Self-watering balcony garden

It’s wonderful having a wide variety of plants herbs and flowers on your balcony. However, knowing me, I would frequently forget to water them causing many of the plants to frequently dry out and eventually die out. Not to mention the occasional two week trip, hated by plants everywhere.

I therefore set out to build a self watering system. But I wanted to to be able to handle a little bit of complexity-that is, to be able to specifically water different parts of the garden differently.

It took several adorations, but I’m very happy with the final version that I ended up with. I have a main tank which is a 5 gallon tank that lasts two weeks at least. That feeds water into a staging tank via a diaphragm pump. Diaphragm pump is able to draw water without priming first.

The system will fill the stating tank with the amount of water needed for a particular section of the garden. A server will then turn a knob to control which of several possible output pipes are selected. A second pump that activates that draws water from the staging tank through the pipe selector, and onto the plants.

Check out the videos below!

Mechanically control a thermostat!

I live in an apartment building, I don’t have the luxury of installing a nest thermostat. But I really wanted to be able to control the temperature via the cloud.

So I 3-D printed a device connected to the cloud using an Arduino compatible chip, and hooked up a servo to mechanically press the buttons of the thermostat.

It actually works really well! See the video below for a demonstration of it working.

$10B vs $68B – Hyperloop must surely force a review of California rail plans

I thought Elon Musk was pretty smart to announce that his Hyperloop idea costs $6 billion for people-carrying pods, and $10 billion for pods that also carry cars. Since California’s high speed rail proposal is projected to initially cost $68 billion, this should force, at minimum, a review of the Hyperloop proposal.

Not that it matters, because there is no way it’s going to happen in the next 5 years. But by god I hope I’m wrong, because the Hyperloop idea is so cool.

The main reason I believe Hyperloop won’t happen anytime too soon is because it’s not an incremental technology change – but rather it’s a proposal of an entirely new technological system. Entire systems simply can’t be implemented overnight.  Systems, by definition, require a lot of modules. And a radical new system such as Hyperloop requires the design of too many modules to implement in one go.

One could argue that the modules already exist, and only incremental changes are needed in those modules to make Hyperloop happen. Air compressors exist, linear induction drives exist, and I’m guessing above-ground tube technology exist. But it seems to me that there still need to be a lot of changes made to all those modules to support Hyperloop.

Think about the modern car. Do you think a top-rate engineering firm could design a modern car such as the Toyota Prius hybrid, in 1920? No way. A million engineering puzzles have been solved since the introduction of the motorcar in order to enable the complex piece of technology that is the modern car.

I also think that there aren’t a whole lot of city-pairs in the world that could use Hyperloop. LA-SF seems particularly well suited due to the desert and farms that lie between the two cities via I-5.

Having said all that, I can very much imagine China implementing this. They are always looking to have a high-tech advantage to showcase their technology. And since Elon Musk open-sourced the plans, they’ve gotten a headstart. So I say build it China! And you can sell it to the world when you have it working nice and smooth. And I bet it might just well cost under $70 billion.

Why the emotional mind matters

A wise mind is one which considers both the rational mind and the emotional mind. Why does the emotional mind matter at all?

I think this is because our experience of the world is an emotional one. We have emotional bodies – feelings course through it like blood.

Think of happiness – it can be a goal in and of itself. And unhappiness can be such a driver of human behavior.

Because we are unable to actually live without emotion, therefore we must take it into consideration. When we feel something, we validate it as it makes sense given our life experience.

The challenge is when emotions become so powerful, that the capability of the wise mind degenerates. Then behaviors can result which affect our functioning in the world. This is when we need special tools to help us regulate those emotions.

Making progress in meaningful work

I really like this one line from a posting today on the Harvard Business Review blog:

Our past research found that, of all the events that can keep people happily engaged on the job, the single most important is making progress in meaningful work.

I think that really nails it. When I get busy with tangential work, I start to get nervous that I’m not really moving the business forward. But when I get solid progress on solving core problems, I feel really good.

Full article here:


Introducing the Reluctant Raconteur – A blog about malaysian family life by my mother

My family is here visiting the San Francisco bay area. Quite the lot of us – my sister and husband, their 2 kids, my mom and my dad. As you can imagine, shepherding 7 people around is not the easiest thing to do! Every drive has its back-seat drivers, and every decision is by public committee!

Two nights ago my mom sent an email to my brother and wife about our trips so far. I read the email, and even though it was just a “travelogue”, I realized how beautiful my mom’s writing can be. And it occurred to me that it would be a great idea for my mom to start writing a blog.

And so the Reluctant Raconteur was born. Head over there to find out a little bit about Malaysian family life.

3+1 bedroom apt for Sale/Rent: Seri Duta 1 near Kenny Hills

Low-density low-rise

With all the development taking place in KL, there are very few apartments remaining that offer this unique combination:

  • Low-density, low-rise
  • Low-noise, green/wooded environment (Kenny Hills-type environment)
  • In KL: only 12 minutes drive to KLCC.
  • High-quality expat-level apartment

If the above combination is what you’re looking for, then come take a look!

Further details:

  • 3+1 Bedroom Apartment.
  • 1920 Sq Ft
  • 1st Floor
  • Pool view on one side, and palace+forest view on the other
  • 2 parking spaces
  • Fully furnished
  • Excellent sound insulation between apartments

Apartment is available for both sale and rent.

Sale Price: RM 1.1 million
Rental Price: RM5000/month

Contact: / Samad: +6012 511 2040 / Marina: +6012 373 7045

Click on a picture below to zoom in. For location map, scroll down to the end.

Location Map

The green arrow shows the property location. Notice the wonderful greenery and forest around the apartment. Note that the words “Seri Duta Apartment” in the map below refers to Seri Duta 2, an adjacent property.


Agents are welcome

If you would like to show the property to a client who is looking for an apartment just like this one, please contact us (owners) as per contact info above.

Experimenting with a new hiring process/way of working

So we’ll be doing a fair bit of hiring in 2012. I’ve decided to experiment with a new hiring process.

One thing I wanted to try was to do away with interviews. Instead, have candidates fill out a fairly difficult online application form that’s technical. This filters our the semi-serious candidates, and gives us a solid first-look at their quality of work.

Then, we’ll offer the candidates we like an opportunity to work on a small, real project. And from there select the ones we’d like to work with on a longer-term basis.

I thought this was pretty radical (esp. the doing away with the interview part), but this is exactly what they do at this company.

I’m a couple of weeks into the process, and so far it’s very promising. I think being able to give the same job to multiple people and actually compare the work side-by-side is revolutionary. This is like website split-testing.

People ask me if this is an expensive way to go about doing things. I say no, because I think the cost of hiring a mismatch is way higher than the initial high cost of giving the same job to multiple people.

Minimum Viable DNA?

Reading about basic molecular biology is pretty mind-blowing. As a computer programmer, I would use the phrase “elegant design” to describe how life seems to work. I discovered a lot of things that I didn’t know about and are truly remarkable.

I knew a little bit about the DNA double helix of course. But what I didn’t know was the impact and/or “reason” such a design is elegant. And I knew about base-pairing, but I hadn’t realized it’s all to do with molecule shapes and molecule binding.

In fact, so much molecular biology seems intrinsically linked to shapes and molecule binding. Given a particular shape, a molecule tends to bind to another molecule. And from there, the engines of life run.

And now I have a question. It seems that DNA can be synthesized pretty easily these days, especially if it has less than 5000 kilo-base pairs. I’ve also read that DNA can be injected into an existing cell, and through some process it can take over the existing DNA. So my question is:

Is it possible to synthesize a minimum-viable DNA that can sustain life in controlled situations?

In computer programming, when we want to understand a complex system, we often try to create something small that tests parts of the system. From there, we can be reasonably confident that we understand a larger system. So, could we synthesize a minimum sequence of genes that can allow a single-cell organism to replicate in controlled situations?

Bacteria DNA is reasonably small but even then it has a lot of inter-gene regions which are not needed. It also has introns, which are regions of gene whih are discarded by the cell during messenger RNA creation.  So in our minimum viable DNA (MVD), we definitely want to strip out all that junk.

Bacteria is around today for good reason. It has a LOT of redundancy to survive through the ages. For example, bacteria can get energy from multiple sugars: lactose, glucose, fructose. However, each of those requires multiple genes. In our “controlled environment”, we could easily give the cell a single food source, say glucose which I think is the preferred food source. We can then strip out all the DNA related to the other food sources in keeping with our goal of minimum-viable DNA. Sure our cell may not survive in the wild, but it could survive indefinitely in a petri dish laded with glucose.

So what would the minimum viable DNA need? Well, I guess it would need all the genes that code for replication and for energy absorption. It could work couldn’t it?

Viruses may come to mind. They have very very few genes after all. However, they can’t replicate on their own, and they need many of the genes of the host cell to survive.  Our MVD needs to survive on its own. No parasites please!

Of course, it would be okay for the petri dish to have amino acids lying around. I think life itself needed amino acids in the beginning to get “kick started”.

The beauty of the minimum viable DNA would then be that it would be possible for a student to completely understand every single gene (assuming the number was reasonably small). And we could put the entire workings of the gene in a 3D computer model that could be analyzed and understood. And there you have it: Life. Created.

Blog at

Up ↑