arkem

Also compartmentalization helps (Keeping products in different javascript origins, e.g. mail.google.com, accounts.google.com, etc)

You can customize it but the UI isn't particularly obvious. You can drag emails between tabs to teach gmail which is the most appropriate tab, you can also designate inboxes as a filter action.

Personally I think it's a great feature but one I won't be using because I already have a labyrinth of filter rules that perform the same function.

Google's advertising revenues are not down 5%. Google's total revenue is up 19% year over year and 1% quarter over quarter. All categories of revenue are up except for network revenue (which is up 7% Y/Y but down 2% Q/Q).

You may be thinking of the share price which in after hours trading dropped 5% but recovered most of that this morning. Or perhaps you're thinking about the Cost-Per-Click metric which decreased somewhere around 5%.

http://investor.google.com/pdf/2013Q2_google_earnings_slides...

This is a discussion on an article that compares an NSA data center to the Stasi archives. I chose to discuss how I find that comparison fallacious.

I am arguing that the NSA is not like the Stasi. They had different goals, different authorization, and different modus operandi. You're arguing that the NSA is bad (because they are like the Stasi). Whether the NSA is bad or not is orthogonal to my discussion.

To directly answer your reply:

1. The Stasi doesn't collect vast amounts of anything any more.

2. Has the NSA used data to ruin innocent lives? Have you any evidence that they will do so in the future?

I am comparing the official missions of the Stasi and NSA which are significantly different.

If your argument is that the NSA is not acting in accordance to its mission, or that abuses are taking place, or that the NSA is acting like the Stasi that's something else entirely, completely separate to what I'm discussing.

I'm pointing out that these are different agencies with different roles and responsibilities, and trying to draw parallels between an NSA data center and a Stasi archive is not really useful.

The National Security Agency is a cryptolgic agency tasked with gathering foreign signals intelligence and providing information assurance services to the United States Government (especially the military).

The Ministry for State Security was a secret police / intelligence service tasked with a broad range of powers focused in keeping the party strong as evidenced by their motto "Schild und Schwert der Partei" (Sword and shield of the Party). They were primarily responsible for internal surveillance of people that were deemed enemies of the party and the personal protection of party leaders.

It's pretty clear that a catch all secret police / security intelligence agency is different to a foreign signals intelligence agency.

There are a few differing definitions of data and information but in general it goes like this:

Data is a collection of facts.

Information is data after context or analysis is applied.

A database of phone records is data, a report on the people a target has called is information.

The article compares the hypothetical amount of data to be stored in a NSA data center to the records storage area of the Stasi.

That is, a guess at the amount of data one facility of a modern signals intelligence agency (belonging to a super-power) is being compared to the amount of record keeping space a cold war era security agency (belonging to a Soviet client-state) possessed. The two agencies have vastly different missions (state security versus signals intelligence), operated at vastly different scales, and in completely different eras. The two facilities being compared are not at all similar. A large computing facility (probably processing raw data) is being compared to a large library archive (largely storing files and documents generated by the agency).

No valid conclusions could possibly be drawn from this comparison. The closest we could come to a valid conclusion is; if we printed out 5 zettabytes of data we would have wasted a lot of paper.

Edit: What I'm trying to say is that this is an apple to oranges comparison made across space and time.

This is as useful as comparing the number of bullets the US Army has today with the number of rifles the 1985 Nationale Volksarmee (army of East Germany) had.

While it's an interesting project (I had fun playing with it) you really shouldn't be relying on remote sources of entropy for security purposes.

Generally the entropy needs to be secret to be effective so sourcing remote entropy reduces your security to that of the transport security (if any). It's hard to imagine a situation where you have a secure transport mechanism but do not have enough entropy since most encryption schemes that you might use to secure the transmission require secure random number generation.

Disclaimer: I am not a cryptographer, cryptology frightens and confuses me.

The amount of effort to decouple Reader from Google's internal infrastructure is not trivial.

I'm an engineer and not a business person but I imagine that finding someone to buy Reader and then executing the deal would also be non-trivial.

Google App Script might be what you're looking for it's an app development platform that can use internal resources (like GMail)

https://developers.google.com/apps-script/

Discussion of leaks takes up the first 15 minutes of the video.

The approach tptacek and Matasano take is explained (for a potential candidate's benefit) on the Matasano site: http://www.matasano.com/careers/

This is very specifically about the San Francisco city area. The peninsular and south bay are much more suburban and family friendly.

Palo Alto isn't a lot cheaper that San Francisco it has a reputation as being much more family friendly and has some good schools. Also, Mountain View, Los Altos, Sunny Vale, Santa Clara are all more affordable and family friendly than San Francisco.

Disclaimer: I'm a single guy with no kids, so what would I know?

For business cards, stickers, greeting cards, labels, etc Moo has an API http://us.moo.com/api/

I haven't used the API but I've bought business cards from Moo and they were great.

Sneakers the movie has my favorite NSA/FBI/CIA jurisdictions quote: http://www.youtube.com/watch?v=1r6y4vjADwY (0:46)

I think you mean EB status not E status, which refers to the priority level of the green card application.

I was really confused because an E3 is a visa very similar to a H1-B for Australians and an E1 is a visa for traders from countries that the US has a free trade agreement with.

The quality of the applicant's educational institution is taken into account by USCIS/DoL. The applicant provides a statement from a US education assessor as to the quality and scope of their non-US degrees.

For example someone with a 4 year Computer Science degree from a disreputable foreign university might only be able to claim equivalence with a US Associate Bachelors in Information Systems degree rather than a 4 year US CS degree.

While it's always fraught to compare educational institutions from across the world, the United States at least acknowledges that it isn't a trivial problem when assessing visa applications.

Statutory minimum wages not rising are not an example of worker exploitation. Even if there were no minimum wage laws worker exploitation does not necessarily follow.

Additionally, in your example while the federal minimum wage might not have increased many states have their own minimum wage laws that have changed over the period you describe.

The closest statement you could make is that an absence of minimum wage laws may make it legal to employ people at an exploitatively low wage. This is not really the name thing.

I don't think you can call it "planned digital obsolescence" since "planned obsolescence" implies that the provider had a specific life span in mind when launching the product.

Google Reader is being shutdown but I doubt that when they launched they said to themselves "We will run this service until 2013 and then shut it down". Without that plan I don't think it can be planned obsolescence.

Some of Google's API do have this, it'll be in their terms of service under "Deprecation Policy". The deprecation policy is a promise that if the service is going to be discontinued you'll have at least X notice.

The examples I can think of off hand are all Google App Engine related but I wouldn't be surprised if some of the Drive APIs have the same thing.

Does anyone else think that his publishing contract is odd? He mentions it being a 50% revenue split after publishing expenses, I was under the impression that any contract that put any of the costs onto the author were considered bad to the point of exploitative[1][2].

It obviously couldn't have worked out terribly for him if he was making $3 a sale but I wonder if this is more common than I thought.

[1] Hydra's Contract: http://whatever.scalzi.com/2013/03/06/note-to-sff-writers-ra...

[2] Yog's Law: http://www.sff.net/people/yog/

[Disclaimer: I work at Google in an unrelated area]

Google platforms like Cloud SQL generally have a deprecation policy that will give you an idea of the minimum length of time that they'll be maintained. For Cloud SQL it's 1 year (it's in the Terms of Service), which means that you'd get at least a year's notice before it could be turned off. It's on par or slightly better than what I know of Amazon's policy (they have 1 year deprecation on their APIs and an undefined deprecation policy on their service offerings).

In my opnion having this deprecation policy would make Google Cloud SQL a lower risk proposition than a 'no-name startup', at least until the startup is in a position to make similar guarantees (and the financial resources to stick to those guarantees).

Edit: I guess my point is that some Google products have service agreements as to how long they'll be maintained and so comparing consumer web services to Cloud SQL isn't really appropriate.

Your employment agreement may include a broad intellectual property assignment clause. If it does and you don't live in a jurisdiction that has legislatively restricted IP assignment clauses the programming language and problem domain of your side project is unlikely to matter.

64 bit Chrome on Windows is a possibility, Justin Schuh is working on it https://plus.google.com/u/0/116560594978217291380/posts/d93X...

I may be wrong but I think that the Starcraft unit is always called 'Marine' and not 'Space Marine'

Early Blizzard employees have given at least one interview where they said that Warcraft and Starcraft were inspired by their love of the Warhammer and Warhammer 40k universes.

You're also not using their service. Is turning your phone off and leaving it in a drawer for a month also breaking the agreement?

This sounds like a run of the mill peering negotiation[1]. The unusual thing here is a company getting press coverage for it.

I have no idea how Google's peering arrangements are set up but I doubt that this is the only transit arrangement Google has ever ended up in.

[1] http://en.wikipedia.org/wiki/Peering