Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
Trust the World's Fastest VPN with Your Internet Security & Freedom - A Lifetime Subscription of PureVPN at 88% off. Also, Slashdot's Facebook page has a chat bot now. Message it for stories and more. ×

Comment Re:People are a problem (Score 1) 88

You're right that the tyranny of the majority could be a big problem, but these days initiative and referendum seems like it has some real benefits. As a safety override for legislatures which are increasingly incapable of only passing legislation beneficial to the moneyed class or so divided by partisanship they are unable to fix issues which the partisans have stakes in but which the electorate sees as non-partisan.

I'd put legalizing recreation marijuana in the category of cases where referendums served the public good. It stays illegal because the existing stakeholders in the security state and big pharma see it as antithetical to their individual interests, and most politicians are too pusillanimous to take a reasonable position on the issue.

Comment Re:Interesting to mull over effect of shapes. (Score 1) 89

You can't really compete with the concept of WWZ zombies -- they're just too fast and aggressive, but I think nearly every other invocation of them would fall away from an elliptical wall.

The other low-tech zombie fighting tool I've always wanted to see employed is a good old demining flail. These look like tanks with a combine attached on front, only the combine part is steel weights the size of melons attached to chains. They rotate and pound the ground to set off any mines.

https://youtu.be/wf6CsvAffHo?t...

If you raised the flail assembly so it just spun in the air, you could literally drive into zombie hoards at low speed and just pulp them.

My guess is that a similar apparatus on a smaller scale could probably be adapted to nearly any vehicle, probably even improvised from hydraulic sweeper attachments for Bobcats.

Comment Re:You all laugh now (Score 2) 89

I always wondered why a slope with an incline that gradually increased to vertical wasn't ever employed in zombie fiction forts. They would shamble forward until their center of mass shifted and then fall back.

With the right slope contour, you could make it so they fell back pretty far.

Another option would be a kind of blind curve, where they shamble in and then just shamble away on the other side.

Comment Re:Doing more with less.. (Score 1) 115

Why aren't these tools built in, though?

IMHO, PKI on Windows is problematic less because PKI is complex but more because the in-built tools suck or are non-existent.

Most IT admins are oversubscribed enough that writing that Powershell script or putting together the third party tools for certificate expiration won't happen, especially when you consider for most organizations the number of certificates that matter is relatively small.

I will grant an exception for Homeland Security, though, as any organization using PKI to that extent ought to have an entire team responsible for managing it, which means they would have the time/tools/experience to deal with it.

Comment Re:Doing more with less.. (Score 1) 115

I think you're basically right, PKI implementations are horribly complex in practice and doubly (or more!) so with Windows.

It seems to get worse as certificate-based security gets added into products as defaults installations. As an example, Exchange 2016 installs a self-signed certificate by default which gets assigned to SMTP and IIS. The normal (spanning back several releases) process of adding and assigning a public certificate to services doesn't change the self-signed certificate assignment and use for the IIS Exchange Back-End site or for transport connectors.

I ran into these are problems recently with a customer who deleted the self-signed certificate after installing and assigning his public certificate. Bam, dead Exchange GUI -- had to re-bind the back-end Exchange site in IIS with the public certificate.

Another customer had "verify certificates" enabled on their spam service and when they switched SMTP delivery to the new server, the self-signed cert was still being used by the front-end receive connector. It took some kludgy, un-documented Powershell to force the connector to use the public certificate -- ie, the attribute has to be built as a compound variable using sub-attributes of the public certificate combined with some text, and then that variable assigned as the TlsCertificateName on the connector.

So even if you're trying to use certificates, application behavior and certificate selection is pretty opaque in many cases and can actually ignore specific certificate assignment options.

I won't even get into the management trainwreck that is Windows certificate server, with its 2003-era dialog boxes and management tools. In my mind at least, all of this could be modernized and made much simpler to manage, but the toolchain remains completely user-hostile.

Comment Isn't this why many people voted for Trump? (Score 1) 440

Or at least semi-intelligent people?

They knew in their hearts he was kind of incompetent, but they also were so cynical about any establishment politician being able to effect meaningful change that the only way to achieve it was to empower an incompetent with the idea that it would break the system.

Of course, breaking the system has lots of unplanned side effects, too.

Comment Re:U.S. profits too??? (Score 1) 170

The solution is not higher taxes, it's closing these gaps that companies exploit.

Doesn't this just end up boiling down to higher effective taxes?

You total up your revenue or profits and divide by what you actually paid in taxes and that's your effective tax rate? I don't think at the scale and complexity of a corporation the size of Apple the notion of a nominal tax rate makes much sense.

So if you close loopholes to increase the absolute amount of tax paid, you're raising the effective tax rate even if the nominal one stays the same. In fact I'd be surprised that its not a rhetorical argument used in lobbying and negotiation -- don't raise our tax rate, close loophole X and we'll pay a higher effective rate (meanwhile, their tax wonks have figured out how to use loophole Y instead).

My general sense is that the larger problem isn't paying or not paying taxes, its the cash hoarding these semi-monopoly companies do. A lot of the money just ends up in short-term treasuries or other semi-liquid investment vehicles and doesn't circulate in the economy. In some ways, taxes can be seen as the economic investment of last resort -- a way to bring hoarded capital into the market.

A better policy would seem to be incentives to spend and not hoard capital so it gets put into motion in the economy.

Comment Re:Cake or death (Score 1) 893

My assumption is that a repeated, long-term pattern of harassing behavior wouldn't be tolerated unless the guy really was the goose that laid the golden egg.

On the other hand, what kind of claim does she have, really? Monetary damages are probably limited to 2x her salary because of how short her tenure was. Let's say that's $300k and it costs Uber another $50k to hire and train her replacement.

I don't know what this guy's job is, but replacing *him* may be something that has a direct cost of $100k (recruiting, signing bonus, etc) and indirect costs due to workflow disruption while he's being replaced. It's not impossible those could add up to another $200k or more -- business disruption is expensive.

Comment Re:Perfect is the enemy of good (Score 3, Insightful) 113

I think the engineering improvement curve for stuff like this is really steep. What's practically impossible today, is practical but outrageously expensive in 9 months and commodity priced in 18 months.

IMHO, all of the VR stuff is so bleeding edge that it's going to make the smartphone cycle look slow and methodical in 5 years. Meanwhile, do you rush out products that are expensive, quickly obsolete and don't grab many buyers in the name of "getting to market first"? Or do you iterate it internally and among select developers until your actual concept is practical and at prices that will gain a high volume of sales?

I don't think they're out of line here, the technology in this stuff is advancing faster than they can integrate it into a coherent product and get it to manufacturing.

Comment Re:Cake or death (Score 1) 893

Uber is a rapidly growing company capable of making many employees extremely wealthy.

The rational choice for Uber may be to be forgiving of a high performing employee with a demonstrated track record when his accuser is a new and unproven hire who has made no contributions.

It may even be that management's cost-benefit analysis is that it's even worth paying off a few people if they get to retain highly productive employees whose short-run value exceeds their long-term liability.

This seems like a case where there's special math involved due to Uber's growth status. At an established, nominal growth company, you're less concerned with high performers and their shorter-term harassment costs exceed their long-term value and they can probably be more easily replaced.

Comment Keyless drive, too (Score 1) 102

I bought a used 2007 model with keyless drive in 2009. The car's menu system showed three keys assigned to the car, and it only came with two actual keyfobs.

The bigger problem with apps seems to be that you can fire up the app anywhere and do stuff with the car. An "extra" keyfob or a poor keyway design is only really a risk if you have physical access to the car.

Although I'd grant you that a weak keyway design with a limited number of unique keys is probably a real big car theft risk due to the fact that thieves can basically shop any large parking area and match a car.

Comment Re:Managed SAP R/3 since 1993... (Score 1) 123

So do you think any ERP systems can work (defined as providing a positive return on investment)?

My guess is the success of ERP systems is probably somewhat inversely proportional to the complexity of the system. The less complex the system, the easier it and the existing business processes can be combined, the easier it will be for management to understand and use the tools and metrics and so on, and the lower the general costs are and the more likely that the technical requirements will be met without cutting corners that compromise functionality.

And there's probably a bunch of complex site-specific factors around the skill of management, their ability to comprehend and use metrics, and so on.

I'd guess if you were to graph it with "usefulness" on the Y axis and "complexity" on the X, it would look like some curve that rises quickly with features but plateaus and then drops off as complexity increases.

Comment Re:Managed SAP R/3 since 1993... (Score 3, Insightful) 123

I think the real problem with ERP systems is that they're so extensive they're almost like fully modeled business plans, but they kind of suffer from the "no one is average" problem where if something is designed to meet an average, it actually fits nobody.

So you end up with this complex system that doesn't actually fit your existing business process, requiring either gobs of customization to match your process and specific business, or change your business processes to match the intricacies of the software.

My guess is that once they realize this, they do both, customize and change business processes and end up doing damage to the business, at best increased expenses and short-term business disruption, or at worst, shrink the business and be saddled with expensive software that can't be shed.

Slashdot Top Deals

The next person to mention spaghetti stacks to me is going to have his head knocked off. -- Bill Conrad

Working...