Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

Comment Re:Bring On The AI (Score 2) 64

You use the words "algorithm based in big data" as if they had intrinsic value. They don't, period. The data - a great foundation, but the algorithm is human-made.

Afraid of machines? Not really. We're not really making any progress towards machines having any sort of free will. I'm afraid of generation gap.

At the moment we have a certain view of the world. This shapes our goals and interpretations. That shapes the algorithms we create. Goal functions. Criteria. Queries.

Now, bugs aside, those algorithms will do exactly what they were designed to do. The point is, we're not nearly infallible. The goals we set now are our best current guesses about what matters. If we're short-term satisfied with the results of passing the responsibility for something to computers, we're going to just let them do it and never look back.

Now, the world of the next generation (not really, could be 5 years apart) includes goals of the previous one implemented as core services, as the ground truth. Any mistakes of the previous one become hard to fix - you'd have to deactivate something that by now is a crucial service and rebuild it from scratch, with new goals. Not likely, there are layers and layers of useful utilities built on this, revenue streams, etc.

In short - passing decisions to algorithms working on big data restricts our future flexibility. The algorithms are there as decision support, that's how it should be. Do not automate strategic decision making. Humans can realize they are wrong, algorithms can't, because, really, they're not - they do what they were designed to do, period. With our tendency to build new technology, processes, etc. on existing solutions if they seem to work well, that creates future dependencies which make error correction very difficult and costly.

The science-fiction scenarios about humans as slaves to machines are likely pure fantasy. Slaves to ancient ideas of how things should be, enforced by machines... now that's much more realistic.

Comment Re:Its a continuation (Score 2) 254

Battery-powered devices are not the most important application. The current batteries are good enough to make them usable. It's not the real goal of this research. Just look at the list - can you imagine a flywheel-powered phone?

The real goal is large-scale energy storage. Cheap per kW solutions with a long life (no, a few thousand charge cycles is not good). That's when "decarbonization" becomes possible. The goal is storage which scales to MWh locally and GWh globally, making unreliable power sources actually useful.

Currently the production and consumption in the grid must match. This is difficult and costly. Burning stuff is the only way to build large powerplant which can quickly adapt to changes in demand. You need to be able to compensate when the demand rises, when it drops it's even worse. That's why solar or wind power are so difficult to add to the system on a large scale. In dynamic systems terms - not enough I in PID...

Large storage is the solution. We have some now - I'm not sure about the English name, but these are hydro plants with pumps (pump at oversupply, dump when demand rises). Problem is, they do not respond that quickly, cost a lot and waste a lot of energy.

Now imagine battery stacks large enough to power a large part of the grid for, say, an hour, or smaller ones at each generator, providing stable output for at least several minutes even when production stops. Suddenly you can run eg. entirely on solar and wind as long as the average production per hour in the entire grid is higher that the average demand per hour even at peak times. Transfer, even on long distances, is easier too - you can plan it, build local reserves over time where the need is expected in a few hours, etc.

That's the holy grail of power grids. And yes, we can actually do this now, it's just much, much too expensive to build solutions that survive long term heavy use, don't explode easily, don't contain lots of acid, etc.

Comment Re: Why is remote desktop on by default? (Score 1) 29

Hey, take back "defended it"! I wasn't defending anything, I was stating a fact and presenting the reasoning behind it. Sure, the more this happens, the more machines go unpatched with serious vulnerabilities due to incompetent or heavily understaffed IT. This was much less of a problem a year ago, when in the same situation such companies just applied all important patches. That trust is now gone and that is entirely MS fault.

Seriously, years of work to get to a situation where either IT has the knowledge and resources to control the patches (a good idea) or the machines autoupdate (worse, but still good)... all wasted in a few months, leading to a situation where IT unable to do the former may actually feel like turning off the latter is better than the alternative. Who cares if they're right or wrong, the point is it used to be very rare and now it really happens. Nightmare.

Comment Re:Why is remote desktop on by default? (Score 5, Informative) 29

That's one thing. The other one is:

"The reason the two Microsoft researchers disclosed this variation of the original attack is to make companies understand the need to keep their systems up to date at all times."

At least one company I know blocked all updates for two reasons entirely under MS control. 1: Win10 is not cleared for use yet for many reasons, updates pushed GWX. 2: High priority updates containing nothing but telemetry. Not enough resources to test & review everything. That's one company looking for other options. Probable outcome - Win cleared for VM use only, under a different host.

MS's feet are like a sieve from all the self-shooting. Future is not looking all that bright. Surprisingly, it's not due to buggy software - they're doing their best ever in that category. That's the price of allowing marketing&sales to touch the security feed.

Comment Re:So the small claims court then? (Score 1) 103

More alike in Poland, I think. The e-court handles small, simple civil cases, purely electronically. Main result - great, huge acceleration of the process, a lot of simple cases solved quickly.

Downside? You really DO want to have an account. This court is used by crooks to collect on nonexistent debts, such as invoces which were already paid - there were too many such cases to consider this rare. It's enough to sue giving a slightly wrong address of the defendant... and the defendant won't even know about the case until the bailiff finds him to collect (by name and PESEL number). By then it's very difficult to fight. If you have the account and a case matching your key ID data is filed, you will at least be notified electronically, even if the address given was wrong. Then it's easy - you protest (electronically), the case is rejected with the option to move to a normal court, the crook usually drops the case at this point.

So... Yeah, it does a lot of good, but many people get screwed. Especially the less-informed ones. I'm not sure about the balance - I think I'd prefer if the e-court didn't exist, but I don't own a company with many small cases to process, so... YMMV. Just be aware of the risks.

Comment Re:OK, I'll bite (Score 1) 106

Yes, and they still would, except this is not just easier, but simply better. It is impossible to keep an up-to-date e-mail contact list of all students that might want to visit. Active subjects are easy, students under my supervision - more dynamic, but still possible. But these are not all students that might want to know! New candidates for my supervision, students of parallel groups in a subject whose teacher is unavailable for a few days, even non-students (e.g. coworkers with a short question) etc, etc.

E-mailing everyone who could potentially be interested is spamming. Static announcements on a web page don't work any better than a card on the door - people only check once they've arrived and I'm not there. Publish-subscribe is the exact right model for this case.

News? Obsolete, and creating a new group just for one person seems overkill. Facebook, etc? Overkill, besides, I refuse to use it. E-mail list server - well... that would actually work and I'll assume that that was what you actually meant by emails.

Except the process of creating a new group at the university server is too bothersome and unclear, running my own is easy but not as robust, the email-controlled subscription interface is not that intuitive to students who rarely use anything like that, etc. Yes, it would work, but using Twitter is much easier for me and most of the students. It's simply the right tool for the job, get over it.

Comment Re:OK, I'll bite (Score 1) 106

E.g. enables me to announce cancellation/rescheduling etc. of my consultation hours to my students. Simple, fast, ubiquitous publish-subscribe platform for short text messages. No need to keep contact list up-to-date anymore. Just works, on any platform. Globally - no problem if e.g. the link to my country is horribly slow (like during my last trip in my hotel). Works great!

That's all I use it for. Same with some of my students, who only created an account to follow my tweets. For that one tweet every couple of weeks. For others it's just one of many uses, they're even more happy that I decided to do it like that.

So... I'm very happy with this service, but I don't think my use case is one that would keep the investors happy...

Comment Re:IT of Commission and Parliament, not University (Score 2) 67

And University researchers are unavailable, unwilling to answer the occasional call?

As someone who has worked for many, many years at a european university (part-time) I'm strongly sceptical about the ability of university staff to do this exact kind of work well. Not to mention the grad students, who will likely be assigned the actual work. Also, it hardly seems like something universities should really do.

By the way, do you have any idea how long this "occasional" call would take? This is EU, with all the regulations. Weeks to prepare the call. At least a month for the call, preferably at least two. A few months for the review and grant agreement preparation. Typically 8-12 months total. Alternative? Public tender. Also months, but not so many. But how do you make sure you can trust the company? It's the era of globalization, if you want to know whether software from eg. a US/russian company is secure (as in some real chance of detecting NSA/FSB modifications), last thing you want is a european branch of another company with ties there. Difficult to ensure with a public tender.

Solution? Have your own small but good team that can do this in less time than a tender or call would take.

Supporting your EU universities and sponsoring research for professors and students does not benefit society?

Yes it does. So, fund it! Pushing routine work like this on us limits our ability to do new things which is the essence of "research". And we will take any work that is called "research" and offers money, that's how universities get money afterall.

I've done my share of work which should never have been given to a university. Routine software development, code review, testing, etc. Practically zero publishable results. Plus, universities do not give the same quality and warranty as a software company in this case. Still, this is a growing trend - throwing such tasks into "research programmes". Expected TRL is growing. Instead of building fascinating prototypes and leaving the conversion to product to spin-offs, universities waste time and talent doing routine work themselves (in consortia, to make things worse). But it's too tempting - instead of allocating budget for something, you just call it a research project and fund it from the science budget. Bingo!

So the internal team is bloated and short on work, but the department/fiefdom must be preserved?

So firefighters should only be recruited when there actually is an emergency? Some jobs have variable workloads, deal with it. And I would be careful with the word "bloat" not knowing how large the team is. For example, having two or three analysts in an organization of this size is hardly bloat.

What makes you think any of this is related to the IT staff's day-to-day work, is within the staff's field of expertise, etc? The person who connects the EUMP's printer to the wifi network may not be the best capable person to analyze malware. All IT jobs/tasks are not equivalent.

What makes you think this would be the same group that runs around installing printers? All IT jobs/tasks are not equivalent. This sort of pro-bono work is exactly a good way of keeping your team of 2-3 security audit guys away from such work and doing exactly what they were hired for. Yes, that team can formally be a part of your "IT services". No, it does not mean they have to be simple support guys with a new task, very much exceeding their competence level.

Comment Re:IT of Commission and Parliament, not University (Score 2) 67

Not necessarily, it depends on their goals.

Looks like they want to keep a strong IT capable of doing effective security audits for them on demand, but the workload is not constant. Projects like this are a great idea. You do something "pro bono", actually useful for you and your society. At the same time you keep the team funded, ready for when you need them more. And, most importantly - you keep them busy doing their actual job, the best form of training there is.

Comment Re:Issue? What issue? (Score 2) 224

I don't know about you, but I have paid for the system I use. It's a paid copy of Windows. If they think it was too cheap to not force a new ad platform (which 10 basically is) on me - they should have priced it higher, too late to change your mind now.

Right now, from my point of view, MS is just not fulfilling their part of the contract. I was promised a feed of security updates. A feed I now cannot use, because it is used to push telemetry and the upgrade with an unacceptable EULA.

I only bought this system as a gaming platform. It did the job. For other uses I was happy with Linux and can easily switch back. There will be no upgrade. Or another purchase. Sorry, you've just lost a client.

Too bad Linux is going through an equally braindead period with systemd taking the role of gwx. Still, there is slack, I've used it for several years, time to get back to it after the bad experiences with ubuntu and others...

Comment Re:Dictator??? (Score 2) 202

Exactly. I don't have a FB account. That doesn't really make anything difficult. It's even getting easier, as companies start realizing that FB-only online presence requiring an account is too limiting. Two years ago things were looking more bleak - everyone was moving to FB and many pages were blocked from non-logged in users.

And a word of advice. If people stop talking to you because you don't have a facebook account... Great! Nice friend-filter. Why would you wan't to talk to people who do that?

Comment Re:Uh... let me think about it (Score 1) 622

It does have its quirks, it can't be blindly trusted. A few years ago it was useless for many directions in my city, due to four facts:

  • for two or so years it kept directing me through one road, which was permanently closed due to construction work. Yet, it was absolutely convinced that another road was closed all this time (due to the same construction work) - it wasn't, they put some temporary tarmac through the middle.
  • yes, there's hardly any traffic on this brigde - perhaps because it's mass transport-only, closed to private cars? But yes, technicly the bridge isn't "closed", just closed to me.
  • no, it is not faster to take the highway if it doesn't exist yet! (it was finished 3 or 4 years later, as planned; this made GPS useless for SO many trips...)
  • no, I'm not flying over the river, there's a new bridge here...

Ok, that last one was just a case of old maps, easily fixed. But the one before - WTF?!?

In short - one function I want on a GPS is the ability to mark a place on the map, or a section of the road, and say "never direct me through this place, period", or "avoid this place for this trip only". No, the "alternative route" button does not work, because it has no idea why I want a different solution. It assumes that the road is blocked for some reason and tries to get you around that and back on the original route ASAP, which is useless if half of that route is in fact closed.

Slashdot Top Deals

When you are working hard, get up and retch every so often.