Forgot your password?

Comment: Re:ok (Score 1) 116

by Dastardly (#47827083) Attached to: Oregon Suing Oracle Over Obamacare Site, But Still Needs Oracle's Help

And, on top of that is the simple truth that most of the time no one knows what they actually want until they see it. So, the exercise of writing requirements for any significant piece of software is an exercise in writing requirements that are at least 50% wrong and even worse having no idea which 50% is wrong. You then put those into a contract and get the wrongness locked in, since changes costs money and have a pain in the ass process to process to get approved. Then, add government contracting which make change even harder and its no wonder that the project fails.

The knee jerk solution is we need more detailed requirements or more analysis or whatever which tends to do little to relieve the problem that 50% of the requirements are still wrong.

Comment: Re:Hi, it looks like you are writing difficult cod (Score 1) 116

by Dastardly (#47648001) Attached to: Wiring Programmers To Prevent Buggy Code

I have seen many books on software development that say that a significant part of a senior developer's job is supposed to be teaching, thereby increasing the overall team's productivity. Of course what an MBA would say is that the senior developer is not doing enough programming and direct the senior developer to stop helping others to the detriment of the team.

Comment: Re:Hi, it looks like you are writing difficult cod (Score 2) 116

by Dastardly (#47647953) Attached to: Wiring Programmers To Prevent Buggy Code

Same here. Usually the coding mistakes occur in the easiest code, and are usually the easiest to detect and fix. The hard and undetected bugs are the ones that are the result of multiple pieces of code interacting in unexpected ways, easy, medium or hard at the individual code chunk level doesn't really matter.

The other source I have found is leaving unspecified paths open to users. You think that you don't have to prevent a user from doing something because it should work. It is actually more effort to prevent that use case. Then, you get bit in the ass because the user expects a different behavior, its not tested very well since it is unspecified, and often no one even really made a conscious choice to "add" the behavior which has effectively become a feature which now needs to be supported.

Comment: Not like bitcoin... (Score 2) 85

by Dastardly (#47618669) Attached to: Ecuador To Forge Ahead With State-Backed Digital Currency

I didn't see anything in the article to indicate this currency would be anything like bitcoin, other than the title saying without any backing evidence "bitcoin-like money". It seems like any other currency except Ecuador avoids the expense of printing paper money or minting coins.

It would be extremely interesting if this is a move by Correa to put into practice Modern Monetary Theory. Correa is an economist by training, and clearly not a neo-liberal. If we see the Ecuador government switch to collecting taxes in the new currency and improving tax enforcement, I think it would be a good sign that is the direction. Assuming the neo-liberals and Washington Consensus types don't assassinate Correa before the transition is complete, it could be a fascinating case study in whether the MMT crowd gets it right. The trick will be figuring out how to get the dollar denominated sovereign debt eliminated by paying it off or conversion to the new currency or possibly fully repudiating it. The problem being that the only real way for Ecuador to get dollars is by having a trade surplus, that they have oil is advantageous. Getting people to stop holding dollars for savings, regular transactions, etc... would move those dollars from private hands to the government where they can use them to pay off dollar denominated bond holders and get out of the business of issuing debt in some other nations currency.

Comment: Re:Bite the bullet / replace the apps (Score 1) 209

An additional concept that might also help is the concept of application strangulation.

In addition to the current functions, try to find out what these groups wish they could do, watch what they do. Because as the grand-parent said,

If you are just going to recreate the apps, you might as well not even bother.

If you just recreate the apps, what will motivate the user to use the ERP or the new app built on top of the ERP? You want to put in something beyond their existing tool to motivate them to stop using it. But, also be careful because you don't want to implement something just because it is in the existing tool, it might not actually have any real value. Applying Agile practices to putting a working product in front of users early, and watch what they actually are doing, you might get lucky and find out that some existing functionality really is not necessary. i.e. if users move completely to the new tool and stop referencing the old one, that would be a sign that any other functions left on the list are not that valuable. Then, you get to do one of those rare events where you come in under budget/save money.

Comment: Re:The next question is... (Score 1) 223

by Dastardly (#47039291) Attached to: Scientists Propose Collider That Could Turn Light Into Matter

Very large stars create matter from light in their core. Although, in order to conserve momentum, it happens most commonly near a nucleus with a single photon. The photon converts to a positoron and electron and the nucleus recoils a bit conserving momentum. I expect that in the giant randomness of a stellar core gamma rays occasionally collide head on thus allowing momentum to be conserved that way, but I expect it is much rarer than the other mechanism, since at those densities a gamma ray probably encounters s nucleus before it can collide with another gamma ray head on.

If the pair production reaches a certain threshold it actually causes the pressure in the core to drop resulting in collapse. This increases the temperature and pair production reducing the pressure further and accelerating the collapse. Since this happens in a non-iron core, run away fusion ensues disrupting the entire star as a pair production supernova. There are only a few candidates observed to date, and not close enough for a very high level of confidence that they really were pair production supernova.

Comment: Re:Hydrogen Vs Batteries (Score 1) 659

by Dastardly (#47003713) Attached to: Future of Cars: Hydrogen Fuel Cells, Or Electric?

A far bigger problem still is the grid that would be needed to charge cars if you replace ALL of them with electric cars. Because the future is one where most cars are one kind of car.

A far bigger problem is the grid that would be needed for electrolysis of hydrogen for ALL of the fuel cell vehicles.

Yes, I know that hydrogen can be extracted from hydorcarbons, but you then go on to say.

Depending on the eventual methods of production is could mostly be local because it comes from water...

Which means electrolysis or thermal extraction of hydrogen from water.

Basically, hydrogen is energy storage same as electricity in a battery. The energy still must come from somewhere in that respect about the only difference with hydrogen is that pipelines and/or trucks are a means of transporting the energy while electricity is pretty much via wire.

Comment: Simple mount point for smart phone... (Score 1) 216

by Dastardly (#46985151) Attached to: GM Sees a Market For $5/Day Dedicated In-Car Internet

I have a Toyota Prius from the year before the Aux port was added (pisser). But, the one thing I really need, is a place to put the freaking phone. My solution is a phone case and some velcro tape on the center console (on the radio) and on the back of the phone. A few seconds to plugin the charger and stick the the phone to the console.

No stupid suction cup phone holders that don't stick properly or block visibility or don't fit the phone right so it falls out or whatever.\

It also works well when I want to use the golf app on my phone for scoring and range finding by sticking some velcro on my pull cart.

If it is good enough to put a man on the moon, it is good enough to hold my smart phone.

Comment: Cheer the boring... (Score 1) 192

Infrastructure should be boring. Cheer the boring. If there is a time period where things just went smoothly, put a big exclamation point on it. And, then list why.

Things like clean failovers.
System patches without downtime.

One other thing that is a pain at my company that you could also show it provisioning speed. When requests come in for VMs, or hardware upgrades how fast a re they served. How many are queued, how many are awaiting management approval, waiting on vendors, waiting on quotes, POs, etc... That is also a great thing to show improvements on. Because if infrastructure is slowing new money making projects from getting off the ground then that is a problem. And, those projects are going to move to cloud services in an attempt to get around infrastructure for good or ill.

Comment: Re:Oh please... (Score 1) 226

by Dastardly (#46774931) Attached to: How 'DevOps' Is Killing the Developer


I run into this similar mentality when explaining Agile team members and full stack development. Usually, the objection I get is "Well, if people are jumping around how are they going to learn anything really well, and not screw up that part of the application." It is people who think we are going to be stupid about assigning work. People and teams specialize, but they have capabilities outside the specialty, but at a lesser effectiveness. Generally, we will assign people work in their specialty. But, if the most important work does not fall into the available team's specialty, then we might give them that work if it is sufficiently more important than the less important work that might fall into their specialty.

Comment: Re:Nothing new here (Score 1) 226

by Dastardly (#46774051) Attached to: How 'DevOps' Is Killing the Developer

I think it misses the point of DevOps. My reading of DevOps is not that developers do operations. It is that operations applies development style processes to their job and works closely if not within the development organizations to make sure the application is designed for operational ease from the beginning. A DevOps specialists will be operations focused with a high degree of skill in the scripting to automate deployment and configuration. They will use version control in order to version the deployment and configurations with the application. That way when a version of the application is deployed to a development environment, testing, or production the exact same software not just for the application but the deployment and configuration is used. The goal is so that when things go wrong it is not due to some untraceable human error, but a fixable automation error.

And, DevOps doesn't mean operations goes away. I think it let's operations focus on delivering a stable platform for applications. i.e. servers, disks, networks, OS, and perhaps the software repository that DevOps can use to configure the platform and deploy applications. In infrastructure as a service cloud situations, servers, disks, networks, and OS become a DevOps responsibility, also.

As far as a DevOps role, I think this falls into the same mistake I get into in my company when trying to explain the concept of Agile team members and feature teams. Just because everyone on an Agile team should be able to take on any task to deliver the software doesn't mean you don't have specialists in Dev, Test, and Analysis. It also doesn't mean that you assign tasks stupidly to those least capable. Most of the time people will be working in their specialty, but they have secondary skills that allow them to take on other types of tasks when over flow occurs in those tasks. A DevOps specialist is focused on the deployment and configuration process and they primarily get that kind of work, but can help with other types of tasks when necessary. Similarly, just because teams are capable of working on any part of the application doesn't mean they don't specialize or that you assign work to teams outside their specialty willy nilly. But, if it looks like there is a long term need for another team in some application area, it is possible to have a team move to that specialty and ramp up their capability there with the help of the existing specialists. It might make sense to have a DevOps specialized team with multiple specialties (SA, DBA, scripting, support) represented in that team, but with flexibility to handle work in other areas, but perhaps at a lower capability that a team specialized in that area.

Comment: What is nothing? Re:Its not nothing (Score 1) 612

I think more physicists need to work on the physics of "nothing". I don't think we really understand "nothing". In addition there are different kinds of "nothing".

We have the "nothing" of empty space-time where particle and energy pop in and out of existence. What happens when there is a lot of nothing? 1 atom per cubic meter, per cubic kilometer over millions of light years?

Even then there is still space-time. What happens when there "nothing" means no matter, no energy, and no space-time?

Or, what happens when the universe expands to the extent that the visible universe contains no matter, and the CMB has cooled to a hundred negative powers of 10 or more. Does space-time lose meaning?

The impression I get from current physics is that "nothing" is unstable. Has anyone studied "nothing" sufficiently to show that there is not some effect proportional to the amount of "nothing". Of course, how do you even talk about a quantity of "nothing"? But, what if the more "nothing" there is, the greater the so called quantum fluctuations, such that something is inevitable.

Comment: If it was related to the car... (Score 2) 329

by Dastardly (#46251931) Attached to: Tesla Model S Caught Fire While Parked and Unplugged

Some rare, but possible causes if it has anything to do with the car.

FOD... (Foreign Object Debris) - shorting power to ground anywhere. Doesn't take much especially on a circuit board somewhere, rapidly heats up and melts solder creating and even bigger short and more heat until fire.

Dendrite formation - Very rare and probably requires more than 4 months to happen, but certain components on a high density BGA array the solder can form tendrils towards other solder balls. As the dendrites get close to each other they will short and break kind of like a fuse, but eventually it can become big enough to hold and sustain current generating enough heat to start the solder balls melting driving more current and heat until fire.

Never trust a computer you can't repair yourself.