Science Addiction

A dormant blog by Devanshu Mehta

Tag: Google

Google AppEngine: SMS, the Namespace and Other Quirks

This weekend I started experimenting with the Google AppEngine.

Wondering what it is? There’s a pretty good answer from Google and one from Wikipedia, but the short answer is that it’s a hosting platform for web applications. Effectively, Google is the sysadmin for your application and provides the ability to scale processing, bandwidth and storage on their infrastructure. For a price.

Quirks

  1. SMS Verification: In order to create an application, you need to provide Google with a valid mobile phone number where you can receive an SMS. Google sends you an SMS with a code, you enter the verification code on the site, bob’s your uncle. The question is: why?

    There is no official answer, but here are a few guesses:
    * AppEngine is freemium service. Basic accounts are free. SMS verification is a complicated CAPTCHA.
    * SMS verification loosely ties one individual to one account. Google only allows one AppEngine account per mobile phone number.

    And no, Google Voice is not one of the options among carriers for verification. Because that would imply people actually use Google Voice.

    I kid. I actually use it. Sometimes.

    I don’t know if Google Voice actually works for verification. There are conflicting reports on the web.

  2. The Google Namespace: Google has a single unified namespace, shared among all Google services. If you have a gmail username, someone else can’t use the same string as an AppEngine app name. This may seem trivial. It’s not. Just like the .com’s, it seems there’s been a gold-rush for Google names.

    Among the many, many I tried, here are some absurd app names that were not available:
    * truetrue
    * truetruetrue
    * truetruetruetrue
    * truetruetruetruetrue
    * truetruetruetruetruetrue
    * truetruetruetruetruetruetrue was finally available. Not that I wanted it. But you get the idea.

    I’m not sure if some deranged squatters are actually holding on to these names on the off chance that they’re worth something, or Google’s availability-checker has a bug.

  3. Programming Languages: The AppEngine only supports Java and Python. I want to use Python (web2py, more specifically), so that’s fine by me.

I have an idea for a nifty app that will help at least one person. Me.

I’m not quite ready to say what it is. And I’m not sure AppEngine is the right home. But we’ll figure it out- the free 500 MB of storage and CPU/bandwidth to support around 5 million page views a month is tempting.

Thoughts on Google Buzz: 120 Days Later

When I last wrote about Buzz, I was cautiously optimistic.

I still am.

I continue to use it after four months, but not many other people do. Of the things that I noted that I did not like about Buzz, only #1 has been resolved. I live with the rest. A few thoughts on Google Buzz usage:

  • Buzz has its own place in my social media landscape– I use it to share links with a known, finite set of people. My wife, my parents, and about 3 friends.
  • Engagement is fairly high among this small group.
  • The only distinguishing feature of this group is that they are all avid GMail users.
  • The only reason I am able to sustain sharing on Buzz is because I am an avid GMail, and more importantly, Google Reader user.
  • The role of Buzz is very different from Twitter, Facebook or others in my social media universe. On Twitter (and my blogs), I talk to the world and the world may or may not listen. On Facebook, I talk to my subset of the world, and the most random embers from my past glow in response. On Buzz, I send specific links curated for a specific set of people. I know who they are, and post links for them. It is so much more like email; no wonder it lives in GMail.
  • I have 41 followers. I know almost all of them personally.
  • I follow 48 people, but only a handful post anything regularly. Almost all the people who post regularly, do so because they share from Google Reader. Put another way, they would be sharing links on Reader even if Buzz did not exist.

Things I Don’t Like About Google Buzz

I generally like Google Buzz, especially things that it may enable with its APIs. Having said that, things I don’t like about Buzz so far:

  1. It brings every message that has a new comment to the top. I have to manually mute each conversation I don’t want to hear from again.
  2. It won’t let me comment on some Google Reader shares. It says “Oops, there was an error posting your comment. Please try again in a few seconds.”
  3. It doesn’t update feeds (twitter, rss, other) in real time. Sometimes delayed as much as 12 hours. Sometimes doesn’t show up at all.
  4. There is no way for me to share to a subset of people without everyone in that subset knowing the members of the subset. Confused?
  5. Finally, I would like more control over what gets posted. e.g. only tweets that are tagged #buzz.
  6. And recognize redundancy. If I post something to twitter and share it in Reader, don’t show my followers both.

Most or all of this can be fixed with subsequent updates, so I’m not worried.

Got a Valentine’s Day Card from Google, seriously!

Google loves me.

“Others will fill your heart this Valentine’s Day
We want to overload your servers.”


Is it love, or is it just advertising? The pen covers my coupon code. 🙂

Google Juice

I’ve run a Star Wars web site called GalaxyFarAway.com for 11 years now. For the first year of its existence, the web site was hosted on tripod.com. It was devan1.tripod.com. GalaxyFarAway.com is on the Google front page of results for many Star Wars related searches, but I just realized that so are some pages of my old tripod site (that I was updating for only about 8 months).

I guess it has seniority in Google’s mind– it is about as old as Google, after all. So last week, I put a notice on the Google-popular page that the site had moved. Something I should have done 10 years ago.

How Do You Say “Irony” in Chinese?

A Chinese Foreign Ministry spokesman recently said at a news conference:

“Many people have a false impression that the Chinese government fears the Internet. In fact, it is just the opposite.”

The news conference was in response to the Chinese government banned all of YouTube, in response to a single video of Tibetans being beaten. [via Arthur Bright at the Citizen Media Law Project]

Change Watch: Say No to YouTube, Mr. President

Chris Soghoian makes an excellent case against using YouTube as the default for the President-elect’s weekly addresses. There are many issues he touches on including the privacy of the viewers from Google, the free Obama-endorsed publicity for YouTube, the embracing of a closed-format, and so forth.

On Community-Based Collaboration: Lesson From the OCLC Debacle

Community-based collaboration or “Crowdsourcing” has become the buzzword in many industries- the idea that by fostering a community, you can solve many major problems through their collective wisdom without actually hiring people with… wisdom. Linux, Wikipedia, the recent Twitter Vote Report and many other projects are often cited as successful examples of this.

The nonprofit OCLC has a membership of over 69,000 libraries around the world. These libraries collaborate to create a database– WorldCat— of bibliographies that all the member libraries can use. It is a great system– or at least it was, until the recent introduction of their upcoming use policy. The two major concerns- via Terry’s Worklog– were:

  1. OCLC would require the license to be placed within the record. This takes the ownership of records away from the library and since it is only a link to the license, the license could be changed at any time without the knowledge of the linking library.
  2. WordCat data could not be used for creation of services– even non-profit– that may compete with it.

The first concern has been largely alleviated in a recent version of the OCLC FAQ, but the second one remains. Who really owns the database? Since it only applies to libraries who are members of OCLC (in contract), what prevents someone else from creating a competing service? And finally, can you really copyright a database?

There are many projects out there, like OpenLibrary, that are trying to create a truly open, non-commercial database of books that would run afoul of this clause. In reality, the problem is not in whether it will be enforced but in that this organization believes it is more than the sum of its parts. That OCLC– not its members– controls how and where the data should be used- data that was created by its members.

This is where OCLC is different from free and open source projects like Linux, Wikipedia and every Creative Commons or GPL licensed copyrighted work. There is no right to fork.

To everyone who contributes to community projects:

Always reserve the right to fork.

That is to say, you should always be able to take the marbles and go home. To fork, in open source projects, means to take all the code/data and create another project. This is made possible by the inherent “free”ness of GPL, CC, GFDL and other licenses. Many open source projects have been forked in the past because a sufficient chunk of the community didn’t like the rules they were being asked to comply with. Nobody controlled the code, so everyone controlled the code.

However, in the case of the OCLC debacle, via Annoyed Librarian:

To use a prison metaphor, it’s clear that librarians dropped the soap decades ago.

Or Stefano’s Linotype:

Basically, by using OCLC’s data you agree to protect their existence. And their monopoly (nobody else in the world does what they do, at the scale they do it). And with data that they didn’t even create.

In a time when everyone is using search engines as their first stop in finding answers, closing WorldCat further is a major step backwards. Like many other old-world companies, the OCLC is trying to remain relevant in the face of major paradigm shifts- in this regard, it is much like the Associated Press, which is losing relevance and support from member libraries (thanks Edward Vielmetti). If this was a commercial enterprise built by a million highly paid employees, it would make no difference what they did with their data. But this is a non-profit built on the backs of its members contributions.

As Princess Leia said:

The more you tighten your grip, Tarkin, the more star systems will slip through your fingers.

Resources:

Is Google Evil?

In the great Hindu mythological tale- the Mahabharata– there is a young prince named Yudhishthira who always speaks the truth. So divine was this virtue that his chariot always remained a few feet above ground. And so implicit that even his enemies, in the heat of battle against him, would trust his word.

Until a fateful day during the great battle, when Yudhisthiras side- the Pandavas- decided that their teacher Drona, who fought for their enemies, must be killed. No one had the skill to do it, unless Drona could be emotionally weakened. And so the plan was hatched.

Drona’s son was named Aswathama. Coincidentally, this was also the name of an elephant. The Pandavas killed the elephant Aswathama and spread the word that Aswathama had been killed. Drona was distraught, assuming it was his son. The only way to confirm the story was to ask Yudhishthira- he who would never lie.

“Is Aswathama dead?” asked Drona.

“Yes,” said Yudhishthira. And then under his breath, he continued: “The man or the elephant.

Drona did not hear the second part as he threw down his weapons and wept. He was quickly killed, and the Pandavas were one step closer to victory.

However, the moment Yudhishthira muttered the half-truth under his breath, his flying chariot came half-way to the ground and stayed that way for the rest of his life. Even today in India, you can simply say “Narova Kunjarova” (man or elephant) and people will know you are referring to a half-truth or a white lie.

The moral: It takes more effort to keep a white sheet white than it does to keep a grey sheet grey. Just ask the formerly-perfect record of last year’s New England Patriots. Or Google.

Three wise monkeys

Google’s motto was “Do no evil.” With a motto like that, they were bound to fall short. They have had many missteps; their chariot is undoubtedly half-way to the ground. But are they evil?

Tomorrow at NPR’s Intelligence Squared debate, Jeff Jarvis, Esther Dyson and Jim Harper will be debating against the motion “Google violates its ‘don’t be evil’ motto“. Siva Vaidhyanathan, Randal Picker and Harry Lewis will be for the motion.

Jeff Jarvis has put up his debate notes on his blog and makes many points that I generally agree with. Google is not evil, if the word ‘evil’ is to retain any meaning. There are evil corporations- ones that have championed wars, economic turmoil, corporations that have hid their toxic contamination of water supplies and even milder forms of evilness, such as consumer unfriendly behavior. There are many seriously evil corporations, but Google is not even close to being in this group.

Of course, when drafting the motto, the founders were probably aiming higher than this kind of evil. What they were probably aiming for was to never adhere to the common corporate evilness- the old Microsoft kind of evil.

Jarvis points to all the good things Google has given us- a way to make money off content in the Internet age, a new platform (maps, services) for a new generation of companies to build tools, using the wisdom of crowds to rank content and the general good that comes from making the world a more connected and smaller place.

Jim Harper on the Tech Liberation Front blog makes similar arguments. But both of them gloss over two areas where Google is venturing in to potential evilness

  1. China: Harper and Jarvis both offer the same argument for Google’s censorship in China: “exiting China would abandon the Chinese people to government-approved information sources only.” But in the current scenario, Google is that government-approved source! By censoring their results, they have become the tool of oppression- which is fine if your a regular corporation out to make a buck, but not when your fundamental motto is to do no evil. With China, Google’s chariot came half-way to Earth.
  2. Privacy: This is a huge debacle waiting to happen. Google has sent out strong signals that they understand the ramifications of a single privacy scandal and have started to craft policies to safeguard search privacy. The advantage for consumers is that there is no brand-loyalty or lock-in with search, so we, the users, would leave Google in droves if it became clear that they are no longer good stewards of our data.

In general, however, I come down on the Jarvis, Harper, Dyson side of the debate. Google is not evil- yet. And the amount of good they do as a company, as a corporate citizen and in philanthropy offsets most of the potential for evil. Their behavior in China has damaged their reputation, but for a company that is aiming for perfection, I will take a near-miss.

So, was there a precise moment when Google’s chariot came half-way to Earth? Maybe it was when Eric Schmidt said the following, on their decision to censor in China:

We actually did an evil scale and decided not to serve at all was worse evil.

Of course, more to the point was what Google’s Marissa Mayer said more recently:

I think that ‘Don’t Be Evil’ is a very easy thing to point at when you see Google doing something that you personally don’t like; it’s a very easy thing to point out so it does get targeted a lot.

Off topic, but in the same ballpark: is Barack Obama setting himself up for a similar backlash? His post-partisan, everything to every progressive, hope, change, peace, net neutrality, end of oil chariot is bound to come flying to Earth. Maybe NPR will hold a debate in 2011 about that?

Google Books Vs. OCLC and WorldCat

Slashdot has a story about the non-profit OCLC trying to tighten its control over the database that libraries around the world use:

“The main source of the bibliographic records that are carried in library databases is a non-profit organization called OCLC. Over the weekend OCLC “leaked” its new policy that claims contractual rights in the subsequent uses of the data, uses such as downloading book information into Zotero or other bibliographic software. The policy explicitly forbids any use that would compete with OCLC. This would essentially rule out the creation of free and open databases of library content, such as the Open Library and LibraryThing. The library blogosphere is up in arms . But can our right to say: “Twain, Mark. The adventures of Tom Sawyer” be saved?”

Of course, the real story here might be the recent resurgence of Google Books as a force to be reckoned with; how they might start competing with OCLC by collaborating with libraries. From the OCLC FAQ about the new policy:

My library has been contacted by a commercial search engine company about contributing our catalog for use in the search engine’s system. Does the Policy permit the transfer of WorldCat-derived records from our catalog to the search engine company?

Since the search engine company is a commercial organization, there must be an agreement in place between OCLC and the search engine company prior to the transfer of WorldCat-derived records. OCLC can let you know if it has an agreement with the search engine company in question. Please submit a WorldCat Record Use Form to OCLC or ask the search engine company to submit a WorldCat Record Use Form to OCLC and we will reply within five business days.

UPDATE: It seems I am not the only one who had this thought. Here at the Disruptive Library Technology Jester blog there is some parsing of the new policy to reach the same conclusion. Also, here is a set of reactions from the librarian community. They’re a passionate bunch.