Tag Archives: startups


Fact checking the hype around mesh networks and FireChat

Being based in Hong Kong, my tech friends and I still occasionally get asked, “So what’s the deal with FireChat?” In case you missed it, FireChat, a smartphone chat app with basic mesh support, made headlines during the recent Occupy Hong Kong protests. Protesters on the ground here turned to it in late September when rumors began to circulate that the government would shut down the internet. It never did, but plenty of people, myself included, downloaded FireChat anyway. Hearing about this, many publications then used FireChat to tell a techno-utopian story about the future of mesh networking. Yet in the same stories, many publications pointed to faulty evidence of, or simply speculated about, people using its mesh features. (The sole exception being TechPresident’s Rebecca Chao, who actually found real evidence of its use.) As the hype has subsided, we at 88 Bar have decided that it’s a good time to review the facts and figures around FireChat.

 

1. FireChat was built for concerts and Burning Man

screen322x572  screen322x572-2

Figure 1: Press images are from its listing in the iTunes store.

 

2. FireChat is “not meant for secure or private communications”

The full quote that FireChat gave Wired last June is: “People need to understand that this is not a tool to communicate anything that would put them in a harmful situation if it were to be discovered by somebody who’s hostile. It was not meant for secure or private communications.” (This is, of course, months before they went on the speaking circuit to market its use during the recent Hong Kong protests.)

The University of Toronto’s Citizen Lab ran a full, tech analysis and concluded: “Messages sent in any of the application’s three messaging modes are sent in the clear without encryption. All messages sent and received, as well as a list of chat channels the user has joined, are stored unencrypted on the device. It is also possible for anyone (regardless of whether they have installed the app or not) to visit an IP address associated with the service in a web browser to see the most recent messages sent by users of the application.”

 

3. Lots of people in Hong Kong signed up for FireChat

Sign up figures for FireChat in Hong Kong during the protests range from 100,000 to 500,000. Assuming only a small number of “troll” accounts and that not every protestor would have downloaded the app, this would imply that there were at least 100,000 to 500,000 people who downloaded it to support the protest. (Estimates say roughly 200,000 people took part in the Hong Kong protests at its peak.)

And according to FireChat’s twitter account, people are still using it: their stats show 208k chat sessions on December 16, 2014.

 

4. But FireChat put Hong Kong citizens in danger during the protests

Not only are messages unencrypted, but it encourages the use of real names. Again from Global Voices: “Once installed, the app requires the user to sign up with her real name (which will be pre-filled with the name she eventually configures on her iOS or Android phone), a username and an email address.”

This is especially troubling because police have arrested at least two people in Hong Kong who posted protest-related messages online. See Quartz’s “police are using Hong Kong’s computer crime law to crack down on pro-democracy organizers.”

 

5. Protesters did use its mesh networking capability, but there’s only one report of it

Of the dozens of articles published on FireChat during Occupy Hong Kong, I only found one report that found evidence of mesh network usage. From Rebecca Chao on TechPresident:

As a volunteer, Ip was among a group of roughly 60 to 120 that monitored the supply stations. He used FireChat to recruit more volunteers. While verification is an issue in the app’s anonymous chat room settings, Ip says that given the proximity needed for off-grid use, “we could meet up instantly to confirm who I was and what volunteers we wanted and to ensure they got the right information.”

The other reports sadly only pointed to evidence of use of its internet (not mesh) chatrooms:

  • In a Wall Street Journal video cited in The Atlantic, The Verge, and TechCrunch, a Hong Kong student protester who shows off FireChat only uses its internet-enabled chatrooms, not its mesh network chatrooms.
  • When FireChat talks about its explosion 2M chatrooms created last month, they are referring to internet chatrooms. As Forbes’ Parmy Olson says, “Once a FireChat user goes off the grid like this, San Francisco-based Open Garden can’t track them anymore. That’s why they have no idea how many people are actually using its off-the-grid feature.”
  • In another Wall Street Journal video, the reporter and FireChat CEO are testing the mesh network function in the middle of the protests. It’s clear they are the only two chatting on the mesh network channel. My own experiments with my friend in the middle of the protest fared similarly.

 

6. FireChat is an IRC-like chatroom app, not a messaging app

2015-01-12 11_14_24

Figure 2: A 12-second clip of FireChat in action (as of 11am, January 12, 2015). Picture shrunk in order to conceal username + locations.

As Nathan Freitas and Oiwan Lam write in Global Voices: “FireChat is not a messaging app. FireChat is a chatroom, a platform to send insecure and public messages to people over the Internet or within your geographical vicinity.” As such it is not a replacement for Facebook Messenger, WhatsApp, Line or WeChat.

During the Hong Kong protests, everyone I spoke to said that Facebook and WhatsApp were by far the most commonly used communication tools. This isn’t to say FireChat wasn’t used, but it’s good to put its popularity into perspective.

 

P.S. The Hong Kong mobile network did not “collapse”

Nellie Bowles writes in Recode that “during the recent enormous pro-democracy protests in Hong Kong, the city’s cellular networks got so jammed they collapsed.” I have seen no reports to that effect (congestion yes, collapse no). As one protester stated in TechPresident, “I found that my LTE [4G] connection was pretty solid in most cases, even with tens of thousands of protesters around me.”

Cheat Sheet for Hardware Startup Accelerators

Here’s the link to the original Google spreadsheet. Please feel free edit! :)

* * *

What’s interesting about the data above is that there are two programs that criss cross China, and a bunch that don’t:

  • HAXL8R starts in Shenzhen and does their final demo day in San Francisco; it’s all about an immersion in China for them.
  • Highway1 is mostly based in San Francisco, but includes a field trip to Shenzhen; there’s a more clear focus on the US market using Chinese manufacturing for them.
  • Berlin Hardware Accelerator and Startupbootcamp (HighTechLX) seem to be about European manufacturing; the subtext is we can do it here in Europe too.
  • Am I missing any Chinese language programs?

P.S. Our friends over at PCH/Highway1 are touring Hong Kong and China this week – see here for more details.

Stats: The kinds of tech startups in HK

Last month, my friend Paul Orlando and I launched Startup Hug*, an experiment in cultivating an early adopter community here in Hong Kong. As part of the process, we evaluated over 130 local startup websites (the list came from Paul’s We Are HK Tech). One side product is that we categorized and annotated this list of 130.

While the data is still up to date, I wanted to share some of the statistics that we found – though do keep in mind the margin of error as this has been one night’s work:

HK startup types
Additional notes:

  • While games only made up 6% of the startups, they actually represent a relatively mature sector here.
  • I was surprised by the number of content startups, given the relative dearth of a reading-blogging culture locally.
  • Only 14% of startups offered (physical) goods, of which only a handful were genuine hardware startups. This seems low given Hong Kong’s proximity to Shenzhen and history of toy/clothing manufacturing.
  • 45 startups (35%) offered free (or freemium) products/services.
  • Productivity tools were nowhere to be found.
  • China-targeting startups were rare (probably less than 10%), which either speaks to the difficulty of doing that from Hong Kong or a lack of trying.

Want to find out more? See the four startups we chose to feature on Startup Hug or see the updated list of startups at We Are HK Tech.

* Disclaimer: Shameless plug.

Shanghai Government to Sponsor Hacking Spaces?

I came across this article recently on hackaday.com, talking up the possibility of new hackerspaces in China:

Government leadership in Shanghai wants to build 100 community hackerspaces funded by the Chinese government. Each space will be at least 100 square meters, open 200 days a year, and come equipped with wood and metal lathes, saws, drills, grinders, mills, and more electronics than we can imagine.

It’s an intriguing prospect.  It’s become commonplace for media to ask when China will have its own Steve Jobs (see here and here).  The most frequent reply is that China, with a censored media environment and an education system that discourages critical thinking, is a long way off from producing innovative thinkers.  Until recently, I also agreed with both points.

But then one day I visited the Makers Carnival at the Beijing Institute for Petrochemical Technologies to give a talk.  I was thrilled by the number of hacked inventions, from swimming robotic fish to airborne weather detectors. During my lecture, I found the students, who came from many universities around China, to be critical and engaged with the ideas I presented on social media and art.  But more importantly, as I toured the different booths, I found the students to be excited to present their creations to me and talk about the pros and cons.  They were critically engaged, active learners and developers, and they were making things I could only imagine doing.

I chatted with David Li, who founded the Shanghai-based Xinchejian (pictured above and below), China’s first official hackerspace. “Most of global hackerspaces are operated as non-profit membership driven to balance between the membership and public access to the space,” he explained. “The government sponsored spaces can be more public accessible.”

Indeed, while hacker communities I’ve seen in both the US and China are ostensibly available to the public, they are in practice usually patronized by a small collective of people.  This is great for community building but not always for the wallet.  Most spaces have to charge a membership fee, thus cutting out a swath of the population that may not be able to afford it.  That swath almost invariably includes students.

Li pointed me to an intriguing piece in MAKE that argues for more publicly-accessible hacker spaces, like a library: “If the only public space where 3D printers, laser cutters, and learning electronics happens is in fee/memberships-based spaces (TechShops, hackerspaces), that will leave out a segment of the population, who will never have access.”

Of course, with government sponsorship comes government oversight. Can true innovation happen in this kind of environment?  My take is that providing tools and resources is a great first step.  Few people have access to large workspaces, a community of like-minded makers and mentors, and expensive tools like laser cutters and 3D printers.  Giving them that access can spark their thinking at the very least and help encourage a culture of making early in life.  Let’s hope this plan pushes through.