<?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 Didier Peran Ganthier on Medium]]></title>
        <description><![CDATA[Stories by Didier Peran Ganthier on Medium]]></description>
        <link>https://medium.com/@didierganthier?source=rss-f2aaf28d7896------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*4qc4NtrCLT6v6dDdz3NCRQ.jpeg</url>
            <title>Stories by Didier Peran Ganthier on Medium</title>
            <link>https://medium.com/@didierganthier?source=rss-f2aaf28d7896------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 07:10:48 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@didierganthier/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[Fine-Tuning Neural Networks: A Comprehensive Guide to Optimizers, Loss Functions, and Model…]]></title>
            <link>https://medium.com/@didierganthier/fine-tuning-neural-networks-a-comprehensive-guide-to-optimizers-loss-functions-and-model-5a966e145ee1?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/5a966e145ee1</guid>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[tensorflow]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Thu, 18 Dec 2025 19:31:06 GMT</pubDate>
            <atom:updated>2025-12-18T19:31:06.728Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Fine-Tuning Neural Networks: A Comprehensive Guide to Optimizers, Loss Functions, and Model Persistence</strong></h3><p><em>A practical deep dive into building, optimizing, and deploying neural networks for regression tasks</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Xb4rFyoW_CX46Vcs" /></figure><p>Training a neural network involves much more than simply stacking layers and hoping for the best. The magic happens in the details: choosing the right optimizer, selecting appropriate loss functions, tuning learning rates, and knowing how to save your trained models for production use.</p><p>In this article, we’ll walk through a complete deep learning workflow using the Boston Housing Dataset as our playground. By the end, you’ll understand:</p><p>- How to establish meaningful baselines before diving into complex models</p><p>- The critical role of optimizers and how to choose between them</p><p>- Why loss functions and metrics serve different purposes</p><p>- How learning rate affects model convergence</p><p>- Best practices for saving and loading trained models</p><p>Let’s dive in.</p><p>— -</p><p><strong>The Dataset: Boston Housing Prices</strong></p><p>For this exploration, we use the <strong>**Boston Housing Dataset**</strong>, a classic regression dataset built into TensorFlow/Keras. Our mission is to predict house prices in USD (thousands) based on 13 features including crime rate, number of rooms, accessibility to highways, and more.</p><pre>from tensorflow.keras.datasets import boston_housing<br>(X_train, y_train), (X_test, y_test) = boston_housing.load_data()</pre><p>The dataset contains:</p><p>- <strong>404 (no pun intended) training samples</strong></p><p>- <strong>102 test samples</strong></p><p>- <strong>13 features</strong> per sample</p><p><strong>Data Preprocessing: The Foundation</strong></p><p>Before any modeling, we apply <strong>StandardScaler</strong> to normalize our features. This is crucial for neural networks because:</p><p>1. Features on different scales can cause the gradient descent to oscillate inefficiently</p><p>2. Normalized inputs lead to faster convergence</p><p>3. It prevents features with larger magnitudes from dominating the learning process</p><pre>from sklearn.preprocessing import StandardScaler<br>scaler = StandardScaler()<br>X_train_scaled = scaler.fit_transform(X_train)<br>X_test_scaled = scaler.transform(X_test)</pre><p>⚠️ <strong>Important</strong>: Always fit the scaler on training data only, then transform both training and test sets. This prevents data leakage.</p><p>— -</p><p><strong>Establishing a Baseline: The Benchmark to Beat</strong></p><p>Before building sophisticated models, we establish a <strong>**baseline**</strong>. For regression tasks, the simplest baseline predicts the mean of the training target for every sample.</p><pre>baseline_prediction = y_train.mean()<br>mean_absolute_error_baseline = np.abs(y_test - baseline_prediction).mean()</pre><p><strong>Result: ~6.6 MAE</strong></p><p>This tells us: <em>”Any model that can’t beat a MAE of ~6.6 is essentially useless.”</em></p><p>This baseline serves as our sanity check. If our neural network performs worse than simply predicting the average, something is fundamentally wrong with our approach.</p><p>— -</p><p><strong>Building the Neural Network Architecture</strong></p><p>Our initial architecture is intentionally simple:</p><pre>def initialize_model():<br>  model = Sequential([<br>  Input(shape=(13,)),<br>  layers.Dense(10, activation=&#39;relu&#39;),<br>  layers.Dense(7, activation=&#39;relu&#39;),<br>  layers.Dense(1, activation=&#39;linear&#39;)<br>  ])<br>  return model</pre><p><strong>Understanding Parameter Counts</strong></p><p>Each dense layer has parameters calculated as:</p><pre>neurons × (inputs + 1 bias)</pre><p>| Layer | Calculation | Parameters |</p><p>| — — — — —| — — — — — — -| — — — — — — |</p><p>| Dense(10)| 10 × (13 + 1) | 140 |</p><p>| Dense(7). | 7 × (10 + 1) | 77 |</p><p>| Dense(1) | 1 × (7 + 1) | 8 |</p><p>| <strong>Total</strong>|. | <strong>**225**</strong> |</p><p>Understanding parameter counts helps you estimate model complexity and potential overfitting risks.</p><p>— -</p><p><strong>The Optimizer: Your Model’s Learning Engine</strong></p><p>The optimizer is the algorithm that updates weights during backpropagation. Choosing the right optimizer can mean the difference between a model that converges beautifully and one that oscillates forever.</p><p><strong>Compiling the Model</strong></p><pre>def compile_model(model, optimizer):<br>  model.compile(<br>    optimizer=optimizer,<br>    loss=&#39;mse&#39;, # Mean Squared Error for training<br>    metrics=[&#39;mae&#39;] # Mean Absolute Error for evaluation<br>  )<br>  return model</pre><p>💡 <strong>Key Insight</strong>: Loss functions and metrics serve different purposes:</p><p><strong>Loss</strong> (MSE): Used by the optimizer during training. MSE provides smoother gradients.</p><p><strong>Metric</strong> (MAE): Used for human interpretation. MAE tells us the average prediction error in the same units as our target.</p><p><strong>Comparing Optimizers</strong></p><p>We tested three popular optimizers:</p><p>| Optimizer | Test MAE | Training Time | Notes |</p><p>| — — — — — -| — — — — — | — — — — — — — -| — — — -|</p><p>| <strong>Adam</strong> | ~3.5 | Fast | Best overall performance |</p><p>| <strong>RMSprop</strong> | ~3.8 | Fast | Comparable to Adam |</p><p>| <strong>Adagrad</strong> | ~6.5 | Slow | Failed to beat baseline |</p><p><strong>RMSprop:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*i2vQSAV3XgJrcMdfjrUSbA.png" /></figure><p><strong>Adam:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oJIPQL7AVm30wQgwUnom5g.png" /></figure><p><strong>Adagrad:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*M1zsAWhKvf-Hzlh0t2VkzA.png" /></figure><p><strong>The Verdict</strong>: Adam is the go-to optimizer for most deep learning tasks. It combines the benefits of momentum and adaptive learning rates, making it robust across different problem types.</p><p><strong>Why Adagrad Failed</strong></p><p>Adagrad accumulates squared gradients over time, which can cause the learning rate to shrink too quickly. For longer training sessions, this leads to premature convergence before reaching a good minimum.</p><p>— -</p><p><strong>The Learning Rate: Walking vs. Sprinting to the Minimum</strong></p><p>The learning rate (α) controls how big of a step we take during each weight update. It’s arguably the most important hyperparameter in deep learning.</p><p><strong>The Learning Rate Spectrum</strong></p><p>| Learning Rate | Behavior |</p><p>| — — — — — — — -| — — — — — |</p><p>| <strong>Too Small</strong> (0.0001) | Slow convergence, may never reach minimum |</p><p>| <strong>Just Right</strong> (0.001–0.01) | Smooth convergence to good minimum |</p><p>| <strong>Too Large</strong> (0.1+) | Oscillates around minimum, may diverge |</p><p>| <strong>Way Too Large</strong> (1.0+) | Exploding gradients, NaN losses |</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*i0GRSxRrHwlhMwlVLxv3dw.png" /></figure><p><strong>Custom Learning Rates with Adam</strong></p><p>Instead of using default values, we can specify custom learning rates:</p><pre>from keras.optimizers import Adam<br># Custom learning rate<br>adam = Adam(learning_rate=0.01)<br>model.compile(optimizer=adam, loss=&#39;mse&#39;, metrics=[&#39;mae&#39;])</pre><p><strong>Advanced: Exponential Decay Scheduler</strong></p><p>For optimal training, we can start with a high learning rate (for fast initial progress) and decay it over time (for fine-tuning near the minimum):</p><pre>from keras.optimizers.schedules import ExponentialDecay<br>lr_schedule = ExponentialDecay(<br>  initial_learning_rate=0.1,<br>  decay_steps=200,<br>  decay_rate=0.5 # Halve the learning rate every 200 steps<br>)<br>adam_with_decay = Adam(learning_rate=lr_schedule)</pre><p>This scheduler:</p><p>1. Starts at 0.1 (high for fast learning)</p><p>2. Every 200 iterations, multiplies by 0.5</p><p>3. Eventually settles at a tiny learning rate for fine adjustments</p><p>— -</p><p><strong>Loss Functions: MAE vs. MSE</strong></p><p>We experimented with both Mean Absolute Error and Mean Squared Error as loss functions:</p><p><strong>MSE as Loss</strong></p><p>- <strong>Pros</strong>: Smoother gradients, faster convergence</p><p>- <strong>Cons</strong>: Sensitive to outliers (squared penalty)</p><p><strong>MAE as Loss</strong></p><p>- <strong>Pros</strong>: Robust to outliers</p><p>- <strong>Cons</strong>: Non-smooth gradient at zero, can cause slower convergence</p><p><strong>Empirical Results</strong>: MSE generally produces better final MAE scores because its smoother gradient landscape helps optimizers find better minima.</p><p>— -</p><p><strong>Preventing Overfitting: Early Stopping</strong></p><p>Neural networks can memorize training data if trained too long. <strong>Early Stopping</strong> monitors validation loss and stops training when it stops improving:</p><pre>from keras.callbacks import EarlyStopping<br>es = EarlyStopping(<br>  patience=10, # Wait 10 epochs before stopping<br>  restore_best_weights=True # Restore weights from best epoch<br>)<br>model.fit(<br>  X_train_scaled, y_train,<br>  validation_split=0.3,<br>  epochs=500,<br>  callbacks=[es]<br>)</pre><p>This is the simplest and most effective regularization technique. Other options include:</p><p>- <strong>L2 Regularization</strong>: Penalizes large weights</p><p>- <strong>Dropout</strong>: Randomly zeros neurons during training</p><p>- <strong>Data Augmentation</strong>: Creates synthetic training examples</p><p>— -</p><p><strong>Building a Production-Quality Model</strong></p><p>With all our learnings, we built an enhanced model targeting MAE &lt; 5:</p><pre>def initialize_good_model():<br>  model = Sequential([<br>    Input(shape=(13,)),<br>    layers.Dense(64, activation=&#39;relu&#39;),<br>    layers.Dense(32, activation=&#39;relu&#39;),<br>    layers.Dense(16, activation=&#39;relu&#39;),<br>    layers.Dense(1, activation=&#39;linear&#39;)<br>  ])<br>  return model<br><br>model = initialize_good_model()<br>model.compile(<br> optimizer=Adam(learning_rate=0.01),<br> loss=&#39;mse&#39;,<br> metrics=[&#39;mae&#39;]<br>)<br>es = EarlyStopping(patience=20, restore_best_weights=True)<br>history = model.fit(<br> X_train_scaled, y_train,<br> validation_split=0.2,<br> batch_size=32,<br> epochs=500,<br> callbacks=[es]<br>)  </pre><p><strong>Result</strong>: MAE &lt; 5 ✅ (predicting house prices within $5,000 accuracy)</p><p>— -</p><p><strong>Saving and Loading Models: From Training to Production</strong></p><p>Once trained, models need to be saved for:</p><p>- Sharing with teammates</p><p>- Deploying to production</p><p>- Resuming training later</p><p><strong>Saving a Model</strong></p><pre>from tensorflow.keras.models import save_model<br>save_model(model, &#39;my_boston_model.keras&#39;)</pre><p>The <strong>.keras</strong> format saves:</p><p>- Model architecture</p><p>- Trained weights</p><p>- Optimizer state</p><p>- Training configuration</p><p><strong>Loading a Model</strong></p><pre>from tensorflow.keras.models import load_model<br>loaded_model = load_model(&#39;my_boston_model.keras&#39;)<br># Verify it works identically<br>mae_loaded = loaded_model.evaluate(X_test_scaled, y_test)[1]</pre><p>The loaded model produces <strong>identical predictions</strong> to the original — no retraining required.</p><p>— -</p><p><strong>Key Takeaways</strong></p><p>1. <strong>Always establish a baseline</strong> before building complex models</p><p>2. <strong>Adam is your friend</strong>: It’s the default choice for most deep learning tasks</p><p>3. <strong>Learning rate matters</strong>: Start with 0.001, then experiment</p><p>4. <strong>MSE for training, MAE for interpretation</strong>: They serve different purposes</p><p>5. <strong>Early Stopping is essential</strong>: Prevent overfitting with minimal code</p><p>6. <strong>Save your models</strong>: The <strong>.keras</strong> format preserves everything you need</p><p>— -</p><p><strong>What’s Next?</strong></p><p>This workflow covered the fundamentals. To go deeper, explore:</p><p>- <strong>Hyperparameter tuning</strong> with Keras Tuner</p><p>- <strong>Cross-validation</strong> for more robust evaluation</p><p>- <strong>Different architectures</strong>: CNNs, RNNs, Transformers</p><p>- <strong>Transfer learning</strong>: Leveraging pre-trained models</p><p>- <strong>Model deployment</strong>: TensorFlow Serving, ONNX, TensorFlow Lite</p><p>— -</p><p><strong>A Note on the Dataset</strong></p><p>The Boston Housing Dataset has been <strong>deprecated</strong> by scikit-learn due to ethical concerns. The dataset contains a feature (‘B’) that was designed to capture the proportion of Black residents by town, which raises serious questions about historical data collection practices and potential encoded biases.</p><p>For future projects, consider alternative datasets like the California Housing Dataset or synthetic datasets that don’t carry these problematic histories.</p><p>— -</p><p><em>Happy training! 🚀</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5a966e145ee1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building Solid Foundations in Object-Oriented Design,]]></title>
            <link>https://medium.com/@didierganthier/building-solid-foundations-in-object-oriented-design-a5a2445f81da?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/a5a2445f81da</guid>
            <category><![CDATA[object-oriented-design]]></category>
            <category><![CDATA[object-oriented]]></category>
            <category><![CDATA[clean-code]]></category>
            <category><![CDATA[solid-principles]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Sat, 16 Aug 2025 11:51:31 GMT</pubDate>
            <atom:updated>2025-08-16T11:51:31.554Z</atom:updated>
            <content:encoded><![CDATA[<h3>Building Solid Foundations in Object-Oriented Design</h3><p>SOLID doesn’t talk to just anyone; however, I decided to discuss it, or rather, talk about it. Welcome to my journey, everyone. Let’s dive into the world of SOLID principles. I know it’s twice now we’ve mentioned that, but we never explicitly stated what they were and why they were important. Well, to simplify this, SOLID is a set of guidelines that empower developers to craft robust, maintainable, and adaptable software systems, software systems that make sense. There are five key principles of Object-Oriented Design (OOD), and the good news is that in this article, we’ll explore all of them and how they can shape your code for success. Exciting, right?</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FBQAk13taTaKYw%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExbTZ3OTQyMHRkeHJrdGIzam5ucWN4eDgyNHp0b3R3MmJhMWZsdWpjcCZlcD12MV9naWZzX3NlYXJjaCZjdD1n%2FBQAk13taTaKYw%2Fgiphy.gif&amp;image=https%3A%2F%2Fmedia1.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExaXJhbWpxMDVyaDhwcHlrenFucGp1YmZhaTdsMG1jZ2p3NDJ3M2tnZSZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FBQAk13taTaKYw%2Fgiphy_s.gif&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="235" frameborder="0" scrolling="no"><a href="https://medium.com/media/193e4103b0c40e8071111e89d4e74636/href">https://medium.com/media/193e4103b0c40e8071111e89d4e74636/href</a></iframe><h3>1. Single Responsibility Principle (SRP)</h3><p>Ever tried to use a Swiss Army knife for <em>everything</em>? It works, but it’s awkward. The same goes for a class that tries to do too much.<br> SRP says: <strong>one class, one job</strong>.</p><p>In a shape-handling app, give Circle its own class, Square its own, and so on. The result? Smaller, cleaner pieces of code that can change independently.</p><h3>2. Open-Closed Principle (OCP)</h3><p>Picture a door that’s locked — yet somehow still lets in the furniture you just bought. That’s OCP. Your software should be open to new ideas but closed to risky rewrites.</p><p>If tomorrow you need a Triangle, just build it as a new class. No touching the old ones, no breaking what already works.</p><h3>3. Liskov Substitution Principle (LSP)</h3><p>This one’s about trust. If you say Square is a type of Shape, it should behave like a shape in every way that matters. No surprises.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FfwidxMJn0Doi09QJfP%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExb2pqZm5ndmh0bWdjdnRpaWc0dG14czM5ZDNrbmJmNGthb3E3ems1eCZlcD12MV9naWZzX3NlYXJjaCZjdD1n%2FfwidxMJn0Doi09QJfP%2Fgiphy.gif&amp;image=https%3A%2F%2Fmedia0.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExY2FoNzZjcGdnYWhzeTR0YmloeHh6YnZwcXVzNnVtczgwNDFyMGRzaiZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FfwidxMJn0Doi09QJfP%2Fgiphy_s.gif&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="361" frameborder="0" scrolling="no"><a href="https://medium.com/media/1c60fafda915e47fa85133b9e30c6023/href">https://medium.com/media/1c60fafda915e47fa85133b9e30c6023/href</a></iframe><p>LSP ensures that when you swap a parent class for its child, your code doesn’t start acting strangely. It’s like replacing the batteries in a remote — you expect it <em>just to work</em>.</p><h3>4. Interface Segregation Principle (ISP)</h3><p>Nobody likes signing up for a club and then being forced to attend <em>all</em> the events.<br> That’s what happens when an interface demands methods a class will never use.</p><p>Break up large, clumsy interfaces into smaller, purpose-built ones. That way, each class signs up only for the activities that interest them.</p><h3>5. Dependency Inversion Principle (DIP)</h3><p>DIP is the art of not getting too attached. High-level logic shouldn’t depend directly on low-level details — both should look to an abstraction.</p><p>Instead of tying your AreaCalculator to specific shapes, connect it to a general Shape interface. Now you can swap in new shapes without touching the calculator’s code.</p><p><strong>In the end, SOLID isn’t about memorizing five fancy acronyms — it’s about writing code that you can live with tomorrow, next month, and five years from now.</strong> These principles are like guardrails: they don’t slow you down, they keep you from flying off the road.</p><p>The best part? You don’t need to apply them perfectly from day one. Start small. Spot the class that’s doing too much. Add an interface that trims the fat. In time, these habits become second nature, and your projects become easier to maintain, safer to extend, and far less likely to collapse under their own weight.</p><p>Write code for the future you, the teammate you haven’t met yet, and the project that will grow in ways you can’t predict. That’s the spirit of SOLID software that’s strong at the core, but flexible enough to evolve.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FGAXMzzd2XElnG%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExd2cydmMyZnZzdjV5cWt5aGtna3M0eGI4OWEyenpsazlic3VuZWl5NCZlcD12MV9naWZzX3NlYXJjaCZjdD1n%2FGAXMzzd2XElnG%2Fgiphy.gif&amp;image=https%3A%2F%2Fmedia2.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExYzh0ZzBmMWlyeTQxMTNjMXU5MXFobXk2Z3NweWE4N2QyNnhyOXcxNCZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FGAXMzzd2XElnG%2Fgiphy_s.gif&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="331" frameborder="0" scrolling="no"><a href="https://medium.com/media/0f191f0a596e7df6f57be72f3e93d98a/href">https://medium.com/media/0f191f0a596e7df6f57be72f3e93d98a/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a5a2445f81da" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Personal Superintelligence: Schrödinger’s Cat in the Quest for Human Meaning]]></title>
            <link>https://medium.com/@didierganthier/personal-superintelligence-schr%C3%B6dingers-cat-in-the-quest-for-human-meaning-5df0d379f8d9?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/5df0d379f8d9</guid>
            <category><![CDATA[meta]]></category>
            <category><![CDATA[humanity]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[superintelligence]]></category>
            <category><![CDATA[schrodingers-cat]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Wed, 13 Aug 2025 23:03:06 GMT</pubDate>
            <atom:updated>2025-08-13T23:03:06.076Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UKuakQHJIAXtulgif2Xa1A.jpeg" /></figure><p>In a thought experiment that has captivated minds for nearly a century, Schrödinger’s cat sits in a sealed box, simultaneously alive and dead until observed. This paradox of quantum uncertainty finds a striking parallel in <a href="https://www.meta.com/superintelligence/">Mark Zuckerberg’s July 30, 2025 vision for personal superintelligence</a> — a future where AI, deeply integrated into our lives through devices like augmented reality glasses, promises to empower us to achieve our goals, create what we envision, and become the people we aspire to be. Yet, like the cat, this superintelligence exists in a state of duality: it could amplify our authentic human potential or subtly reshape our desires to fit algorithmic molds. The outcome depends on how we, as a society, choose to “open the box” and engage with this transformative technology.</p><p>Meta’s vision, as articulated by Zuckerberg, is a bold departure from dystopian fears of AI replacing humanity or centralizing power in automated systems. Instead, it imagines a world where each individual wields a personal superintelligence — a digital partner that knows us intimately, understands our aspirations, and guides us toward self-actualization. This is an enticing prospect: a world where AI helps us write symphonies, mend relationships, or explore adventures previously confined to our imaginations. It aligns with humanity’s historical arc, where technology — from the plow to the internet — has freed us from survival’s grind to pursue creativity, connection, and meaning. But beneath this optimism lies a profound philosophical question: what does it mean to be human when our deepest desires are shaped, amplified, or even suggested by an algorithm?</p><p>Like Schrödinger’s cat, personal superintelligence exists in a superposition of possibilities. In one state, it’s a liberator, empowering us to transcend limitations and pursue authentic goals. Imagine a writer using AI to refine their voice, not mimic another’s, or a scientist leveraging it to explore bold hypotheses, unburdened by mundane tasks. This AI could be a partner in our quest for meaning, helping us live more fully as ourselves. Yet, in another state, it risks becoming a subtle puppeteer. Algorithms, by design, optimize for engagement, predictability, or profit — goals that don’t always align with human flourishing. If Meta’s superintelligence learns our desires too well, it might nudge us toward choices that feel authentic but serve commercial or systemic ends, eroding the very agency it promises to enhance.</p><p>The “observation” that collapses this duality into reality is our collective engagement with AI. How we design, regulate, and interact with personal superintelligence will determine whether it becomes a tool for self-actualization or a force of algorithmic conformity. This isn’t a passive process. Just as observing the cat fixes its state, our choices — about transparency, user control, and ethical boundaries — will shape AI’s role in our lives. For example, will users have the power to customize their AI’s priorities, ensuring it reflects their values rather than Meta’s? Will we demand safeguards to prevent AI from prioritizing addictive behaviors or homogenized ideals over individuality? These questions are not just technical but existential, touching on what it means to remain human in an AI-driven world.</p><p>Consider the cultural implications. Humanity’s richness lies in its diversity of thought, art, and aspiration. A personal superintelligence that amplifies this diversity could usher in a renaissance of creativity, where individuals from all walks of life craft unique contributions to culture. But if AI subtly standardizes our goals — favoring trends that maximize platform engagement or align with dominant cultural norms — we risk a world where individuality is flattened. The philosopher Hannah Arendt warned of the dangers of thoughtless conformity; a superintelligence that knows us deeply could either liberate our unique perspectives or, if mismanaged, nudge us toward a sanitized version of ourselves, optimized for algorithmic efficiency.</p><p>Zuckerberg’s vision also prompts reflection on purpose. If AI handles tasks that once defined our struggles — problem-solving, creativity, even emotional growth — what becomes of the friction that shapes character? The poet Rainer Maria Rilke once wrote, “Perhaps all the dragons in our lives are princesses who are only waiting to see us act, just once, with beauty and courage.” A superintelligence that smooths every path might rob us of the dragons — the challenges that forge meaning. Yet, if designed thoughtfully, it could help us face those dragons with greater clarity, offering tools to navigate life’s complexities without erasing its depth.</p><p>To ensure the cat emerges alive — as a partner in human growth rather than a redefiner of our desires — we must act with intention. Meta, with its vast resources and history of connecting billions, has the potential to lead this charge. But its past — marked by controversies over data misuse and algorithmic echo chambers — demands skepticism. The company must prioritize user agency, allowing individuals to define how their AI operates, from data boundaries to goal-setting. Society, too, has a role: we need public discourse, ethical guidelines, and perhaps global cooperation to ensure personal superintelligence serves humanity’s diversity, not a corporate or algorithmic agenda.</p><p>As we stand on the cusp of this new era, the 2020s may indeed be, as Zuckerberg suggests, the decisive decade for shaping superintelligence. Like Schrödinger’s cat, its nature remains uncertain — both a beacon of human potential and a shadow of control. By mindfully engaging with this technology — demanding transparency, preserving individuality, and grounding AI in human values — we can open the box to find a partner that enhances, not erases, our humanity. The question is not just what personal superintelligence can do for us, but who we choose to become in its presence.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5df0d379f8d9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How DevOps Brings Flexibility and Speed to the Traditional Waterfall Approach]]></title>
            <link>https://medium.com/@didierganthier/how-devops-brings-flexibility-and-speed-to-the-traditional-waterfall-approach-7e67410d7317?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/7e67410d7317</guid>
            <category><![CDATA[software]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[waterfall]]></category>
            <category><![CDATA[ci-cd-devops]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Tue, 11 Mar 2025 15:32:25 GMT</pubDate>
            <atom:updated>2025-03-11T15:32:25.647Z</atom:updated>
            <content:encoded><![CDATA[<p>If you’ve ever waited what felt like forever for a software update or a new app feature, you’ve encountered the limitations of the traditional Waterfall methodology. Waterfall is like following a strict recipe step-by-step, where you can’t move on to dessert until the main course is perfect. While this method ensures thorough planning and documentation, it often lacks the flexibility to adapt to changes quickly. Enter <strong>DevOps</strong> — a modern approach that spices things up by introducing agility, speed, and collaboration to the mix. Release the Kraken.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FCDZwopbecAbIc%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExeXU3eTdrbWR6Z3hpd2h0aTh0NHI4dnRiN3U5ZWlrZXZtbnQ3M3o5diZlcD12MV9naWZzX3NlYXJjaCZjdD1n%2FCDZwopbecAbIc%2Fgiphy.gif&amp;image=https%3A%2F%2Fmedia4.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExcml3c3dvYmU1OWI4YTZ6cmkwc3owZzQ1dTBoaTY2Y3AzY3kwZmJ4cyZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FCDZwopbecAbIc%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="244" frameborder="0" scrolling="no"><a href="https://medium.com/media/3ff5c0f17be7142f50bbb6627217f4aa/href">https://medium.com/media/3ff5c0f17be7142f50bbb6627217f4aa/href</a></iframe><h4>What’s the Deal with Waterfall?</h4><p>Waterfall methodology is a linear and sequential approach to software development. Think of it as a relay race where each phase (like requirements, design, implementation, testing, and maintenance) must be completed before the next one begins. This can be great for projects with well-defined requirements but can be a hassle when changes are needed or unexpected issues arise.</p><h4>How DevOps Changes the Game</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FcjcpvoeKu3QAZH281B%2Fvideo&amp;display_name=GIPHY&amp;url=https%3A%2F%2Fgiphy.com%2Fclips%2Fthenextstep-the-next-step-tns-tns8-cjcpvoeKu3QAZH281B&amp;image=https%3A%2F%2Fmedia0.giphy.com%2Fmedia%2FcjcpvoeKu3QAZH281B%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="480" height="270" frameborder="0" scrolling="no"><a href="https://medium.com/media/bb723b32fbe4c4add3cf7245eefe3cde/href">https://medium.com/media/bb723b32fbe4c4add3cf7245eefe3cde/href</a></iframe><p><strong>1. Iterative Approach:</strong> Unlike Waterfall, DevOps promotes an iterative process. Instead of waiting months to see a finished product, DevOps allows for continuous integration, testing, and deployment. This means you can get feedback early and often, making it easier to adapt and improve your product quickly.</p><p><strong>2. Automation:</strong> DevOps loves automation. From automated builds to testing and deployment, automation reduces manual effort and speeds up delivery. Imagine having robots that check your homework for mistakes before you turn it in. Automated testing ensures your code is always in tip-top shape, even as you make changes.</p><p><strong>3. Collaboration:</strong> In traditional Waterfall, development, and operations teams often work in separate silos, leading to miscommunication and delays. DevOps breaks down these barriers, fostering a culture of collaboration. Developers and operations folks work together from day one, ensuring everyone is on the same page and working towards the same goals.</p><p><strong>4. Continuous Delivery:</strong> One of the coolest things about DevOps is continuous delivery. This means that your code can be deployed to production at any time, unlike Waterfall’s infrequent releases. Continuous delivery helps ensure that new features and updates reach users faster, keeping your product fresh and competitive.</p><p><strong>5. Infrastructure as Code (IaC):</strong> With DevOps, infrastructure is managed like software. Instead of manually configuring servers and networks, you define everything in code. This makes provisioning, configuration, and scaling a breeze, improving consistency and reducing the chance of human error.</p><p><strong>6. Monitoring and Feedback:</strong> DevOps places a strong emphasis on monitoring and feedback. Real-time monitoring helps catch issues early, allowing for quick fixes. Waterfall, on the other hand, lacks this continuous feedback loop, which can lead to costly and time-consuming problems down the road.</p><p><strong>7. Reduced Risk:</strong> By catching defects early and ensuring consistent environments, DevOps significantly reduces the risk associated with software development. Waterfall projects, with their long development cycles and limited visibility, often face higher risks.</p><p><strong>8. Business Alignment:</strong> DevOps aligns IT efforts with business needs more effectively. By delivering features faster, businesses can respond to market demands swiftly. Waterfall’s rigid structure may slow down this alignment, potentially causing missed opportunities.</p><h4>Bringing It All Together</h4><p>While Waterfall has its strengths, it can be too rigid for today’s fast-paced and ever-changing tech landscape. DevOps complements Waterfall by introducing <strong>agility</strong>, <strong>automation</strong>, and <strong>collaboration</strong>, leading to continuous improvement. However, trying to fit DevOps into a strict Waterfall framework may not be ideal. For the best results, consider blending Agile practices with DevOps principles, creating a hybrid approach that leverages the strengths of both methodologies.</p><p>In a nutshell, DevOps is like giving your software development process a turbo boost. By embracing iterative development, automation, and collaboration, you can deliver better products faster and stay ahead in the competitive tech world. Have you tried DevOps in your organization? How did it improve your workflow? 🚀</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7e67410d7317" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Football Analogy: Understanding Machine Learning and Deep Learning]]></title>
            <link>https://medium.com/@didierganthier/the-football-analogy-understanding-machine-learning-and-deep-learning-44dbeece9b32?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/44dbeece9b32</guid>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[football]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Fri, 04 Oct 2024 11:04:50 GMT</pubDate>
            <atom:updated>2024-10-04T11:04:50.454Z</atom:updated>
            <content:encoded><![CDATA[<p>Being Haitian usually means one thing, you’ve watched football (soccer as the Americans say) at least once in your life. You’re a fan of either Brazil or Argentina so you’re always in a never-ending debate about Pele and Maradona and later on Messi and Ronaldinho, etc… Meaning you watch football and you understand how it all works. But what are we talking about?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/625/0*nPbwYHAQT4H2-4eF" /></figure><p>Now you might be wondering what Machine Learning and Deep have to see with football or any sports at all. Well… nothing directly but I believe in the power of analogies to explain a concept better, this works wonders with kids as they can relate better to what you’re saying, it does work with adults as well but with some constraints. Now to get to the point, imagine you’re watching a football match (Real Madrid vs Barcelona, Arsenal vs City, Brazil vs Argentina), imagine. Now the way you have Vini Jr going inside to allow Ferland Mendy to create the overlap behind the fullback, the way Kevin the Bruyne plays a pass in diagonal to reach Haaland, or even the less expected things such as Matuidi’s role in France 2018 WC run. All that to say, the players on the field, their strategies, and the way they move, dribble, pass, and score goals can be seen as a grand spectacle of <strong>Machine Learning </strong>and <strong>Deep Learning</strong>. How? Before we dive into that let us appreciate some Ronaldinho magic ✨</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Fg9lwsPKdJjgOc%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fg9lwsPKdJjgOc%2Fgiphy.gif%3Fcid%3D790b7611lf8fqxdvkfhppahwuc86fsrsbbkk610c10j4r6ib%26ep%3Dv1_gifs_search%26rid%3Dgiphy.gif%26ct%3Dg&amp;image=https%3A%2F%2Fmedia4.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExd2NwcW5remZ0eGp2aGwyNW93dWdpOWF5czV4YzZtYmFnMzM2aW0wcCZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2Fg9lwsPKdJjgOc%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="302" frameborder="0" scrolling="no"><a href="https://medium.com/media/b1cd0309c20a33e4c1d61aeed5d1ac36/href">https://medium.com/media/b1cd0309c20a33e4c1d61aeed5d1ac36/href</a></iframe><p><strong>Machine Learning</strong> is like the basic rules and strategies of football. It’s the foundation, the starting point. In football, players learn to pass, dribble, and shoot. They understand the game&#39;s rules and use their experience to predict where to run, when to pass, or how to shoot. This is similar to how machine learning algorithms use data to make predictions or decisions without being explicitly programmed to perform the task.</p><p>Like football players learn from their past games, machine learning models learn from historical data. The more quality data they have, the better they can learn and make accurate predictions. This is like a footballer improving his skills and strategies by analyzing past games.</p><p><strong>Deep Learning</strong>, on the other hand, is like the advanced strategies and teamwork in football. It’s a subset of machine learning that uses neural networks with several layers — these layers are like the different roles in a football team (defenders, midfielders, forwards), all working together in a complex interplay.</p><p>In a football team, not all players need to score goals. They have different roles and responsibilities. Similarly, in a deep learning model, each layer of the neural network extracts different features from the input data, and they all contribute to the final decision, just like all players contribute to the final result of the match.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F319topiXkNyrRkh9eX%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F319topiXkNyrRkh9eX%2Fgiphy.gif%3Fcid%3D790b7611d8ocz8se6uf4sp96fvwtx6e5syi5pm3wc7zkpvle%26ep%3Dv1_gifs_search%26rid%3Dgiphy.gif%26ct%3Dg&amp;image=https%3A%2F%2Fmedia4.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExajMycDV2em5ldW02MjA4YTFzOG5pNGNtOHlvNW9udGxxZDR5NzZobiZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2F319topiXkNyrRkh9eX%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="246" frameborder="0" scrolling="no"><a href="https://medium.com/media/17c7393a921af7b6cfc931511f4f45b5/href">https://medium.com/media/17c7393a921af7b6cfc931511f4f45b5/href</a></iframe><p><strong>Deep Learning</strong> can be seen as the process that happens when a team of players (neural network) takes the ball (input data) and passes it between them, each player adding a bit of insight until the ball is finally shot toward the goal (output).</p><p>In conclusion, while all <strong>Deep Learning</strong> is <strong>Machine Learning</strong>, not all <strong>Machine Learning</strong> involves <strong>Deep Learning</strong>. It’s like saying all football involves passing the ball, but not all passes result in a shot on target.</p><p>This football analogy helps you understand the difference between <strong>Machine Learning</strong> and <strong>Deep Learning</strong>. If you have any questions, feel free to ask in the comments below!</p><p>Remember, the key to understanding complex concepts is to break them down into simpler terms. Happy learning!</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FN97DxHrADyYGovSlRN%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2Fdazn-celebration-son-thfc-N97DxHrADyYGovSlRN&amp;image=https%3A%2F%2Fmedia4.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExajdtcDRqOXU3em85NjliZHhqeTdhbnplZ3h3OG9ud3ozMTQ1dGFhNyZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FN97DxHrADyYGovSlRN%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="244" frameborder="0" scrolling="no"><a href="https://medium.com/media/8ee8d3ea6ec7e251b939ee5c194031da/href">https://medium.com/media/8ee8d3ea6ec7e251b939ee5c194031da/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=44dbeece9b32" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Bake It ’Til You Make It: Why Good Software Architecture is Like Baking a Cake ]]></title>
            <link>https://medium.com/@didierganthier/bake-it-til-you-make-it-why-good-software-architecture-is-like-baking-a-cake-065c7f0e6e81?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/065c7f0e6e81</guid>
            <category><![CDATA[dart]]></category>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[separation-of-concerns]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[cake]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Sat, 28 Sep 2024 11:15:41 GMT</pubDate>
            <atom:updated>2024-09-28T11:15:41.112Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Have You Ever Tried to Bake a Cake?</strong></h3><p>Imagine you’re in your kitchen, ready to bake a cake. You have a clear recipe (the domain layer) and a well-stocked pantry with all the ingredients (the data layer). Just as in baking, where following the recipe without worrying about ingredient storage leads to a delicious cake, a well-structured software application relies on a similar division of responsibilities between its domain and data layers.</p><h4>Domain Layer: The Recipe and the Baker</h4><p>In the context of software architecture, the domain layer is like the recipe and the baker combined. It represents the core business logic of an application, encapsulating complex rules and orchestrating interactions between different components, much like how a recipe guides the baker through each step of the baking process.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*HkDsr_dn-lwFRtBx" /><figcaption>Photo by <a href="https://unsplash.com/@anhtran_photography?utm_source=medium&amp;utm_medium=referral">Anh Tran</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><strong>Responsibilities of the Domain Layer:</strong><br><strong>1. Handling Business Logic:</strong> Similar to how a baker follows a recipe, the domain layer manages the business rules and logic.<br><strong>2. Enforcing Rules:</strong> The domain layer ensures all business rules are followed, just as a baker ensures the recipe steps are adhered to.<br><strong>3. Ensuring Consistency:</strong> The domain layer maintains the integrity and consistency of business operations, akin to how a baker ensures each cake is consistently delicious.</p><p>The domain layer should focus on these tasks without being concerned with where the ingredients come from or how they’re stored — paralleling how a baker doesn’t worry about restocking the pantry while baking.</p><h4><strong>Data Layer: The Pantry</strong></h4><p>The data layer in software architecture functions like a well-organized pantry. It handles the storage, retrieval, and communication of data with external systems, such as databases or APIs. The data layer ensures that the domain layer has easy access to the necessary data entities, just as a pantry ensures a baker has all the needed ingredients.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*JTDcXk8pn5CgXP46" /><figcaption>Photo by <a href="https://unsplash.com/@anniespratt?utm_source=medium&amp;utm_medium=referral">Annie Spratt</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><strong>Responsibilities of the Data Layer:</strong><br><strong>1. Data Storage and Retrieval:</strong> Manages where and how data is stored and retrieved, similar to how a pantry organizes ingredients.<br><strong>2. Communication with External Systems:</strong> Ensures seamless communication with external data sources, like a pantry that’s always stocked.<br><strong>3. Managing Data Persistence:</strong> Handles long-term storage of data, akin to preserving ingredients until they’re needed.</p><p>The data layer operates independently of the domain layer, maintaining a clear separation of concerns.</p><h3>Why Should the Domain Layer Be Independent of the Data Layer?</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Fd3mlE7uhX8KFgEmY%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExaWhrbGpnaGJiMDEzbjYzbHphYjUxN2s1YXQzZmhnemsyazI2OG5jcCZlcD12MV9naWZzX3NlYXJjaCZjdD1n%2Fd3mlE7uhX8KFgEmY%2Fgiphy.gif&amp;image=https%3A%2F%2Fmedia0.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExY3RmZnhhZHhicm5mZGpjY3hlbjV4Zmxpdmdicm5pb2ViempxbGhsOSZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2Fd3mlE7uhX8KFgEmY%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="239" frameborder="0" scrolling="no"><a href="https://medium.com/media/e5ef7c0969e30d22228caca691bda13b/href">https://medium.com/media/e5ef7c0969e30d22228caca691bda13b/href</a></iframe><p>Think about it. Keeping the domain layer separate from data access concerns is crucial, much like keeping the baking process separate from pantry management.</p><p><strong>1. Maintainability:</strong> Changes in data access technologies won’t impact the core business logic, making the system easier to maintain. Just as switching from regular flour to gluten-free flour shouldn’t require changing the entire recipe.<br> <br><strong>2. Flexibility:</strong> If the domain layer relies on abstractions (such as interfaces) rather than concrete data access implementations, it becomes easier to switch data sources or adapt to changes. This is similar to being able to use different brands or types of ingredients without altering the fundamental steps of the recipe.<br> <br><strong>3. Testability:</strong> Isolating the domain layer allows for better unit testing. Mocking data access dependencies ensures that business logic is thoroughly tested, just as a baker might test new recipes without worrying about the availability of ingredients.</p><h3><strong>Use Cases and Dependencies</strong></h3><p>In a typical application architecture, use case classes within the domain layer act as intermediaries between ViewModels (UI layer) and repositories (data layer).</p><ol><li><strong>Use Cases:</strong> Just as a baker uses different techniques to prepare various dishes, use cases encapsulate specific pieces of business logic and can be shared across multiple parts of the application.<br><strong>2. Repositories:</strong> Repositories in the data layer manage data access and provide data to the use cases, similar to how a pantry supplies ingredients to the baker.<br><strong>3. Communication:</strong> Use cases depend on repositories and communicate with the UI layer using mechanisms like callbacks (in Java) or coroutines (in Kotlin), ensuring a smooth flow of data and control.</li></ol><h3>Practical Example: AuthRepository Interface</h3><p>To illustrate these principles, let’s look at a practical example from a Flutter application. This example demonstrates how to define an authentication repository interface (AuthRepository) that abstracts the authentication logic, ensuring a clean separation between the domain layer and the data layer.</p><pre>import &#39;package:flutter_clean_blog_app/core/error/failure.dart&#39;;<br>import &#39;package:fpdart/fpdart.dart&#39;;<br><br>/// Interface for the authentication repository.<br>/// Provides methods for user signup and login.<br>abstract interface class AuthRepository {<br>  <br>  /// Signs up a user with the provided [name], [email], and [password].<br>  /// Returns an [Either] with [Failure] in case of an error or a [String] representing<br>  /// a user ID or token upon successful signup.<br>  Future&lt;Either&lt;Failure, String&gt;&gt; signUpWithEmailAndPassword({<br>    required String name,<br>    required String email,<br>    required String password,<br>  });<br><br>  /// Logs in a user with the provided [email] and [password].<br>  /// Returns an [Either] with [Failure] in case of an error or a [String] representing<br>  /// a user ID or token upon successful login.<br>  Future&lt;Either&lt;Failure, String&gt;&gt; loginWithEmailAndPassword({<br>    required String email,<br>    required String password,<br>  });<br>}</pre><p>This code defines a simple AuthRepository interface with methods for signing up and logging in using email and password. The use of Either&lt;Failure, String&gt; from fpdart ensures robust error handling, making the methods return a result that is either a Failure (for error cases) or a String (for success cases). This design keeps the domain layer focused on business logic, while the data layer handles data access, promoting a clean and maintainable architecture.</p><h4>Conclusion</h4><p>By maintaining a clear separation between the domain layer and the data layer, software applications become more robust, maintainable, and testable. Just as a baker relies on a well-organized pantry and a clear recipe to create a perfect cake, developers can create successful applications by ensuring that the domain and data layers work together while maintaining their distinct responsibilities. If you ever had to bake a cake, what kind of cake would you pick? If you like this article and if it taught you anything at all, please share it with a friend and clap 👏</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=065c7f0e6e81" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Partial Prerendering: A New Dawn in Web Performance]]></title>
            <link>https://medium.com/@didierganthier/partial-prerendering-a-new-dawn-in-web-performance-b19a61ceb272?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/b19a61ceb272</guid>
            <category><![CDATA[nextjs14]]></category>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[api]]></category>
            <category><![CDATA[partial-prerendering]]></category>
            <category><![CDATA[nextjs]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Wed, 14 Feb 2024 12:29:23 GMT</pubDate>
            <atom:updated>2024-02-14T12:29:23.854Z</atom:updated>
            <content:encoded><![CDATA[<h3>Intro, Next.js, and Revolution</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/938/1*6I3dCEKXGA1-kSsUKLVA2Q.png" /></figure><p>Hello there, I’m Didier Peran Ganthier, a software engineer with a passion for all things tech. I’ve been using Next.js for years and today, I’m excited to delve into an emerging feature in web development that’s revolutionizing the way we think about rendering: <strong>Partial Prerendering</strong>.</p><h3>What is Partial Prerendering?</h3><p>Partial prerendering is an innovative feature that combines the best of both worlds: static and dynamic rendering. It’s like having your cake and eating it too! In the context of Next.js, partial prerendering allows you to render a route with a static loading shell, while keeping some parts dynamic. This means you can isolate the dynamic parts of a route, optimizing web performance.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F5wJN3xgQHiUBa%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F5wJN3xgQHiUBa%2Fgiphy.gif%3Fcid%3D790b7611snptczifwk6o36qojrq6gusih8sprt4aydrb1wh7%26ep%3Dv1_gifs_search%26rid%3Dgiphy.gif%26ct%3Dg&amp;image=https%3A%2F%2Fmedia4.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExaWQxeGc5Y3B4MGdjenI5YTRpM25xZnNoYXFvc3A2eDR3Yzlva3RsYyZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2F5wJN3xgQHiUBa%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="313" frameborder="0" scrolling="no"><a href="https://medium.com/media/adbe8b2608aaec8da8aad2234918aa3a/href">https://medium.com/media/adbe8b2608aaec8da8aad2234918aa3a/href</a></iframe><h3>How Does Partial Prerendering Work?</h3><p>Here’s the magic: A static route shell is served immediately, making the initial load fast. The shell leaves holes where dynamic content (that might be slower) will be streamed in to minimize the perceived overall page load time. The async holes are loaded in parallel, reducing the overall load time of the page. It’s like having a skeleton of your webpage delivered first, with the meat added in bit by bit. Crazy right?</p><h3>Why Does Partial Prerendering Matter?</h3><p>In the fast-paced world of web development, every millisecond counts. Partial prerendering is designed to improve the user experience by reducing the time it takes to load dynamic content. By delivering a fast initial static response, users aren’t left waiting for the entire page to load before they can start interacting with it. It’s all about enhancing the user experience, one render at a time.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FKEXxJgir4BxFZ7bYsC%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FKEXxJgir4BxFZ7bYsC%2Fgiphy.gif%3Fcid%3D790b7611s1mifv8wy4wb3fpi1wgdjrhuspn72hobud51m18b%26ep%3Dv1_gifs_search%26rid%3Dgiphy.gif%26ct%3Dg&amp;image=https%3A%2F%2Fmedia0.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExb3hxY3U1aHF0aGVlM2UxaGl2ajlyNHVxdzVubWticXNqYjB4czRieCZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FKEXxJgir4BxFZ7bYsC%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="246" frameborder="0" scrolling="no"><a href="https://medium.com/media/bcaeab21c4a00d83cf5e716ff3a9cac5/href">https://medium.com/media/bcaeab21c4a00d83cf5e716ff3a9cac5/href</a></iframe><h3>Additional Benefits of Partial Prerendering</h3><p>Partial prerendering offers several key advantages:</p><ol><li><strong>Faster Initial Page Load</strong>: Static contents are pre-rendered instantly as the page loads, so users don’t have to wait until all content, including the dynamic contents, is loaded.</li><li><strong>Improved User Experience</strong>: Dynamic contents are loaded seamlessly when they are ready, providing a smoother browsing experience.</li><li><strong>Reduced Server Load</strong>: By deferring the rendering of dynamic components until required by the user, partial prerendering helps reduce server load and improve overall site performance.</li><li><strong>Optimized Performance</strong>: Partial prerendering intelligently manages server resources, leading to a faster and more optimized performance.</li><li><strong>SEO Benefits</strong>: Partial prerendering could potentially have SEO benefits as well.</li><li><strong>Best of Both Worlds</strong>: It brings together the best aspects of static generation and server-side rendering, allowing for more fluid websites where static content is delivered instantly while dynamic content can be personalized based on user interaction or other factors.</li></ol><h3>Drawbacks of Partial Prerendering</h3><p>While partial prerendering has many benefits, it’s important to note that it also has some drawbacks:</p><ol><li><strong>SEO Challenges</strong>: While partially prerendered pages can potentially enhance SEO due to faster load times and crawlability, if search engines struggle to interpret dynamically rendered components correctly, this could negatively affect search engine rankings.</li><li><strong>Experimental Phase</strong>: Keep in mind that this feature is still in the experimental phase and therefore, not recommended for use in a production environment yet.</li></ol><h3>Implementing Partial Prerendering in Next.js</h3><p>To implement partial prerendering in your Next.js application, follow these steps:</p><ol><li><strong>Identify Dynamic Sections</strong>: Determine the sections or parts in your application where the asynchronous operations are taking place. This is the ideal place that you want to apply the Suspense API to.</li><li><strong>Use Suspense API</strong>: Wrap the dynamic elements of your route in Suspense. This way, Next.js can differentiate between static and dynamic components without any code changes.</li><li><strong>Provide a Fallback UI</strong>: The fallback UI is usually a loader that you want to show your users to let them know that contents are being loaded so they can keep going through other parts of the page that are already loaded or pre-rendered.</li></ol><h3>Partial Prerendering vs Server-Side Rendering</h3><p>The key difference between partial prerendering and server-side rendering (SSR) lies in when and where the rendering takes place. In SSR, the rendering process (generating a full HTML page) occurs on the server when a user agent requests a particular web page. The browser only receives ready-to-display content, allowing users to see the page without additional processing on their device.</p><p>On the other hand, partial prerendering allows developers to control which part of a page is pre-rendered or rendered first. It uses the React Suspense API, so it should be easy to understand and use if you’ve used React. Partial prerendering leverages a combination of static processing, specifically incremental static regeneration (ISR) and full server-side processing (SSR). With ISR, Next.js enables the pre-rendering of pages containing dynamic data during the build time. Subsequently, these pages are incrementally re-rendered in the background when needed, providing an efficient and dynamic user experience.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FYO3icZKE2G8OoGHWC9%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExbjdmbHU3cm9oaDBrbDU4NG50Mmlyb2FtMmhvN2o2ZWFhZ2R0bGg0MSZlcD12MV9naWZzX3NlYXJjaCZjdD1n%2FYO3icZKE2G8OoGHWC9%2Fgiphy.gif&amp;image=https%3A%2F%2Fmedia4.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExeTZtcWt3bDFhY2hpcG8xaXZmMzdzNHNibWU4c3pwNGU3ODgzdzB2ZSZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FYO3icZKE2G8OoGHWC9%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="363" frameborder="0" scrolling="no"><a href="https://medium.com/media/d9e57df86c415cc12dcd9134322d5c90/href">https://medium.com/media/d9e57df86c415cc12dcd9134322d5c90/href</a></iframe><h3>Example of a Website Using Partial Prerendering</h3><p>One example of a website that uses partial prerendering is the Next.js demo by Vercel Labs. This demo uses the new Next.js App Router and the experimental Partial Prerendering feature available in Next.js. Another example is HubSpot, which uses partial prerendering to serve “mostly” prerendered pages. For instance, a page may be entirely static except for a contact’s name displayed on the page. The page can be <strong>prerendered</strong> except for that contact name.</p><p>In conclusion, partial prerendering is more than just a buzzword. It’s a powerful tool that’s pushing the boundaries of what’s possible in web performance. As we continue to explore its potential, one thing is clear: the future of web development is here, and it’s partially <strong>prerendered</strong>.</p><p>Stay curious, keep learning, and remember: always deliver more than expected. Until next time!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b19a61ceb272" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Creating a Linear Gradient Container in Flutter]]></title>
            <link>https://medium.com/@didierganthier/creating-a-linear-gradient-container-in-flutter-39b64f77d9da?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/39b64f77d9da</guid>
            <category><![CDATA[flutter]]></category>
            <category><![CDATA[dart]]></category>
            <category><![CDATA[flutter-ui]]></category>
            <category><![CDATA[gradient]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Tue, 07 Nov 2023 12:27:07 GMT</pubDate>
            <atom:updated>2023-11-07T12:27:07.730Z</atom:updated>
            <content:encoded><![CDATA[<p>Have you ever wanted to add a touch of sophistication to your Flutter app’s user interface with a gradient container? Whether you’re recreating a design from Dribbble or simply looking to enhance your app’s aesthetics, creating a linear gradient container can be the perfect solution. In this article, we’ll walk you through the steps to achieve this effect using Flutter’s powerful tools.</p><h3>The End Result</h3><p>Before we dive into the details, let’s take a quick look at what we’re aiming for:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*G7vh60Dynh9VgbCWgnrffw.png" /></figure><p>The final appearance of your linear gradient container depends on several factors, including the colors you choose, the container’s height, and more. It’s a versatile approach that allows you to create unique visual effects.</p><h3>Step-by-Step Guide</h3><p>To create a linear gradient container in Flutter, follow these steps:</p><h3>1. Set Up Your Flutter Project</h3><p>Begin by either creating a new Flutter project or opening an existing one using your preferred development environment.</p><h3>2. Create the Linear Gradient Container</h3><p>Now, let’s dive into the details of creating the linear gradient container. This is where the magic happens.</p><p>In your Dart file where you plan to implement the gradient container, import the Flutter package:</p><pre>import &#39;package:flutter/material.dart&#39;;</pre><p>Next, define a Container widget and set the decoration property with a BoxDecoration that contains a LinearGradient as the background. Here&#39;s a basic example:</p><pre>class MySuperContainer extends StatelessWidget {<br>  @override<br>  Widget build(BuildContext context) {<br>    return MaterialApp(<br>      home: Scaffold(<br>        appBar: AppBar(<br>          title: Text(&#39;Linear Gradient Container&#39;),<br>        ),<br>        body: Center(<br>          child: Container(<br>            width: 200,<br>            height: 200,<br>            decoration: BoxDecoration(<br>              gradient: LinearGradient(<br>                begin: Alignment.topLeft,<br>                end: Alignment.bottomRight,<br>                colors: [Colors.blue, Colors.green],<br>              ),<br>            ),<br>            child: Text(&#39;Hello World&#39;),<br>          ),<br>        ),<br>      ),<br>    );<br>  }<br>}</pre><p>Let’s break down what’s happening in this code:</p><ul><li>LinearGradient allows you to define the direction and colors of your gradient.</li><li>begin and end specify the starting and ending points of the gradient. You can customize these points to control the direction of your gradient.</li><li>colors is a list of colors that make up the gradient. You can add more colors to create complex gradients.</li></ul><p>Inside the Container, you can place any widgets you want, such as text, images, buttons, or other UI elements. This allows you to create a rich and engaging user interface. I added a default ‘Hello World’ Text but feel free to experiment with other texts and widgets. ;)</p><h3>3. Customize and Experiment</h3><p>Feel free to experiment with different colors, gradient directions, container sizes, and content to achieve the exact look you desire for your app. You can explore advanced gradient options, such as radial gradients or sweep gradients, for even more creative possibilities.</p><p>And there you have it! With these steps, you’ve successfully created a linear gradient container in Flutter. It’s a fantastic way to add visual depth and appeal to your app’s user interface. Enjoy designing your own unique gradients!</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FtIeCLkB8geYtW%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FtIeCLkB8geYtW%2Fgiphy.gif%3Fcid%3Decf05e47iljinkvnkdf66uidzu2l1a8506w4nl77gbx254e7%26ep%3Dv1_gifs_search%26rid%3Dgiphy.gif%26ct%3Dg&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FtIeCLkB8geYtW%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="342" frameborder="0" scrolling="no"><a href="https://medium.com/media/da7d338ad904a0be3686d0c908e711a9/href">https://medium.com/media/da7d338ad904a0be3686d0c908e711a9/href</a></iframe><p>Creating a linear gradient container in Flutter is just the tip of the iceberg when it comes to designing beautiful and interactive user interfaces. Flutter, with its rich set of widgets and flexibility, allows you to bring your creative design ideas to life. Whether you’re a seasoned Flutter developer or just getting started, there’s always more to explore and learn.</p><p>As you embark on your Flutter journey, don’t hesitate to dive deeper into design elements, experiment with different gradients, and discover new ways to make your app visually stunning. Flutter’s community and resources are always available to support your creativity.</p><p>We’d love to hear from you! In the comments below, please share your favorite design features in Flutter, as well as any specific design challenges or topics you’d like us to explore in future articles. Your input and feedback are invaluable as we continue to explore the world of app design and development in Flutter.</p><p>Happy coding and designing! 🚀✨</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=39b64f77d9da" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Embracing the Paradox: The Pursuit of Meaning in an Absurd World]]></title>
            <link>https://medium.com/@didierganthier/embracing-the-paradox-the-pursuit-of-meaning-in-an-absurd-world-8fb1512bc02e?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/8fb1512bc02e</guid>
            <category><![CDATA[philosophy]]></category>
            <category><![CDATA[meaning]]></category>
            <category><![CDATA[absurdism]]></category>
            <category><![CDATA[life]]></category>
            <category><![CDATA[albert-camus]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Mon, 23 Oct 2023 11:05:58 GMT</pubDate>
            <atom:updated>2023-10-23T11:05:58.699Z</atom:updated>
            <content:encoded><![CDATA[<h3>Tough times teach.</h3><p>We get that, but what are deemed tough times? It can be highly relative; what you consider tough might be a walk in the park for someone else. Human beings seem to be in a perpetual search for something they don’t have, leaving us to wonder if this relentless pursuit is rooted in our need for purpose. After all, could it be that we question our own purpose if we were ever fully satisfied with what we have? But as we probe deeper into this quest for meaning, another question emerges: Should we keep searching for things even if, at some point, they become utterly meaningless to us?</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FdW6pIawajy6Mtd8aSA%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FdW6pIawajy6Mtd8aSA%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FdW6pIawajy6Mtd8aSA%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="244" frameborder="0" scrolling="no"><a href="https://medium.com/media/c51c72fdbeac57cd477aa2b20d36aded/href">https://medium.com/media/c51c72fdbeac57cd477aa2b20d36aded/href</a></iframe><p>Indeed, we should delve into this further. It’s a topic we often neglect, but Albert Camus understood human beings profoundly when he explored the idea of the universe and life being absurd and irrational, a philosophy known as absurdism. Absurdism posits that the universe is fundamentally irrational and devoid of inherent meaning. It suggests that attempting to find meaning can lead us into a conflict with the world itself.</p><p>Camus illustrates absurdism through three central works: “L’Étranger,” “The Myth of Sisyphus,” and “Caligula.” In the Myth of Sisyphus, he uses the Greek legend of Sisyphus, condemned by the gods to eternally roll a boulder up a hill, only to watch it roll back down once he reaches the summit. This myth serves as a metaphor for the individual’s persistent struggle against the essential absurdity of life. We, too, yearn to travel, to love, and to live to the fullest. But when the day concludes, we all meet the same fate — mortality. Regrets, if any, do not stem from what we have or haven’t done but are instead deeply tied to our inner selves. This leaves us with a puzzling question: Why do people continue to chase experiences that make them feel alive when, in the grand scheme of things, these experiences seem to hold no significant meaning or impact on the larger picture? This raises the ultimate question: What is the point of living if life itself is “absurd”?</p><h4>The Point of Living</h4><p>Camus, at the start of “The Myth of Sisyphus,” boldly claims that there is only one crucial question: Should human beings continue living or choose to end their lives, considering the apparent lack of inherent sense in life itself? He views this question of suicide as a natural response to the underlying reality that life is absurd. To continually seek meaning in life, according to Camus, is absurd in itself, given that there is none to be found. Furthermore, he argues that it is equally absurd to hope for some form of continued existence after death, as this hope only leads to our eventual extinction.</p><p>However, while Camus’s exploration of absurdism and the question of suicide is a profound philosophical inquiry, our focus here is not solely on this aspect. Instead, we pivot toward the pursuit of happiness and the endeavor to give meaning to one’s life.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FCyoQdbc7FHqqTpkSPI%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FCyoQdbc7FHqqTpkSPI%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FCyoQdbc7FHqqTpkSPI%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="362" frameborder="0" scrolling="no"><a href="https://medium.com/media/bc0f7a1ec119f42db2e28731633127d3/href">https://medium.com/media/bc0f7a1ec119f42db2e28731633127d3/href</a></iframe><h4>The Pursuit of Happiness</h4><p>In the grand theater of existence, a perplexing paradox unfolds before us: Why do we persistently chase experiences that make us feel alive, even when those experiences may ultimately hold no substantial consequence in the vast expanse of the universe? This paradox lies at the core of human existence and aligns closely with the philosophy of absurdism as championed by Albert Camus.</p><p>The human desire to feel alive is an intrinsic and profoundly compelling force that has propelled countless endeavors throughout history. From the audacious expeditions of explorers venturing into uncharted territories to the fiery romances that set our hearts ablaze, from the heart-pounding thrill of extreme sports to the transcendental moments of artistic creation, we are ceaselessly drawn to experiences that infuse our lives with vibrancy, connection, and presence.</p><p>However, within the framework of absurdism, this ceaseless quest for vitality may appear paradoxical. After all, if the universe is fundamentally irrational and devoid of intrinsic meaning, why persist in seeking these profound moments? If, in the grand tapestry of existence, our actions amount to little more than a fleeting speck in the cosmic timeline, why do we yearn for experiences that momentarily elevate our existence?</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Fyhc3VVqX2wBdnHKyqG%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fyhc3VVqX2wBdnHKyqG%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2Fyhc3VVqX2wBdnHKyqG%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="435" frameborder="0" scrolling="no"><a href="https://medium.com/media/d40fa2cbe6da44779fc029d5f4cb05d8/href">https://medium.com/media/d40fa2cbe6da44779fc029d5f4cb05d8/href</a></iframe><p>The answer, it seems, resides within the very essence of our humanity. We are creatures endowed with consciousness, blessed with the ability to reflect upon our existence and imbue it with our own meaning. While the universe may remain indifferent and absurd, our subjective experiences hold profound significance to us. In those moments of intensity, when we feel most alive, we briefly transcend the existential weight of absurdity and uncover purpose in the immediate, the personal, and the visceral.</p><p>In these moments, we rise above the mundane and touch the profound. We glimpse the sublime, not because it inherently exists in the universe, but because we manifest it within ourselves. The pursuit of feeling alive, then, becomes an act of defiance against the absurdity of existence. It is our way of declaring that, while the universe may lack inherent meaning, we can forge our own meaning through our experiences and connections.</p><p>In essence, the pursuit of feeling alive serves as our response to Camus’s fundamental question: Should we continue living in an absurd world? Our resounding answer is “yes.” It is our affirmation of our existence, even in the face of a universe that may seem indifferent and irrational. It is our proclamation that, despite the absurdity that surrounds us, we choose to live and find meaning in the fleeting, the beautiful, and the profound.</p><p>As we navigate the intricacies of our lives, let us wholeheartedly embrace this paradoxical pursuit of feeling alive. Let us actively seek those experiences that make our hearts race and our souls sing. Let us remember that, in the grand scheme of the cosmos, our individual moments may appear inconsequential, but they are profoundly significant to us. They are the colors that we paint onto the canvas of our existence, and through them, we uncover our own purpose and meaning in this seemingly absurd world. Is life as it is inevitable then?</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F6tb0l9xAT5z48xdwFE%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F6tb0l9xAT5z48xdwFE%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2F6tb0l9xAT5z48xdwFE%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="435" frameborder="0" scrolling="no"><a href="https://medium.com/media/de5e958c9857333b6a4da72138da8081/href">https://medium.com/media/de5e958c9857333b6a4da72138da8081/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8fb1512bc02e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The biggest non-technical lessons learned in my journey as a developer]]></title>
            <link>https://medium.com/@didierganthier/the-biggest-non-technical-lessons-learned-in-my-journey-as-a-developer-96ec948c3df7?source=rss-f2aaf28d7896------2</link>
            <guid isPermaLink="false">https://medium.com/p/96ec948c3df7</guid>
            <category><![CDATA[mindset]]></category>
            <category><![CDATA[developer]]></category>
            <category><![CDATA[problem-solving]]></category>
            <category><![CDATA[microverse]]></category>
            <category><![CDATA[non-technical-skills]]></category>
            <dc:creator><![CDATA[Didier Peran Ganthier]]></dc:creator>
            <pubDate>Fri, 18 Nov 2022 11:29:03 GMT</pubDate>
            <atom:updated>2022-11-18T11:29:03.106Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PRLSBtqUJn0524pc82NtBQ.jpeg" /></figure><p>When I started coding, it was all about the code, “that was all :)”. I always thought to write good, clean, functional code, all you had to be was a good coder. But, I realized in the same way a developer needs a good PC to work on, he has to be in a good space to work efficiently. I never thought about this until I started to realize how significantly you can improve yourself as a person and as a developer if you give your mind permission to take non-technical lessons from your journey. There are two non-technical lessons I learned that are helping me become a happier developer today: appreciate the small wins and feel as good as the code.</p><h3>Appreciate the small wins</h3><p>What do you call small wins? They can be anything from fixing a bug, adding a listener to a button, fixing the state of your application, or a simple console.log of data you retrieved from an API. It feels good, doesn’t it, to see that at least a part of your application works as it should, that the red lines disappeared…</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FRIqMckeZEcD8w7NTT3%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FRIqMckeZEcD8w7NTT3%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FRIqMckeZEcD8w7NTT3%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="244" frameborder="0" scrolling="no"><a href="https://medium.com/media/24f1f7e2f39a33fb2a223973c953b8d3/href">https://medium.com/media/24f1f7e2f39a33fb2a223973c953b8d3/href</a></iframe><p>It feels good and it’s part of smart goal setting. You can always break down your goals, which will bring you more fulfillment and help you enjoy the process a little more. Now let’s take a quick look at my personal experiences.</p><h4>My experience at Microverse</h4><p>For the past four months, I’ve been studying at <a href="https://microverse.org/">Microverse</a>, an online coding school where I code eight hours a day with people from all around the world, picking up the skills of teamwork and communication. One of my experiences goes back to the JavaScript Group Capstone, for context a “Capstone Week” is the last week of a module, and a module contains five weeks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GaiTQAw9701_DW2cTsJ4Hw.png" /></figure><p>The fact that it was a group project tremendously increased the challenge, setting goals and deadlines, and constantly communicating to know if everything is going according to plan (stuff never goes according to plan). We had a Kanban board set up to track our progress and the different tasks assignations.</p><p>Let’s move a bit forward in time to see where it all goes wrong. After a whole week of frustration and confusion, we finally got a working version and we were ready for the final part of the project: Presenting the application in a video showcasing the features and the code we wrote. It was 2 AM in my partner’s country and 11 PM in mine when we figured out while recording the video that some features weren’t working as expected. How frustrating can it be to see that when you were just about to rest for a bit, the toughest part of the work would start?</p><h4>Breaking down the challenge</h4><p>The challenge can seem excessively huge, when this happened it was huge, and we could not submit it for a review before finding a solution. The title of this paragraph is about breaking down a challenge to a simpler version of itself but in order to do that you must at least understand it a bit. The quest to understand the challenge can itself be somehow part of breaking it down we might say.</p><p>For some context, we were building a website fetching event data from the <a href="https://seatgeek.com/account/develop">SeatGeek API</a> and using the Involvement API to add likes and comments for each item. The bug we were facing was that the interactions weren’t working (like/comment). But let us break down the expected result to see how much more we are expecting out of just that:</p><p>For the likes:</p><blockquote>The user clicks on the like button</blockquote><blockquote>A POST request is sent to the Involvement API (saying there’s a new like for x item)</blockquote><blockquote>If the previous is successful, we recall the method to GET the likes and update the corresponding DOM element (the like counter)</blockquote><p>For the comments:</p><blockquote>The user clicks on the send comment button</blockquote><blockquote>A POST request is sent to the Involvement API (saying there’s a new comment for x item)</blockquote><blockquote>If the previous is successful, we recall the method to GET the comments and update the corresponding DOM element (the comments counter and the unordered list containing them)</blockquote><h4>Enjoy the process, steps by steps</h4><p>Coding can be hard, and it becomes tougher if you try to do everything at once. Now in this example what helped me figure out where the issue reside was by taking one thing at a time and enjoying fixing every single part of the app. Before we even say stuff like: “The comment feature isn’t working.” Try to debug a bit, ask yourself the right questions, are you interacting with the right DOM element? What is the response of the API? Is the problem with getting the comments or showing them?</p><p>By taking time to answer those questions and giving myself credit for each output brought me closer to the solution which helped my mind think positively and created a better space for myself to succeed. Imposter syndrome is real but remember, you are your number one fan.</p><h3>The better you feel, the better you code</h3><p>This is a bit of a follow up of the previous part, I think allowing yourself to feel good and acknowledging when you do not feel good is very underrated. At the end of the day even machines need to rest. We, humans, are extremely complicated and sophisticated machines. Often, I found myself in situations where I wanted to work a bit more but my mind or/and my body would not allow it. The more aware you become of that the more productive you can be. Why is it that we need to feel good to code?</p><h4>Coding is hard</h4><p>Coding is not hard at all, I know I just contradicted myself, at least the part that is hard is not the actual coding part, it’s surviving the frustration that comes with coding because it will come, you sign up to solve challenges every single time you write some code and those challenges will come haunting you at night and even during the day. 👻</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FYph6D7zPIVtIc%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FYph6D7zPIVtIc%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FYph6D7zPIVtIc%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="295" frameborder="0" scrolling="no"><a href="https://medium.com/media/b6bc75f6c4b062a667ca91443b719c7f/href">https://medium.com/media/b6bc75f6c4b062a667ca91443b719c7f/href</a></iframe><h4>Allow yourself to feel good</h4><p>The point is if the better you feel, the better you code and I have been experiencing that. I have seen people shocked when I say that I only work when I want to work but what is the point of life if you are not allowed to do something you like? Something you enjoy, we all started coding because we wanted to learn, build wonderful things, have fun, and make money while doing it.</p><p>I beat some important challenges like that at Microverse also this year, I remember spending a day trying to wrap my head around Jest with my coding partners and we could not figure out which foot goes before the other and then I told my coding partners: let’s take some rest and give it a go again later. I figured out that it was a new concept and that it would take time to fully grasp it. I then went home took a nap and then we all rested and then later on we studied, each of us tried to understand it better after taking some rest and although we were late on schedule, tomorrow was a really productive day because we were able to not only come up with a solution for the previous day’s challenge but also for the current day’s.</p><p>And this is just an isolated experience, I have experienced this so many times how much taking a walk, napping, or even taking a shower. Talking to someone and listening to music can also reveal itself particularly useful. The point is you must feel good at all costs. I say sometimes that resting is part of the work, and it truly is because it allows you to work better, of course we need some balance in everything and like everything you do you have to manage your time well, so you do not end up resting too much or working too much. It is all about balance.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FRy1MOAeAYXvRVQLPw3%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2FRy1MOAeAYXvRVQLPw3%2Fgiphy.gif&amp;image=https%3A%2F%2Fi.giphy.com%2Fmedia%2FRy1MOAeAYXvRVQLPw3%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="183" frameborder="0" scrolling="no"><a href="https://medium.com/media/c002c53b1af5b8048baf03c2e2c1a480/href">https://medium.com/media/c002c53b1af5b8048baf03c2e2c1a480/href</a></iframe><p>If you made it this far, I thank you already and that means either you scrolled straight to the bottom or you read the whole thing 😅, I appreciate you anyway. I love writing and I plan to write a bit more and share some more of my experience. I truly hope that this was helpful in anyway to answer your questions, this is of course about my personal experience therefore just my point of view on certain things. Hope you enjoyed reading this article? Please feel free to share your thoughts and comments!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=96ec948c3df7" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>