That is somewhat misleading, thier mistake wasn't really funding a couple of students on the outreach program, that was well within their resources.

Thier mistake was managing the program without insisting on pre-payment from the other funders and hence screwing up their cashflow. They could have got into a very similar mess with an outreach program that was not discriminatory.

Also once you start having pay for licenses you have to start tracking said pay for licenses, even if they are so cheap that the license cost itself is a rounding error the cost of tracking those licenses, ensuring software can only be downloaded/used by people who have licenses and so-on often is not.

There is a world beyond your shitty little home or small buisness "broadband" connection.

While theoretical maximum speeds make good headlines the real purpose of advances in wireless communications is not so much supporting higher speeds to a single user as supporting more users of a given speed in a given area.

Other working groups in the IEEE and otherwise have been working on both getting more out of existing infrastructure and producing standards for new infrastructure.

Unfortunately slow home/small buisness broadband speeds are mostly an economic and/or regulatory problem. We have tapped out what conventional phone lines to the phone exchange can can deliver given their attenuation and noise/interference characteristics and while cable TV cables can carry much more data than phone lines they are fundamentally broadcast networks so available bandwidth per user is limited. To significantly improve from the current point in areas with conventional copper ADSL requires some or all of the infrastructure to be replaced with fiber.

New last mile infrastructure providers face large hurdles both economic and regulatory and existing monopoly/duopoly providers often see little gain in doing a large scale forklift upgrade of their infrastructure when only a minority will pay significantly more for the higher speeds and when the technology is still improving (so the more they delay the upgrade the better the system they will get in the end).

Don't get me wrong, I understand your frustration but it's not really something that the IEEE working groups can solve.

Noise and interference limit how many bits you can have per symbol on a channel and the lack of independence in the channels limits the gains from MIMO*. There are still gains to be had but you get into diminishing returns in terms of bandwidth delivered to a given location (bandwidth delivered to many different locations is another matter, there are major gains to be had from cross-cell mimo in cellular systems but also major implementation difficulties).

* MIMO relies on some clever maths to create inddependent channels out of non-independent channels but that math has the side affect of amplifying the noice.

AIUI the trouble with those cards is that the only interface between the card and the thing it's plugged into is a shared storage array and said shared storage array is rather lacking in terms of good mechanisms to handle writes from both sides without corruption.

This kind of limits their utility for anything beyond their intended use of making it slightly easier to get the pics off your camera.

The problem isn't so much "speaking IP", a PIC18 can do that.

But when you want TCP, SSL, HTTP, support for several types of VPN* and so-on it gets harder and harder to implement on a small microcontroller and something a bit more powerful that can run a proper OS (albiet a pretty stripped down one) looks more attractive.

* VPN support is useful in "internet of things" type applications because you often want to deploy them where you don't have a public IP and/or you don't control the network and yet you would quite like to be able to remote admin them.

If you can keep the vulnerable people evenly spread out and at a low enough density then the disease will gradually die out. Many diseases that used to be endemic are now almost unheard of. One has even been wiped out.

But if the overall rate of vulnerable people rises too high or if you get clusters of vulnerbale people (e.g. a fundamentalist church) then the virus can start to spread again.

Now this won't work for every disease, some diseases have a combination of vaccine success rate, mutation rate and transmission rate that even with 100% vaccination they will still be endemic, just at a lower rate than without the vaccination.

The problem with crypto is it's really easy to end up with something that works but isn't secure.

Libressl and openssl take different approaches. Openssl's appoach is to rely on the OS as little as possible. Libressl is targetted at openbsd and relies heavilly on library features provided by openbsd.

http://insanecoding.blogspot.n...
http://insanecoding.blogspot.n...

If and when there is a linux port of libressl that is blessed by the libressl team then it might be worth considering but a bad port of libressl could easilly end up being a lot worse than openssl.

Be careful with libressl though, the developers have been stripping out what they see as the wrong approach to portability and currently only support openbsd. There have been third party ports to other operating systems but great care is needed when making such ports as assumptions that hold on openbsd but not on other platforms may go unnoticed.

A badly done port of libressl could easilly end up significantly worse than openssl.

IIRC recent versions of gnutls are under a LGPLv3+/GPLv2+ dual license.

I belive mozilla NSS is under the standard mozilla tri-license but i'm not positive on that.

I'd say the horrendous state of ssl certiciate security has aspects in all three categories.

Specification failure: Certificates can only be signed by a single CA, no mechanism for multiple signatures on a cert to give a greater assurance level. No mechanism to limit a CA to a subset of the dns heirachy.
Implementation failure: Major implementations include an insane default CA list*, poor handling of certificates of different trust levels (clever use of redirects can allow interception of form data for an EV site without making the green bar disappear)
Deployment failure: noone looks at the list of CAs and makes an informed descion on what to accept, they just leave it at the software vendors defaults.

but yeah none of these are openssl's fault. By the time they came along the bad descisions had already been made.

* Seriously mozilla why the heck did you include the chineese government.

And you also need them because sometimes your demographic predictions say that a local increase in the number of schoolkids is temporary, or because your demographic predictions were wrong and you need to add capacity quickly.

There is also the question of what will your district look like in a decade or so? will it still be full of families with school-age children? or will it be full of older people whose kids have moved on to a new area?

They could do it the other way arround, make an update to the updater that (among other things) checks more thouroughly what edition of the OS it's running on, then make it so you have to update the updater to get any other updates.

I dunno about the rest of europe but here in the UK car tax for ordinary cars registered since 1 march 2001* is determined from official CO2 emmision figures. The ammount of CO2 that comes out of the exhaust is almost directly proportional to the ammount of carbon in the fuel that goes from the tank to the engine**. I don't know if they directly measure the CO2 or just calculate it from the fuel input but either way it isn't going to make a massive difference.

* Older cars and new cars for which official CO2 figures are not available (scratch built custom cars and such) are taxed based on engine size. Very old cars are classed as historic vehicles and exempt from road tax.
** Carbon that comes out in any form other than CO2 is bad since it represents incomplete combustion and can be a nasty polloutant in it's own right.