost:3000/">http://localhost:3000/</a>.</p><p>Voila!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/964/1*6v6U9xkZf3GCdh10gwlGsw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cKouMGEdmIrJ45zqWNGMCA.png" /></figure><p>If your result looks like this, your gulp set up is complete.</p><p>I hope this is helpful for anyone trying out gulp for the first time and might have a little struggle setting this up properly.</p><p>Happy styling your prototype!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7917f5f5d2c5" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/setting-up-gulp-4-0-2-for-bootstrap-sass-and-browsersync-7917f5f5d2c5">Setting up Gulp (4.0.2)for Bootstrap, SASS, and BrowserSync</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Making the Most out of Remote Ideation]]></title>
            <link>https://medium.com/swlh/making-the-most-out-of-remote-ideation-495a40e615ef?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/495a40e615ef</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[ux-strategy]]></category>
            <category><![CDATA[ideation]]></category>
            <category><![CDATA[remote-working]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Tue, 04 Jun 2019 22:25:13 GMT</pubDate>
            <atom:updated>2019-06-05T10:08:23.868Z</atom:updated>
            <content:encoded><![CDATA[<blockquote>“The way to get good <strong>ideas</strong> is to get lots of <strong>ideas</strong>, and throw the bad ones away.” — Linus Pauling</blockquote><p>In the process of ideation, the most important part is to be mindful, non-judgemental, and think of crazy/ out of the box ideas. As a moderator of an ideation session, it is very important to create a safe space for your peers, making sure that we encourage even stupid ideas so we can extract as much as we can. Make it a norm for our peers to not judge other.</p><p>Our goal is to generate as many ideas as we can, regardless of them being possible or not. Then filler out the bad ones later.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/786/1*dODv3JEfamPes04RKpav8Q.png" /></figure><p>What we want to achieve from an ideation session is the green asterisk. If you judge your ideas before you share them, we will go as far as the second circle. Whereas, if we are encouraged to share all types of idea even if it’s crazy, we will go as far as the outer most circle. The green asterisk is where innovation happens.</p><p>Having to facilitate an in-person interaction is already an everyday challenge, because we want to put all of our peers in the right mindset and time box our exercise to make the most out of everyone’s time. However, when working in remote offices, we cannot rely on in-person interactions. We need to rely on technology. The difference is you never know if in that one hour meeting there will be problems with the internet connection on a certain day or if someone gets kicked out of a session while posting their ideas on a virtual ideation tool or if the sound is working right.</p><p><strong>Problems with Remote Ideation:</strong></p><p><strong>Scenario 1:</strong> (6 mins ideation) tool: <a href="http://miro.com/">Realtimeboard (Miro)</a> Participants: 5</p><p>After the time starts, only 3 people were able to access the board, the other 2 had to wait to post their ideas. This creates inconsistency and by the time they meant to type the idea, the stimulation was already gone. At this point, time boxing doesn’t really help.</p><p><strong>Scenario 2:</strong> (10 mins discussion) tool: <a href="http://miro.com/">Realtimeboard (Miro)</a> Participants: 5</p><p>Because it was a virtual co-ordination, not everyone understands the big picture of our problem that we are trying to solve. Which in turn, provided us with very little solutions to work with because our peers don’t fully understand the context.</p><p><strong>Some Solutions:</strong></p><p>Having to do remote sessions with the New York office, here are some of the things that I’ve tried over the course of one year and it finally worked!</p><p><strong>PRE-GAME:</strong></p><ol><li><strong>Have someone on the other side to help you.</strong> In my current Scenario, I’m based in Berlin, but our HeadQuarter and most client facing departments are in New York. Since, I won’t be able to control anything in New York it was really nice to have someone who understands your process enough to help you facilitate on the other side.</li></ol><ul><li>Try to explain the concept of ideation and what do you want to achieve out of this session. A normal ideation session will take about 30 mins, if it’s done in-person, but in this case we have to rely on technical installations and some virtual set up so time it as a 1 hour session.</li><li>Whether that person on the other side is the Lead Engineer, the Product Manager, or a team lead, we need to establish a relationship where we truly understand the intent of this session. Let him or her know that we want our peers to generate crazy ideas and not judging other. Let them be the one to come up with some crazy ideas before hand, in case we get stuck during the ideation. Brief each other about the problems we are solving in this session.</li><li>Make sure you leave as little room as possible for curiosity, because this person needs to understand your process well enough to help you deliver the context.</li></ul><p><strong>2. Always know your problem statement. </strong>Never start an ideation session without doing enough research.</p><ul><li>You want to be able to tell your peers why are we doing this and why this is a problem. So prior to this session, make sure you have enough information about our users and clearly identify our problems. In this session we can split it into two sessions: Define + Ideate or Just Ideation. However, if you want to do Define+Ideate, make sure that the problem statement is small enough to generate many “How might we….?” questions.</li><li>More times that not, I will try to define the opportunities from our problem statements with our Product Managers before hand and pick the “How might we….” questions that aligns with our business problem the most. This will shorten the time spend on categorizing our focus problems and allow our PMs to walk our peers through certain problems in case of sound defects.</li></ul><p><strong>3. Prepare a visual User Journey or a Visualization of our Problem. </strong>So we want our peers to get up to speed with the problems we have discovered, but we want to pick their brains for certain solutions. The best way to get make the most out of this session is when everyone understands the problem the same way as us.</p><ul><li>The easiest way to make that easier on our peers is to show them what we’ve discovered in the easiest way possible. A good rule of thumb is if someone can understand it in 10 seconds, it’s good enough to communicate our problem.</li><li>Use minimum amount of words in our visualization. We can also show an empathy map that provides information on the SAY:DO , THINK:FEEL basis.</li><li>Use a timeline journey or a step-by-step Journey and add some percentage if possible.</li><li>Highlight some IMPORTANT quotes from our users. Only the important quotes.</li></ul><p><strong>IDEATION TIME:</strong></p><p>Remember, we only have one hour for this session and we should not drag on the meeting to waste our peers’ time. Take your time to prepare documents or resources you need from above and get ready for the ideation.</p><ol><li><strong>Create a safe space. </strong>Remind your peers that this is a safe space, every solution is a good solution, when one person talks all of us should be respectful and listen, and we will not judge anyone’s ideas.</li><li><strong>Think of Crazy Solutions before hand. </strong>I would think of crazy solutions before hand just to handle awkward situations when people run out of ideas and they need some guidelines to go further. My solution doesn’t have to be possible, but it has to be crazy enough that it helps our peers think outside the box.</li><li><strong>Explain your problem statements carefully. </strong>Again, we want to be on the same page and make sure we are solving the same problem. This step is just as essential as empathizing our users.</li><li><strong>Be Patience. </strong>Anything can go wrong during the ideation process, especially, when it is done remotely. Whether it’s the internet connection, the sound from the chrome box, the post-it notes on the board that you cannot read from your screen, etc. Just be patience and trust your co-ordinator on the other side.</li></ol><p><strong>Wrapping up the Ideation:</strong></p><p>The first session will always be the hardest, but simple feedbacks such as “I like” and “I wish” from the first session will give us an idea of how we can improve them next time.</p><p>After the ideation session, you should anticipate to get many useful ideas. Group them into relevant categories and vote on ones that we want to experiment with the most.</p><p>In the end, in-person interaction will always be easier. However, whether it’s in-person or remote, we want to achieve the same thing, coming up with smart solutions in a short period of time.</p><p>Of course, when remote ideation gets too hard, you can always try to do it alone. <strong>But remember, if you want to go fast, go alone. But if you want to go far, go together.</strong></p><p>Happy ideating everyone!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=495a40e615ef" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/making-the-most-out-of-remote-ideation-495a40e615ef">Making the Most out of Remote Ideation</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to make dumplings like a product designer.]]></title>
            <link>https://medium.com/swlh/how-to-make-dumplings-like-a-product-designer-a31baeaf51da?source=rss-512851396056------2</link>
            <guid isPermaLink="false">https://medium.com/p/a31baeaf51da</guid>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[food]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[user-research]]></category>
            <dc:creator><![CDATA[Amy Lily]]></dc:creator>
            <pubDate>Sun, 12 May 2019 23:20:47 GMT</pubDate>
            <atom:updated>2019-06-05T09:55:34.436Z</atom:updated>
            <content:encoded><![CDATA[<p>When someone asks me “If you can eat one menu for the rest of your life, what would that be?” One the of my first answers would be “dumplings”.</p><p>My love for dumplings was shared amongst my colleagues in the office last year, when I brought dumplings to the office on my birthday. Our Berlin office is the best place to share my dumplings because it’s a multinational office with more than 10 nationalities. So the feedback for my dumpling project would be unbiased and distributed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3gSTpqrUOUmDIfWDu6qRFw.jpeg" /><figcaption>Figure 1 Freshly made raw dumplings with pork and leeks stuffing.</figcaption></figure><p><strong>Prototype zero:</strong> Not knowing my audiences’ preference, I decided to create the first set of dumplings with the conventional “Japanese Gyoza” method with the Chinese “pork and leeks” stuffing.</p><p><strong>Ingredients:</strong></p><p><strong>Dumpling skin:</strong> flour and water (2:1)</p><p><strong>Stuffing:</strong> Pork + leek/green onions + soy sauce</p><p><strong>Method:</strong></p><ol><li>Fry in flat not stick pan with a small amount of plant-based oil for 2–3 mins.</li><li>Pour ~2 cups of water into the pan and cover the pan with a lid until the water is reduced to dry.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VLjndA7rA17ucOo-TaI0MA.jpeg" /><figcaption>Step 1: Place your dumplings in the pan with plant-based oil.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*710GX3SIH6pTP06uHzGdkg.jpeg" /><figcaption>Step 2: Pour water into the pan and close the lid for all dumplings to be fully cooked.</figcaption></figure><p>The output are as below. Dumplings looked almost brown and crisp on the bottom but soft on the outside. My favorite dumpling dipping sauce is the simple soy sauce + vinegar (2:1) with thin slices of ginger.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Mhr2L8fpj54wZtM-nkdbGA.png" /><figcaption>Figure 2 After pan fried all dumplings</figcaption></figure><p><strong>Test 0 Results:</strong></p><p>Likes</p><ul><li>Received 90% positive reviews.</li><li>The stuffing taste great!</li><li>Nice texture and tastes really well with the dip.</li><li>Dumpling skin is very thin but strong enough to whole the stuffing without breaking.</li></ul><p>Wishes</p><ul><li>A few people wish it was less soggy and more crispy</li></ul><p>Problem</p><ul><li>Temperature needs to be monitored for the dumplings because you want the stuffing to be cooked without burning the dumpling skin.</li><li>When pouring water into the pan, you need to swirl the pan to make sure dumplings don’t stick to the pan.</li></ul><p>After gathering feedback from my birthday I brought my 1st prototype of dumplings on my colleague’s up and coming birthday 3 weeks later.</p><p><strong>Prototype 1</strong>: what is different ?</p><ol><li>How might we make our dumplings less soggy and more crispy without burning them?</li></ol><ul><li>Solutions : 1. Increase frying time with oil and reduce water. 2. Dip folded dumplings in the flour before frying.</li></ul><p><strong>Results for Prototype 1:</strong></p><ul><li>Dumplings are less soggy</li><li>Positive feedback on the taste as usual.</li></ul><p><strong>Problem:</strong></p><ul><li>The current method still sticks to the pan and needs to be monitored closely.</li><li>When packing dumplings the night before, they stick to each other if it is placed too closely. Hence, some dumplings were broken in shape and the stuffing is revealed when I brought them to the office.</li></ul><p>At this point, we know that these dumplings taste great and the texture is acceptable by multicultural standards. No problem with our dip sauce. However, we need to find a way to pack our dumplings nicely without breaking its shape before it arrives to the office.</p><p><strong>Prototype 2:</strong></p><p><strong>How might we make our dumplings unstick to the pan, but soft enough on the outside?</strong></p><p><strong>Solution:</strong></p><ol><li>Add flour to the water before pouring them into the pan. (voted for this)</li><li>Add broth instead plain water. (This would not make too much of a difference)</li><li>Loose the water and only fry them in oil.( The end result will be too crispy.)</li></ol><p>How much flour needs to be added to the water?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OzRTrgI5G95onKbW-p3e9Q.jpeg" /><figcaption>Flour and water mixture to add to the frying pan.</figcaption></figure><p>I started testing prototype 2.1 with 3 teaspoons of flour and 1 cup water.</p><p><strong>Result: </strong>It adds a huge flour layer on the bottom of the dumplings and took twice as long to fry.</p><p>So I decided to adjust the portion to 1 cup water and 1 teaspoon water for prototype 2.2</p><p>As a result, our dumplings are not stuck to the bottom anymore and it adds a thin and crispy layer in the bottom just enough to feel the right texture for gyoza.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_Vuv9fjnj5-6rADttDMhFg.jpeg" /><figcaption>After the flour is bottom is cooked, it doesn’t stick to the pan and you can remove all dumplings at once.</figcaption></figure><p>Voila!</p><p>We finally had a solution for non-stick bottom dumplings!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bQgMZXBPhySbmjtm2IulVg.jpeg" /><figcaption>Dumplings prototype 2</figcaption></figure><p><strong>Summary:</strong></p><p>Dumplings are referred to as a symbol of team work. You need to create a thin enough skin that it cooks through the meat in a short time but strong enough to hold your stuffing together, which is the reason I like to bring dumplings on my colleagues birthdays. It symbolizes our strong team work.</p><p>However, as a product designer, the crucial part is the ability to receive feedback and understand your audiences preference. You always need to know which audience you are designing your products for. For this reason, different cultures have different forms of dumplings. They come with different user experience, but they are all made of similar components.</p><p>What’s your ultimate favorite menu?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a31baeaf51da" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/how-to-make-dumplings-like-a-product-designer-a31baeaf51da">How to make dumplings like a product designer.</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>