Reading the bug report commentary, it appears there's an error in the specification: http://www.x.org/docs/XKB/XKBp... that Peter Hutterer propagated into the code. The specification should be fixed as well as the code. Peter's comments about the change also discuss a null-pointer dereference problem - I'm not clear how that is related to the change - and therefore whether reverting the change is the complete solution.

The specification appears to be dated 1997-12-15, so all this is blowback from 16-year-old specification error.

Having seen plenty of serious bugs sitting unfixed in bug reports for years and years, I don't think the problem of enormous bugfix latency is particularly related to or limited to acccessibility issues.

Tweets from Scott Henderson, who drew the diagram you cite, clarify what information we do and don't have: https://twitter.com/_AntiAlias...

Bzzt. http://i1.minus.com/iPcccu2MDL... does not show the factual location of the pings. Read the caption. It shows "Examples" of pings that could have given the tracks that the NTSB released. The actual location of the pings has not been publicly released, even though the ping data must have strongly influenced the NTSB tracks that have been published. This image from minus.com was drawn by Scott Henderson, who has explained that the pings shown in the diagram were drawn to illustrate the process that the NTSB presumably employed. This artifice got some strong negative reactions, such as http://willyloman.wordpress.co...

Knowing the actual ping locations, particularly the 3:11 and 4:11 pings, could help clarify when the turn to the south took place and better pin down the complete track.

TFAbstract says that WPA2 can be cracked with brute force search, and that long passwords are more secure than short ones. Looking up the home pages of these internationally renowned researchers http://www.brunel.ac.uk/bbs/pe... http://issel.ee.auth.gr/people... http://www.research.lancs.ac.u... reveals that these three claim no other security-focused publications. But perhaps I'm too quick to judge. Somebody pay the man and read their paper. Or is this the two-step get-rich-quick scheme?: - (1) Publish Paywalled Article Exposing Security Holes in Commonly-Used Security Protocol (2) Profit! (PPAESHiCUSP-P)

The link is to an article that's a week old, and if you read the article carefully, the plot described ends at 2:15 AM, without knowing whether IGREX was reached. The Inmarsat angle data leads toward the southern Indian Ocean.

You might as well consider the similarities to Oceanic 815 - though that flight was Australia to US, apparently ending up at Wallis Island - or not. Unfortunately for your fantastic theory, the current SAR effort is focused on the southern Indian ocean, not Himilaya. While the 8:11 angle estimate has a northern segment, all attention is leading toward the southern segment, at the extremum of the supposed fuel range of the plane.

The plane might have traveled to the west of Indonesia before turning south. The intermediate satellite angle information could help distinguish when and where the plane turned southward.

Unfortunately, this map has non-factual locations for the circles other than 8:11. The angle information for the earlier pings has not been released, but artwork was drawn up that estimated these earlier pings from the reported estimated tracks attributed to the NTSB. This artwork, drawn by Scott Henderson, was likely the source for the map on theaviationist.com's site. See http://willyloman.wordpress.co... for details.

Inmarsat has been coy about the exact value of the ping angles. They issued a press release that said that the information had been given to the Malaysian government, and that anyone who wanted details should contact Malaysia. See http://www.inmarsat.com/news/i... IMHO, they have been doing this because the earlier ping data may make clearer that the aircraft track takes it over Malaysia, where the lack of detection may be a source of official embarrassment.

The earlier ping data may also indicate whether MH370 overflew Indonesia, or whether it flew west to avoid Indonesia, and that has an effect on the plane's remaining range and the estimate of the flight's bearing when it presumably turned southward toward the 90E/45S region where the SAR operations have been focused lately. It would appear that this data was factored into the NTSB track estimates, but the lack of an official release of the angle data has hampered the armchair/amateur speculation about the location. IMHO, if MH370 avoided overflying Indonesia, it may have been a deliberate attempt to lay a false track in a west or northwest direction.

That's not true of patents filed prior to mid-1995. They get a seventeen-year term from issue no matter how long it takes to be granted.

Article seems to be talking about patent application 05/302771, and the status of the case is miles away from the way it's described in the article. This patent has been through several levels of non-final and final rejections, appeals, and court actions. Through the USPTO's public PAIR (Patent Application Information Retrieval) system, you can access hundreds of pages of information and history on the case, including what are now several hundred pending claims. Even if the application itself hasn't been published, the file history is ripe with lots of information. You can see the patent examiners' rejections and there's a 494-page appeal brief filed on behalf of the Applicant, from which you can see many of the pending claims. The patent office rejections appear mainly under section 112 on the basis that the claims aren't adequately supported by the patent disclosure. It's not as if he just applied for the patent and waited 43 years - he's been trying hard not to take NO for the answer.

In addition, there appear to be about 150 additional patent cases filed as continuations on dates between 1977 and 1995 - some still pending and some abandoned. Most of them aren't accessible under the public PAIR system because of the pre-1995 filing dates. Presumably there's no continuations filed after 1995 because under the post-1995 rules, the application would expire 20 years from the earliest filing date, so they'd expire before granting. If many of these continuations have hundreds of claims like the parent case, there could be tens of thousands of claims that he's trying to get granted.

Yes, I understand that when you "deposit" bitcoin to a ledger account, you're not going to get back your original bitcoin on a "withdrawal," but that's not the point. When Mt Gox attempted to send a bitcoin from a Mt Gox wallet, then was told it didn't work, they should have re-sent that same bitcoin that they sent the first time.

It's like if someone told you to hand them a dollar, and you reached into your pocket, grabbed a dollar and hand it to them, then when they say they didn't get it, you don't just reach deeper into your pocket and grab a different dollar - you look at your palm and hand them the dollar that still sticking to your fingers. If it's not in sticking to your fingers, you'd better figure out where the dollar went!

What I fail to understand is how MtGox managed to lose all this money with "transaction malleabiliity." My question is simple and stupid, but I haven't been able to discover the answer from articles discussing this fiasco.

Depositors initiate a transfer of bitcoin, then complain that it didn't go through, then MtGox transfers additional bitcoin.

Why didn't they simply send the original bitcoin again?

The 1/4" jack is embedded in the bottom of the guitar, along with the power button, USB port, and pick connector.
There are no pickups because the strings don't even get tuned - the thing on the end, I think, senses the amplitude of the string vibrations so you can play it with your fingers. This means no note bending, etc. The implication is that it figures out what note you've played by sensing which if the segmented frets are connected to which strings.

That what's I've gleaned from the website information. What I don't get it why the frets are segmented, as it should be easy to demultiplex the frets and strings by scanning either of them sequentially, perhaps that would degrade the note timing a little bit, though you can scan which frets are touching which strings in advance so when the string vibration is sensed you know which note to play.

There is NO pickup. The strings are not tuned. (Yes, I was surprised to read this too.)

That's the purpose of the extension of the helmet forward over the forehead.