<?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 omerfeyzioglu on Medium]]></title>
        <description><![CDATA[Stories by omerfeyzioglu on Medium]]></description>
        <link>https://medium.com/@oomerfeyzioglu?source=rss-947c603be518------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*7tYF2t33SUR__6GfKG1Ydg.png</url>
            <title>Stories by omerfeyzioglu on Medium</title>
            <link>https://medium.com/@oomerfeyzioglu?source=rss-947c603be518------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 17:04:22 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@oomerfeyzioglu/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[Building a Real-Time Crypto Pipeline with Redpanda and ClickHouse]]></title>
            <link>https://blog.dataengineerthings.org/building-a-real-time-crypto-pipeline-with-redpanda-and-clickhouse-0e5d00d09174?source=rss-947c603be518------2</link>
            <guid isPermaLink="false">https://medium.com/p/0e5d00d09174</guid>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[clickhouse]]></category>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[data-engineering]]></category>
            <category><![CDATA[red-panda]]></category>
            <dc:creator><![CDATA[omerfeyzioglu]]></dc:creator>
            <pubDate>Fri, 17 Jul 2026 19:45:30 GMT</pubDate>
            <atom:updated>2026-07-25T17:40:40.846Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nUyxHJnLGDaImeTMd5MZ9w.png" /></figure><p><strong>Introduction</strong></p><blockquote>Why I Built This Project</blockquote><p>I wanted to see what is really behind a streaming pipeline. While we are going on our happy path what are the rough edges? In today’s world reaching any information or learning how to use a tool is the easiest part. The real challenge is becoming aware of silent problems or possible failure causes of the pipeline. At the beginning, the pipeline looked like :</p><p><strong><em>WebSocket → Redpanda →ClickHouse</em></strong></p><p>However, while I was working on it ; I realized everything could keep running without any error when data was missing, duplicated or stale. The article is about the design decisions made to handle possible causes of failure.</p><h4>The Project</h4><p>The pipeline consumes Binance aggregate trade events from a <a href="https://developers.binance.com/docs/binance-spot-api-docs/web-socket-streams">WebSocket stream</a>, publishes them into Redpanda to decouple live ingestion from the rest of the pipeline, and stores the records in ClickHouse for analytical queries.</p><p>The consumer detects missing aggregate trade IDs, fills bounded gaps from the Binance REST API, and republishes recovered records back to Redpanda with a Kafka header.</p><p>Airflow periodically runs data quality checks on the data written to ClickHouse.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_ksDCzMhhF0_eiKGvszJXw.png" /><figcaption>High-level Architecture Diagram</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8F_b-WapgO2qIZ2zutNorQ.png" /><figcaption>Data Flow Diagram</figcaption></figure><h4>Reconnecting to the WebSocket Does Not Recover Missed Trades</h4><p>A WebSocket connection can drop for unexpected reasons: a short network issue, a client-side interruption, or the server closing a long lived stream. The client may reconnect and the pipeline may start receiving live events again. But that only means the <strong>connection recovered</strong>. It does not mean the trades that happened during the disconnected interval were recovered.</p><p>That is the dangerous part. After reconnecting, everything can look healthy: new messages arrive, the process keeps running, and there may be no visible error. But the data can already be incomplete.</p><p>To detect this, I relied on <a href="https://developers.binance.com/docs/binance-spot-api-docs/web-socket-streams">Binance aggregate trade IDs</a>. For each symbol (trading pair), the consumer keeps track of the last aggregate trade ID it has processed. If the next incoming ID jumps forward, the system can be aware of missing events.</p><blockquote>The consumer detects ID gaps per symbol and fills bounded missing ranges from <a href="https://developers.binance.com/en/docs/catalog/core-trading-spot-trading/api/rest-api/market#agg-trades">Binance REST</a>.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/385/1*ZgRupJwGI3MIzJQ8sDDO5w.png" /></figure><h4>Filling One Gap Can Create Another</h4><p>After consumer detected the gaps occurred , next step is obvious : Fill the gaps via REST API. We gave missed event’s first aggregate trade ID and limit to API payload. But there is a critical point : while we are sending request and waiting to fetch data , <strong>live market data is still streaming</strong>. If the same flow is waiting for REST response and cannot continue reading live events, new incoming data may be missed or processed too late. So, gap filling may cause <strong>data loss reaching downstream</strong>.</p><p>To solve this tightly coupled behavior, I needed a <strong>buffer between live ingestion and downstream processing</strong>.</p><p>The live WebSocket reader should not wait for gap recovery. Its job should be simple: keep reading live market data and publish it somewhere. The consumer can then read from that buffer, detect gaps, call the REST API when needed, and write the data to ClickHouse.</p><blockquote>This is where <strong>Redpanda came in</strong>.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/774/0*Xp7-onlsQgBLlhCk" /></figure><p>Redpanda gave the live stream a place to continue flowing while the consumer handled slower recovery work.</p><p>I chose Redpanda because it was easy to run locally and Kafka compatible. I could use Kafka concepts such as topics, keys, partitions and headers without running a heavier Kafka setup for this project.</p><p><strong>Where Should Recovered Data Go?</strong></p><p>After missing trades are fetched, another question appears: where should this recovered data go?</p><p>First option is writing directly to ClickHouse. But that only fixes the storage layer. If <strong>another consumer</strong> listens to the Redpanda topic later, it should also need the recovered data. Then the only way to get recovered data would be reading it from ClickHouse. At the end of the day, that is not a sustainable solution. So I published the backfilled data back to the <strong>same Redpanda topic</strong>.</p><blockquote>But this decision creates another problem.</blockquote><p>If the records go to the same topic, the consumer will <strong>read them again</strong>. But, these records are older than the live message in the topic, so the consumer should not treat them like normal live event. Therefore, I marked those recovered records with a “source=backfill” header.</p><blockquote>With this header , consumer skips gap detection for that message and avoids false gaps.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5jWB7-3_DdlG80bySMSA-w.png" /><figcaption>Recovered Event Instance</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c0TRc_Uv4B6U7lK-gxo2zA.png" /><figcaption>Live Event Instance</figcaption></figure><blockquote><strong>A Small Detail</strong></blockquote><blockquote>Live records and recovered records were produced by different Kafka clients. The WebSocket producer used <strong>aiokafka</strong>, while the backfill side used <strong>confluent-kafka</strong>. Since both paths published records with the symbol (trading pair)as the key, I needed both clients to use the <strong>same partitioning behavior</strong>. Otherwise, the same symbol could land on different partitions depending on which client produced the message. To avoid that, I aligned the partitioning behavior with Kafka’s Murmur2 hashing on backfill side.</blockquote><pre>    producer = Producer({<br>        &quot;bootstrap.servers&quot;: kafka_config.bootstrap_servers,<br>        &quot;linger.ms&quot;: kafka_config.linger_ms,<br>        &quot;compression.type&quot;: kafka_config.compression_type,<br>        &quot;partitioner&quot;: &quot;murmur2_random&quot;,<br>    })</pre><p><strong>What If the Gap Is Too Large?</strong></p><p>For small gaps, the consumer can recover missing trades from the REST API. But I did not solve large gap recovery problem yet.</p><p>If the pipeline is down for a long time, trying to fill a large gap inline could block live processing and make recovery itself the new bottleneck. To avoid that, I added a maximum gap size. Larger gaps are logged and skipped, so they are visible, but they still require a separate recovery strategy.</p><p>A better approach would be a separate batch backfill job or a manual recovery workflow for those cases.</p><h4>Deduplication Is Not Immediate</h4><p>After the records reached ClickHouse, there was one more point to be careful about : duplicates. This can happen because of a consumer retry, or because a trade that already arrived from the WebSocket path is written again during backfill.</p><blockquote>I used <a href="https://clickhouse.com/docs/engines/table-engines/mergetree-family/replacingmergetree">ReplacingMergeTree</a> in ClickHouse and defined the trade identity with: <strong>ORDER BY (symbol, agg_id)</strong> , <em>symbol represents the trading pair.</em></blockquote><p>That means the records will be deduplicated on symbol and agg_id columns. I chose these two columns because their combination uniquely identifies an aggregate trade. But ReplacingMergeTree engine doesn’t guarantee duplicates will immediately disappear. So when running SQL queries, you may still see duplicate records.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/367/1*rlNWFx1SyRl3BrUGUyq8NA.png" /><figcaption>Duplicate records can still be visible before ClickHouse merges parts.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/397/1*4ZmSYXz7TEHWXbjb0CXFow.png" /><figcaption>After forcing a merge with OPTIMIZE FINAL, duplicate records are collapsed.</figcaption></figure><blockquote>I used OPTIMIZE FINAL only to demonstrate this behavior. It is not part of the normal ingestion path.</blockquote><h4>The Table Can Be Stale While Everything Looks Fine</h4><p>At this point, records were reaching ClickHouse successfully. But writing data into a table and storing it is not enough for a data pipeline. A problem can still happen in the background in any component of the pipeline. Also you have to <strong>validate the data</strong> ,<strong>check the quality</strong> and business rules.</p><p>That is why I added Airflow checks on top of ClickHouse. The checks query the final table directly and look for problems such as stale data, missing aggregate trade IDs, duplicate records, invalid values and unusual backfill ratios.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*B35gXuaNZr3tswd4sGkN6A.png" /><figcaption>Airflow fails the freshness check when ClickHouse stops receiving recent data</figcaption></figure><h4>What I Would Improve Next</h4><p>This project helped me to see different perspectives of a data pipeline. I realized design decisions can be more critical especially for real time data streaming.</p><p>There are still some important parts I would improve in the next iteration:</p><p>For example, last seen aggregate trade ID. If the consumer restarts, the state easily can be lost. Better approach would be storing that “last_seen” variable somewhere won’t be lost.</p><p>Large gap recovery is a limitation in this project. A separate batch backfill would be better solution.</p><p>I would also improve offset handling. Offsets should be committed after the data is written to ClickHouse.</p><p>The important part is seeing the failure points and improving the design step by step.</p><h4>Conclusion</h4><p>At the beginning, the pipeline looked like :</p><p><strong><em>WebSocket → Redpanda →ClickHouse</em></strong></p><p>But while building it , I realized that a pipeline can keep running while the data is missing or duplicated. Choosing tools or connecting them is not the most important part of creating a data pipeline. Real value is caring about the data : what is the data flow, what are the possible failure points, how we keep our pipeline live , how do we validate data quality and tons of other questions. That was the main thing that I wanted to see with this project.</p><p>github link: <a href="https://github.com/omerfeyzioglu/real-time-crypto-flow">https://github.com/omerfeyzioglu/real-time-crypto-flow</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0e5d00d09174" width="1" height="1" alt=""><hr><p><a href="https://blog.dataengineerthings.org/building-a-real-time-crypto-pipeline-with-redpanda-and-clickhouse-0e5d00d09174">Building a Real-Time Crypto Pipeline with Redpanda and ClickHouse</a> was originally published in <a href="https://blog.dataengineerthings.org/">Data Engineer Things</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Google Business Profile API Guide:Access, Auth & Reviews]]></title>
            <link>https://medium.com/dogus-teknoloji/google-business-profile-api-guide-access-auth-reviews-1a7d9585f842?source=rss-947c603be518------2</link>
            <guid isPermaLink="false">https://medium.com/p/1a7d9585f842</guid>
            <category><![CDATA[google]]></category>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[business-automation]]></category>
            <category><![CDATA[data-integration]]></category>
            <category><![CDATA[google-api]]></category>
            <dc:creator><![CDATA[omerfeyzioglu]]></dc:creator>
            <pubDate>Mon, 25 Aug 2025 16:28:07 GMT</pubDate>
            <atom:updated>2025-08-25T16:28:07.900Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FexdQonz6fHu5qQ2ufAilQ.jpeg" /></figure><p>In today’s world, where data is invaluable, using your business data with Google to <strong>take your company a step further</strong> is truly essential. Discovering the current state of each branch, comparing performance, measuring <strong>customer satisfaction</strong> and identifying possible improvements is just one API away.</p><p>With this integration, businesses can track dealer performance, monitor response times, and follow rating trends. Thanks to AI support, customer reviews can be summarized, analyzed and even answered automatically.</p><p>This way, managers not only see the big picture but also take quick action. <strong>The result is higher customer trust and better decisions based on real data.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T0rldJFW4kHKC6gyhAF1Ow.png" /></figure><h3>1. Installation</h3><p>As the first step you need to fill out <a href="https://support.google.com/business/contact/api_default">Google’s Business Profile API Support form</a> to request access to the entire Business Profile API suite and wait 3–5 days for approval. Once approved, you can enable the necessary APIs for your business.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/914/1*q9huITlx_P1ZByEyhHIfeQ.png" /><figcaption>Google Business Profile API Support Form</figcaption></figure><p>Meanwhile, you’ll need your client to grant <strong>manager access</strong> to their business locations through the Google account you’ll be using for the integration.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*w4vjrcR8irVd2e53zPA_ng.png" /><figcaption>Enabled APIs</figcaption></figure><p>It’s time for enabling three specific APIs in Google Cloud Console: <strong>Google My Business API</strong> for accessing reviews and insights, <strong>My Business Business Information API</strong> for location details and business data, and <strong>My Business Account Management API</strong> for retrieving the account ID that you must own to use your APIs. You are able to see all the APIs you enabled for your profile.</p><h3>2. Integration</h3><pre>import json<br>from google_auth_oauthlib.flow import Flow<br><br>CLIENT_SECRET_FILE = &quot;client_secret.json&quot;<br><br>with open(CLIENT_SECRET_FILE, &quot;r&quot;, encoding=&quot;utf-8&quot;) as f:<br>    client_json = json.load(f)<br><br>redirect_uris = client_json.get(&quot;installed&quot;, {}).get(&quot;redirect_uris&quot;, [])<br>REDIRECT_URI = redirect_uris[0]<br><br>SCOPES = [&quot;https://www.googleapis.com/auth/business.manage&quot;]<br><br>flow = Flow.from_client_secrets_file(<br>    CLIENT_SECRET_FILE,<br>    scopes=SCOPES,<br>    redirect_uri=REDIRECT_URI,<br>)<br><br>auth_url, _ = flow.authorization_url(https://medium.com/%3Cbr%3E%20%20%20%20access_type=&quot;offline&quot;,%3Cbr%3E%20%20%20%20include_granted_scopes=&quot;true&quot;,%3Cbr%3E%20%20%20%20prompt=&quot;consent&quot;%3Cbr%3E)<br><br>print(auth_url)<br>redirect_response = input(&quot;Paste the full redirect URL here: &quot;).strip()<br><br>flow.fetch_token(authorization_response=redirect_response)<br>creds = flow.credentials<br><br>print(&quot;\nAccess Token:&quot;, creds.token)<br>print(&quot;Refresh Token:&quot;, creds.refresh_token)<br>print(&quot;Token Expiry:&quot;, creds.expiry)</pre><p>The next step is critical for authentication. We used the<strong> OAuth2 web application</strong> method, which might seem unconventional but proved effective for our use case. In <strong>Google Cloud Console</strong>, under the same project where we enabled the APIs, we created OAuth2 credentials and downloaded the<strong> client JSON file</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y1tCKAyMwJ44UnrgSGimmA.png" /><figcaption>client_secret.json</figcaption></figure><p>After running our authentication code, we copy the given URL and open it in our browser, then <strong>sign in with the Google account</strong> that has manager permissions to the business locations. After successful authentication, Google redirects us to our specified redirect URI with an<strong> authorization code </strong>in the URL parameters. We copy this final <strong>redirected URL</strong> from the browser and <strong>paste it back into</strong> our terminal to complete the authentication process and obtain our tokens.</p><pre>import json<br>import requests<br><br>with open(&quot;client_secret.json&quot;, &quot;r&quot;, encoding=&quot;utf-8&quot;) as f:<br>    client_json = json.load(f)<br>client_info = client_json.get(&quot;web&quot;, client_json.get(&quot;installed&quot;, {}))<br><br>client_id = client_info[&quot;client_id&quot;]<br>client_secret = client_info[&quot;client_secret&quot;]<br>token_uri = client_info[&quot;token_uri&quot;]<br><br>refresh_token = &quot;Refresh Token&quot;<br><br>token_response = requests.post(<br>    token_uri,<br>    data={<br>        &quot;grant_type&quot;: &quot;refresh_token&quot;,<br>        &quot;client_id&quot;: client_id,<br>        &quot;client_secret&quot;: client_secret,<br>        &quot;refresh_token&quot;: refresh_token<br>    }<br>)<br>token_response.raise_for_status()<br>access_token = token_response.json()[&quot;access_token&quot;]<br><br>#call api<br>headers = {&#39;Authorization&#39;: f&#39;Bearer {access_token}&#39;}<br>accounts_url = &#39;https://mybusinessaccountmanagement.googleapis.com/v1/accounts&#39;<br>response = requests.get(accounts_url, headers=headers)<br>print(&quot;Accounts:&quot;, response.text)</pre><p>Once we have a valid access token, the next step is retrieving the <strong>account ID</strong> using the My Business Account Management API. This account ID belongs to the customer’s Google account that granted us manager permissions, and it’s essential for all subsequent API calls. We’ll need this account ID when fetching both location data and reviews in the following steps.</p><blockquote>Note :The account ID serves as the primary identifier that links all business locations and their associated data under the customer’s Google Business Profile. Without it, we can’t access any of the business information we need.</blockquote><pre>{<br>  &quot;accounts&quot;: [<br>    {<br>      &quot;name&quot;: &quot;accounts/123456789012345678901&quot;,<br>      &quot;accountName&quot;: &quot;ExampleBusiness&quot;,<br>      &quot;type&quot;: &quot;LOCATION_GROUP&quot;,<br>      &quot;role&quot;: &quot;MANAGER&quot;,<br>      &quot;verificationState&quot;: &quot;VERIFIED&quot;,<br>      &quot;vettedState&quot;: &quot;NOT_VETTED&quot;,<br>      &quot;accountNumber&quot;: &quot;1234567890&quot;,<br>      &quot;permissionLevel&quot;: &quot;MEMBER_LEVEL&quot;<br>    },<br>    {<br>      &quot;name&quot;: &quot;accounts/987654321098765432109&quot;,<br>      &quot;accountName&quot;: &quot;PersonalAccount&quot;,<br>      &quot;type&quot;: &quot;PERSONAL&quot;,<br>      &quot;verificationState&quot;: &quot;UNVERIFIED&quot;,<br>      &quot;vettedState&quot;: &quot;NOT_VETTED&quot;<br>    }<br>  ]<br>}</pre><h3>3. Fetching</h3><p>Finally, we have all the credentials required for fetching APIs. Firstly, we need to <strong>get locations and their location ids’</strong> to be accessed reviews.With location ids you can get <strong>the reviews of your business</strong>. The code below shows getting locations and automatically fetching reviews for all locations.</p><pre>import requests<br><br>access_token = &quot;Access Token&quot;<br>account_id = &quot;accounts/123456789012345678901&quot;<br><br>headers = {&quot;Authorization&quot;: f&quot;Bearer {access_token}&quot;}<br>params = { &quot;readMask&quot;: &quot;name,title&quot; }<br><br>#fetch location<br>locations_url = f&quot;https://mybusinessbusinessinformation.googleapis.com/v1/{account_id}/locations&quot;<br>locations_response = requests.get(locations_url, headers=headers, params=params)<br>locations_response.raise_for_status()<br>locations_data = locations_response.json()<br><br># iterate over locations and fetch reviews for each<br>for location in locations_data.get(&quot;locations&quot;, []):<br>    location_resource = location[&quot;name&quot;]  # e.g. &quot;locations/233213213&quot;<br>    reviews_url = f&quot;https://mybusiness.googleapis.com/v4/{account_id}/{location_resource}/reviews&quot;<br>    reviews_response = requests.get(reviews_url, headers=headers)<br>    print(f&quot;Reviews for {location_resource}:&quot;)<br>    print(reviews_response.text)</pre><p>My Business Business Information API returns<strong> all accessable locations</strong> like the JSON below.</p><p>After that , our code gets these names and fetch My Business API.</p><pre>import requests<br><br>access_token = &quot;Access Token&quot;<br>account_id = &quot;accounts/123456789012345678901&quot;<br><br>headers = {&quot;Authorization&quot;: f&quot;Bearer {access_token}&quot;}<br>params = { &quot;readMask&quot;: &quot;name,title&quot; }<br><br>#fetch location<br>locations_url = f&quot;https://mybusinessbusinessinformation.googleapis.com/v1/{account_id}/locations&quot;<br>locations_response = requests.get(locations_url, headers=headers, params=params)<br>locations_response.raise_for_status()<br>locations_data = locations_response.json()<br><br># iterate over locations and fetch reviews for e<br>for location in locations_data.get(&quot;locations&quot;, []):<br>    location_resource = location[&quot;name&quot;]  # e.g. &quot;locations/233213213&quot;<br>    reviews_url = f&quot;https://mybusiness.googleapis.com/v4/{account_id}/{location_resource}/reviews&quot;<br>    reviews_response = requests.get(reviews_url, headers=headers)<br>    print(f&quot;Reviews for {location_resource}:&quot;)<br>    print(reviews_response.text)</pre><h3>What’s next?</h3><p>You can store business reviews in your database, build meaningful <strong>dashboards</strong>, or deliver it back to your customer in a way that helps them act fast.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*KC-uwYoWpKQBtImx" /></figure><p>Going a step further, AI can help summarize thousands of <strong>reviews into clear, actionable insights: </strong>“customers loved the staff’s friendliness, but complained about waiting times.” This turns scattered feedback into practical guidance. Whether it’s a car dealership, a coffee shop or a service center, this data is the critical information that helps businesses improve <strong>customer satisfaction</strong>, <strong>refine their strategy</strong> and grow stronger every day.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1a7d9585f842" width="1" height="1" alt=""><hr><p><a href="https://medium.com/dogus-teknoloji/google-business-profile-api-guide-access-auth-reviews-1a7d9585f842">Google Business Profile API Guide:Access, Auth &amp; Reviews</a> was originally published in <a href="https://medium.com/dogus-teknoloji">Doğuş Teknoloji</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Interface in Java]]></title>
            <link>https://medium.com/@oomerfeyzioglu/interface-in-java-e5439ed10d5f?source=rss-947c603be518------2</link>
            <guid isPermaLink="false">https://medium.com/p/e5439ed10d5f</guid>
            <category><![CDATA[java]]></category>
            <category><![CDATA[interfaces]]></category>
            <dc:creator><![CDATA[omerfeyzioglu]]></dc:creator>
            <pubDate>Fri, 21 Oct 2022 17:34:39 GMT</pubDate>
            <atom:updated>2022-10-21T17:35:08.138Z</atom:updated>
            <content:encoded><![CDATA[<p>Interfaces are used to set the standard for a class. To access the interface methods, the interface must be “implemented” by another class with the implements keyword .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/880/0*13ou1RrcNoVg9Eq2.png" /></figure><blockquote>Interfaces are used for multiple inheritance.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f4-dv9vRvJViFn7WRXTSTg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*N_o5QvSJ-oKlgYZEY0gZig.png" /></figure><blockquote>We have to use “eat” and “travel” methods because of you have implemented “Animal interface”.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2flzA-crN49IHuF7s_KBsA.png" /></figure><blockquote>Both cat and dog eat and travel.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e5439ed10d5f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Value and Reference Type in Java]]></title>
            <link>https://medium.com/@oomerfeyzioglu/value-and-reference-type-in-java-d3a0ae824378?source=rss-947c603be518------2</link>
            <guid isPermaLink="false">https://medium.com/p/d3a0ae824378</guid>
            <category><![CDATA[reference-type]]></category>
            <category><![CDATA[value-type]]></category>
            <category><![CDATA[java]]></category>
            <dc:creator><![CDATA[omerfeyzioglu]]></dc:creator>
            <pubDate>Fri, 21 Oct 2022 16:06:57 GMT</pubDate>
            <atom:updated>2022-10-21T16:06:57.670Z</atom:updated>
            <content:encoded><![CDATA[<p>There are 2 kind of data types in Java, Primitive and Reference .Reference type can be named non-primitive.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*9URkMr1CV-X8P_oJ.png" /></figure><p>Primitive data type : it just specifies the size and type of variable values.</p><p>— When we send a variable we created in primitive data type to a method, the value of variable is sent and we can operate on this variable with operators.</p><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/0*5j16m-3UtLEaCurv.gif" /></figure><p>Reference data type : it stores the addresses of objects in memory. Therefore, it doesn’t specify any value of variables.</p><p>— We cannot operate with operators in reference data types, we use the methods of the relevant class to operate these types.</p><blockquote>As we know “int” is a value data type. Therefore, when “number2” is equal to “number1”, number2 is directly defined as “numbers1” s first value that is 10, since it does not take up space in the address. Although we print “number2” after changing “number1” to 50, “number2” will be printed as 10.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iwIci5o8t59wO99xTIsHxA.png" /></figure><blockquote>As for reference type, things change completely. Array “[]” is a reference data type. We defined an array which has elements 1,2 and 3.And we defined another array which has elements 4,5 and 6. “numbers2” (second array) is equal to “numbers1” (first array). We changed “numbers1”s first element as 10 and “numbers2”s first element is changed as 10 too because of array is a reference type. Therefore, “numbers2”s first element will be printed as 10.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*plTJlIdj1O-HiyweVl8siw.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d3a0ae824378" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>