<?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 ezgineer on Medium]]></title>
        <description><![CDATA[Stories by ezgineer on Medium]]></description>
        <link>https://medium.com/@ezgineer?source=rss-6ba83902d090------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*rH-UFkinu6eU9DVvyt7Sqw.jpeg</url>
            <title>Stories by ezgineer on Medium</title>
            <link>https://medium.com/@ezgineer?source=rss-6ba83902d090------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 11:34:17 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ezgineer/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[Usable Security and Privacy: Introduction]]></title>
            <link>https://medium.com/@ezgineer/usable-security-and-privacy-introduction-d676abc8c61d?source=rss-6ba83902d090------2</link>
            <guid isPermaLink="false">https://medium.com/p/d676abc8c61d</guid>
            <category><![CDATA[humancomputer-interaction]]></category>
            <category><![CDATA[usable-security]]></category>
            <category><![CDATA[security]]></category>
            <dc:creator><![CDATA[ezgineer]]></dc:creator>
            <pubDate>Mon, 13 Jul 2020 12:28:01 GMT</pubDate>
            <atom:updated>2020-07-20T12:56:29.911Z</atom:updated>
            <content:encoded><![CDATA[<blockquote><em>Humans are incapable of securely storing high-quality cryptographic keys, and they have unacceptable speed and accuracy when performing cryptographic operations (They are also large, expensive to maintain, difficult to manage, and they pollute the environment. It is astonishing that these devices continue to be manufactured and deployed. But they are sufficiently pervasive that we must design our protocols around their limitations).</em></blockquote><blockquote>Kaufman, Perlman and Speciner, 2002</blockquote><p>The above quote from a book about <a href="https://books.google.co.uk/books/about/Network_Security.html?id=m9pQAAAAMAAJ&amp;redir_esc=y">Network Security</a> embodies the goal of a discipline called Usable Security and Privacy (USEC). In 1999, two papers called “<a href="https://www.usenix.org/conference/8th-usenix-security-symposium/why-johnny-cant-encrypt-usability-evaluation-pgp-50">Why Johnny can’t encrypt</a>” and “<a href="https://discovery.ucl.ac.uk/id/eprint/20247/2/CACM%20FINAL.pdf">Users Are Not the Enemy</a>” were published, which started raising questions about usability of security systems. This area deals with very similar problems with Human Computer Interaction (HCI), except the problem space is well defined around security and privacy. This includes authentication, permission setting, phishing, software updates, warnings (which this post has examples of), key management, privacy policies, pairing and mobile security. The reason it’s split from HCI is because effective communication is especially important in security, if you intend to do no harm. Human factor is the center of the research in this area, because even if you design the securest system in the world, if you have the human in the loop, it boils down to their behaviour how secure your system is. It’s easy to blame the user for their action, but are you <em>responsibly</em> designing your system for its intended purpose?</p><h3>What does “human-in-the-loop” really mean for secure design?</h3><p>I took the below diagram from a quite interesting <a href="https://www.usenix.org/legacy/events/upsec08/tech/full_papers/cranor/cranor.pdf">paper</a>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/702/1*WukcJFOGEkxy77TEJls9Sw.png" /></figure><p>What does the above diagram tell us? For example, imagine we are evaluating the below warning according to the above framework (I’m not going to explain everything in that model for the sake of a shorter blog post):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/0*FkGjslkWv-fr6sBh" /></figure><p><strong>Communication: </strong>The above Starbucks cup is trying to communicate that the contents of itself is very hot and customers who are drinking it should proceed with caution.</p><p><strong>Communication impediments: </strong>The text is illegible because you spilled some drink on it or the cup holder thingy/your hand is in the way.</p><p><strong>Human receiver:</strong></p><ul><li>You were concentrating at your code before having your first morning coffee sip, so you didn’t pay attention to the warning that much <strong>(attention switch/maintenance)</strong>.</li><li>You drank a McDonald’s coffee earlier which had the same message on it, and it wasn’t hot at all <strong>(knowledge transfer)</strong>.</li><li>You think that warning is there simply because Starbucks wants to avoid lawsuits <strong>(attitudes and beliefs)</strong>.</li><li>The translation, which happens to be the only language you understand in that photo, is not as powerful and alarming as the English one (food for thought for i18n supporters out there!) <strong>(demographics)</strong>.</li><li>You just… <a href="https://www.reddit.com/r/mildlyinteresting/comments/6vk476/i_put_my_coffee_cup_in_the_microwave_and_it/">don’t live by the rules</a> <strong>(motivation)</strong>.</li></ul><p><strong>Behaviour: </strong>You do not proceed with caution, and burn yourself.</p><p><strong>What could the cup do better to prevent this behaviour? </strong>I don’t really know (what an anticlimactic ending… I bet you were waiting for me to list possible fixes 😅). That’s what user studies are for. It’s not necessarily a badly designed warning either. It shows how warnings can be ineffective due to human factor, though, even if the intention is clear for the warning writer.</p><p>Here’s another example of a warning, more related to tech:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/253/1*vBuzaqH3uNk_nCwUNVd2gQ.png" /></figure><p>Which essentially means this for the end user:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/253/1*NsIbiRYWRaLggZl7Xh7kzA.png" /></figure><p>What’s Akamai? What mismatch? Who is trying to intercept me? What would happen if they did? Why are you even showing me this if it’s unlikely?</p><p>Sometimes we work on something for so long as engineers or product owners that it becomes very obvious what we are talking about to ourselves, so we do not communicate effectively with our users. Note that the user does not have to be travelers in our case — my product’s users are the service owners in Skyscanner, because I am from the infrastructure team. I also write supposedly non-technical communications, and I’m pretty sure I sound like the second window above (something something we served 5xxs instance healthcheck cattywampus bumbershoot).</p><p>Anyway, back to the topic at hand…</p><h3>How do we go with designing secure communication?</h3><p>Microsoft has (had?) a <a href="https://adam.shostack.org/ReederEtAl_NEATatMicrosoft.pdf">Usable Security team in Microsoft Trustworthy Computing</a>, which acts as a central source of truth for secure UX design. They came up with a guidance called NEAT for their internal designers to use. Since the E part is very important, they further split it up to SPRUCE (formerly CHARGE US, which I guess didn’t gain enough traction).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/809/1*2lfaDeSeuV4xDpV8KhA4qw.png" /></figure><p>Let’s see NEAT and SPRUCE in an example:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/489/1*IcOEUw7aZ_rVJcfmjMKkzg.png" /></figure><ul><li><strong>N</strong>ecessary? (read: is involving the human <em>absolutely</em> necessary? Can we please not?): Seems like a yes, see below heading*.</li><li><strong>E</strong>xplained?</li></ul><p><strong>— S</strong>ource: Firefox.</p><p><strong>— P</strong>rocess: “What can you do about it” section asks the user to check the correct date in system settings and refresh the page. If it’s correct, it states that there is no further action except notifying the web page’s admin.</p><p><strong>— R</strong>isk: “Attackers could do bad things”</p><p><strong>— U</strong>nique knowledge user has: “Your computer clock is set to 11/03/2020 — is that correct? Can you check?”</p><p><strong>— C</strong>hoices: Going back is <em>clearly</em> recommended, both with colour and text, and proceeding is hidden in the “advanced” button</p><p><strong>— E</strong>vidence: If the user is making an informed decision, they should know some information that can factor into their decision, e.g. the owner of the website. I guess it’s not really easy to prove that the site they are trying to connect to is legitimate and trusted if the certificate is expired.</p><ul><li><strong>A</strong>ctionable? The user can choose “go back” or “advanced” (Would they choose the right action? No matter if the website is benign or malicious?)</li><li><strong>T</strong>ested? This spotlights the human-in-the-loop framework again, especially the intention bit. Testing includes visual inspection by many eyes in the company from diverse fields and formal usability testing.</li></ul><h3>*Why do we give the user the option to proceed with the unsafe choice at all?</h3><p>Sometimes it’s not realistically possible to prevent the unsafe choice, as in the coffee example above. Other than that, this is somewhat a philosophical question, because this research field is dealing with humans which are “difficult to manage”. I believe that the human nature would get in the way (remember “attitudes and beliefs” from the above framework). If you do not allow people to perform an action, humans would try to find a loophole to perform it anyway, which would be even less secure than what you’re trying to prevent originally.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/599/1*9aC1bYyM-hTzuXev-qSTMA.png" /></figure><p><strong>That’s all from me for the introduction!</strong> If you’re interested in this sort of studies, there are conferences on Usable Security (e.g. CHI, SOUPS, USEC, EuroUSEC) or check out <a href="http://vaniea.com/">vaniea.com</a> (Kami Vaniea, the USEC/HCI lecturer I took this class from) to see what she and her current lab TULIPS have been working on.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d676abc8c61d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Positive Discrimination in Tech: Good or Bad?]]></title>
            <link>https://medium.com/@ezgineer/positive-discrimination-in-tech-good-or-bad-c018cafb3e77?source=rss-6ba83902d090------2</link>
            <guid isPermaLink="false">https://medium.com/p/c018cafb3e77</guid>
            <category><![CDATA[women-in-stem]]></category>
            <category><![CDATA[diversity-in-tech]]></category>
            <category><![CDATA[discrimination-positive]]></category>
            <category><![CDATA[affirmative-action]]></category>
            <category><![CDATA[women-in-tech]]></category>
            <dc:creator><![CDATA[ezgineer]]></dc:creator>
            <pubDate>Sun, 19 Jan 2020 22:20:16 GMT</pubDate>
            <atom:updated>2020-10-23T09:30:27.022Z</atom:updated>
            <content:encoded><![CDATA[<h3>Positive Discrimination in Tech: Opinions</h3><p><em>This is a highly opinionated post, and depends so much on personal experience. That’s why it usually focuses on women in tech, not the other minorities. If you think that anything I am saying here is wrong or if I am missing a point just leave a comment and we can discuss.</em></p><p>As a Turkish woman in the tech industry, I sometimes cannot help but think that I am the token hire, although most of the workplaces I’ve ever worked for have not treated me as such, at least not on purpose (I am very lucky!). This does not help my <a href="https://en.wikipedia.org/wiki/Impostor_syndrome">imposter syndrome</a> at all, since I tend to think that if I did not tick the right boxes in the equal opportunity section, I would not be given an interview. If you do not get an interview then it’s actually worse, since people make dumb comments such as “How did you not get one? Every other woman in the university did”. I feel like I have a responsibility for both Turkish people and women. I am scared of making a mistake because I would not want people to stereotype others as “oh the Turkish, always so lazy” or “yeah, she is the diversity hire <em>(whatever that means)</em>, let’s not give her a hard time”.</p><h4>Despite the title of this post, positive discrimination is such a gray area.</h4><p><strong>It is needed, unfortunately, until we find a better solution.</strong> I cannot tell you how many times I’ve been in a tech related thing or a school or a job as one of the handful women and everybody just instantly assumed that I am studying sociology or an HR person or from marketing, since we are so rare.</p><p><strong>It should not make the person feel like they are a makeweight, though. </strong>Furthermore, I don’t think it actually is helping the root cause of why people are getting unequal access to opportunities in the first place.</p><p>The equal opportunity section should be removed from the CV elimination stage, or the section should be worded in an alternative way. If the diversity of a company needs to be monitored, it should be done <em>after</em> the person is hired, right? There are definitely good enough non-male and non-white people who are worth their CVs to be looked at. The name and photo of the person should be removed from the CV (which many diverse and large tech companies do at this point). If there are applicants who have similar qualifications, only then taking an affirmative action should be considered.</p><p>Instead of having this section to prove you are an equal opportunity employer, sharing successful minority stories from your company and using inclusive words in your ad is way more valuable. A company should not make people think that they are hiring a minority just for the sake of hiring a minority.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/910/1*ZGJQfSVDUbSA6bhGl7aezg.png" /><figcaption>A very straight-forward “equal opportunity” section. Although the title says “monitoring”, no information is there about whether this will be used in the hiring process or not.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*edfH4TQvrXW3AELxQcb-lg.png" /><figcaption>A better way to word the equal opportunity section (although change the word gender to sex, maybe?)</figcaption></figure><p>The scholarships given to women such as Anita Borg should stay. I also support Young Girls in Code type of events <em>to some extent</em>, since it is important to raise girls from a younger age and maybe give them a bit of a head start before they go to the university, especially in the areas where women have significantly less access to opportunities, such as Middle East. However, I think having a separate Women in Code event is like having bantamweight boxing. I believe that these kind of events polarise people even more, and allow people to not leave their safe spaces. Also, directly quoting a friend: “it also doesn’t prepare (women) for the fact that (unfortunately) they’ll be likely working in a male dominated environment.”</p><p>Women should not feel like they are not good enough to have coding events together with men. We should incentivise women to code or compete in other ways, such as different swag or targeted ads towards women. I swear, in one or two of the hackathons I have been to, the women’s bathroom was not even operating since there weren’t any women, which is super sad! We should investigate the root causes for this and come up with solutions.</p><p>(By the way, if you are facilitating/organising a Women in Code type of event, please don’t hate me after reading this post! I still think it is valuable since in the end more women are coding, but I am just saying that it is not the optimal and the long term solution. The effort could go into encouraging women in mixed workshops instead).</p><p>Another strong opinion: companies should not drop the word “diversity” every time they can when they talk about how awesome their company is. Of course diversity should be celebrated, but it should not feel forced and the company should be honest about it. Having surveys about how people feel about diversity, sharing the results of them and trying to find a solution to any negativity is definitely more valuable than having bleak numbers and percentages. I personally do not want to be a pixel of a color in a pie chart, I want to be hired because my opinions and work bring value to the business. My current employer is doing an awesome job at this, so kudos to them.</p><p>Final motivational push, also to myself I guess: we should not care about what people think of us and give ourselves a break. There will always be people who will have either conscious or unconscious bias, no matter how the world is trying to change that. We are awesome, and these companies obviously would not want to hire us if we were untalented. Everybody has a different way of thinking, and having diverse people coming from different cultures is so valuable, for both the differences in our way of technical thinking but also ideas we have about the products we are working on.</p><p>Keep on doing what you do!</p><p><em>(shameless self plug: </em><a href="https://ezgineer.wordpress.com/2015/12/22/women-in-computer-science/"><em>here’s the post</em></a><em> I wrote in 2015 about why the tech industry might be male dominated)</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c018cafb3e77" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>