<?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 Geng yong on Medium]]></title>
        <description><![CDATA[Stories by Geng yong on Medium]]></description>
        <link>https://medium.com/@coolgeng?source=rss-85f5f94b9996------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*BaATFl-bLljIQvd5.jpg</url>
            <title>Stories by Geng yong on Medium</title>
            <link>https://medium.com/@coolgeng?source=rss-85f5f94b9996------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 04:08:15 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@coolgeng/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[Tipranks]]></title>
            <link>https://coolgeng.medium.com/tipranks-45819206bcc?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/45819206bcc</guid>
            <category><![CDATA[tipranks]]></category>
            <category><![CDATA[stocks]]></category>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Mon, 25 Jan 2021 23:15:51 GMT</pubDate>
            <atom:updated>2021-01-25T23:15:51.953Z</atom:updated>
            <content:encoded><![CDATA[<p>How to use tipranks to pick stocks</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=45819206bcc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to design a twitter]]></title>
            <link>https://coolgeng.medium.com/how-to-design-a-twitter-791890ec0536?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/791890ec0536</guid>
            <category><![CDATA[interview-questions]]></category>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Wed, 06 Jan 2021 06:42:11 GMT</pubDate>
            <atom:updated>2021-01-06T08:33:07.152Z</atom:updated>
            <content:encoded><![CDATA[<p>Interview questions</p><p>Requirements:</p><ol><li>Tweet</li><li>Timeline/Feed</li><li>Follow a user</li><li>Like a tweet</li><li>Search tweets</li></ol><p>TO DO MORE</p><p>postTweet(userToken, String tweet)</p><p>deleteTweet(userToken, String tweetId)</p><p>likeOrUnlikeTweet(userToken, String tweetId, bool like)</p><p>(maybe can support different moji like suprise or sad etc, then change bool like to int effect)</p><p>readHomeTimeLine(userToken, int pageSize, String pageToken)</p><p>readUserTimeLine(userToken, int pageSize, String pageToken, String userId)</p><p>Pull mode:</p><p>Push mode:</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=791890ec0536" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Stock invest list]]></title>
            <link>https://coolgeng.medium.com/stock-invest-list-c6dff9d70695?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/c6dff9d70695</guid>
            <category><![CDATA[stock-market]]></category>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Wed, 16 Dec 2020 08:29:32 GMT</pubDate>
            <atom:updated>2020-12-16T08:29:32.993Z</atom:updated>
            <content:encoded><![CDATA[<ol><li>NIO</li><li>XPEV</li><li>API</li><li>VLDR</li><li>TDOC</li><li>REAL</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c6dff9d70695" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Docker and Kubernetes (1)]]></title>
            <link>https://coolgeng.medium.com/docker-and-kubernetes-1-a4fbae96fffc?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/a4fbae96fffc</guid>
            <category><![CDATA[docker]]></category>
            <category><![CDATA[kubernetes]]></category>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Wed, 24 Jul 2019 07:06:01 GMT</pubDate>
            <atom:updated>2019-07-24T07:06:01.229Z</atom:updated>
            <content:encoded><![CDATA[<p>Talking about the history of Docker and Kubernetes</p><p>In 2013 year, AWS and OpenStack is the most popular technology in the cloud computing area. But Cloud Foundry is an exception and bring some new idea about the cloud PaaS project.</p><p>At that time, if you ask anyone who are working on cloud computing industry. They definitely will tell you PaaS time are coming.</p><p>The core feature of Docker is Docker image which is to compress the OS ISO file and application together which Cloud Foundry doesnt have this before. In another words, most of PaaS application don’t support this. I think that’s the reason Docker project become so popular in couple months and other PaaS projects exit the competition.</p><p>The Docker image can make sure the local OS and application can be the exactly the same when operation engineer deploy the code to production environment which not like other PaaS project engineer need to prepare the special script to deploy to cloud. Sometimes engineer only can debug on cloud side since it’s still a little bit different from the local environment.</p><p>Docker image change the development process as well. Software development can be done in the sandbox which created by Docker image.</p><p>Also Docker use the same concept to create sandbox which come from Cgroups and Namespaces mechanisms.</p><p>The “Drawdown Strike” that the Docker project brings to the PaaS world actually provides a very convenient packaging mechanism. This mechanism directly packages the entire operating system required for the application to run, thus ensuring a high degree of consistency between the local environment and the cloud environment, and avoiding the pain of the user trying to match the differences between the two different operating environments through “trial and error”.</p><p>Some small companies start to provide the support Container-as-a-Service like Deis and Flynn. They call themselves CaaS.</p><p>At the end of 2014, DockerCon voluntarily released its own “Docke native” container cluster management project Swarm, which not only pushed this wave of “CaaS” to an unprecedented climax, but also pinned the whole Docker redefines the grand wishes of PaaS.</p><p>In the peak years of 2014, Docker is really one step away from its ideals.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a4fbae96fffc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Some useful command in iTerm]]></title>
            <link>https://coolgeng.medium.com/some-useful-command-in-iterm-8b748ad5195b?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/8b748ad5195b</guid>
            <category><![CDATA[linux]]></category>
            <category><![CDATA[iterm2]]></category>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Sun, 14 Jul 2019 08:31:37 GMT</pubDate>
            <atom:updated>2019-07-14T08:33:09.956Z</atom:updated>
            <content:encoded><![CDATA[<p>Same command on all panes.</p><ol><li>If you find the need to enter the same commands into multiple panes at the same time, simply hit “⌘Command + ⇧Shift + I”.</li></ol><p>Continue updating ….</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8b748ad5195b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Linux command line output]]></title>
            <link>https://coolgeng.medium.com/linux-command-line-output-4b507a41d15f?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/4b507a41d15f</guid>
            <category><![CDATA[linux-command-line-basics]]></category>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Wed, 24 Apr 2019 18:26:29 GMT</pubDate>
            <atom:updated>2019-04-24T18:26:29.285Z</atom:updated>
            <content:encoded><![CDATA[<p>&gt;/dev/null redirects the command standard output to the null device, which is a special device which discards the information written to it</p><p>2&gt;&amp;1 redirects the standard error stream to the standard output stream (stderr = 2, stdout = 1). Note that this takes the standard error stream and points it to same location as standard output at that moment. This is the reason for the order &gt;/some/where 2&gt;&amp;1 because one needs to first point stdout to somewhere and <em>then</em> point stderr to the same location if one wants to combine both streams in the end.</p><p>In practice it prevents any output from the command (both stdout and stderr) from being displayed. It’s used when you don’t care about the command output.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4b507a41d15f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Ray VS Dask]]></title>
            <link>https://coolgeng.medium.com/ray-vs-dask-d0154a774f2a?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/d0154a774f2a</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[uc-berkeley]]></category>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Fri, 04 Jan 2019 00:21:38 GMT</pubDate>
            <atom:updated>2019-01-04T00:21:38.854Z</atom:updated>
            <content:encoded><![CDATA[<p>Get some comments from the article:</p><p>Dask also aims to improve the ecosystem for parallel/distributed Python. There are lots of practical differences and design decisions.</p><p>Dask supports something very similar to our remote functions. The actor abstraction doesn’t have an equivalent and is pretty important for building stateful services like parameter servers.<br>- Ray uses a distributed scheduling scheme to allow high task throughput (e.g., millions of tasks per second), whereas Dask uses a centralized scheduler.<br>- Ray focuses a lot on latency, so the latency to submit a task and get the result is about 30x lower in Ray than in Dask.<br>- Ray has focused more on libraries for machine learning and reinforcement learning, whereas Dask has built more distributed collections libraries (distributed arrays, DataFrames).<br>- We handle data differently from Dask (using shared memory and zero-copy serialization). Dask does serialization with pickle (with some optimizations).<br>- Ray handles failures (e.g., transparent recovery from machine failures).</p><p>I will look into these difference between these two projects.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d0154a774f2a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[面试算法]]></title>
            <link>https://coolgeng.medium.com/%E9%9D%A2%E8%AF%95%E7%AE%97%E6%B3%95-f3d1b9c1f0eb?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/f3d1b9c1f0eb</guid>
            <category><![CDATA[美国经历]]></category>
            <category><![CDATA[计算机]]></category>
            <category><![CDATA[内功修炼]]></category>
            <category><![CDATA[面试]]></category>
            <category><![CDATA[代码人生]]></category>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Thu, 03 Jan 2019 05:49:08 GMT</pubDate>
            <atom:updated>2019-01-03T05:49:08.803Z</atom:updated>
            <content:encoded><![CDATA[<p>在我看来准备算法是必要的环节，因为作为一个程序员你的内功就是算法和数据结构，真的在你去看很多核心代码的时候，会发现很多高级算法在里面，但是面试的时候基本上只需要一些基本的算法，因为一个是时间的原因，你不可能去实现很复杂的算法，太过于复杂的算法使用场景不是经常遇到，所以也不能考察一个程序员的基本功，但是这些基本的算法却至少可以过滤掉一些基本功不过硬的程序员。</p><p>这里面有 10 个数据结构：数组、链表、栈、队列、散列表、二叉树、堆、跳表、图、Trie树；</p><p>10个算法：递归、排序、二分查找、搜索、哈希算法、贪心算法、分治算法、回溯算法、动态规划、字符串匹配算法。</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f3d1b9c1f0eb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[July 27 2018]]></title>
            <link>https://coolgeng.medium.com/july-27-2018-2a3befc56614?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/2a3befc56614</guid>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Fri, 27 Jul 2018 07:36:43 GMT</pubDate>
            <atom:updated>2018-07-27T07:36:43.337Z</atom:updated>
            <content:encoded><![CDATA[<p>Start to learn <a href="https://www.cs.cornell.edu/projects/ladis2009/talks/dean-keynote-ladis2009.pdf">https://www.cs.cornell.edu/projects/ladis2009/talks/dean-keynote-ladis2009.pdf</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2a3befc56614" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Get ready]]></title>
            <link>https://coolgeng.medium.com/get-ready-b6675488b10b?source=rss-85f5f94b9996------2</link>
            <guid isPermaLink="false">https://medium.com/p/b6675488b10b</guid>
            <dc:creator><![CDATA[Geng yong]]></dc:creator>
            <pubDate>Mon, 23 Jul 2018 21:51:50 GMT</pubDate>
            <atom:updated>2018-07-23T21:51:50.004Z</atom:updated>
            <content:encoded><![CDATA[<p>Everyday I plan to go throught one of the topics in the list below:</p><p><a href="https://github.com/binhnguyennus/awesome-scalability">GitHub - binhnguyennus/awesome-scalability: The Patterns of Scalable, Reliable, and Performant Large-Scale Systems</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b6675488b10b" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>