* Posts by dwrolfe

9 publicly visible posts • joined 17 Mar 2016

That position you just applied for might be a 'ghost job' that'll never be filled

dwrolfe
Unhappy

This was endemic in Ireland the last time I checked...

Haven't been on the job market here for some time (thankfully!), but recruiters would routinely either invent jobs or steal job listings from a company they had no relationship with.

The latter is arguably worse, because they are injecting themselves into a conversation you could have had with a prospective employer yourself, and tacking on their fees in the process. This means that if you do get the job you'll be paid less, because those fees have to paid by [i]someone[/i].

Db2 is a story worth telling, even if IBM won't

dwrolfe

Millions of TPS on DB2...

Can you point me at a source for the claim "These customers are running millions of transactions a second and need absolute reliability."?

* Process millions of rows? Not a problem.

* Do lots (but not millions!) of TPS? Also possible.

* Millions of discrete transactions? On DB2? Or Oracle? Under real world conditions?

I'd like to see a source and description of the use case.....

Research finds electric cars are silent but violent for pedestrians

dwrolfe

Report is a bit dubious...

No hard evidence that EV's take noticeably longer to stop - A Tesla M3 has more or less the same stopping distance as a BMW 3 series:

https://www.latimes.com/business/autos/la-fi-hy-tesla-model3-consumer-reports-20180529-story.html

Increased mass of EV of limited relevance when hitting pedestrians - An 85KG person will lose against a car, whether it weighs 1,200KG or 1,900KG. Basic applied maths at work here.

Father of SQL says yes to NoSQL

dwrolfe
Meh

SQL can and does scale to the hundreds of thousands of TPS...

First full disclosure: I am an employee of www.voltactivedata.com, who build a high performance OLTP SQL platform.

I'm going to have to respectfully disagree with parts of this article.

The concerns about scale and performance apply to [b]older[/b] RDBMS products, all of which were designed in the late 80's on the assumption there was one CPU and not a lot of RAM. There is nothing inherently slow or unscalable about SQL. It's just you need a modern architecture. We have customers running north of 500K TPS, with SQL, ACID transactions and zero eventual consistency. While NoSQL is one way of getting lots of transactions done, it's not the only way.

As for JSON? It's a blessing and a curse, and is needed. Traditional SQL struggles.- like everything! - with application customization. You'd end up with thousands of tables and comically complicated schemas. JSON makes it much easier to add 'extra' items to a record. But the ability to add arbitrary stuff using JSON is a double edged sword. You could end up with 10 million records, with 19 different JSON schemas, one of which you nobody told you about. So JSON is useful, but not a replacement for everything.

David Rolfe

Google co-founder Brin named a defendant in wrongful death complaint

dwrolfe

Re: So, let me see if I get this

This is *not* what happened. Most airplanes have standard ways of installing temporary additional fuel tanks so you can ferry them long distances for one-off flights.

Installing the tank involves form 337, which documents an approved, temporary modification to an airframe.

So while it's Google's airframe it's not Google's fault...someone screwed up somewhere and the crew waiting too long before turning and running for home...

Flying car biz Alef claims 3K preorders, still hasn't done a proper demo

dwrolfe

Ever googled "Parasitic Drag"? Because this lot haven't...

I saw a half scale model of this at MWC last week. If - and thats a big 'if' - it works at all, it will never go faster than about 40 mph. the bizarre "multiplane as bodywork" will create so much drag that it simply can't...

Why?

The amount of parasitic drag it will create will be farcial...

David Rolfe

Aviation regulators push for more automation so flights can be run by a single pilot

dwrolfe
WTF?

So how will this single pilot get trained?

Let's assume we can overcome the safety and cultural problems and create a single pilot airliner that's safe to fly and people will pay to travel on.

Who trains the single pilot? How?

Will we still insist of 1,500 hours before letting them be a pilot? Hours on what?

Investor tells Google: Cut costs now and stop paying staff so much

dwrolfe

So what happens the year after 30K involuntary google alumni hit the streets?

Has TCI considered how Google's long term interests will be harmed by putting 10's of thousands of highly qualified people who hold a grudge against The Chocolate Factory on to the streets?

Google obviously pay more than they need to *hire*.

But maybe they also pay the right amount to *retain* people who could be future competitors if they leave?

I'm not a fan of Google, but they aren't idiots, and spend this money for a reason.

Maybe TCI want a short term boost so they can declare 'victory', even if every other shareholder loses in the long run?

David Rolfe

FAA's 'drone smash risk to aircraft' is plane crazy

dwrolfe

Drones are far more dangerous than birds, here's why...

Folks,

I've a Private Pilots licence and IMHO drones are a far bigger problem than this study claims:

1. Birds actively avoid aircraft, generally by diving down. Since aircraft don't make radical changes in altitude the bird will rapidly so far below the airplane that the pilot couildn't hit it if they wanted to. Drones on the other hand don't avoid anything.

2. Drones fly in straight lines which makes them harder to see. Any object on a collison course with you will remain in the exaxt same place in your field of view, while getting bigger slowly, until suddenly it's in your face. Birds wander and flap their wings, which makes them easier to see.

3. The study looks looks at the risks posed by direct collisions. IMHO a far bigger issue will be the panic and distraction they can cause, especially to student and less experienced pilots. The vast majoirty of aviation accidents are caused by a chain of events, not a single catastrophic incident. An unanticipated ecounter with a drone would be an excellent start to such a chain of events.

4. What makes anyone think drones aren't going to get miuch, much bigger?

5. Though experiment: If we can have drones in the air why can't we have drone cars that get given to kids as xmas presents and share the roads with us? Oh wait, that'd be a terrible idea...