Month: January 2017

Fake news: a failure

The easy solution to dealing with fake news is to detect it, block it and prevent it from spreading.

That’s fraught with danger, as is any censorship.

How do you distinguish fake news from extreme positions held sincerely? The motive of a post can make it troublesome, but extracting the motive behind content is difficult. For a human to separate satire from deception can be unreliable. To believe that misinformation can be reliably detected by software is trusting technology more than it deserves.

How can you trust the institutions that are determining what is fake news? The organizations that determine what that should limit or block can be influenced by money and power to support someone’s agenda. Bullies on twitter can push people around and outplay the cards that a businessperson keeps close.

The problem of destructive news being propagated is a failure.

Ideally, a successful democracy has informed citizens who make rational decisions based on information that they thirst for.

The internet dopes up any such thirst with quick answers that don’t have any knowledge behind them… let alone wisdom.

 

YYJ – a historian

One of my current projects is yyj (for lack of a better name. It’s the archive file’s suffix.). It’s a file historian system. It disclaims any aspiration to be a version control system. It is primarily meant to be used by a single user.

The goal of the system is to efficiently maintain a history of documents. It is not based on a check-in model. Instead, the history is updated continuously in the background. If a file was saved every half hour, each update would be available without any intervention from the user.

It make use of the fact that .docx and .ods files are actually compressed with zip. I believe it would be efficient Java .jar files. It is optimized for XML files.

I’ve been using variants of yyj for many years and find it useful. The versions I’ve been using aren’t useful by anyone else because there is no UI. Variants have existed since 1989.

My inspiration for polishing it was listening to a student describe his Capstone project at IUPUI. He mentioned having to make copies repeatedly and struggle to keep the copies organized and up to date.

yyj would make that organization trivial. The student could retrieve any past version if it was needed. He would only need to save a single file to keep his work safe from software and hardware failures. Intermediate versions could be deleted when they aren’t relevant any more.

I think that if I succeed, yyj could be useful to very many people.