Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror

Comment Re:What about "legitimate" use? (Score 1) 155 155

Yes there is a formal procedure you have to follow, just having a prescription is not enough.

You need to have that kind of medical evidence for need; and request in advance and gain approval from your sporting bodies testing organisation.

Such as https://en.wikipedia.org/wiki/World_Anti-Doping_Agency and their TUE (Therapeutic Use Exemption) process https://www.wada-ama.org/en/what-we-do/science-medical/therapeutic-use-exemptions

Comment Re:Heisenberg (Score 1) 98 98

So many like 0.3% of all people needing such treatment ? I'm making my numbers up like your comment attempts to incite there is some systemic concern.

Yes people do elsewhere as with all services, maybe it is because they are also the ones who are able to afford private medical expenses from the best in the world.

Comment Re:No filter is truly effective (Score 1) 269 269

The solution is easy you make it an economic problem of needing the sender to use computing power as a cost.

When an SMTP client offers a messages (during the dialog and protocol exchange) the server announces a mathematical problem to the sender (that will allow the message to be accepted in the first place).

This mathematical problem needs to be easy for the server to generate. The server withholds the answer and other information from the client and presents the problem to solve in a way that the client is forced to brute-force the answer consuming CPU time. The amount of CPU time needs to scale both linearly and exponentially (so we are talking a quadratic scaling mechanism).

The server can decide per SMTP transaction to offer no problem to solve (like SMTP right now) or an easy problem to well behaved systems and a harder problem to untrusted systems.

Now the client has the option to decide if it can afford the cost of sending at the moment of delivery (allow a bounce for HAM). Yes the spammers can go out and buy server farms to solve these problems, let them do it. You are forcing the cost of sending spam up in the process. Yes they use botnets but if these botnets start consuming 100% CPU people notice faster and get it fixed sooner and it rate limits what one bot in a botnet can send per hour. Power consumption goes up on sever farm botnets etc.. all noticeable metrics to someone to fix the problem.

Now the question is how is a mathematical / cryptographic boffin who can propose such a mathematical problem. Generate random number, decide on problem scaling size (how hard it will be to compute answer), do something with these numbers and output a question and answer. The important point is that should take an short instant to generate while scaling takes it from a longer instant to solve to many 100s years to solve.

Comment Re:So what? (Score 1) 179 179

Surely recruiters love job hopping ? If its contract periods of time by definition you are job hopping after. If it is for permanent positions then people stay at least a year or two so the recruiter probably got paid after the first 3 months. But now that recruiter can earn his commission on you again to place you somewhere else. Recruiters love job hopping.

Comment Re:Who cares? (Score 1) 86 86

Ah, they (Google) need a contract with one or more GSM providers, this is where the plan will fail.

Mobile phone network operators here in UK are already providing wifi calling apps for smartphones.

I guess they know the situation is coming and are seeking to be the ones in control of it. Here in the UK once you spend enough per month (not a lot of money about the same as decent cable or landline Internet access) you already get a mobile plan with unlimited SMS and voice calling,

So the network operators are reducing their network load, this does not necessarily mean lower monthly fees in the future. Because they still have the monopoly of the expensive bit, the towers/equipment/network and mobile operator licenses. Not something google will be able to muscle in on anytime soon.

Submission + - SourceForge Joins the Bundle Wagon

An anonymous reader writes: The irony of submitting this on /. is not lost on me.
http://arstechnica.com/informa...
"Apparently, SourceForge's mysterious "sf-editor1" has also claimed ownership of a number of other accounts for open source and other software projects."
SF is claiming ownership of these projects for the specious reason of them being "abandoned" when in fact these project simply stopped using SF (apparently for good reason).

Submission + - SourceForge grabs GIMP for Windows' account, wraps installer in bundle-pushing -> 1 1

shanehiltonward writes: SourceForge, the code repository site owned by Slashdot Media, has apparently seized control of the account hosting GIMP for Windows on the service, according to e-mails and discussions amongst members of the GIMP community—locking out GIMP's lead Windows developer. And now anyone downloading the Windows version of the open source image editing tool from SourceForge gets the software wrapped in an installer replete with advertisements.

Update: In a blog post issued shortly after this story posted, an unidentified member of SourceForge's community team wrote that, in fact, "this project was actually abandoned over 18 months ago, and SourceForge has stepped-in to keep this project current." That runs counter to claims by members of the GIMP development community.

The GIMP project is not officially distributed through SourceForge—approved releases are only posted on the GIMP project's own Web page. But Jernej Simoni, the developer who has been responsible for building Windows versions of GIMP for some time, has maintained an account on SourceForge to act as a distribution mirror. That is, he had until today, when he discovered he was locked out of the Gimp-Win account, and the project's ownership "byline" had been changed to "sf-editor1"—a SourceForge staff account. Additionally, the site now provided Gimp in an executable installer that has in-installer advertising enabled. Ars tested the downloader and found that it offered during the installation to bundle Norton anti-virus and myPCBackup.com remote backup services with GIMP—before downloading the installer authored by Simoni (his name still appears on the installer's splash screen).

Link to Original Source

Comment Re:danger vs taste (Score 1) 630 630

You are still not getting it. The scientific claims can only be made for the substance used in the experiment. Since you have not verified the mechanism of action you can not be sure it is a blanket claim to anything that tastes sweet.

Since the claim is so interesting can you cite any other experiments over the past 8 years (since the papers date) to confirm the mechanism of action ? Surely someone would have tried it with many other sweet tasting substances since then ?

Taste buds may know more than you think (literally). As in your brains ability to detect sweet and sour in concious thought is one thing, but the taste buds might be able to detect a lot more than your concious thought is capable of discerning.

It is also very difficult to have a substance interact with taste buds and then remove 100% of the substance. One way maybe to evaporate it off the tongue with a blow torch? But still there maybe particles that won't evaporate, so how do you remove ?

Comment Re:Former Google Engineer - my internal perspectiv (Score 1) 265 265

You don't have to present a certificate to the server?

You can initiate SSL/TLS where by the only party presenting a certificate is the server to the client.

Do you think that all HTTPS clients present a certificate to the HTTPS server ? This is not how HTTPS usually works, only rare systems that are using client side SSL certificate for authentication use it. But your standard credit card transaction or login portal does not present any certificate to the server.

With STARTTLS sending you start unencrypted, enable TLS via STARTTLS command, then perform some kind of authentication inside the secure TLS channel (this can be plaintext authentication inside TLS). Now you proceed to use the SMTP have both setup a secure channel and authenticated.

Comment Re:That's the WRONG way to do it (Score 1) 265 265

Yes you are correct.

The problem is simple to fix, make it cost them CPU computation time.
Implement an SMTP Client <> SMTP Server cookie system, where by an adhoc association can be established between two systems, that the client can represent an arbitrary token to help build trust and reputation around it (or simply use IP address or SSL certificate hash).
Next define a mathematical problem that is cheap (in CPU cost) to setup and verify, but hard for the SMTP client to compute, forcing it to brute force the problem (this making the client pay the greater CPU cost). This needs to scale both linear and exponential.
Allow the server to define the problem to solve and the scale of the challenge, this more trusted clients have a cheap problem, brand new clients get hit with a harder CPU problem.
Built it all into the SMTP protocol.
Now the server is in complete control of the cost a particular client must pay to send the message, the client can decide to accept the cost or bounce the message.

Now sending from a ADSL link, from a foreign country or from a well known virtual host provider can all be scaled accordingly until the point SPAM becomes too expensive to rent enough server capacity.

"Ignorance is the soil in which belief in miracles grows." -- Robert G. Ingersoll

Working...