Can we please shun the term "IP"?

Call me a Stallmanite grouch, but I still find the term "IP" unhelpful, preferring the more specific terms: patents; trademarks; copyrights; trade secrets.

This story is about patents and patent applications.

So is it like this?

So does it work like this?

a) There are loads of naturally occurring bacteria, some of which are susceptible to antibiotics, others are resistant

b) These bacteria have reproduced for millennia, using humans and other species as breeding grounds

c) The antibiotic susceptible ones have been by far the more widespread, outcompeting the resistant ones

d) 70 years ago we waded in with antibiotics, and started hitting out at the susceptible ones in their key breeding grounds (firstly humans, then other species)

e) The anti-biotic resistant ones have found their competitors' population falling, and have started to spread out into the extra space.

That's all very well, doctor^H^H^H^H^H^Hfireman, but shouldn't you be at work?

Awesome

This is an awesome public service, making one of history's most significant finds publicly available together with authentic interpretation. Long may Google perform this sort of service.

Nothing to see here

A service in the very early stages of field trials had a fault, the fault was fixed. The end.

+1

I was going to +1 this article, but there isn't a button, only a legacy "like" button

Come on Reg, get with the programme. This isn't 2010 you know.

29th June Oracle Reply

I would encourage you to read the submission Oracle sent on June 29th, references in TPM's article. It is the first time I've read a legal submission which is laugh out loud funny - got some strange looks from my colleagues in the office, mind, when I told them what I was reading.

In particular, their use of "partnership" in inverted commas brings to mind Dr Evil and the "Laser". I have no idea of the legal merits of their submission, but as knockabout stuff it is top notch.

Hashing

And that, boys and girls, is why you should use a long password. Modern GPUs can calculate hashes at an extraordinary rate, making brute force attacks on 6, 7 and even 8 character passwords eminently doable.

In the cinema

And for those who live in the Midlands and want the full cinema experience, it is showing at Warwick Arts Centre on the 18th and 19th of August - I expect the remastered version will pop up in arts house cinema across the country.

MtGox != Bitcoin

The currency wasn't compromised, but the most widely used exchange was. The integrity of the block chain was not compromised. Indeed, the conversion rates at other exchanges stayed pretty much unchanged.

Mt Gox is not a financial grade conversion facility, and the willingness of people to entrust large sums of money to an "exchange" about which so little is known is testament to people's stupidity (and greed). At the same time, the bitcoin block chain will keep running onwards, an interesting little backwater, where both techies and ne'er-do-wells have entries in a fascinating peer-to-peer replicated ledger, whose entries are secured with strong encryption.

This episode may well be the beginning of the end of Mt Gox, but Mt Gox != Bitcoin

I might be missing something here...

So have we essentially re-invented server based write-through block level cacheing where a single server addresses the LUN?

The dilemma for HP-UX shops.

The dilemma for HP-UX shops is this:

a) Keep HP-UX as your mission critical platform, risking early obsolescence for what should be the most long-lived systems and applications

b) Form a contingency plan. In which case Larry wins

In the midst of this awesome gladiatorial struggle between two titans, the livelihoods of thousands of honest decent IT professionals are "collateral damage". And this is saddening.

API changes

While the users are humans rather than applications, this is just like changing an API: something you previously did has a different consequence

Usually a change in API is publicised well in advance, and accompanied by Release Notes. Given that FB are positioning themselves as a general provider of online identity services, they need to start taking APIs seriously, whether used by software or humans.

Seems ironic that they are moving into the buildings vacated by Sun in Menlo Park - Sun's Binary Compatibility Guarantee (maintained under Oracle) is the gold standard of stability for user software APIs.

Reminds me of Sun kit around year 2000

I remember a spate of similar thefts of Sun kit around the year 2000. The wrong-uns would steal the processor/memory and I/O boards from running E3500/4500/6500 systems - as there were no machine readable serial numbers inside, they could be used in any unpopulated system chassis. I'm guessing there were plenty of empty base units sold in the target territories. The advanced version, used in one particularly large heist, was to leave just enough hardware that the system would reboot and come up, but with little capacity.

I'd be interested to know the make and generation of kit which disappeared here. Still, I supposed it makes a change from the trains being late because the copper in the signalling system was stolen.

My Mum and Dad

Well, for my Mum who rarely writes a document, and whose primary usage is email, browsing and simple games (card games), I love the idea. She can turn it on when she wants to use it, print stuff out using cloudprint, and then turn it off again at the switch. And I don't need to set her up a backup regime, coz nothing's held on the machine. And if the very worst comes to the very worst, and she drops it, the machine is a FRU.

Good solution for this class of application

For an application where:

a) There is no personal data involved, and

b) A loss of data would be a nuisance but not disastrous, and

c) Where there are no performance constraints beyond "good enough"

this is an ideal solution. Indeed, it delivers on pretty much every promise of "the cloud" - and the three week time to market for new customers underlines the point.

The challenge is coming up with solutions where one of (a), (b) or (c) is not the case.

Carbon footprint is Power times Time

Whether the windmills are generating at the time of maximum demand will effect the amount of non-windmill capacity required. But they are still generating many kilowatt hours every year, which replaces burnt carbon, whether it is at times of maximum demand or not. The carbon footprint isn't about the instantaneous rate at which it is emitted, rather about the total quantity emitted each year.

Even if the windmills are becalmed for a significant proportion of the time, that doesn't mean they "don't work".

10% for men and 3% for women

Lewis's interpretation isn't how it was presented this morning.

It's *not* that 10% is down to former alcohol consumption and 3% down to current, rather that for men it's 10% and for women it's 3%, in both cases down to "current and former alcohol consumption". It's not entirely clear as to whether this correlation is directly down to the alcohol or whether there are other co-correlating variables.

But, on the whole, this is actually a rather better study than most of the "health stories", involving a very large cohort being followed over more than a decade. The findings, as always, need caution but it's not as rubbish as most of them are.

Support is very costly

It's pretty simple really. Each different platform on which you support your software costs money: you have to compile and test it, and create installers, and documentation sets, and then you need to create and distribute patches and retain teams of support staff in each global region for the extended life of the product, so for products launching next year this will take you towards 2020. If you only have a few hundred sales, or even a few thousand, it will cost you way more than you will make in sales.

With Exadata, P-series and SPARC at the top end, and Wintel and GNU/Linux at the bottom end, the potential market share for Itanium based systems is pretty small. For Intel and HP it is a blow, but for Oracle it is sensible housekeeping.

Public perception of risk unrelated to probability

Why are people having a go at Lewis?

We live surrounded by radiation: it comes from the sky, from the ground, and people take actions all the time which expose themselves to radiation - getting on an aircraft; moving to a city made of granite; having an X-ray or CT scan. There is very good collected data about the harmful levels of radiation, not least from 50 years of running nuclear power stations, and international safety levels have been set with reference to this collected evidence, well within the bounds where potential harm might happen.

What has happened at Fukushima is a right mess, and it will take them years to decommission. And there have been short releases of radioactive matter into the air: matter with short half lives, blown out to sea by the wind, with an intensity which is detectable but has no health consequences.

Away from Fukushima, we have a country where large numbers of people have no homes to go back to, thousands are dead, large areas of cultivation have been ruined and the land contaminated with salt, and a significant proportion of the country's electricity generating capacity has been lost.

The focus on Fukushima is understandable, but its consequences will minimal.

Exadata

For the database layer, we're doing this already with Exadata: there's no choice to have external storage, but you can plumb racks together by joining the Infiniband domains together.

At the server layer I'm with Chris - a Cisco UCS installation has all the I/O usage from potentially tens of servers aggregated at the Fabric interconnect. You then configure two or four or six uplinks into the storage fabric - you could put this directly into a storage subsystem and, if you don't, you're massively reducing the number of connected physical ports on your SAN.

Of course, manging all the virtualisation and abstraction layers places new requirements for organisation and discipline on the system administrators.