Three weeks have again flown by since my last update. Currently (to give you some context, since life is all about context) I'm sitting in a coffee shop come bakery - "Speciality's" - in Santa Clara. This afternoon was a final round interview at a great startup. Running through my head for the last few weeks has been a series of elaborate decision making processes as I try to determine what job offer to accept and which to reject.

As I've mentioned earlier, I knew job offers would come more easily to a software engineer in Silicon Valley. In particular, my strategy of selecting employers whose product I know and love and primarily applying to smaller employers has paid off.

As a brief segue, current students looking for graduate entry jobs seem to apply to the big name companies. This makes their job hunt harder since these employers have their pick of graduates and can be more selective. Additionally, I'm skeptical that new engineers in these companies have much leeway to work on projects that interest them.

My friends who know me well know that I have been planning to come study a Master's (in the US) since my last year at Cambridge. While the decisions that I've made since regarding employment, living arrangements and significant others have been made to this objective, I now face a growing amount of uncertainty. There is no obvious next goal and I have several potential routes to achieving the various things that are high up on my list of life ambitions.

My options are: work for an established company, work for a startup, start my own company. The latter of which is the riskiest and brings with it the most financial uncertainty. It is also the most exciting. Within the first two categories, I need to pin down several decisions: whether to work for a consumer or enterprise technology company, whether to work on a product or a service, and whether to work in South Bay or in San Francisco.

Overwhelmingly it feels as if the more exciting consumer focussed companies are in San Francisco - while the enterprise companies are based in South Bay (i.e. Silicon Valley). I love living in Berkeley and working in San Francisco would make it possible to commute in from Berkeley. Working in South Bay would mean I would have to live in South Bay (or commute in from San Francisco). Besides rent being cheaper in Berkeley, it is surrounded by beautiful scenery, I have many friends there and, despite being fairly city-like, it remains very peaceful.

On the other hand, the harder computer science problems seem to be with those companies in South Bay. Working on an actual tech product/problems is something that I've often tried to do in a professional context in the past but struggled to find. Building services in Java often felt like virtual plumbing - taking one library and piping its output into another library or database.

There are a number of sub-decisions to be made here and I need to pick carefully since it's going to affect my life for the near to medium term!

We were forwarded the blurb below. Essentially, build our product for us and we'll take 82% of the equity because we came up with the idea. I was much amused-

HAAS TEAM IN SEARCH OF ENGINEER!

We are a group of 5 Haas students who are currently in a process of developing a p2p mobile app. We have already recruited a team of interns who are eager to start working under the leadership of a new VP of Engineering.

We have skills in marketing, interface design, business development, investment research, and statistical analysis.

New *** venture, ******, seeks Cofounder/VP Engineering to spearhead development of a mobile software platform with peer-to-peer functionality. This is a straightforward project and we are aiming to get an efficiently working prototype completed by the week of Nov 22. We have proof of concept and market research. Now we simply need to build and sell. We are advised by a founder of Berkeley's CET and a venture capitalist, who also teaches at Cal.
Compensation is a 1/6 share of total equity in a pool of five other founders!

You should be passionate about programming, able to demonstrate your skills, and ready to work as a team to grow a successful startup in the sharing economy (i.e. Lyft, AirBnB, ZipCar).

Skills needed:
- demonstrable experience with coding mobile software
- experience with peer-to-peer functionality
- ability to develop application for both Android and iOS
- ability to build and lead a software development team (of interns to start)
- willingness to follow Agile development process
- self-motivated, sense of urgency
- team mentality, results oriented, goal driven
- strong time management skills and ability to commit to co-located work hours
- as cofounder, ability to work in a business development team to grow the company quickly
- excellent communication skills
- true passion for the startup environment and building mobile apps

As I mentioned previously, I was offered the fantastic opportunity to give Neal Stephenson a lift to Oakland airport on his way home after a panel session at UC Berkeley. Rather than asking him about something particular to his work, I thought I'd ask him about how he gets stuff done. It's a topic that I'm continuously fascinated with - being both a complete data junkie and a pathological procrastinator. As an author who is clearly very productive, I was curious as to how he does it. Accompanying me was Constantin, a PhD student at UC Berkeley.

Work Environment

Turns out the treadmill desk in REAMDE wasn't entirely fictional - Neal actually uses one of these to work on. Apparently a low walking speed makes it possible to type and work. A wrist rest is necessary to dampen the side to side rocking of your wrists as you walk. For work where he's handwriting, he uses a standing desk (since it's not possible to handwrite while walking).

I previously noted the curious coffee shop culture in the US and was curious if Neal adhered to the cafe-author stereotype. He mentioned that he didn't - preferring an atmosphere of quiet and to stay in one place (I recall libraries being mentioned). If he needs to use a lot of his notes and materials, he needs to be at home with a desk to lay things out.

Schedules

Generally he doesn't like travelling to speaking arrangements since it can be quite disruptive to getting work done. Not only the actual time spent travelling and speaking but also the communication overhead that goes into organising an event.

He works in the morning normally - and stops as soon as he feels his alertness tailing off. Most of the time he writes a paragraph well enough the first time. As he said, 'editing a work of literature is like performing surgery on a human body. There are always scars and seams left over.'

Medium

All of his work eventually hits a computer - either in LaTeX or, lately, using Mac writing application Scrivener. Occasionally he'll handwrite work - work that is handwritten tends to be better thought through since there's a longer buffer between his thoughts and putting them to paper (since handwriting is slower).

On Notifications

I asked him how he felt about this world of push notifications - where we are pinged quickly with every new bit of information - tweet, news article or email. He says that he just has those turned off since his job doesn't require continuous notification. I would suggest that that's probably true of most jobs - we're just addicted to the small dopamine hit that accompanies each notification. Perhaps that will be the next life hack I try to implement.

Organisation

With such incredibly intricate story lines and characters, I was curious how he organises his research. Supposedly he doesn't have a definitive scheme, preferring to keep notes organised randomly. This is apparently useful to allow ideas and notes to cross-pollinate others as he searches for the notes he wants.

Backup

Finally, I was curious how he keeps his work from getting eaten by his computer. Supposedly he backs up to a RAID in the basement and to a USB stick every two weeks.


On a final note - it was a pleasure to meet the man himself. I took a hiatus from reading serious amounts of fiction for most of my late teens, as I became more immersed in my cycling hobby. As I left Cambridge, I started reading fiction again and started off with his novels. I've not stopped since.

It's been just over 3 weeks since my last update and I've evidently failed to blog weekly as promised. It's hard to believe October is almost over. Fall (or Autumn) is definitely here now - leaves are falling and there are, on average, fewer hours of sunlight each day. The cold is sweeping in. As we started October the Berkeley 'Indian summer' carried over to give us a HOT first few days. This didn't persist much beyond the first weekend though and it quickly moved into 'hoody' weather.

Soon after that, I managed to accidentally smash my bedroom window while fitting my new amplifier and the nights became much cooler. This led to some very chilly nights, which became better when Ryan used his superior duct tape to actually seal the window with cardboard (versus just slotting it in - which was my approach).

Broken glass aside, it's been a monumentally busy month. The day after my last update, I went sea kayaking on the bay with Cal Adventures, the outdoor centre that is run by (affiliated with?) the university. This was a one day introductory course that allows me now to go out on the bay alone (within line of sight of the centre). While the pace of the course was a little slow for my liking, we were extremely lucky to catch such beautiful weather and being out on the bay was a peaceful experience that helped mitigate some of the stress of our second Advanced Robotics assignment.

The calm didn't last for long though because I was soon onto an assignment for Computer Vision. Once this was completed, I had an Advanced Robotics assignment that was due in a week. It had come out late beacuse our Professor had recently changed the questions and had been trying to solve them himself. This took longer than expected...perhaps an ominous sign. Logistics made it impossible to heed his advice that we start immediately and so I tried my best to do it in a week. This proved to be difficult - in the end taking 9 days (using all of my remaining late days).

During those 9 days, I put in about 50 hours of work, getting stuck on the last part of the first question for three days. Despite going to see our teaching assistant three times about it, I got no further. Bad strategy which resulted in me turning in the assignment 70% complete. Whoops. Looks like trying to juggle the job hunt with a graduate degree at Berkeley is a difficult optimisation problem!

This past week has been a bit of a rollercoaster for a number of reasons. After receiving one job offer, the other employers are reacting much more quickly and this next week I have 6 interviews planned. I'm wishing I hadn't scheduled my Google phone screen first - now that I've had numerous telephone screens, I might have screwed that one up much less badly.

A few weeks ago a professor sent around a request for volunteers to pick up/drop author Neal Stephenson from/to the airport. He came to a panel on campus, vaguely entitled 'On The Future: Beyond Computing'. Other members of the panel included Peter Norvig, Director of Research at Google and Jaron Lanier, author and coiner of the term 'virtual reality'. I capitalised on my free Zipcar credit and replied - a few hours after the email was sent around. It looks like my response time was favourable and I was tasked with dropping him off to Oakland airport on the day after the talk. Equally as exciting, Professor Katz invited me to dinner with the panelists afterwards. (Googling Professor Katz revealed him to be one of the creators of RAID. Incredible. I was very tempted to ask him if he knew where my data all went, that one time...)

This was an interesting experience, being both simultaneously over and underwhelming for reasons best explained in person. I'll post a separate note with my takeaways from my conversation with Neal Stephenson. Still, I was buzzing with excitement for a good two days after that. What struck me as most admirable was how normal these tremendously acclaimed people were. I'd always imagined that reaching their heights of 'notoriety' came with some ego inflation but meeting these three demonstrated how untrue that assumption was.

The next day we had the interview at the culmination of our application to Steve Blank's Lean Launchpad class. As I mentioned previously, this was one of the classes that helped me decide to come here and I was quite looking forward to it. The interview itself was short - which indicated that either we had enough information on our application or that they had already made up their mind. We'd put a lot of effort into our application, spending quite a bit of effort researching the space around our proposed 'startup' (similar to my capstone project). They asked questions around our idea and around our team for a total of 5 minutes before we were let free.

A double gin and tonic (or club soda, tonic is expensive here) plus 8.95 hours of sleep later, I woke up to a depressing rejection email from the teaching assistant for the class. We've yet to receive feedback as to our rejection which is strange, because they're normally very quick to queries. The silver lining is that I'll now have significantly more time next semester to sleep, exercise and submit applications for accelerators.

The last couple of days have been moderately alcoholic; after the last few weeks, I was desperate for some mindless social interaction. Yesterday we had our MEng Halloween party and it occurred to me that the primary reason people hold Halloween parties as adults here is to take photos. Despite buying $10 worth of cardboard sheets, some cyan spray paint and duct tape, I ran out of time and wasn't able to recreate a costume based on the Hype Dark robot. One day.

This coming 12 days is going to be the toughest of the semester yet. In addition to the 6 interviews, I have a couple of social engagements (including seeing LTJ Bukem and Bachelors of Science live!), two homework assignments and a midterm to prepare for. On the other hand, once this fortnight is over - life will become a lot easier.

My Lepai amp is dying a slow death and so I decided to replace it. I've tried to bootstrap my hi-fi system here with cheap but high sound quality components - starting with the Lepai amp (which I bought to drive my cheap outdoor speakers in the garden). If you haven't come across this, it's possibly the cheapest 2 channel amplifier you can buy - I paid ~ £20 for it, they go for about $20 on Amazon.com.

It was reasonable and the reviews tend to rave about it. I worried that, just as audiophiles tend to assume more expensive products sound better than cheaper products, people thought the Lepai amp was better than it was because it was so cheap. After moving to Berkeley, I eventually picked up a set of the critically acclaimed Andrew Jones SP-BS22-LR speakers for $75 (they normally retail for $130 ish but Amazon has some amusing pricing fluctuations).

This combination sounded great but would distort significantly if you tried to drive the speakers at any sort of volume. I put this down to the amp.

A few weeks ago the left channel would fail. A quick jiggle (for lack of a more technical term) of the speaker cable would help but eventually it would cut more frequently and this wouldn't help. It still works sporadically but for the most part - it wouldn't. This means the amp failed after just 2 months of use. At £20, that makes it relatively expensive. I normally expect to at least a year out of my electronics (my Acoustic Energy Aego-M speakers at home are still going strong, 5 years later).

After some googling, it was a split between the Dayton DTA-100a and the Fiio A1, which were both cost effective good quality amplifiers.

In the end I chose the A1. It was slightly cheaper and put out less power but several reviews questioned the reliability of the Dayton amp (having had amps fail after a few months). So far, the A1 has proven to be significantly more reliable than my Lepai amp - but only time will tell if this persists.

The power concern seems to not be such an issue either, it will easily go loud enough for my music to be heard on the street outside our flat (again, a very scientific metric). As for quality, I've found it is significantly clearer at higher volumes than the Lepai amplifier. Tonality seems to be better - in particular, the bass seems more pronounced.

You could buy 4 Lepai amps for the same as Fiio but you really do get what you pay for. Other people don't seem to have had the same reliability issues (or seem to have overlooked them, based on the low cost of the Lepai) but if you're after a quality product, you can't go wrong with the Fiio. (Plus, I'm hopeful that the warranty/support situation is better, since Fiio appears to be a more reputable business.)