Skip to main content
 

Just amazing work by @chankfonts including design, content (and fonts!) and photography by @midwestartbuyer and tons of other people...

Will upload to @internetarchive when I'm happy with the results..

 

Actually with some tweaks I got all the grunge list content I was looking for!

Still think it would be awesome if the backup tapes were found. @herf wrote a great article about the lyrics part of the site:

http://stereopsis.com/lyrics.html

 

So has anyone made a poster with Udi Manbar's "It's our problem" statement?

It was one of the more inspiring things I had seen when I was watching old TFIFs (when we could do that..)

See the attached Noogler training deck and consider how things have changed today. (Also a reminder of go/gthink and the insight reports on https://sites.google.com/a/google.com/gthink/g/resources/analyst-insights-reports)

if users can't spell, it's our problem

if they don't know how to form the query, it's our problem

if they don't know what words to use, it's our problem

if they can't speak the language, it’s our problem

if there is not enough content on the web,
it’s our problem

if the web is too slow, it’s our problem

https://docs.google.com/presentation/d/1ZBLK7WfoLHRveZs8FiSlMvgVlkKWNKDhiEgi2Ocy4W8/edit#slide=id.i1...





 

Godwin's law; but for content recommendations..

h/t to Ray Cromwell who had the movie Iron Sky recommended in Google Feed..... (which has hitler riding a dinosaur on the moon....)





 

go/tacotown is now deprecated code wrapped in a dead product (Buzz) and contained in another deprecated product (consumer G+).

content warning: excessive food.


https://www.youtube.com/watch?v=evUWersr7pc

And I sure hope it doesn't get wrapped in a big Apps deprecated blueberry pancake.





 

A ton of great content here including Vint's talk on Digital Preservation.

As a Googler you'll want to have a thick skin and open mind for many of these.

https://decentralizedweb.net/videos/





 

This is a very well written article about how the film/content industries are attempting (and failing) to keep up with archiving their data.

Reading about LTO I now understand why gtape is moving to http://go/chronicle and http://go/timelock (Also interesting that timelock has a Q2 OKR for 500PB, which is 10x the size of the film archive mentioned in the article.)

I'm also curious if Google has had to delete data that we've regretted. Do we have all of the old street view data?



http://spectrum.ieee.org/computing/it/the-lost-picture-show-hollywood-archivists-cant-outpace-obsole...






 

Found some weekend reading based on the abstract alone.


_Although it has been a part of the dark underbelly of the Internet since its inception, recent events have brought the discussion board site 4chan to the forefront of the world's collective mind. In particular, /pol/, 4chan's "Politically Incorrect" board has become a central figure in the outlandish 2016 Presidential election. Even though 4chan has long been viewed as the "final boss of the Internet," it remains relatively unstudied in the academic literature.
In this paper we analyze /pol/ along several axes using a dataset of over 8M posts. We first perform a general characterization that reveals how active posters are, as well as how some unique features of 4chan affect the flow of discussion. We then analyze the content posted to /pol/ with a focus on determining topics of interest and types of media shared, as well as the usage of hate speech and differences in poster demographics. We additionally provide quantitative evidence of /pol/'s collective attacks on other social media platforms. We perform a quantitative case study of /pol/'s attempt to poison anti-trolling machine learning technology by altering the language of hate on social media. Then, via analysis of comments from the 10s of thousands of YouTube videos linked on /pol/, we provide a mechanism for detecting attacks from /pol/ threads on 3rd party social media services._


https://arxiv.org/abs/1610.03452






 

_IPFS - Content Addressed, Versioned, P2P File System_ - Research

Pretty well written paper that covers the basics of IPFS.


https://raw.githubusercontent.com/ipfs/papers/master/ipfs-cap2pfs/ipfs-p2p-file-system.pdf






 

Don't forget to takeout your Orkut content. I almost lost my Vox.com content way back and would have regretted that.

Takeout available until "September 2016".


https://brasil.googleblog.com/2014/06/adeus-ao-orkut.html#__sid=md1






 

_If you create some content but there is no one to see it, does it exist?_

from +112427496122904544974


https://paul.kinlan.me/ephemeral-social-network/






 

Blockchain based content publishing system is full of content about making money with content.


https://steemit.com/






 

Anyone have a contact at npr.org?

Their pages are generating invalid Content-Type headers:

Content-Type: text/html;;charset=UTF-8

Fixing on our end, but someone should let them know..









 

A great read, but they missed out on another big issue - tons of content is semi-private and will be lost forever.



http://www.theatlantic.com/technology/archive/2015/10/raiders-of-the-lost-web/409210/?single_page=tr...






 

Here's a 30 day challenge.

Only consume Youtube and Play Music content in a language you barely know.

This is how much of the world interacts with our products.

All of a sudden you understand why lyric videos are so popular on Youtube. It also explains weird comments on G+ from people that are just playing around with English.

I've been doing something like that for a few days. It's eye opening.

Oh and if you're interested in music discovery from Suisse Romande see the link below.


https://www.youtube.com/user/LesReperagesCouleur3






 

Direct link: https://www.tumblr.com/tv

Nice intro with TV static, but I do wonder if anyone born after 1990 knows what that is..

- Supports next/prev, but no pause
- Content is limited, not very customizable.
- No context cues presented, no intros/bumpers other than the initial splash screen.


http://venturebeat.com/2015/06/25/tumblr-launches-tumblr-tv-a-platform-for-surfing-gifs/






 

I find the Blogger decision baffling given the effort in Social for Project R

http://go/projectr

What's the story here?  @103643500382916616176 or @106642279061626308531  can you weigh in?

I was hopeful that we'd offer a way for adults to have adult conversations around adult topics and content.  Hell, sometimes I'd like to restrict my youtube comments to only adults (*yes* we have the ability in the ASBE backend to do this!)

I'm also sad that we're spending any amount of time to help Tobacco companies peddle their wares, which it seems is the main driver for a lot of this...









 

Search the html, not the content.  I can think of a bunch of ways that this is useful...


http://nerdydata.com/






 


_"embedded restaurants and taxis"_

This is why I think we need to do more, not less with live content, especially on mobile....

http://ben-evans.com/benedictevans/2014/8/1/app-unbundling-search-and-discovery






 

 

So I was just saying to myself wouldn't it be great if we could have incognito iframes and behold! there it is..  iframe sandbox's unique origin solves that problem...

Anyone using this in practice?  I'd love to use this as a way to serve up 3P content and supply the appropriate user information directly from the parent via postMessage to avoid user-overlap confusion.


http://www.html5rocks.com/en/tutorials/security/sandboxed-iframes/






 

Pluck spammed me about their new events/hangouts-like functionality.

Hello Paul,

Can you do this today?

Hold a live event, on your website or Facebook page. Invite an executive, topical expert or celebrity to host it. Let hundreds of thousands of your fans not only follow the event, but participate with comments, questions and photo and video uploads. Manage your event’s tone, pace and energy by choosing the content your fans see. Compel them to spend an average of 40 minutes on your website or Facebook page.

How about this?

Enhance a virtual or in-person event with a dynamic social wall experience. Collect, curate and display related social media content directly on your website. Moderate fan content – comments, questions, photos, videos, tweets, Facebook posts – so that it best represents your brand. Add your own original content or editorial commentary.  Create a complete, highly visual story around a particular topic or event.

Creating a Real-time Experience is about starting a conversation, not joining it. It’s about listening and engaging. Real-time Experiences are about collaborative content, and leveraging unique audience consumption opportunities on your owned media.

Pluck’s CoveritLive Real-time Engagement Platform gives you everything you need to design and produce highly interactive real-time experiences. More than 1,500 leading brands, retailers and publishers like Subway, EA, Jeep, Ford and Anthropologie have created real-time experiences with CoveritLive. They have connected with their fans – and generated new ones – through Live Blogs, Q&A Sessions, Social Walls, Second Screen Experiences, Featured Galleries and Facebook Chats.

For digital marketers seeking customer and shopper engagement, real-time experiences are highly differentiated and very sticky. CoveritLive routinely supports events with hundreds of thousands of attendees, who on average are engaged for 40 minutes.

And here’s the best part. You can download CoveritLive and run trial events of your own at no cost. What are you waiting for?









 

Paradigm shift.  Post->Create, View->Search.

I think part of the YT problem is one of perception that reminds of the early days of Gmail.  Remember when you 'put' messages in 'folders'?  That's what people think of when they post.  Instead you have 'Content' that is created that has certain attributes.

When you have content with attributes you can use search to surface it, rank it, and filter it.

So my minor suggestion for Youtube comments is to add a really simple thing: Add _Searching for Comments..._ to the loading indicator.









 

T-Mobile user with a grandfathered plan?

Install this before you upgrade to KitKat...


https://play.google.com/store/apps/details?id=com.pdanet#body-content






 

A song about gentrification in San Francisco...  It pains me to see Google vilified in this way when it enables so much (including this content itself)

Sadly I don't hold out much hope for technology fixing the dysfunctional housing situation in the Bay Area.

[And where's my soundcloud embed  @112474028169763877889 ??]

Also if you decide to watch the youtube video remember that _This video has been age-restricted based on our Community Guidelines_


http://soundcloud.com/persia415/persia-google-google-apps-apps






 

Looks like we're not quite   for image filtering.

TL;DR -- SIte displays porn content from freebase/knowledge graph that violates AdSense ToS.


_A quite strange thing happened recently. I have a news website and I'm using Freebase API to provide readers for more in depth information about the content of news articles.  The site runs Google's AdSense ads._

_Following the publication of a local news, on the sidebar containing text and images downloaded through Freebase API, an image of a woman with her bare breasts was displayed._

_Well, that image, made available by Freebase, which is a Google product, was considered obscene by the team of AdSense (also a Google product), which decided to suspend the service. A sort of vicious circle of Google, exercising a censorship on itself._

_Given the impossibility to make Google people understand that the visualization of such kind of images was occasional and not intentional, I'd prefer to avoid such problems in the future. Hence, I'd like to know whether there's some way, when using Freebase API, to filter out images considered obscene by the AdSense team._









 

Everyone.  Please add sensible canonicals.
- Especially if you have a 'enable SSL' bit.
- or if you have content that spans multiple domains
- or if you respond to /foo and /foo/ without a 301 redirect on one of them.

I've noticed two canonicalization issues (project hosting and sites) in the past few days.  I'm guessing there are many more out there..


http://support.google.com/webmasters/bin/answer.py?hl=en&answer=139394&ctx=plusone






 

The secret truth -- the ITU wants deep packet inspection to fix their Content-Type charset headers..


https://plus.google.com/photos/107786897865850743842/albums/5823819709617866049/5823819712652075282






 

Interesting  points.

Caught up with an old friend.  His teenage sons are both big kik users on their iPod touches.  They don't want an iPhone since they're not allowed in school but the iPod is.

The son of my wife's stylist asked me if we could add file transfer to Hangouts or messenger.  I'm somewhat dubious on this (seems like a good way to spread virii -- however adding Drive sharing to chat might be interesting (and safer..) and is orthogonal with Photo sharing in Messenger.

Send file == upload to drive, share file from drive to friend, share link over chat channel, render drive content using embed.


http://kik.com/






 

Teach Me How to -Dougie- Viral Loop.

What you see here is a work of art for making viral loops.

The page leading up to this asked me to _confirm_ that someone had endorsed me.  I accepted the endorsement (adding it to my profile).

This page uses some decent semantic analysis to generate 'Skills you might Endorse', and a one-click way to generate a 4x viral loop (not to mention all the other mechanisms via 'see more' and loading up more endorsements.

Performing actions here also help improve the algorithms.  If one person tags a user with a skill then you can present the same tag once added to the profile to other users and have a high likelihood of success.

And at the end of the day LinkedIn will have a much better way of finding strong candidates with specific skills; users will have more profile content, etc etc.


https://plus.google.com/photos/107786897865850743842/albums/5795440289929294673/5795440287701076610






 

It's been a while since I've encountered geo-restricted content.   Man it sucks.  Take my money dammit!

Oh, and if anyone wants to help me buy something from the France iTunes store LMK before I go off and try to find gift cards from the far corners of the internet.
Happy to reciprocate too!









 

Embeds United!

I'm somewhat alarmed that there really isn't a "Life of a Widget/Embed" that has best practices on how to embed Google Content on third party sites.

Has anyone tackled this before? Obviously +1, adsense, youtube, maps, analytics and others all use their own mechanisms.

Is it time to consider deploying common infrastructure for these use cases? Especially for systems that require the ability to identify a Google user?

So here's my proposal.

* Use oauth2 semantics everywhere: tokens and api requests.
* Use a single bootstrap system to identify js domain, collect API keys as needed.
* Generate two new types of OAuth tokens that do not contain user info.
- 'Anonymous-by-domain'
- 'Anonymous-by-domain-by-api-key'
* Allow an account widget that can move you from anonymous -> identified -> authorized. (including multilogin, signin and account registration)

If all the various embeds used this common system we'd have a way to enable personalization on third party sites more easily and provide a clear way to go beyond.

Am I barking up the wrong tree?









 

.. random ES thought on content-based circle classifications..

Imagine if we could infer the type of circle based on either the content you post there (via NLP), the linguistic name, or user specification?

We could:
-- boost/dampen the volume of a post based on circle matching:
I have you in my _Photos_ circle, you share to your _heavy_metal_ circle == dampen.
-- You post your high scores to _public_ + _chess_, I follow you in a circle named _games_ == boost.
-- My wife posts to _family_, I have her in _family_ == boost
-- My wife posts to _Tennis_ and public, I don't have her in a sports circle == dampen.


The reason I've been thinking about this is another what-if --
-- what if Apps had a one-to-one relationship with a plus page?
-- When I add an App, it can circle me back.
-- If I perform an action on the third party site the app can plus-mention me, or it can group together users into relavant circles to send updates.

-- I can put noisy apps in a circle and lower the volume.
Note that App could be a Game (like Angry Birds), an App (like Remember the Milk) or even a web site (like 8tracks or Soundcloud..)

Maybe this is stretching the model too much, but it's certainly fun to ponder!









 

A new future for Buzz?

Instead of getting rid of Buzz maybe we can turn it into what it's defacto become -- an echo of activity on the web.

Just as we have the games stream we can have the buzz stream -- the collection of imported content from you and your friends. Two simple changes are needed:

* Don't allow new posts to Buzz.
* Add +1 and share buttons to imported content -- allowing you to share the best stuff to your main stream.

Thoughts?









 

LOL - comment from a thread on consumer + :

*Darren Bounds* - I hope the day never comes where I'm able to import RSS, Atom, Twitter or Facebook content directly into +. Give a Hoot. Don't Pollute.









 

RT @google: We're releasing +1 buttons to the whole web, making it easier to recommend content to your friends & contacts http://goo.gl/ ...

 

Interesting bumping+unread side-effect.

When you add a new person to your circles you get all of their posts at the top of the stream. I like that, but it's a little jarring since you see older content on top of the newer stuff...









 

Cross posted from a buzz thread about implicit graph matching.



creepy/yuck factor? At one time organ transplants were considered frankenstein creepy. Street view was (and some say is) considered creepy. The facebook news feed was once considered creepy. People's views change over time, especially if there's an underlying value that one can understand. For all of the above I'd say that the value is there and folks have adjusted.


Having a benevolent trusted partner that can search all of your content and knows you better than you know yourself using heuristics and AI? That stuff is science fiction magic. We can make it happen, and it forms the basis for serendipitous search.


I'd say we're going to have a lot more moments like this where we surface stuff to the user that they didn't expect nor ask for. Making everything opt-in is not the answer -- there has to be a middle ground here where value is given before it is asked for.









 

Had a great lunch with @macduff@google.com and @rkhare@google.com ..

Based on our conversations I'm going to claim this idiom *social referer*

For a number of reasons it might make sense to send the originating social profile in an API request. For example a nytimes +1 could include an sgn URL with the API request.

POST /plusone/create
Content-Type: text/json
X-Social-Referer: sgn://nytimes.com?pk=12345

This would help with spam mitigation (we could throttle individual accounts at a third party instead of an entire site) It would also allow us to have a reciprocal relationship with our partners since they can see the profile information in one way only. Of course we should be willing to send social referers to sites we make API calls to as well.

POST /api/v2/contacts
Host: api.yahoo.com
X-Social-Referer: sgn://google.com?pk=OBFUSCATEDGAIAID

Stupid idea? Good idea? Sound off in the comments!