The Halfway Point

Firefox 4.0 beta - initial thoughts

I just tried the new Firefox 4 beta, and I would say the Linux version is for power users only, due to install issues. It has a lot of features which are not easy to explore, and one of the most annoying demo videos I've used in ages.

The install

I initially tried to install on a Linux (FC13) 64 bit desktop, and found that there is no 64 bit Linux version. There's a 64 bit Windows version I bet, but I'm too lazy to install a 64 bit Windows VM to see. Since I didn't want to mess up my 64 bit system with bunches of 32 bit libraries, I started a 32 bit VM with Fedora 12 installed and used that. The install went quickly, and I was able to start the browser. Seems like Firefox, if I was a regular user I wouldn't find it much of a learning curve.

I noted that the browser supports HTML5, so I went to youtube to view some HTML5 content. Seems not to have any, or not to notice that your browser can support it, so I was stuck with flash. Flash didn't work. It told me I needed to install a plugin, which I told it to do. After much thrashing of disk and network, and entering the root password two or three times, it completed. Then it still wouldn't play flash. I restarted it, still would play flash. So I won't test it for normal business use, too many people put things in flash.

The video

There's a "see video" link, and clicking it got a video in some format the browser could play. A very video, indeed, the player started with the volume all the way up. Not nice to override my settings. After watching the commercial for testing the beta, I tried the "download in mpeg4" link. It does not download an mpeg4 version, it plays an mpeg4 version, again after turning the volume all the way up! At this point I was less than thrilled with the demo, and if they people setting it up didn't know the difference between dowload and view, clearly the developers were not in the loop with the marketing. The whole video tells how much they want beta user feedback, but...

Sending feedback

The link to report feedback gives you one line for your feedback. So my feedback was
Problem #1: one line is not enough to report all the things I found
It doesn't wrap, after 100 characters or so it stops listening. Don't want people wasting our time with telling us how to reproduce the problem, do we?

Overall

I couldn't find a site with any of the new video formats it supports, I thought youtube detected HTML5 capability, but it just kept trying Flash. The install of flash didn't work and I am not overly interested in spending a lot of time to find out why, the libraries needed to play flash work fine with the old Firefox, I'm not clear why they didn't just work with the beta. If the install of Flash happened at all, it reset the mtime and ctime so "find" didn't identify it.

The things I did seemed fast and solid, I tried a few sites which have been a challenge, and they all worked fine as long as they didn't need flash. As usual the conflict between Adobe and people who have other formats only hurts the user. Rather choice we are left with the need to make all the conflicting formats work because we have to use them with customers or suppliers.

Changing tax preparers - data security issues

I had to change income tax preparers this year, our long time tax advisor is retiring, and it's something like changing doctors. You know you have to do it, it's not something you should do for yourself, but no matter what recommendations you get and how well certified and recommended the new "tax guy" is, you always have issues of trust. And like getting a new doctor, people with minor problems tend to think they should try to solve them without professional help. This is particularly true of taxes, people are tempted by the simplicity of the short form, the boasts of their friends who do their own (and occasionally offer to help with yours), the ads for tax software, and they feel that they can use their past returns as a guide to do it themselves and save. They tend to forget that the law may have changed in a year, and ignore the ads for Tax Masters and others similar services who specialize in helping those who have been negnigent, incompetent, or worse yet creative in their tax preparation.

In my case, perhaps more than usual, the feeling of trepidation was strong. The person we have been using has not only become a good personal friend, but she and her husband won a weekend getaway from the IRS, given annually to the preparer with the lowest error rate in the state. That is a very high standard to meet. This provided a feeling of being in good hands far more than a few recommendations and certificates on the wall. How could I ever find someone deserving of that same level of trust again?

Our tax preparers and advisors have always been competent although not always satisfactory, our first preparer was a man we called Jaba the Hut, because of a resemblance in both appearance and demeanor. He sat behind his desk in a gloomy basement office, treated us like idiots, told dirty jokes, sharpened his pencil after every field of the return, and responded to my questions with comments like "I explained that last year," and "only if you want to go to jail." In spite of that, he was thorough, found a number of things I missed doing a "test run" on my own to be sure I had the required information, and he impressed the hell out me one night by calling the District Director of Internal Revenue at home on a Sunday night, and getting a binding ruling on "involuntary liquidation" which saved us big money by turning regular income into capital gain. In spite of his good fiscal services, we were somewhat relieved when he passed away and we found a new CPA. Incidentally, the gruff manner didn't run in the family, his lawyer nephew did the incorporation of our first company, and was quite pleasant as well as competent.

So when we needed a new tax preparer, we did "due diligence" and got a clean bill of health on both competence and past performance. Because it was a tax service (local, not one of the national chains), I asked about things like information being taken home (no), preparers being qualified (yes), all unusual items checked by a CPA (yes), and good data security (about as good as Windows gets). So we felt reasonably secure, our return was given a once-over by our retiring friend, and the security felt pretty good.

Then I found this information. It seems that while the tax preparation is reasonably secure against hacking and casual physical access, once your tax information is submitted, the IRS fails to take even simple precautions which I would think are necessary on any business system, much less the extraordinary measures which I think should be used to protect the personal information of millions of people. When I was working for SBC, if an employee left the company for any reason, access codes for the appropriate computers and networks were changed then, day, night, or weekend, and new credentials were immediately distributed where they were needed by secure channels. On at least one occasion two of us waited until long after normal hours for the results of a meeting with an employee who had violated policy, so that if termination was appropriate 38 servers could be changed before the employee left the interview.

It's a sad commentary on the whole tax collection system, when the least secure guardian of our personal data is the one piece of the process you can't replace with a more competent alternative.

Toyota throttle problems

When I first heard about unintended acceleration in Toyota cars, particularly the Prius, I thought of my fried and his wife, both in their 80s, who have an early Prius. Not the people most likely to do the right thing by instinct. But some other parts of the issue came up, some not thoroughly covered by the mainstream media, so I'll mention them here.

Driver error induced by human interface

When I first heard of unintended acceleration years ago, it was not in a Toyota at all. It was a rash of accidents caused by drivers who claimed they had been hard on the brake. And that was when cars had real brake and throttle controls, unhindered by "fly by wire" style software deciding that the driver didn't really mean what the control input said. After investigation, it appears that many of the drivers had recently gotten cars which were not offered with a manual transmission, and which had the brake and throttle pedals more or less centered under the steering wheel, rather than in the traditional locations, throttle on the right, brake in the middle, clutch pedal (or space for it) on the left. It seems likely that the drivers, in an emergency, hit the pedal in the slightly right of middle location, and that this turned out to be the throttle.

Like many analyzes, this was "likely" rather than definitive, it may or may not apply to some of the Toyota incidents. But see the next item.

Driver population

In trying to blame all the incidents on driver error, Toyota and the press have noted that the majority of drivers having accidents were young men and older drivers. This could be caused by any of several things, Toyota would like you to believe that older drivers make mistakes and young men drive too fast. While both of these are true to some extent, I find it odd that the population distribution isn't mirrored in other vehicles, at least as far as I could tell. And regarding the pedal location issue I previously mentioned, those drivers cited are also the drivers who are more likely than average to have experience in driving vehicles with manual transmissions. This suggests an independent analysis of pedal position as a contributing factor.

Cruise control

If there is a software problem related to the throttle behavior, it could be related to use of cruise control. The phrase "I sped up to pass a car and it wouldn't slow down" has appeared in several reports recently. Since these occurred on highway settings, use of cruise control is likely. Note that the founder and designer of the Apple computer, Steve Wozniac, states that he can generate this failure at will in his own Prius, and Toyota has not gotten back to him to investigate, even months after he contacted them.

Are anti-lock brakes involved?

In spite of Toyota repeating over and over that the brakes will stop the car even with the engine at full throttle, the California Police report shows that the police could smell the brakes and see the brake lights, and yet the car didn't slow. This raises an interesting possibility. The anti-lock braking system in a car is designed to make the car stop in a straight line, not to make it stop faster! It does this by pumping the brakes so the wheels keep turning. But in these cases having the wheels stop is exactly what the driver wants, and if the anti-lock system were somehow reducing the effectiveness of the brakes, thinking there was some kind of skid happening, then the driver would be unable to stop. And that scenario matches the observed facts. I have yet to see a driver statement about anti-lock which would address this possibility.

Finally, the cover-up

Reports from Consumer Reports, Reuters, and DailyTech indicate that Toyota has only a single laptop capable of reading the black box recorder in all of the USA, and that there is dispute about whether the black box even reports if the brake is on (much less if the anti-lock is limiting it). There seems to be no reliable information about the crashes, what little there is comes from company spokesmen after they have done the data extraction.

Now I admit I'm a suspicious person, but it seems to me that if the black box data showed that the accelerator was pressed and the brake was not, and that if the accelerator pedal was not pressed that the actual throttle was also at idle, or anything else to indicate that this really was driver error, then I would have a jet full of those laptops flown in and hand one to every law enforcement officer I could find, and give a truckload to the NTSB so they could tell everyone that it was really driver error. But if the data showed a problem, and I had no fix, I would make sure that any black box showing a problem was "damaged in the crash." And if I had a well known expert on computer hardware and software claiming that he could prove my software was bad, I'd totally ignore him as long as I could.

I suspect that this is going to wind up being a reenactment of the the Nixon tapes, the damage from the cover up was far worse than the actual act. If it were shown that Toyota covered up a problem and let people continue to die, it may be a corporate cerimonial suicide.

Hiring politicians is hard

We have odd choices when it comes to politicians, and the latest controversy about NY Gov David Patterson is just one example, but because it is current let me start there.

David Patterson became Governor when Elliot Spitzer resigned following the scandal regarding his patronage of high priced call girls. He started by telling the people and the legislators that he was bringing a new era of working with the legislators against them. Spitzer was known for quotes like "I'm a steamroller, baby" to threaten legislators who failed to do things his way. It all looked very promising.

The next day Patterson and his wife revealed that both had dated other people during a separation, "to avoid rumors." I personally think people are way too interested in politicians sex lives, and they, in return, are far too interested in controlling mine.

Now there is a story that Patterson took free World Series tickets and called a woman who was filing a complaint against one of Patterson's aides. While I don't approve of either, and one may have been illegal, I can still remember that the Senate didn't pass, or even meet to discuss, any legislation for weeks last summer, while they were putting their efforts into who shall control the Senate and not into who shall faithfully do the people's business?
Now they are trying to claim the high moral ground against Patterson, when their high ground is really a a moldering pile of barnyard waste. The legislators have not proposed any reasonable fund raising measures, or practical cuts, while staring at a multi-billion dollar budget deficit. From here it looks as if they are afraid to offend their contributors the special interest lobbys, or their voting block unions, so the burden is being droped on the middle class. Paterson has said that he will not run for reelection, but that he intends to stay and balance the budget.

Here's the dilemma. If Patterson resigns, Lt Governor Richard Ravitch, an elderly and unelected politician steps in as Governor. Ravitch was appointed Lt Governor by Patterson, and following a legal battle affirming the Governor's right to appoint a Lt Governor, he was recognized in July of 2009. Patterson's choice for Lieutenant Governor has no intention of running for office in the future, and he is totally a lame duck, having neither ambition nor constituents, so NY politicians have no reason to fear reprisal before or after the election. Do I, do we the people, decry Patterson for his apparent moral lapses, or do we seek to retain him as the last latch holding the door against a massive budget deficit?

If this seems familiar, consider embattled President Bill Clinton, who had a relationship with a woman other than his wife. While that was certainly a moral lapse, a lapse of judgment, the more we know about past Presidents the less it seems shocking. Clinton left the White House to a nation at peace with a balanced budget, and Patterson stays in office declaring that he will keep making cuts until the budget is balanced, unless the legislature finds extra funds. And given that the legislators are talking about funding by raising taxes on sugary drinks like soda (including diet) and orange juice, which puts the burden on the middle class, instead of raising the tax on those with the highest incomes, keeping Patterson suddenly looks better than the alternative.

Finally, to show the other side of the issue of morals vs. competence in politicians, consider George W. Bush. He failed to protect the country, got it into a war by lying about non-existent WMD in Iraq, demanded tax cuts in a time of war, leaving the country deep in debt to our competitors, approved or at least tolerated torture of prisoners and denying the rule of law to them. No matter how you feel about his goals, you can't think he showed competence in attempting to reach them. Yet he is a Christian, reportedly faithful to his wife, and beloved of his dogs and children.

And Barak Obama, who seems a most moral man, and fully capable of getting elected, seems to lack the fortitude to forsake bipartisan compromise or offend the extremist fringe members of his own party. Lyndon Johnson was not afraid to offend people, step on toes, make threats (and mean them), and otherwise take care of business, leaving a legacy of civil rights legislation and pissed off friends and opponents alike. Johnson made enemies when necessary, when compromise and persuasion didn't work, Spitzer appeared to make enemies because compromise and persuasion were lacking in his skill set. At least for the moment President Obama seems unwilling to offend, to make blunt threats in an election year.

Given a choice I'll take a competent person in any job, the nation, the state, and the localities can't afford any more professional politicians who avoid the hard choices, nor any more well-intentioned bumblers who are simply incapable of doing the job at all.

Windows7 app presents possible security issues

I see that once again the "what could possibly go wrong" syndrome has struck. It seems that Windows7® has a feature allowing multiple systems to share a single network connection using an application called "Virtual Wi-Fi" for connection. This allows a computer to be connected via hardwaire or Wi-Fi link, and at the same time serve as an Access Point (AP) for other systems to use. It's not clear from descriptions if this is done by making the system an independent AP on its own, or by use of the 802.11s "mesh networking" protocol.

And it doesn't matter! The point is that this allows multiple systems to share a connection, even if the connection is to a private internal network. The Network World article describes this, there are other articles such as this one at livescience.com presenting other ways of looking at the capability.

The original Microsoft project was aimed at allowing connections to multiple APs, or at least that's my reading of the description at the Microsoft® site. It's not that the idea is bad, but that the security implications are not obvious, and people may use it without realizing that it opens a network connection, which might be heavily secured, to pretty much unrestricted access. Worse yet, the packets injected to the network have the IP address and MAC address (hardware identification) of the system running the application, with no trace back to the machine getting the data. Imagine for a minute that a proprietary document is downloaded and leaked, and your laptop was the only machine accessing that data in several weeks.

Linux users have had support for the 801.11s (mesh) protocol for a while, over a year, but

if there's a simple application to enable I didn't find it
Linux users tend to be a bit more technical than Windows users
packet forwarding is not enabled in any default user release I've seen
Linux also has easy to use support for encrypted VPN connections, making it easier to limit who can connect and what can be seen by unauthorized parties.
Note: don't read that as "no possible problem," just "easier to avoid the problems."

If you do find this application useful, don't just start it in your startup folder and do think about where you're using it.

Things that work - cat litter products

It always nice to find something that works as intended, so here are two things I've tried and found to work as well as claimed, somewhat to my surprise.

The test laboratory

The test lab for cat products consists of a group (currently five) cats of assorted size, gender, age, and disposition. The lab is an 1886 Victorian house, three stories high, and 14 rooms, not counting various alcoves and walk-in closets which would be called rooms in little modern houses.

Product 1 - Fresh Step "multi cat clumping formula with Carbon odor control"

If you have cats and use something else for litter, you probably have an idea how much I trusted the claim of odor control. A multi cat sandbox (we have two) even on the third floor is likely to make everyone in the house aware that you have cats. I had a business partner who built a room with forced air circulation to the outside and only a cat door in, and declared it the minimum solution. But I tried, I emptied both sandboxes, refilled with the test product, and left for 24 hours. At the end of the test time I took a walk to do the olfactory equivalent of "cleanse my palette," and entered the house. And sniffed. And walked to the second floor and sniffed, and up to the landing and sniffed, and to the top of the stairs and sniffed. Until I got to the doorway of the actual room the cats use, I wondered if it had constipated them, and even then the aroma was barely detectable.

"Well," I thought, "I had cats for years, maybe I'm just used to it." So over the next few days as friends who don't own cats came over, I used them as testers. The verdict was that the new litter works, at least with the minimal isolation provided by a separate room. And I noticed that if one cat failed to completely cover a gift, with multiple cats another cat will soon be along to complete the job, covering litter is apparently a survival trait, so everyone helps.

Notes: this was done in the winter, with the house all closed up, litter odor doesn't seem to be heat activated, so with windows open there should be less problem, not more.

Product 2 - the quick clean litter pan

I can't give you a brand name for this, there wasn't any, but it was sold at national chain Pet Smart for about $20. If that sounds like a lot to pay you like changing the box better than I do.

The product consists of four parts: two litter pans, one screen bottom pan, and a top "fling stopper" guard which reduces expulsion of the contents by enthusiastic burying efforts. The box is large enough to eliminate (so far) the practice of a cat standing in the box with rump outside the box and dropping a few off target loads. It also has prevented anything more objectionable than a few grains of littler from leaving the box and hitting the floor. The object is to eliminate the time spent bent over the litter pan with a little slotted scooper in hand. If you can afford to just dump everything every day, this product won't save you a thing, and you probably don't worry about clumping litter, either.

Hare's how it works. The two identical solid pans nested inside one another, so from the floor up you have pan, pan, screen, blast shield. The litter is poured into the top, the cats use the equipment as normal, and at cleaning time you:

snap off the top "flinging guard"
lift the screen and full pan off the empty bottom pan
lift the screen with all gifts and clumps off the full pan
dump the screen contents into bag, can, out the window,
down the toilet, whatever your normal practice might be.
set the screen pan in the empty (formerly bottom) pan
pour the clean litter into the screen pan
set the ready to use pans in the now empty pan
snap the guard back on the screen pan

The whole operation takes about as long to do as it does to read. Snap, lift, lift, dispose, set, pour, set, snap. Yes it's really that fast and easy.

Well, except for step four... you know, you really should think about the details of step four before you're standing in the middle of the floor with a big pan of cat shit souvenirs planning your next step. Note that the pan size is just smaller than the mouth of a supermarket plastic bag, if you go to a market that still provides those. I doubt that's coincidence, works for me.

So there you have it, two cat products tested by a dedicated group of cats and a lazy human always looking for an easier and/or better way.