Subscribe: Don Marti
Added By: Feedage Forager Feedage Grade B rated
Language: English
ads  advertising  based  bot  change  don  information  marketing  news  open  people  search  site  sites  twitter  user 
Rate this Feed
Rate this feedRate this feedRate this feedRate this feedRate this feed
Rate this feed 1 starRate this feed 2 starRate this feed 3 starRate this feed 4 starRate this feed 5 star

Comments (0)

Feed Details and Statistics Feed Statistics
Preview: Don Marti

Don Marti

personal blog feed

Last Build Date: Tue, 22 Aug 2017 19:22:26 GMT


List-based and behavior-based tracking protection

Tue, 22 Aug 2017 07:00:00 GMT

In the news...

User privacy is at risk from both hackers and lawyers. Right now, lawyers are better at attacking lists, and hackers are better at modifying tracker behavior to get around protections.

The more I think about it, the more that I think it's counterproductive to try to come up with one grand unified set of protection rules or cookie policies for everybody.

Spam filters don't submit their scoring rules to ANSI—spammers would just work around them.

Search engines don't standardize and publish their algorithms, because gray hat SEOs would just use the standard to make useless word salad pages that score high.

And different people have different needs.

If you're a customer service rep at an HERBAL ENERGY SUPPLEMENTS company, you need a spam filter that can adjust for your real mail. And any user of a site that has problems with list-based tracking protection will need to have the browser adjust, and rely more on cleaning up third-party state after a session instead of blocking outright.

Does your company intranet become unusable if you fail to accept third-party tracking that comes from an internal domain that your employer acquired and still has some services running on? Browser developers can't decide up front, so the browser will need to adjust. Every change breaks someone's workflow.

That means the browser has to work to help the user pick a working set of protection methods and rules.

0. Send accurate Do Not Track

Inform sites of the user’s preferences on data sharing. (This will be more important in the future because Europe, but privacy-crazed Eurocrats will not save us from having to do our share of the work.

1. Block connections to third-party trackers

This will need to include both list-based protection and monitoring tracking behavior, like Privacy Badger, because hackers and lawyers are good at getting around different ones.

2. Limit data sent to third-party sites

Apple Safari does this, so it's likely to get easier to do cookie double keying without breaking sites.

3. Scramble or delete unsafe data

If a tracking cookie or other identifier does get through, delete or scramble it on leaving the site or later, as the Self-Destructing Cookies extension does. This could be a good backup for when the browser "learns" that a user needs some third-party state to do something like a shopping cart or comment form, but then doesn't want the info to be used for "ads that follow me around" later.

How is everyone's tracking protection working? An update

Sun, 20 Aug 2017 07:00:00 GMT

When I set up this blog, I put in a script to check how many of the users here are protected from third-party tracking.

The best answer for now is 31%. Of the clients that ran JavaScript on this site over the past two weeks, 31% did not also run JavaScript from the Aloodo "fake third-party tracker".

The script is here: /code/check3p

This is not as good as I had hoped (turn on your tracking protection, people! Don't get tricked by ad blockers that leave you unprotected by default!) but it's a start.

The Information Trust Exchange is doing research on the problem of third-party tracking at news sites. News industry consultant Greg Swanson:

All of the conversations on the newspaper side have been focused on how can we join the advertising technology ecosystem. For example, how can a daily newspaper site in Bismarck, North Dakota deliver targeted advertising to a higher-value soccer mom? And none of the newspapers them have considered the fact that when they join that ecosystem they are enabling spam sites, fraudulent sites – enabling those sites to get a higher CPM rate by parasitically riding on the data collected from the higher-value newspaper sites.

More info: Aloodo for web publishers.

SEO hats and the browser of the future

Sat, 19 Aug 2017 07:00:00 GMT

The field of Search Engine Optimization has white hat SEO, black hat SEO, and gray hat SEO. White hat SEO helps a user get a better search result, and complies with search engine policies. Examples include accurately using the same words that users search on, and getting honest inbound links. Black hat SEO is clearly against search engine policies. Link farming, keyword stuffing, cloaking, and a zillion other schemes. If they see you doing it, your site gets penalized in search results. Gray hat SEO is everything that doesn't help the user get a better search result, but technically doesn't violate a search engine policy. Most SEO experts advise you not to put a lot of time and effort into gray hat, because eventually the search engines will notice your gray hat scheme and start penalizing sites that do it. Gray hat is just stuff that's going to be black hat when the search engines figure it out. Adtech has gray hat, too. Rocket Fuel Awarded Two Patents to Help Leverage First-Party Cookies to More Meaningfully Reach Consumers. This scheme seems to be intended to get around existing third-party cookie protection, which is turned on by default in Apple Safari and available in other browsers. But how long will it work? Maybe the browser of the future won't run a "kangaroo cookie court" but will ship with a built-in "kangaroo law school" so that each copy of the browser will develop its own local "courts" and its own local "case law" based on the user's choices. It will become harder to predict how long any single gray hat adtech scheme will continue working. In the big picture: in order to sell advertising you need to give the advertiser some credible information on who the audience is. Since the "browser wars" of the 1990s, most browsers have been bad at protecting personal information about the user, so web advertising has become a game where a whole bunch of companies compete to covertly capture as much user info as they can. Today, browsers are getting better at implementing people's preferences about sharing their information. The result is a change in the rules of the game. Investment in taking people's personal info is becoming less rewarding, as browsers compete to reflect people's preferences. (That patent will be irrelevant thanks to browser updates long before it expires.) Adfraud is the other half of this story. Fraudbots are getting smarter at creating human-looking ad impressions just as humans are getting better protected. If you think that a web publisher's response to harder-to-detect bots, viewing more high-CPM video ads, should be "pivot to video!!1!!" I don't know if I can help you. And investments in building sites and brands that are trustworthy enough for people to want to share their information will tend to become more rewarding. (This shift naturally leads to complaints from people who are used to winning the old game, but will probably be better for customers who want to use trustworthy brands and for people who want to earn money by making ad-supported news and cultural works.) Bonus links One of the big advertising groups is partnering with Digital Content Next’s trust-focused ad marketplace Partisanship, Propaganda, and Disinformation: Online Media and the 2016 U.S. Presidential Election ANA Endorses TrustX, Encourages Members To Use Programmatic Media-Buying Stamp Of Approval Call for Papers: Policy and Internet Special Issue on Reframing ‘Fake News’: Architectures, Influence, and Automation Time to sink the Admiral (or, why using the DMCA to block adblockers is a bad move) I'm a woman in computer science. Let me ladysplain the Google memo to you. Easylist block list removes entry after DMCA takedown notice Will Cities Ever Outsmart Rats? Uber drivers gang up to cause surge pricing, research says Google reveals sites with ‘failing’ ads, including Forbes, LA Times Koch group, Craigslist founder come to Techdirt'[...]

cdparanoia returned code 73

Fri, 18 Aug 2017 07:00:00 GMT

Welcome, people googling for the above error message.

I saw the error

cdparanoia returned code 73

and it turns out I was trying to run two abcde processes in two terminal windows. Kill the second one and the error goes away.

Hope your problem was as simple as that.

ePrivacy and marketing budgets

Wed, 16 Aug 2017 07:00:00 GMT

(Update 18 Aug 2017: this post is also available at Digital Content Next.) As far as I know, there are three ways to match an ad to a user. User intent: Show an ad based on what the user is searching for. Old-school version: the Yellow Pages. Context: Show an ad based on where the user is, or what the user is interested in. Old-school versions: highway billboards (geographic context), specialized magazines (interest context). User identity: Show an ad based on who the user is. Old-school version: direct mail. Most online advertising is matched to the user based on a mix of all three. And different players have different pieces of the action for each one. For user intent, search engines are the gatekeepers. The other winners from matching ads to users by intent are browsers and mobile platforms, who get paid to set their default search engine. Advertising based on context rewards the owners of reputations for producing high-quality news, information, and cultural works. Finally, user identity now has a whole Lumascape of vendors in a variety of categories, all offering to help identify users in some way. (the Lumascape is rapidly consolidating, but that's another story.) Few of the web ads that you might see today are matched to you purely based on one of the three methods. Investments in all three tend to shift as the available technology, and the prevailing norms and laws, change. Enough background. Randall Rothenberg of the IAB is concerned about the proposed ePrivacy Regulation in Europe, and writes, The basic functionality of the internet, which is built on data exchanges between a user’s computer and publishers’ servers, can no longer be used for the delivery of advertising unless the consumer agrees to receive the ads – but the publisher must deliver content to that consumer regardless. This doesn't look accurate. I don't know of any proposal that would require publishers to serve users who block ads entirely. What Rothenberg is really complaining about is that the proposed regulation would limit the ability of sites and ad intermediaries to match ads to users based on user identity, forcing them to rely on user intent and context. If users choose to block ads delivered from ad servers that use their personal data without permission, then sites won't be able to refuse to serve them the content, but will be able to run ads that are relevant to the content of the site. As far as I can tell, sites would still be able to pop a "turn off your ad blocker" message in place of a news story if the user was blocking an ad placed purely by context, magazine style. Privacy regulation is not so much an attack on the basic functionality of the Internet, as it is a shift that lowers the return on investment on knowing who the user is, and drives up the return on investment on providing search results and content. That's a big change in who gets paid: more money for search and for trustworthy content brands, and less for adtech intermediaries that depend on user tracking. Advertising: a fair deal for the user? That depends. Search advertising is clearly the result of a user choice. The user chooses to view ads that come with search results, as part of choosing to do a search. As long as the ads are marked as ads, it's pretty obvious what is happening. The same goes for ads placed in context. The advertiser trades economic signal, in the form of costly support of an ad-supported resource, for the user's attention. This is common in magazine and broadcast advertising, and when you use a site with one of the (rare) pure in-context ad platforms such as Project Wonderful, it works about the same way. The place where things start to get problematic is ads based on user identity, placed by tracking users from site to site. The more that users learn how their data is used, the less tracking they tend to want. In one survey, 66% of adult Americans said they do not want mark[...]

Moral values in society

Tue, 08 Aug 2017 07:00:00 GMT

Moral values in society are collapsing? Really? Elizabeth Stoker Bruenig writes, The baseline moral values of poor people do not, in fact, differ that much from those of the rich. (read the whole thing).

Unfortunately, if you read the fine print, it's more complicated than that. Any market economy depends on establishing trust between people who trade with each other. Tim Harford writes,

Being able to trust people might seem like a pleasant luxury, but economists are starting to believe that it’s rather more important than that. Trust is about more than whether you can leave your house unlocked; it is responsible for the difference between the richest countries and the poorest.

Somehow, over thousands of years, business people have built up a set of norms about high-status and low-status business activities. Craftsmanship, consistent supply of high-quality staple goods, and construction of noteworthy projects are high-status activities. Usury and deception are examples of low-status activities. (You make your money in quarters, gambling with retired people? You lend people $100 until Friday at a 300% interest rate? No club invitation for you.)

Somehow, though, that is now changing in the USA. Those who earn money through deception now have seats at the same table as legitimate business. Maybe it started with the shift into "consumer credit" by respectable banks. But why were high-status bankers willing to play loan shark to begin with? Something had to have been building, culturally. (It started too early to blame the Baby Boomers.)

We tend to blame information technology companies for complex, one-sided Terms of Service and EULAs, but it's not so much a tech trend as it is a general business culture trend. It shows up in tech fast, because rapid technology change provides cover and concealment for simultaneous changes in business terms. US business was rapidly losing its connection to basic norms when it was still moving at the speed of FedEx and fax. (You can't say, all of a sudden, "car crashes in existing fast-food drive-thrus are subject to arbitration in Unfreedonia" but you can stick that kind of term into a new service's ToS.) There's some kind of relativistic effect going on. Tech bros just seem like bigger douchebags because they're moving faster.

Regulation isn't the answer. We have a system in which business people can hire lobbyists to buy the laws and regulations we want. The question is whether we're going to use our regulatory capture powers in a shortsighted, society-eroding hustler way, or in a conservative way. Economic conservatism means not just limiting centralized state control of capital, but preserving the balance among all the long-standing stewards of capital, including households, municipalities, and religious and educational institutions. Economic conservatism and radical free-marketism are fundamentally different.

People blame trashy media for the erosion of norms among the poor, so let's borrow that explanation for the erosion of norms among the rich as well. Maybe our problem with business norms results from the globablization and sensationalism of business media. Joe CEO isn't just the most important corporate leader of Mt. Rose, MN, any more—on a global scale he's just another broke-ass hustler.

More random links

Sun, 06 Aug 2017 07:00:00 GMT

Not the Google story everyone is talking about, but related: Google Is Matching Your Offline Buying With Its Online Ads, But It Isn’t Sharing How. (If a company becomes known for doing creepy shit, it will get job applications from creepy people, and at a large enough company some of them will get hired. Related: The Al Capone theory of sexual harassment) Least surprising news story ever: The Campaign Against Facebook And Google's Ad "Duopoly" Is Going Nowhere Independent online publishers can't beat the big surveillance marketing companies at surveillance marketing? How about they try to beat Amazon and Microsoft at cloud services, or Apple and Lenovo at laptop computers? There are possible winning strategies for web publishers, but doing the same as the incumbents with less money and less data is not one of them. Meanwhile, from an investor point of view: It’s the Biggest Scandal in Tech (and no one’s talking about it) Missing the best investment advice: get out of any B-list adtech company that is at risk of getting forced into a low-value acquisition by a sustained fraud story. Or short it and research the fraud story yourself. Did somebody at The Atlantic get a loud phone notification during a classical music concert or something? Your Smartphone Reduces Your Brainpower, Even If It's Just Sitting There and Have Smartphones Destroyed A Generation?, by Jean M. Twenge, The Atlantic Good news: Math journal editors resign to start rival open-access journal Apple’s Upcoming Safari Changes Will Shake Up Ad Tech: Not surprisingly, Facebook and Amazon are the big winners in this change. Most of their users come every day or at least every week. And even the mobile users click on links often, which, on Facebook, takes them to a browser. These companies will also be able to buy ad inventory on Safari at lower prices because many of the high-dollar bidders will go away. A good start by Apple, but other browsers can do better. (Every click on a Facebook ad from a local business is $0.65 of marketing money that's not going to local news, Little League sponsorships, and other legit places.) Still on the upward slope of the Peak Advertising curve: Facebook 'dark ads' can swing political opinions, research shows You’re more likely to hear from tech employers if you have one of these 10 things on your resume (and only 2 of them are proprietary. These kids today don't know how good they have it.) The Pac-Man Rule at Conferences How “Demo-or-Die” Helped My Career [...]

Pragmatists for copyleft, or, corporate hive minds don't accept software licenses

Sun, 06 Aug 2017 07:00:00 GMT

One of the common oversimplifications in discussing open-source software licenses is that copyleft licenses are "idealistic" while non-copyleft licenses are "pragmatic." But that's not all there is to it.

The problem is that most people redistributing licensed code are doing so in an organizational context. And no human organization is a hive mind where those who participate within it subordinate their goals to that of the collective. Human organizations are full of of people with their own motivations.

Instead of treating the downstrem developer's employer as a hive mind, it can be more producive to assume good faith on the part of the individual who intends to contribute to the software, and think about the license from the point of view of a real person.

Releasing source for a derivative work costs time and money. The well-intentioned "downstream" contributor wants his or her organization to make those investments, but he or she has to make a case for them. The presence of copyleft helps steer the decision in the right direction. Jane Hacker at an organization planning to release a derivative work can say, matter-of-factly, "we need to comply with the upstream license" if copyleft is involved. The organization is then more likely to do the right thing. There are always violations, but the license is a nudge in the right direction.

(The extreme case is university licensing offices. University-owned software patents can exclude a graduate student from his or her own project when the student leaves the university, unless he or she had the foresight to build it as a derivative work of something under copyleft.)

Copyleft isn't a magic commons-building tool, and it isn't right for every situation. But it can be enough to push an organization over the line. (One place where I worked had to a do a source release for one dependency licensed under GPLv2, and it turned out to be easist to just build one big source code release with all the dependencies in it, and offer that.)

Hey kids, favicon!

Sat, 05 Aug 2017 07:00:00 GMT

Finally fixed those 404s from browsers looking for favicon.ico on this blog.

  1. Google image search for images where "reuse with modification" is allowed.

  2. Found this high-quality lab mouse SVG image.

  3. Opened it in GNU Image Manipulation Program, posterized, cropped to a square. Kept the transparent background.

  4. Just went to and did what it says, and added the resulting images and markup to the site.

That's about it. Now there's a little mouse in the browser tab (and it should do the right thing with the icons if someone pins it to their home screen on mobile.)

Why surveillance marketers don't worry about GDPR (but privacy nerds should)

Tue, 01 Aug 2017 07:00:00 GMT

A lot of privacy people these days sound like a little kid arguing with a sibling. You're going to be in big trouble when Dad gets home!

Dad, here, is the European Union, who's going to put the General Data Protection Regulation foot down, and then, oh, boy, those naughty surveillance marketers are going to catch it, and wish that they had been listening to us about privacy all along.


But Internet politics never works like that. Sure, European politicians don't want to hand over power to the right-wing factions who are better at surveillance marketing than they are. And foreign agents use Facebook (and other US-based companies) to attack legit political systems. But that stuff is not going to be enough to save GDPR.

The problem is that perfectly normal businesses are using GDPR-violating sneaky tracking pixels and other surveillance marketing as part of their daily marketing routine.

As the GDPR deadline approaches, surveillance marketers in Europe are going to sigh and painstakingly explain to European politicians that of course this GDPR thing isn't going to work. "You see, politicians, it's an example of political overreach that completely conflicts with technical reality." European surveillance marketers will use the same kind of language about GDPR that the freedom-loving side used when we talked about the proposed CBDTPA. It's just going to Break the Internet! People will lose their jobs!

The result is predictable. GDPR will be delayed, festooned with exceptions, or both, and the hoped-for top-down solution to privacy problems will not come. There's no shortcut. We'll only get a replacement for surveillance marketing when we build the tools, the networks, the business processes, the customer/voter norms, and then the political power.

Extracting just the audio from big video files

Sat, 29 Jul 2017 07:00:00 GMT

Got a big video, and want a copy of just the audio for listening on a device with limited storage? Use Soundconverter.

soundconverter -b -m mp3 -s .mp3 long-video.webm

(MP3 patents are expired now, hooray! I'm just using MP3 here because if I get a rental car that lets me plug in a USB stick for listening, the MP3 format is most likely to be supported.)

Soundconverter has a GUI but you can use -b for batch mode from the shell. soundconverter --help for help. You do need to set both the MIME type, with -m, and the file suffix, with -s.

Online ads don't matter to P&G

Fri, 28 Jul 2017 07:00:00 GMT

In the news: P&G Cuts More Than $100 Million in ‘Largely Ineffective’ Digital Ads

Not surprising.

Proctor & Gamble makes products that help you comply with widely held cleanliness norms.

Digital ads are micro-targeted to you as an individual.

That's the worst possible brand/medium fit. If you don't know that the people who expect you to keep your house or body clean are going to be aware of the same product, how do you know whether to buy it?

Bonus link from Bob Hoffman last year: Will The P&G Story Bring Down Ad Tech? Please?

Got a reply from Twitter

Wed, 26 Jul 2017 07:00:00 GMT

I thought it would be fun to try Twitter ads, and, not surprisingly, I started getting fake followers pretty quickly after I started a Twitter follower campaign. Since I'm paying nine cents a head for these followers, I don't want to get ripped off. So naturally I put in a support ticket to Twitter, and just heard back. Thanks for writing in about the quality of followers and engagements. One of the advantages of the Twitter Ads platform is that any RTs of your promoted ads are sent to the retweeting account's followers as an organic tweet. Any engagements that result are not charged, however followers gained may not align with the original campaign's targeting criteria. These earned followers or engagements do show in the campaign dashboard and are used to calculate cost per engagement, however you are not charged for them directly. Twitter also passes all promoted engagements through a filtering mechanism to avoid charging advertisers for any low-quality or invalid engagements. These filters run on a set schedule so the engagements may show in the campaign dashboard, but will be deducted from the amount outstanding and will not be charged to your credit card. If you have any further questions, please don't hesitate to reply. That's pretty dense San Francisco speak, so let me see if I can translate to the equivalent for a normal product. Hey, what are these rat turds doing in my raisin bran? Thanks for writing in about the quality of your raisin bran eating experience. One of the advantages of the raisin bran platform is that during the production process, your raisin bran is made available to our rodent partners as an organic asset. I paid for raisin bran, so why are you selling me raisin-plus-rat-turds bran? Any ingredients that result from rodent engagement are not charged, however ingredients gained may not align with your original raisin-eating criteria. Can I have my money back? We pass all raisin bran sales through a filtering mechanism to avoid charging you for invalid ingredients. The total weight of the product, as printed on the box, includes these ingredients, but the weight of invalid ingredients will be deducted from the amount charged to your credit card. So how can I tell which rat turds are "organic" so I'm not paying for them, and which are the ones that you just didn't catch and are charging me for? (?) Buying Twitter followers: Fiverr or Twitter? On Fiverr, Twitter followers are about half a cent each ($5/1000). On Twitter, I'm gettting followers for about 9 cents each. The Twitter price is about 18x the Fiverr price. But every follower that someone else buys on Fiverr has to be "aged" and disguised in order to look realistic enough not to get banned. The bot-herders have to follow legit follower campaigns such as mine and not just their paying customers. (I call them "sleepers." They do all sorts of natural things (following suggested accounts, tweeting quotes) aging into "trusted" zone.)— Andréa López (@bluechoochoo) July 21, 2017 If Twitter is selling those "follow" actions to me for nine cents each, and the bot-herder is only making half a cent, how is Twitter not making more from bogus Twitter followers than the bot-herders are? If you're verified on Twitter, you may not be seeing how much of a shitshow their ad business is. Maybe the're going to have to sell Twitter to me sooner than I thought. [...]

Incentivizing production of information goods

Wed, 26 Jul 2017 07:00:00 GMT

Just thinking about approaches to incentivizing production of information goods, and where futures markets might fit in.

Artificial property

Article 1, Section 8, of the US Constitution still covers this one best.

To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries;

We know about the problems with this one. It encourages all kinds of rent-seeking and freedom-menacing behavior by the holders of property interests in information. And the transaction costs are too high to incentivize the production of some useful kinds of information.

Commoditize the complement

Joel Spolsky explained it best, in Strategy Letter V. Smart companies try to commoditize their products’ complements. (See also: the list of business models in the Some Easily Rebutted Objections to GNU's Goals section of the GNU Manifesto)

This one has been shown to work for some categories of information goods but not others. (We have Free world-class browsers and OS kernels because search engines and hardware are complements. We don't have free world-class software in categories such as CAD.)


Release a free information good as a way to signal competence in performing a service, or at least a large investment by the author in persuading others that the author is competent. Works at the level of the individual labor market and in consulting. Don't know if this works in other areas.

Game and market mechanisms

With "gamified crowdsourcing" you can earn play rewards for very low transaction costs, and contribute very small tasks.


Higher transaction costs are associated with "crowdfunding" which sounds similar but requires more collaboration and administration.

In the middle, between crowdsourcing and crowdfunding, is a niche for a mechanism with lower transaction costs than crowdfunding but more rewards than crowdsourcing.

By using the existing bug tracker to resolve contracts, a bug futures market keeps transaction costs low. By connecting to an existing cryptocurrency, a bug futures market enables a kind of reward that is more liquid, and transferrable among projects.

We don't know how wide the bug futures niche is. Is it a tiny space between increasingly complex tasks that can be resolved by crowdsourcing and increasingly finer-grained crowdfunding campaigns?

Or are bug futures capable of achieving low enough transaction costs to be an attractive incentivization mechanism for a lot of tasks that go into a variety of information goods?

My bot parsed 12,387 RSS feeds and all I got were these links.

Sun, 23 Jul 2017 07:00:00 GMT

Bryan Alexander has a good description of an "open web" reading pipeline in I defy the world and go back to RSS. I'm all for the open web, but 40 separate folders for 400 feeds? That would drive me nuts. I'm a lumper, not a splitter. I have one folder for 12,387 feeds. My chosen way to use RSS (and one of the great things about RSS is you can choose UX independently of information sources) is a "scored river". Something like Dave Winer's River of News concept, that you can navigate by just scrolling, but not exactly a river of news. with full text if available, but without images. I can click through if I want the images. items grouped by score, not feed. (Scores assigned managed by a dirt-simple algorithm where a feed "invests" a percentage of its points in every link, and the investments pay out in a higher score for that feed if the user likes a link.) I also put the byline at the bottom of each item. Anyway, one thing I have found out about manipulating my own filter bubble is that linklog feeds and blogrolls are great inputs. So here's a linklog feed. (It's mirrored from the live site, which annoys everyone except me.) Here are some actual links. This might look funny: How I ran my kids like an Atlassian team for a month. But think about it for a minute. Someone at every app or site your kids use is doing the same thing, and their goals don't include "Dignity and Respect" or "Hard Work Smart Work". Global network of 'hunters' aim to take down terrorists on the internet It took me a few days to figure things out and after a few weeks I was dropping accounts like flies… Google's been running a secret test to detect bogus ads — and its findings should make the industry nervous. (This is a hella good idea. Legit publishers could borrow it: just go ad-free for a few minutes at random, unannounced, a couple of times a week, then send the times straight to CMOs. Did you buy ads that someone claimed ran on our site at these times? Well, you got played.) For an Inclusive Culture, Try Working Less As I said, to this day, my team at J.D. Edwards was the most diverse I’ve ever worked on....Still, I just couldn’t get over that damned tie. The Al Capone theory of sexual harassment Initially, the connection eluded us: why would the same person who made unwanted sexual advances also fake expense reports, plagiarize, or take credit for other people’s work? Jon Tennant - The Cost of Knowledge But there’s something much more sinister to consider; recently a group of researchers saw fit to publish Ebola research in a ‘glamour magazine’ behind a paywall; they cared more about brand association than the content. This could be life-saving research, why did they not at least educate themselves on the preprint procedure.... Twitter Is Still Dismissing Harassment Reports And Frustrating Victims This Is How Your Fear and Outrage Are Being Sold for Profit (Profit? What about TEH LULZ??!?!1?) Fine, have some cute animal photos, I was done with the other stuff anyway: Photographer Spends Years Taking Adorable Photos of Rats to Break the Stigma of Rodents [...]

the other dude

Sat, 22 Jul 2017 07:00:00 GMT

Making the rounds, this is a fun one: A computer was asked to predict which start-ups would be successful. The results were astonishing.

  • 2014: When there's no other dude in the car, the cost of taking an Uber anywhere becomes cheaper than owning a vehicle. So the magic there is, you basically bring the cost below the cost of ownership for everybody, and then car ownership goes away.

  • 2018 (?): When there's no other dude in the fund, the cost of financing innovation anywhere becomes cheaper than owning a portfolio of public company stock. So the magic there is, you basically bring the transaction costs of venture capital below the cost of public company ownership for everybody, and then public companies go away.

Could be a thing for software/service companies faster than we might think. Futures contracts on bugs→equity crowdfunding and pre-sales of tokens→bot-managed follow-on fund for large investors.

Stupid ideas department

Tue, 18 Jul 2017 07:00:00 GMT

Here's a probably stupid idea: give bots the right to accept proposed changes to a software project. Can automation encourage less burnout-provoking behavior?

A set of bots could interact in interesting ways.

  • Regression-test-bot: If a change only adds a test, applies cleanly to both the current version and to a previous version, and the previous version passses the test, accept it, even if the test fails for the current version.

  • Harmless-change-bot: If a change is below a certain size, does not modify existing tests, and all tests (including any new ones) pass, accept it.

  • Revert-bot: If any tests are failing on the current version, and have been failing for more than a certain amount of time, revert back to a version that passes.

Would more people write regression tests for their issues if they knew that a bot would accept them? Or say that someone makes a bad change but gets it past harmless-change-bot because no existing test covers it. No lengthy argument needed. Write a regression test and let regression-test-bot and revert-bot team up to take care of the problem. In general, move contributor energy away from arguing with people and toward test writing, and reduce the size of the maintainer's to-do list.

Playing for third place

Mon, 17 Jul 2017 07:00:00 GMT

Just tried a Twitter advertising trick that a guy who goes by "weev" posted two years ago. It still works. They didn't fix it. Any low-budget troll who can read that old blog post and come up with a valid credit card number can still do it. Maybe Twitter is a bad example, but the fast-moving nationalist right wing manages to outclass its opponents on other social marketing platforms, too. Facebook won't even reveal how badly they got played in 2016. They thought they were putting out cat food for cute Internet kittens, but the rats ate it. This is not new. Right-wing shitlords, at least the best of them, are the masters of database marketing. They absolutely kill it, and they have been ever since Marketing as we know it became a thing. Some good examples: The 1920s version of the KKK Richard Viguerie's direct mail operation All the creepy surveillance marketing stuff they're doing today is just another set of tools in an expanding core competency. Every once in a while you get an exception. The environmental movement became a direct mail operation in response to Interior Secretary James G. Watt, who alarmed environmentalists enough that organizations could reliably fundraise with direct mail copy quoting from Watt's latest speech. And the Democrats tried that "Organizing for America" thing for a little while, but, man, their heart just wasn't in it. They dropped it like a Moodle site during summer vacation. Somehow, the creepier the marketing, the more it skews "red". The more creativity involved, the more it skews "blue" (using the USA meanings of those colors.) When we make decisions about how much user surveillance we're going to allow on a platform, we're making a political decision. Anyway. News Outlets to Seek Bargaining Rights Against Google and Facebook. The standings so far. Shitlords and fraud hackers Adtech and social media bros NEWS SITES HERE (?) News sites want to go to Congress, to get permission to play for third place in their own business? You want permission to bring fewer resources and less experience to a surveillance marketing game that the Internet companies are already losing? We know the qualities of a medium that you win by being creepier, and we know the qualities of a medium that you can win with reputation and creativity. Why waste time and money asking Congress for the opportunity to lose, when you could change the game instead? Maybe achieving balance in political views depends on achieving balance in business model. Instead of buying in to the surveillance marketing model 100%, and handing an advantage to one side, maybe news sites should help users control what data they share in order to balance competing political interests. Bonus links Rupert Murdoch ‘could use Sky data trove for political ends’ Newspapers’ Stand Against Tech Giants Won’t Save Them Publishers and the Pursuit of the Past Book Review: Twitter and Tear Gas, by Zeynep Tufekci ★ New on Daring Fireball: Display Ads The News Business Sinks Ever Closer to Rock Bottom [...]

Smart futures contracts on software issues talk, and bullshit walks?

Fri, 14 Jul 2017 07:00:00 GMT

Previously: Benkler’s Tripod, transactions from a future software market, more transactions from a future softwware market

Owning "equity" in an outcome

John Robb: Revisiting Open Source Ventures:

Given this, it appears that an open source venture (a company that can scale to millions of worker/owners creating a new economic ecosystem) that builds massive human curated databases and decentralizes the processing load of training these AIs could become extremely competitive.

But what if the economic ecosystem could exist without the venture? Instead of trying to build a virtual company with millions of workers/owners, build a market economy with millions of participants in tens of thousands of projects and tasks? All of this stuff scales technically much better than it scales organizationally—you could still be part of a large organization or movement while only participating directly on a small set of issues at any one time. Instead of holding equity in a large organization with all its political risk, you could hold a portfolio of positions in areas where you have enough knowledge to be comfortable.

Robb's opportunity is in training AIs, not in writing code. The "oracle" for resolving AI-training or dataset-building contracts would have to be different, but the futures market could be the same.

The cheating project problem

Why would you invest in a futures contract on bug outcomes when the project maintainer controls the bug tracker?

And what about employees who are incentivized from both sides: paid to fix a bug but able to buy futures contracts (anonymously) that will let them make more on the market by leaving it open?

In order for the market to function, the total reputation of the project and contributors must be high enough that outside participants believe that developers are more motivated to maintain that reputation than to "take a dive" on a bug.

That implies that there is some kind of relationship between the total "reputation capital" of a project and the maximum market value of all the futures contracts on it.

Open source metrics

To put that another way, there must be some relationship between the market value of futures contracts on a project and the maximum reputation value of the project. So that could be a proxy for a difficult-to-measure concept such as "open source health."

Open source journalism

Hey, tickers to put into stories! Sparklines! All the charts and stuff that finance and sports reporters can build stories around!

Blind code reviews experiment

Thu, 13 Jul 2017 07:00:00 GMT

In case you missed it, here's a study that made the rounds earlier this year: Gender differences and bias in open source: Pull request acceptance of women versus men:

This paper presents the largest study to date on gender bias, where we compare acceptance rates of contributions from men versus women in an open source software community. Surprisingly, our results show that women's contributions tend to be accepted more often than men's. However, women's acceptance rates are higher only when they are not identifiable as women.

A followup, from Alice Marshall, breaks out the differences between acceptance of "insider" and "outsider" contributions.

For outsiders, women coders who use gender-neutral profiles get their changes accepted 2.8% more of the time than men with gender-neutral profiles, but when their gender is obvious, they get their changes accepted 0.8% less of the time.

We decided to borrow the blind auditions concept from symphony orchestras for the open source experiments program.

The experiment, launching this month, will help reviewers who want to try breaking habits of unconscious bias (whether by gender or insider/outsider status) by concealing the name and email adddress of a code author during a review on Bugzilla. You'll be able to un-hide the information before submitting a review, if you want, in order to add a personal touch, such as welcoming a new contributor.

Built with the WebExtension development work of Tomislav Jovanovic ("zombie" on IRC), and the Bugzilla bugmastering of Emma Humphries. For more info, see the Bugzilla bug discussion.

Data collection

The extension will "cc" one of two special accounts on a bug, to indicate if the review was done partly or fully blind. This lets us measure its impact without having to make back-end changes to Bugzilla.

(Yes, WebExtensions let you experiment with changing a user's experience of a site without changing production web applications or content sites. Bonus link: FilterBubbler.)

Coming soon

A first release is on a.m.o., here: Blind Reviews BMO Experiment, if you want an early look. We'll send out notifications to relevant places when the "last" bugs are fixed and it's ready for daily developer use.