fryol

Curiosity Trip
Comments: (0) - Author: Rishi

Chatting with my 5 yr old, I realized that she had a particular resistance to the words “I don’t know.” It was an indication that I had to put something to an end for her, and “I don’t know” if it was fear or something else. I just don’t know. What I knew was, she had to become comfortable.

We started with the concept of asking “why” more. She prides herself heavily in knowing answers, and like most kids, takes a lot of joy in the praises therefore. Once we settled, she quickly asked me, “why is it ok to say I don’t know?” My job was half done, and I was sufficiently pleased with Vasvi. If anything, I told her, “you will never know everything, I don’t know the answers to most questions.” For brevity here, I did explain that “I don’t know” is the beginning of knowledge. There is no end, and when no one knows the answer, you have the intense potential to make meaningful impact. The trick is to know what is already known to get to the limits of available knowledge.

The conversation was sparked when we read an article on a quote from Einstein, that was about spending most of the time on a problem, than thinking about its solutions. In a recent interview, this was my biggest peeve. The candidate jumped to solving the problem in the first few minutes, without analysis or questions on the problem itself. At the end of the interview, when the algorithm chosen didn’t solve the corner cases of the data structure provided, I did mention, “I wish you asked me about ‘why’ I was even posing this problem.”

Most people presume that the world around them is in some harmony. This is what I want to prevent my daughter from sliding into. The harmony is a good thought, not exactly reality. Making your kids understand that the last generation doesn’t exactly have their ducks in a row, is what drives us forward. The sooner they get it, the more curious they become.

Ultimately, curiosity evolves the cat. A couple of hours I spent chatting with my father this weekend brought back some intense curiosity about how do ATM machines verify currency – that was easy to answer. What stumped me was, “how did they do it in 1980s? There were no scanners.” After a minute of googling it, the term ‘magnetic ink’ lit the bulbs. One thing led to another and I had a huge spike of curiosity, meaningful enough to keep the rest of the weekend interesting.

The only challenge is, curiosity to me, is not to be driven in the direction of what cannot be known because it is dependent on human reaction. In other words, if sound and unemotional logic cannot be tied back to a reason, the trip is unworthy. Vasvi agreed, but did not ask why. That is a discussion for to tomorrow, and it may turn out that she enlightens me to prove me wrong.

Mars Mission Decoded To A Four Year Old
Comments: (0) - Author: Rishi

Vasvi declared that she wanted to go to Mars, and I had to decode a lot of things for her to explain the differences between San Francisco and where she wants to go. The best way to overcome challenges is to face them and start thinking about how this generation needs to prepare.

Feature Simple Speak
Gravity (3.71 m/s2) You’ll walk as if you’re light-hopping, you’ll weigh only 1/3rd; Jumping from a rock will feel slow motion (you’ll take about thrice as more time to fall)
Diameter (6794 km) Will take half the time to get around Mars, map scales will be larger
Sun’s luminosity (40%) Brightest days will look like very early mornings
Atmospheric Pressure (600 Pa) Water will boil at about 0 degrees C, right where it freezes here on Earth (see this table)
Temperature (-60C) Really cold as compared to most of earth
Two Moons As fascinating as this sounds, these are really very small rocks around Mars, as compared to our moon and have minimal effect on Mars. In reality, you can’t see a large moon on Mars (its pretty empty).
Orbital Period (687 Days) Summers will last about 6 months, and every other season will be longer than here. Calendars will have longer months and longer years there.
Life (unknown) There are no trees, no bees, no butterflies, no birds and no flowers. Its pretty empty from a living creature perspective.

To a child’s mind, the differences are not as extraordinary as it is to adults. They are still linking to what is supposedly normal.

Getting There

It was not funny for me any more after Vasvi repeated that she is going to Mars, and will on the way stop on Moon and Venus. I asked her what challenges will she have to overcome to get there. She had no idea. I decided to document some challenges in simple speak for a Vasvian brain.

What Simple Speak
Distance (60 to 400 million km) Mars and earth are constantly moving around. It would be best to start when Mars is closer to Earth. If I call you when you’re around Mars, my “hello” will take anywhere between 5 and 20 minutes to reach you. It is that far.
Launch Window About every two years, Mars and Earth are somewhere close that it will be fastest to get there. Our next launch window is Mar/2016. (cosmic schedule)
Packing Food, water, recycle gadgets (to convert gases and liquids back to being usable), space suits, solar panels/batteries, a telephone, an exercise machine and your favorite Alouette..
Delta-v Budget How much money we will spend? This will depend on how much “gas” (propellent) is required, and how fast we can go. We’ll have to use the gravity (or the pull force) of other planetary bodies to gain speed (gravity assist), helping reduce the money we have to spend.
Spacecarft A big giant car, with lots of buttons inside. This is where you will live for about 7-8 months, so it should be nice inside.
Radiation Like very bright sunlight bothers, there are other types of “lights” that are dangerous and can be harmful. Sunscreen lotions can’t protect, so your spacecraft will need to have special outer coatings to protect from this radiation. Otherwise, it can be similar to you being inside a microwave
On the way It will be dark outside, and the sun would be suddenly very bright. Lots of stars to see, the moon will show itself on the way, and depending on when the launch date, Venus could be seen from close enough. Earth will look like a blue ball, that gets smaller and smaller, every time there’s actually a chance to see it. Lots of exercise will be needed on the way so that bones and muscles stay strong.
Mars Orbit Go around Mars, round and round, instead of landing down. Mars’ air and atmosphere will help a slow down to stay above it. Going around the planet, mornings and nighs may be shorter – ranging from a few to many many hours of day and night. Going around will be strange, since it will be elliptical – very close and then very far. Two not so big rocks will be seen, one very close (Phobos) to Mars and another one further away (Deimos). Earth won’t be seen as it used to be seen. 

Landing

At this time, exhausted with the information overload, the decision was not to land on this trip. We’ll land some other time, time to come back.

Coming Back

This is another complex task that I couldn’t get any reliable data for, at least as of this writing. Challenges of coming back had to be simple-spoken.

Homeward Simple Speak
Distance (60 to 400 million km) The same distance needs to be covered as earlier. At best, the journey can start only about a year after it was accomplished, since launch windows occur every two years, remember?
Launch Window The next launch window from Mars will be in Feb/2018 – about 15 months after reaching there. This will mean a lot of waiting while going around the planet.
What to pack Just eat very slowly, don’t waste any chocolate bars and keep calm.
Delta-v Budget Significantly higher amount of money is needed to pack double the fuel for getting back (to achieve the same speeds). Multiply everything you want to get there by 2. This may not be completely true if you are above Mars.
Spacecraft Much bigger spacecraft, since you will be taller and supposedly heavier, you need more playground. This increases the delta-v budget, meaning more money.
Psychological Impact Again, it will be very dark outside. Time will seem like it will never end. Feelings of “no one cares” will develop. Alouette, of everyone else, may suddenly seem like the only one who cares and loves. This will be hard.
Close to Blue Being back on Earth doesn’t look any further. Sitting in a space capsule to fall into an ocean sounds easy, but an additional space capsule in the craft is needed. More delta-v. Alternatively, achieve a low earth orbit and then someone comes to pick up. Super-alternatively, somehow attach to the International Space Station – though no one would like the possibility of destroying a $100 billion thing if something went wrong.

Will we get there?

This was the big question. After all this explaining over a few days, this question always rang with a super-optimistic spin at the end – “Yes?” I am just as optimistic as the little one. Yes, we’ll get there, and possibly, some would make it back.

Interviewing Right
Comments: (0) - Author: Rishi

There are countless online articles that get hits when someone is preparing to get interviewed: “10 things to keep in mind”, “10 don’ts while interviewing”, “How to interview well..”, and so on. This works well for the times we live in. You get people in the door for a typical 4-6 hour interview, they act nice and everyone tries to play by the rules.

What’s wrong here?

Interviewing is a tough skill. Think about the things that the interviewer is doing, within a 45-60 minute window:

  • assessing technical capabilities
  • assessing success potential
  • assessing seniority
  • assessing team work capabilities
  • assessing <insert buzzword here>

That is a lot to assess. Relevant signals from the candidate can get lost. Depending on the interviewer, they can become too critical or too positive and amplify signals that are one-offs for the candidate, or play down those that give more insight into a candidate’s nature.

Additionally, the clock. Less than an hour to go, and so much to discover in an interview. Possibly, more time is spent on the right car to buy (reviewing, comparing, discussing etc.), than on a likely co-worker.

Then again, not all interviewers will take extensive notes. This leads to lost information, and sometimes, to distorted memory about the candidate. A common trap to fall into is the gist-based memory and associative memory errors (see this article).

Interviewing skills get better with experience. Not being true for everyone, some are good at assessing others while others will just never be good at forming reliable opinions in small time windows. Since there is a learning curve here, it will have to unfortunately come at the cost of hiring misfires.

Given that interviewing is conducted by a panel, the idea rooted here is that judgments passed by a group are less likely to be at fault. Though, more senior or well-respected interviewers tend to dominate the jury.

Referrals by current employees, of past co-workers, is a popular hiring method. Referrals (mostly) get interviewed in the standard ways.  Yet, employee referrals have an added data point. This makes a  difference.

To summarize,

  • interviewees may not be themselves
  • too many skills / behaviors to assess
  • not enough time to appropriately assess
  • not enough record of the interviews conducted
  • interviewers gain experience at a cost
  • interview panels may be dominated by seniors
  • referrals get a bias

Avoidances

What follows is a list of some of my personal practices to avoid these problems.

Pre-interview notes on the candidate

  • Decipher the candidate’s resume, to highlight their strengths, !strengths and anything else of note to brief the interviewers prior to the interview. Your help with this will set the stage and prime interviewers’ experience with the candidate. It is important to get it right and not (bias it).
  • Job descriptions help, but often, the panel needs to be reminded of what exactly is it that you are looking for. Think outside the job description – personality traits, seniority levels (for real), specific personas (like someone with a customer oriented mindset, or someone who’d be more heads down).
  • Keep “referral” biases out, as much as possible. Instead, talk to the person who referred this candidate, and include notes in your pre-interview notes.

Develop focus areas for interviews

  • As a hiring manager, a lot of thought needs to go into the kind of person needed on the team for a position. Some times, when you are building a team, you need to create a few different personas.
  • Every interviewer gets a couple different focus areas. The tricky part is to match the focus area for the interviewer as well. I typically read through past interviews from these interviewers to determine what they naturally focus on.
  • Although interviewers give additional feedback, appreciate it, but remind them to “concentrate” on their areas
  • Keep them real. Write out focus areas for what the team is actually doing over 3-6 months, instead of asking for things that lead to optimizing the number of triangles within 5 overlapping circles.

Share focus areas with the candidate

  • Well, why  not? You want to earn the trust of the candidates so that they can be themselves.
  • What if they make up things? If someone can make up a fantasy about an architecture that they never worked on, or only read about, and explain it well, you probably want someone like that.
  • Won’t everyone just know what we ask? Every interviewer would have their own set of problems and will be discussing very different scenarios. The idea is to make the candidate comfortable and forthcoming.

Make note-taking easy

  • I’ve printed out focus areas in the past, on paper, and given a page of blank space that has multiple of following markers (to interviewers) under each focus area:
    • Asked – what the interviewer asked. 
    • Response –  what the candidate talked about.
    • Analysis – refers to notes that the interviewer is encouraged to write out their analysis for the asked/response pair (under that focus area).
  • Interviewers should be encouraged to go back and fill in the analysis after the interview is complete so that they can fill out as much information as they can before their cache expires.

Keep information flowing

  • Most importantly, during the interview. Having brief discussions while the interview is in progress with the interviewers, helps to adjust focus areas. Additional context is added for incoming interviewers to re-evaulate certain areas. Orchestrating the feedback so that it flows to and fro, helps.
  • The slight downside to this is that bias from earlier interviewers can flow to and fro too. Conscientious effort is required to avoid this.

 Post-interview debrief and go/no-go

  • Many organizations have a central committee that decides the outcome. This committee’s judgment is only as good as the feedback captured, the hiring manager’s overall note and the focus on what you need for the position v/s what the candidate has to offer.
  • Sitting down with all the interviewers for a brief amount of time and discussing everything with the panel is very important. It highlights the panel’s own performance, along with the interviewee’s.

Other Problems

Reading through interview reviews on sites like glassdoor, I’ve often seen many other problems that interviewees bring up. Common themes are, interviewers that…

  • …aren’t empathetic
  • …dominate airtime
  • …don’t give enough context
  • …lean into the candidates
  • …never smile & keep things tense

My approach is to find the best way on how to use such personalities (with these unique qualities), such that the interview process is not affected. Often, it means, not using them.

Final Note

After a few cycles of doing this rigorously, this was easier (mostly cloning prepared documents for ongoing interviews). In the beginning, I had detailed focus areas . Later on, I would compress the focus areas to be brief.

There is a lot that can be improved, and I’ve not yet seen all the pitfalls. There’s a possibility that some of the techniques above have serious downsides, but this has worked better than anything else thus far. Well, there’s definitely a better way…

This is all good “for the times we live in.” Interviewing today, is still constrained in a short time window that an interview is conducted in. The future of interviewing will probably not look like any of this. Take home assignments, short projects lasting 1-2 days, contractual work leading to permanent roles and may be even (crazy idea) separate organizations acting as frontiers for people who could to try things out.

I’ve been surprised a couple of times on how certain people who were hired, turned out to be really good at what they did despite shaky panel feedback and an almost “no hire”. There are probably many more that were never interviewed, not hired due to process flaws or hired but weren’t the right fit. This has to change.

Flying From Twitter
Comments: (0) - Author: Rishi

Twitter Journey

Twitter Journey

This last Friday, after just over two years at Twitter, I decided to drive down the street to Uber.

The most difficult part of this decision was the fact that I only spent, what seems like an extremely short time at Twitter. The experience was amazing – the methodologies, the technologies and above all, the wonderful people I’ve had the honor to work with.

Will truly miss the relationships I’m leaving behind, especially my immediate team (Aras, DRob, DavidJ, Jeff, Joe, Mike, Toby and Zeke). And then, there’s Ian – who has become a very close friend in this very short time.

Really looking forward to the next adventure.

Sanity
Comments: (0) - Author: Rishi

About 20 or so years ago, I had the privilege of having a guest at our place. Amazingly enough, we would almost always have dinner on the dinner table, unlike how I’m used to with my kids (on the couch, bed, standing and sometimes, on the table meant for this exclusive function).

This guest was from a part of India, now known for its exceedingly high PM2.5 (particulate matter). The air pollution itself is responsible for lung and neural diseases. Add to that the fact that you have sewage seepage into drinking water pipes and we’re talking serious long term damage to the brain and lungs.

Cut: back to the guest at our dinner table. As is customary, you serve your guests (with generous portions of food), and acting the good boy, I started doing so. Taking the lid off of one of the bowls, I kept it facing down on the table. “WHAT are you doing?” said the person, very perplexed and a bit chidingly. Looking confused, I asked, “What am I doing wrong?” – as that was obviously the question instead. “Don’t put that lid facing down on the table, ‘germs’ will get back into the ‘daal’ from the table when you re-lid it.”

This was a lesson that I haven’t unlearned since. The transfer of biological hazard from a dining table that looks clean, but really isn’t – through a lid’s internal surface. It made total sense. My father did laugh in a way that made me think that he disagreed with the phenomenon, but didn’t want to argue unnecessarily.

Many many years on, an offspring of this guest was infected with Typhoid. If you don’t know, the only reason you can get Typhoid, is when you have consumed human fecal matter. The only sane way for that to happen, is through contaminated food or water. The only places with the highest reported typhoid cases in the world, are Africa and India. Note, “reported.”

As I was reading an article on Delhi’s PM2.5 levels (close to 1000 now), and the mention of contaminated drinking water, I realized something.

People often obsess over things they’ve been brought up with. There is a lot more to be obsessed about, but the key learning for me and my kids is to constantly evaluate your obsessions. This guest’s freak out, made a lot more sense in 1970s may be, but there are way more things to worry about in 1996 – and the perspective changes even further, in 2015.

The last place to get bacteria in your food is from the dining table. FWIW, those bacteria are probably going to help strengthen your immunity. Worry about the lavatory more, or, if you’re in Delhi, about the sewage pipe running next to your water supply. Welcome to sanity.

Setup and Play Unreal Tournament 2004 Natively on Ubuntu 14.04 LTS (x64)
Comments: (0) - Author: Rishi

Recently learned from a colleague that UT2004 runs perfectly on Linux – natively. Figured from Ubuntu forums, that it does indeed work perfectly.

What is this post going to tell you that others don’t? How to buy UT2004 online, download it and set it up on Ubuntu (without having to worry about a CD, because it’s 2014 and not 2004).

  1. Install wine:
    sudo apt-get install wine
  2. Get wintricks:
    wget http://winetricks.org/winetricks
  3. Make it executable:
    chmod u+x ./winetricks
  4. Install steam:
    ./winetricks steam
  5. You’ll see something like this downloaded:
    /home/$USER/.cache/winetricks/steam/SteamSetup.exe
  6. Run setup with wine:
    wine /home/$USER/.cache/winetricks/steam/SteamSetup.exe
  7. If you didn’t change the path to install, launch steam with wine:
    wine ~/.wine/drive_c/Program\ Files/Steam/Steam.exe -no-dwrite

    (That -no-dwrite is important because the text disappears without that on Steam. We don’t want to use this for gaming any way, so this is fine).

  8. Create account or log in to your existing account on Steam.
    (space)
  9. Search for Unreal Tournament 2004.
    (space)
  10. You’ll get an option to buy UT for about $8 (real cheap right now) – buy it.
    (space)
  11. Download the game and install it (took me about 30 minutes).
    (space)
  12. Now copy (or move) this game from ~/.steam/drive_c/Program\ Files\ \(x86\)/ into another location.
    (space)
  13. Download the mega pack and linux patch bundle from http://www.utzone.de/forum/downloads.php?do=cat&id=19 (titled UT2004 Mega Pack Linux + LinuxPatch)
    (If this link is dead, search the Internets).
    (space)
  14. Overwrite the mega pack and patch files into this other location where you moved your game to.
    (space)
  15. Install this:
    sudo apt-get install libstdc++5
  16. Also:
    cp /usr/lib/x86_64-linux-gnu/libSDL-1.2.so.0.11.4 /Wherever/Your/Unreal/Dir/System/libSDL-1.2.so.0
  17. Finally, get the CD key by running:
    wine regedit
  18. Search for UT2004 (and you should find a key with the CD key in it)
    (space)
  19. Copy this CD key into a file named cdkey that is placed in /Wherever/Your/Unreal/Dir/System/
    (space)
  20. Launch /Wherever/Your/Unreal/Dir/System/ut2004-bin-linux-amd64
    (S
    ome of these steps will change if you are running this on a 32bit system.)
    (space)

(space)
I did not come up with all these steps, so I’m linking all the resources below in case you need additional help/instructions:

  1. https://developer.valvesoftware.com/wiki/Steam_under_Linux (for steps 1 through 6)
  2. https://bbs.archlinux.org/viewtopic.php?id=146223 (for “no fonts/text appearing” – the -no-dwrite in step 7)
  3. https://help.ubuntu.com/community/Games/Native/UnrealTournament2004 (for the mega pack installation steps 13 through 16 and 19)
  4. https://support.steampowered.com/kb_article.php?ref=1572-EOPC-1358 (for the CD key step 17)

 

 

 

 

 

 

 

Fryol Twitterized
Comments: (0) - Author: Rishi

Fryol is now integrated with Twitter to show up as a card summary in tweets. Took very little hacking to adapt Twitter meta tags to enable this and then get the domain approved at dev.twitter.com.

Beaming Out of Yahoo
Comments: (1) - Author: Rishi

After 7+ years, I am leaving Yahoo. Going to join “the flock” at Twitter.

This is big and then, not so big for me. My role and responsibility remains somewhat unchanged as I go from Yahoo to Twitter, but I’m moving to San Francisco (the movers finished that yesterday). So, living in the city will be new. I’ve not taken very kindly to the traffic there so far, but this will change I’m sure.

Yahoo is an amazing place to work for. I’ve repeated this, that I’m not that good at transactions. When it is not time to buy a house, I go and buy one. This, was probably not the best time to leave Yahoo. The share price is kicking and the products coming out are simply amazing. I can’t say here, but there are some amazing surprises on the way for people using Yahoo products. I don’t say this because I work(ed) there, I really mean it. If you want to join Yahoo, this is the time – and it is not going to be easy. It is well worth it.

Then why did I decide to move? I was getting very comfortable. Things were  smooth, stuff was getting done and all the wheel-work needed was just some oiling here and there. I felt that this was the best time for me to get out of the groove as well as the team and people I would transition things out to/for wouldn’t be as affected. Taking a lot of time and careful thinking for the transitioning, I am happy that it has been very successful. Much better than leaving a ship that’s about to sink.

People in my team have extended amazing support for this decision. I am very moved by their kind words and actions during the last few days. This has been a big surprise, since a big part of me always felt that I was not doing enough for them. I want to call out (and in no particular order): Alex, Deepthi, Hitesh, Sudip, Ram, Mike, Muthu, Tom, Jason, Sachin and Srinivas for everything they’ve said and done. Thank you folks, you all are simply awesome. Selling Yahoo is extremely convincing when I give examples from interactions with people like you.

Twitter is an equally amazing place to go to. Seeing the amazing energy in the Twitter office, I’ve been very excited about what they’re doing and about to do. I am yet to find out more, but from what little I’ve seen, the people there are super nice (just like Y) and the culture is alive and kicking. Their technology platform stack is either mostly open or about to open up, something that excites me even further.

All said and done, it is no lie that if you cut me, you shall see that I bleed purple.

Hacking fryol.net
Comments: (0) - Author: Rishi

So I see that there’s some bots coming in to ssh into fryol.net. I can’t stop those guys. As long as there are doors with locks, there will be attempts to break in. Even though this site has everything turned inside out, bots don’t care, they’ll try to get in anyway.

Anyway, the purpose of this post is to warn folks out there that you shouldn’t keep your usernames (on your ssh port) that are a part of the list in this text file: http://fryol.net/b/?attachment_id=1264. Actually, there are a lot of ways to make sure that your ssh port is well protected, and disabling all users but the one that you would actually only use (AllowUsers setting in /etc/ssh/sshd_config) is a great way to do this. Also, please disable password authentication. Use keys only. If you dunno what I’m talking about, it might just be a good idea to not run sshd at all. Heh.

Some very interesting names in there. Who the hell is praktikant that made it into that list? Lucky bastard.

Me Cook Food
Comments: (1) - Author: Rishi

When hunger strikes, you have to finally get cooking. For the first time ever, I served food out for myself like a civilized being (instead of devouring from the pots and pans).

On the table

  • Arhar Dal (Lentil)
  • Matar Pulao (Basmati Rice with Peas)
  • Aloo Pyaz ki Sabzi (Potatoes and Onions)
  • Sliced Tomato Salad

Took about 35 minutes to do this.