Yes, this is an "important" person. But, he is 77. Dementia?
Granted, there are a lot of unexplained events that suggest the possiblity of extraterrestrial visitation. And, that might warrent investigation of the phenomenom, and a government bent on covering it up, perhaps with the assistance of "aliens" with "advanced technology" night be able to do so easily. (A leak, with proof? No problem! Go back in time and "fix" it.)
But, science does not work that way.
It requires repeatable, peer-reviewed, observation of the purported effect, and strong correlation to the purported cause.
We may have been visited by extraterrestrials. It's a secuctive thought for any geek, but one VIP's say so does not make it so, to anyone with a scientific mind.
And every so often, they find an employee who is a "self-starter" and takes responsibility for the systems they manage.
And, this leads to levels of reliability and efficiency in infrastructure that are unstable, in that they can't be sustained with the available budget, once the altruistic and generous admin is no longer available.
So, is the correct thing to do not build out to unsustainable levels (that the residents would then rely upon), or build out to them and hope for the best?
Politicians will always promise more than can be realistically delivered, and sometimes "bad things" happen because of it. It's the nature of government.
In this case, it would probably be relatively cheap (compared to other projects) to DOUBLE or TRIPLE the support staff (one to three), make their wages market competetive (double them?) AND provide an incentive to the current lead to make himself replacable (bonus based on drills that his subordinates have to pass). The cost would go from what? $100k/year to $600k/year? Peanuts for a city like SF, even with its budget woes.
But, it is cheaper to not do that, and have a ready-made scapegoat in case this mismanagement causes a "failure": "It was his job."
Anyone having spent that much effort creating a network - and succeeding - would become paranoid and protective of it. I challenge anyone to invest so much in any project and then happily see it messed up by people who are less competent.
Even more so when it's one's job to make sure it runs correctly. IOW, if he does provide access to others, and they screw up, it's his fault.
While paranoia may have something to do with it, I've found that a reasonable dose of it is healthy. I think, however, that it's more of a case of normal admin paranoia (which is a good thing), coupled with "damned if you do, damned if you don't" workplace policy (e.g. "give him access and you will suffer if he screws up"), that likely led to this paranoia growing to the degree it has.
Sane, logical, people are often driven "crazy" when forced to work in an illogical environment (where no matter what one does, one is "wrong", and the "correct" choice is the one which has the lowest product of "consequences" multiplied by "likelihood of discovery").
I recently took the "high road", and paid a hefty price for it: I had been hired to be subcontracted to a large "three letter" telecom firm. The project manager at this firm wanted my SSN. They had no legitimate reason for it: I was being paid (and had SS withholdings taken) by my employer, not them. The manager claimed that it was necessary for me to get an access badge. I responded that I didn't mind signing in every day and getting a temp badge. The manager then claimed it was necessary for a "background check". I responded that I would happily provide my SSN and any other identifying data (date of birth, drivers' license, etc.) to any reputable, well known, background-checking company of their choice. I was reported as "difficult" and reqested that I be removed. While my employer strongly defended my position, in the end, it was not reasonable that I be paid to be idle, we parted ways amicably, and I quicky found (much better!) work elsewhere.
Turns out the manager in question was allegedly driving those working "for" her as slaves and threatening them with derogatory credit reports from bogus lenders if they did not comply, using one instance of prior theft to justify such "background checks" requiring the SSN.
It was easy, (though expensive), to walk away from that job. But, in this case, lives may very well hang in the balance no matter what choice Childs makes: refusing raises the possibility of the network "going down". Complying, where he has reasonable belief that others will have a good chance of making the network go down, causing havoc in the city, could border on criminal negligence. As far as criminal charges, what could stick? "Theft?" (of access codes).
Should he be fired? Perhaps, for insubordination. But, if the management of the city is so bad that backup systems aren't in place, and properly trained IT staff to run networks, this might be the only way to raise awareness of the problem that could really impact lives of the city's residence. He may have very well done a very good deed -- as the story unfolds we'll know more. At this point, like with all accused, he should receive the benefit of the doubt.
But, regardless of whether he did the "right thing": No good deed goes unpunished, and he should realize this.
There are two issues: code layout (style), and code/documentation/naming consistency (standards).
There are a lot of different styles. What matters is that a consistent style is used for code that is checked in, and that that style be reasonably readable by all. Otherwise source code control system diffs become cumbersome. (Yes, one can diff on retrieval, but that gets expensive and not all source code control systems support it.) Beyond that, it would be really nice to reformat to whatever style the individual is accustomed when one spends a lot of time with the code).
My personal irritants are styles which are inconsistant. For example, indenting function argument lists right beyond the left bracket, but not aligning a right bracket and statements or intra-struct/class declarations the same way, It's not done in the latter case because it is "space wasteful". Is it not space-wasteful in the function call case as well?
if (foo) {
blah
blah }
then why not void fooFunction(...arg,...arg )
?
Personally, I like my bracketing constructs to line up, brackets as well as braces. Furthernore, I find opening braces on a line by themselves wasteful. I don't object as much as I used to, but I come from the days of 24x80 text terminals, and seeing more code at once was a plus.
SO,
void fooFunc
( arg1,
arg2
)
{ stmt1;
stmt2;
}
(Note the indent for the bracketing elements, and the additional indent for the items bracketed. I'm not pedantic about the former.)
In practice, if anything within bracketing constructs fits on a single line, I'd collapse it to that. So,
void fooFunc(arg1, arg2)
{ stmt1;
stmt2;
}
Or, even:
void fooFunc(arg1, arg2) { stmt1; stmt1; }
That works great for little code snippets which are idiomatic, and extend to short switch cases with breaks.
As for complex compound expressions that don't fit on one line, I break them up on the basis of multiplicative and additive operators, and compound variable references are broken on the basis of direct (.) and indirect (->) selection operators. The common rule is to break arbitarily with an extra indent. I place diadic selectors or operators in the indent margin of the second line, consistent with the precedence of the operation. So, turn the page sideways and you see the compiler parse tree -- which is what you really need in the case of a complex compound expression, particularly in an if statement. So,
if
( (a < b)
&& (b < c)
)
{ stmt1;
val1
= a + b + c;
}
The assignment is really odd here, and I have played with
val1
= a + b + c;
as well, but the former is consistent with my diadic rule, and th latter is an exception.
You want ME to power and host part of YOUR datacenter to distribute data, some of which might be questionable, like CHILD PORNOGRAPHY, for which I can be held liable?
</hyperbole>
I don't think so.
I knew there was a reason I don't watch TV: no need for a set top box that I don't control.
Cowardus Anonymous asks: "Why the fuck would you ever take the card out of the car?"
It also controls my access to my office space and building elevator: it's an access card for areas of the whole building, not just the parking garage. So, I need to keep it on my person when at work. Rather cool, actually. Though, as with all mono-controls, when it is misplaced it is a real PITA.
It remains clipped to my slacks, therefore. The biggest danger is when it is removed at pre-laundry time and set aside instead of clipped to tomorrow's work slacks. Though, AC is correct in implying that the safest "resting place" from a convenience perspective would be the car... if that didn't violate corporate policy. The unattended car is not considered a safe place for access controls, corporated computers, or media, and correctly so.
The better question would be, "Why the heck can't we easily backup such mono-controls, even to a limited degree (to counter the risks of theft multiplied by the number of access devices)?"
Where I work, we have the option of monthly parking passes which are simply smart cards. The daily rate otherwise is $15.
If you are a monthly parker, and forget your smart card, however, you are STILL charged the full $15 daily parking rate, even though you paid for monthly parking.
I used to think this was simple extortion, until I realized that monthly parking is handled electronically, but daily parking is handled by a human and therefore actually costs more to implement. (why I can't just swipe a debit or credit card, as I can at many downtown unattended lots, I don't know).
Still, I'd think a montlhy parker who has forgotten his smart card should be refunded a good chunk of paying the daily rate upon presentation of a daily parking receipt and his active smart card.
Of course, don't get me started on automated car washes that don't warn in advance that (a) their debit card reader is down, and (b) their cash reader only takes exact change -- I was once stuck for 15 minutes in a car wash line with people honking behind me because the stupid reader with broken debit card handling wouldn't even take a $10 bill (and keep the damn change for a $7 wash). Automatec cash readers should ALWAYS allow an option to pay more for convenience if they can't provide change. Those that give receipts could easily indicate the overpayment so subsequent refund could be arranged.
The world is populated by morons. Some design stuff.
I usually explain that various NDAs prevent me from disclosing code I've written of significance, and suggest that I be asked to complete a programming exercise.
Most employers have a set at the ready these days, and I usually respond with the 1 hour answer and the 1 day answer, the later showing an evolution of the former, with polish and usually a more generic solution.
Er, Via's nano-ITX offerings with a C3 processor, CN400, and VT1625 has been capable of this for at least two years now. The OpenChrome project provides for the accelleration: hardware MPEG2 decoding with XvMC. glxgears does about 500 fps on an 800 Mhz fanless cpu, closer to 700 on a 1.0 GHz CPU.
Been running this for a couple of years now in a Silverstone LC08 case.
A Libertarian is just a Republican who doesn't support the war
More correctly, a libertarian is an individual that values individual liberty, and decries the initiation of force. A Libertarian is a member of a recognized political party that pursues libertarian socio-political goals.
Libertarians align with Republicans on many economic issues (against increased taxation, though you wouldn't know it from looking at the NeoCon's record), and Democrats on many social issues (getting government out of the definition of marriage, prohibitions against same-sex marriage and poluygamy, and relaxation of drug laws).
Compare and contrast "Objectivist," ons subscribing to Ayn Rand's view of "Objectivism," which recognizes only three legitimate roles for government: 1) settling of disputes when private arbitration has failed, 2) protection from criminals, and 3) protection from foreigh invaders. While there is close agreement that less government is better, it isn't clear that any "government", except by concent of ALL of those governed, is necessary. See also, anarcho-capitalist, and less relevently, anarchist (since anarchists do not necessarly reject the inmitiation of force in achieving their goals, even as many of them do). All, however, can be classed as "minarchists": supporting the smallest government necessary.
I wonder how the Founders would have reacted to such proposed restrictions on their conduct.
Hmm. Did they not become founders (of a new nation), because of such restrictions on their conduct (among other things)?
We don't have to ask how they would have acted: we know how they did act.
Of course the Honorable House Majority Leader would claim that right-wing spin, in the absence of "fairness" is the source of that historical account.
If public funds are used to pay for a political message from an elected representative, it stands to reason that some allotment of public funds are available to all elected representatives for such purposes. Let the comm^H^H^H^Hdemocrats pay for their own damn propaganda.
(And, for the record, I am libertarian, and take equal issue with many things the right does as well).
I guess that XDR wasn't good enough, then, or ASN.1 (which supports multiple abstract encodings to boot).
XML, as an interchange format?
I suppose one could load source code into memory, and compile it every time, too. Even Java compiles to bytecode.
Bloated formats are fine for human interpretation (I rather like one kind of structure for my config files), or occasional parsing (which is why most of the stuff in/etc is human-readable, for small data sets (I do remember when "the internet" was one big/etc/hosts file), but for interchange? Just cause you're big-endian and I'm little-endian?
The trick to making non-human readable formats acceptable is the prevelence of wide-spread encoding and decoding tools.
Yes, XML is self-describing, at least syntactically (and formally with an XSD), and specific encoding semantics can be tagged, but the same can be achieved with means for type encoding. The big thing with XDR and related formats is that types are implicit -- both ends need to know what is being serialized. For RPC, with well-defined interfaces, this is not a problem, but it does make type-checking a remote service a bit of a challenge.
However, types can be encoded as data, and serialized as well: this happens for variant types naturally. Thus, there is no reason to not have a type-encoding and type-exchange protocol to permit dynamic type-checking. The advantage over self-describing data serializations is that it can be done on an as-required basis, instead of with every damn serialization.
I quit-claimed it to my ex in a divorce, and it would be worth about $360k today, even in the soft market,... if she hadn't trashed it. For various reasons, I now have a power of sale over it, and about $31k in liens against it. With around $200 in mortgages on it, I will be making a reserve offer of $231 when I market it for 60 days.
But, it is quite trashed: a few holes in the drywall, and a non-functional microwave and stove. As well, she ripped out carpets in four bedrooms. The hardwood floors could stand to be refinished as well. She really did a number on it in the year since we divorced.
Nothing $5k - $10k can't fix, and restore the house to around $360k value. If someone wants a fixer-upper, it's a great deal. But, most buyers don't want fixer-uppers. In any case, I'll either wind up with the home, or my $31k.
Realtors have told me that in its present condition, it might fetch $260k - $270k if a willing buyer who wanted a fixer-upper could be found.
SSL certificates provide one thing, and one thing only: Encryption between the two ends using the certificate.
Well, technically it is possible to use them for endpoint authentication without encryption (which might be useful on a physically secure point to point link where one can't tell if the endoint has been compromised, and endpoint posession of private keys is transient, though why not encrypt as well, then?) but browsers are not configured to work that way.
But, yes, endpoint verification is only as good as the trust you place in whoever signed the certificate the endpoint presents.
In a closed community, with a secure means of distributing certificates, self-signed certs are fine. In fact, they are used to set up VPN links all the time.
Well-known CAs serve the purpose of vouching for the identities third parties present -- they are rather like government issued drivers licenses, or other photo ID. Bars use them all the time for age verification. Of course, the checking that most CAs do, compared to drivers' licensing departments is laughable, but that's a completely different issue from the potential usefulness of a signed X.509 cert.
The whole point of certificate authorities is scalability: one does not have to verify every single certificate individually -- one need only have trust in the (few) authorities that sign them. Even if one accepts self-signed certs from one's correspondent, it would make far more sense for the correspondent to generate one, and use it to sign the certificates used by all their servers, instead of having a self-signed cert per server.
PGP decentralizes the notion of a CA, but makes establishing individual trust webs more cumbersome. Some would argue this is a good thing. But, ultimately, who's signature you trust on a certificate is up to you, and no, it should not be blind.
As for vouching for the repeatability of reaching a given endpoint, even without CA signatures, this is only true
when the endpoint's private key has not been compromised. So, you have to trust the care your correspondent places in their servers. This is also an argument for accepting self-signed certs as siging certs from known conterparties, but not as endpoint identification-serving certs: if the endpoint is compromised, but the CA cert isn't (managed by the same entity), CRLs and OCSP servers can still be trusted to say "certificate X has been compromised". If the signing cert is compromised, ALL certs signed by it can no longer be trusted, nor any related CRLs or OCSP servers.
Yeah. RMS said that provision was to encourage distributing the source with the binaries. If you do that, you don't have to worry about giving the source to anyone but your customers.
A bit more than that -- in some places, telephone charges (analog modem), or internet access via other means (DSL, cable), might be prohibitively expensive.
All is fine if you originally distributed binaries that way (since then source was available "with" binaries), but not if you didn't.
I had one vendor argue with me that "FTP was a medium for data exchange" per the GPL, but of course the GPL was addressing physical media (we needed their mods to the Tulip ethernet driver in the kernel so we could apply subsequent patches -- and more importantly ALL their modified source to their binaries that we were redistributing -- which was impractical to FTP from them since their server died frequently). We did eventually get a source CD but only after a hostile email exchange.
Remind your bean counters that simply distributing the source with the binaries is, by far, the best way to avoid costly litigation like in TFA.
That may be the case today, but in the 90s, the likelyhood of GPL enforcement was largely regarded as a joke by many. Often it was those of us who respected it that went the extra mile to ensure that our employers complied by investing our time to make source distros, etc.
I was responsible for GPL complience at a former employer, and had RMS come and give a lecture on the subject to our devs and set up a build process where source RPMs would be packaged for all the GPL-tagged RPMS in our software install CDs automaticaly (so long as the SPEC files had the right license tags). It did not guarantee complience, of course (since it didn't check linking GPL and non-GPL code), but it helped avoid mistakes of omission.
But, for many products, like routers that sell for $20, including a $0.50 CD IS an exorbitant expense in a commodity product market of razor-thin margins. So, complience does have costs associated with it that some might try to illegally avoid.
Is is that hard to whack a tarball onto a server and give out the link.
Actually, yes.
The software may not be in a form for easy building and consumption. Think about dependency hell and configuration issues. A lot of shops have a golden "build machine" that does the right thing. This poor practice is less common today, but was du rigeur in a lot of shops where I worked in the 90s -- scrambling to get product out. Often, there is a desire to not publicize the "lack of polish".
Furthermore, there might not be a budget to support an internet-facing server if you don't already have one. Often, the "web presense" is managed by a different department than the engineering group. Again, less of an issue than it used to be, but still a problem. One place I worked (which was not exactly known for it's support of free software, and a large presence in Redmond, WA) would have to "internally lease" a server from the department that did this, and get billed $10k/month for the privilege.
Then, there is the issue of "mail requests". If you don't distribute source with binaries, and are not an academic institution, you must be willing to provide an offer, good for three years, to provide source to anyone who asks! -- not just those you distributed binaries to. A lot of businesses are not set up to do this. [This is for GPLv2. I have not checked if GPLv3 is as onerous.]
It actually is a lot easier to ship source with product if you don't already have the infrastructure to distribute to any and all comers. If you're not an agile shop with a dynamic internet presence, you don't have the infrastructure. Even if you do, the department that makes stuff and the department that related "publicly" often do not communicate efficiently enough.
But, then the bean counters that try to shave every penny on the cost of an item will balk at including a CD that does nothing for functionality -- arguing that a web server is cheaper in the long run. But, you lose the market window in the time it takes you to set one up in a manner consistent with corporate policies, and might not have the budget in the short term.
Finally, you might have customers who do not want to receive source, and balk at the CDs you send them (because getting rid of them is now an expense for them). Can you treat them in the "offer good for three years" manner? No, because you are not set up for the transferrable clause in that offer.
This is a case of "what is easy for the individual is hard for the corporation" because of political and scalability issues.
I rather like it too... the spicier the better! It's actually quite healthy, with all the vitamins.
But, if you prepare your own, yes, it does reek to high heaven as it's fermenting.
You can purchase it in Korean groceries, and the smell is nowhere near as rank as the homemade stuff.
My father used to like upenceller cheese. It made limburger smell like roses, by comparison. He swore it was delicious, but, to my nose, it had such a fecal odor as to be wretched beyond belief. He'd smear it quite thickly on rye bread -- "shitloaf" I think I called the result.
I once built a SBC based on a 6809 CPU with a 6883 (IIRC) Video/DRAM refresh controller and 64 KB RAM, but programmed a PAL wrong, and had no refresh. Would stay stable for about 30 seconds after reboot. Easy to fix, though.
Slashdot Mirror
Yes, this is an "important" person. But, he is 77. Dementia?
Granted, there are a lot of unexplained events that suggest the possiblity of extraterrestrial visitation. And, that might warrent investigation of the phenomenom, and a government bent on covering it up, perhaps with the assistance of "aliens" with "advanced technology" night be able to do so easily. (A leak, with proof? No problem! Go back in time and "fix" it.)
But, science does not work that way.
It requires repeatable, peer-reviewed, observation of the purported effect, and strong correlation to the purported cause.
We may have been visited by extraterrestrials. It's a secuctive thought for any geek, but one VIP's say so does not make it so, to anyone with a scientific mind.
And, this leads to levels of reliability and efficiency in infrastructure that are unstable, in that they can't be sustained with the available budget, once the altruistic and generous admin is no longer available.
So, is the correct thing to do not build out to unsustainable levels (that the residents would then rely upon), or build out to them and hope for the best?
Politicians will always promise more than can be realistically delivered, and sometimes "bad things" happen because of it. It's the nature of government.
In this case, it would probably be relatively cheap (compared to other projects) to DOUBLE or TRIPLE the support staff (one to three), make their wages market competetive (double them?) AND provide an incentive to the current lead to make himself replacable (bonus based on drills that his subordinates have to pass). The cost would go from what? $100k/year to $600k/year? Peanuts for a city like SF, even with its budget woes.
But, it is cheaper to not do that, and have a ready-made scapegoat in case this mismanagement causes a "failure": "It was his job."
Even more so when it's one's job to make sure it runs correctly. IOW, if he does provide access to others, and they screw up, it's his fault.
While paranoia may have something to do with it, I've found that a reasonable dose of it is healthy. I think, however, that it's more of a case of normal admin paranoia (which is a good thing), coupled with "damned if you do, damned if you don't" workplace policy (e.g. "give him access and you will suffer if he screws up"), that likely led to this paranoia growing to the degree it has.
Sane, logical, people are often driven "crazy" when forced to work in an illogical environment (where no matter what one does, one is "wrong", and the "correct" choice is the one which has the lowest product of "consequences" multiplied by "likelihood of discovery").
I recently took the "high road", and paid a hefty price for it: I had been hired to be subcontracted to a large "three letter" telecom firm. The project manager at this firm wanted my SSN. They had no legitimate reason for it: I was being paid (and had SS withholdings taken) by my employer, not them. The manager claimed that it was necessary for me to get an access badge. I responded that I didn't mind signing in every day and getting a temp badge. The manager then claimed it was necessary for a "background check". I responded that I would happily provide my SSN and any other identifying data (date of birth, drivers' license, etc.) to any reputable, well known, background-checking company of their choice. I was reported as "difficult" and reqested that I be removed. While my employer strongly defended my position, in the end, it was not reasonable that I be paid to be idle, we parted ways amicably, and I quicky found (much better!) work elsewhere.
Turns out the manager in question was allegedly driving those working "for" her as slaves and threatening them with derogatory credit reports from bogus lenders if they did not comply, using one instance of prior theft to justify such "background checks" requiring the SSN.
It was easy, (though expensive), to walk away from that job. But, in this case, lives may very well hang in the balance no matter what choice Childs makes: refusing raises the possibility of the network "going down". Complying, where he has reasonable belief that others will have a good chance of making the network go down, causing havoc in the city, could border on criminal negligence. As far as criminal charges, what could stick? "Theft?" (of access codes).
Should he be fired? Perhaps, for insubordination. But, if the management of the city is so bad that backup systems aren't in place, and properly trained IT staff to run networks, this might be the only way to raise awareness of the problem that could really impact lives of the city's residence. He may have very well done a very good deed -- as the story unfolds we'll know more. At this point, like with all accused, he should receive the benefit of the doubt.
But, regardless of whether he did the "right thing": No good deed goes unpunished, and he should realize this.
Damn. There were extra spaces in places. I hope people get the idea.
The bottom like is that, given a style standard, I should be able to take ANY valid piece of code and format it according to that style, correctly.
There are two issues: code layout (style), and code/documentation/naming consistency (standards).
There are a lot of different styles. What matters is that a consistent style is used for code that is checked in, and that that style be reasonably readable by all. Otherwise source code control system diffs become cumbersome. (Yes, one can diff on retrieval, but that gets expensive and not all source code control systems support it.) Beyond that, it would be really nice to reformat to whatever style the individual is accustomed when one spends a lot of time with the code).
My personal irritants are styles which are inconsistant. For example, indenting function argument lists right beyond the left bracket, but not aligning a right bracket and statements or intra-struct/class declarations the same way, It's not done in the latter case because it is "space wasteful". Is it not space-wasteful in the function call case as well?
if (foo) {
blah
blah
}
then why not
...arg, ...arg
void fooFunction(
)
?
Personally, I like my bracketing constructs to line up, brackets as well as braces. Furthernore, I find opening braces on a line by themselves wasteful. I don't object as much as I used to, but I come from the days of 24x80 text terminals, and seeing more code at once was a plus.
SO,
void fooFunc
( arg1,
arg2
)
{ stmt1;
stmt2;
}
(Note the indent for the bracketing elements, and the additional indent for the items bracketed. I'm not pedantic about the former.)
In practice, if anything within bracketing constructs fits on a single line, I'd collapse it to that. So,
void fooFunc(arg1, arg2)
{ stmt1;
stmt2;
}
Or, even:
void fooFunc(arg1, arg2) { stmt1; stmt1; }
That works great for little code snippets which are idiomatic, and extend to short switch cases with breaks.
As for complex compound expressions that don't fit on one line, I break them up on the basis of multiplicative and additive operators, and compound variable references are broken on the basis of direct (.) and indirect (->) selection operators. The common rule is to break arbitarily with an extra indent. I place diadic selectors or operators in the indent margin of the second line, consistent with the precedence of the operation. So, turn the page sideways and you see the compiler parse tree -- which is what you really need in the case of a complex compound expression, particularly in an if statement. So,
if
( (a < b)
&& (b < c)
)
{ stmt1;
val1
= a + b + c;
}
The assignment is really odd here, and I have played with
val1
= a + b + c;
as well, but the former is consistent with my diadic rule, and th latter is an exception.
You want ME to power and host part of YOUR datacenter to distribute data, some of which might be questionable, like CHILD PORNOGRAPHY, for which I can be held liable?
</hyperbole>
I don't think so.
I knew there was a reason I don't watch TV: no need for a set top box that I don't control.
It also controls my access to my office space and building elevator: it's an access card for areas of the whole building, not just the parking garage. So, I need to keep it on my person when at work. Rather cool, actually. Though, as with all mono-controls, when it is misplaced it is a real PITA.
It remains clipped to my slacks, therefore. The biggest danger is when it is removed at pre-laundry time and set aside instead of clipped to tomorrow's work slacks. Though, AC is correct in implying that the safest "resting place" from a convenience perspective would be the car... if that didn't violate corporate policy. The unattended car is not considered a safe place for access controls, corporated computers, or media, and correctly so.
The better question would be, "Why the heck can't we easily backup such mono-controls, even to a limited degree (to counter the risks of theft multiplied by the number of access devices)?"
Where I work, we have the option of monthly parking passes which are simply smart cards. The daily rate otherwise is $15.
If you are a monthly parker, and forget your smart card, however, you are STILL charged the full $15 daily parking rate, even though you paid for monthly parking.
I used to think this was simple extortion, until I realized that monthly parking is handled electronically, but daily parking is handled by a human and therefore actually costs more to implement. (why I can't just swipe a debit or credit card, as I can at many downtown unattended lots, I don't know).
Still, I'd think a montlhy parker who has forgotten his smart card should be refunded a good chunk of paying the daily rate upon presentation of a daily parking receipt and his active smart card.
Of course, don't get me started on automated car washes that don't warn in advance that (a) their debit card reader is down, and (b) their cash reader only takes exact change -- I was once stuck for 15 minutes in a car wash line with people honking behind me because the stupid reader with broken debit card handling wouldn't even take a $10 bill (and keep the damn change for a $7 wash). Automatec cash readers should ALWAYS allow an option to pay more for convenience if they can't provide change. Those that give receipts could easily indicate the overpayment so subsequent refund could be arranged.
The world is populated by morons. Some design stuff.
I usually explain that various NDAs prevent me from disclosing code I've written of significance, and suggest that I be asked to complete a programming exercise.
Most employers have a set at the ready these days, and I usually respond with the 1 hour answer and the 1 day answer, the later showing an evolution of the former, with polish and usually a more generic solution.
I think the CN700 does MPEG4 but is otherwise braindead compared to the CN400. Google VIA Unichrome.
Er, Via's nano-ITX offerings with a C3 processor, CN400, and VT1625 has been capable of this for at least two years now. The OpenChrome project provides for the accelleration: hardware MPEG2 decoding with XvMC. glxgears does about 500 fps on an 800 Mhz fanless cpu, closer to 700 on a 1.0 GHz CPU.
Been running this for a couple of years now in a Silverstone LC08 case.
A Libertarian is just a Republican who doesn't support the war
More correctly, a libertarian is an individual that values individual liberty, and decries the initiation of force. A Libertarian is a member of a recognized political party that pursues libertarian socio-political goals.
Libertarians align with Republicans on many economic issues (against increased taxation, though you wouldn't know it from looking at the NeoCon's record), and Democrats on many social issues (getting government out of the definition of marriage, prohibitions against same-sex marriage and poluygamy, and relaxation of drug laws).
Compare and contrast "Objectivist," ons subscribing to Ayn Rand's view of "Objectivism," which recognizes only three legitimate roles for government: 1) settling of disputes when private arbitration has failed, 2) protection from criminals, and 3) protection from foreigh invaders. While there is close agreement that less government is better, it isn't clear that any "government", except by concent of ALL of those governed, is necessary. See also, anarcho-capitalist, and less relevently, anarchist (since anarchists do not necessarly reject the inmitiation of force in achieving their goals, even as many of them do). All, however, can be classed as "minarchists": supporting the smallest government necessary.
I wonder how the Founders would have reacted to such proposed restrictions on their conduct.
Hmm. Did they not become founders (of a new nation), because of such restrictions on their conduct (among other things)?
We don't have to ask how they would have acted: we know how they did act.
Of course the Honorable House Majority Leader would claim that right-wing spin, in the absence of "fairness" is the source of that historical account.
If public funds are used to pay for a political message from an elected representative, it stands to reason that some allotment of public funds are available to all elected representatives for such purposes. Let the comm^H^H^H^Hdemocrats pay for their own damn propaganda.
(And, for the record, I am libertarian, and take equal issue with many things the right does as well).
I guess that XDR wasn't good enough, then, or ASN.1 (which supports multiple abstract encodings to boot).
XML, as an interchange format?
I suppose one could load source code into memory, and compile it every time, too. Even Java compiles to bytecode.
Bloated formats are fine for human interpretation (I rather like one kind of structure for my config files), or occasional parsing (which is why most of the stuff in /etc is human-readable, for small data sets (I do remember when "the internet" was one big /etc/hosts file), but for interchange? Just cause you're big-endian and I'm little-endian?
The trick to making non-human readable formats acceptable is the prevelence of wide-spread encoding and decoding tools.
Yes, XML is self-describing, at least syntactically (and formally with an XSD), and specific encoding semantics can be tagged, but the same can be achieved with means for type encoding. The big thing with XDR and related formats is that types are implicit -- both ends need to know what is being serialized. For RPC, with well-defined interfaces, this is not a problem, but it does make type-checking a remote service a bit of a challenge.
However, types can be encoded as data, and serialized as well: this happens for variant types naturally. Thus, there is no reason to not have a type-encoding and type-exchange protocol to permit dynamic type-checking. The advantage over self-describing data serializations is that it can be done on an as-required basis, instead of with every damn serialization.
It was a $250k home at the time.
I quit-claimed it to my ex in a divorce, and it would be worth about $360k today, even in the soft market, ... if she hadn't trashed it. For various reasons, I now have a power of sale over it, and about $31k in liens against it. With around $200 in mortgages on it, I will be making a reserve offer of $231 when I market it for 60 days.
But, it is quite trashed: a few holes in the drywall, and a non-functional microwave and stove. As well, she ripped out carpets in four bedrooms. The hardwood floors could stand to be refinished as well. She really did a number on it in the year since we divorced.
Nothing $5k - $10k can't fix, and restore the house to around $360k value. If someone wants a fixer-upper, it's a great deal. But, most buyers don't want fixer-uppers. In any case, I'll either wind up with the home, or my $31k.
Realtors have told me that in its present condition, it might fetch $260k - $270k if a willing buyer who wanted a fixer-upper could be found.
$100K got me into a 4BDR with central air 2400 ft^2 house 45 minutes from main campus, though, in 2004.
And the market is really soft now.
Well, technically it is possible to use them for endpoint authentication without encryption (which might be useful on a physically secure point to point link where one can't tell if the endoint has been compromised, and endpoint posession of private keys is transient, though why not encrypt as well, then?) but browsers are not configured to work that way.
But, yes, endpoint verification is only as good as the trust you place in whoever signed the certificate the endpoint presents.
In a closed community, with a secure means of distributing certificates, self-signed certs are fine. In fact, they are used to set up VPN links all the time.
Well-known CAs serve the purpose of vouching for the identities third parties present -- they are rather like government issued drivers licenses, or other photo ID. Bars use them all the time for age verification. Of course, the checking that most CAs do, compared to drivers' licensing departments is laughable, but that's a completely different issue from the potential usefulness of a signed X.509 cert.
The whole point of certificate authorities is scalability: one does not have to verify every single certificate individually -- one need only have trust in the (few) authorities that sign them. Even if one accepts self-signed certs from one's correspondent, it would make far more sense for the correspondent to generate one, and use it to sign the certificates used by all their servers, instead of having a self-signed cert per server.
PGP decentralizes the notion of a CA, but makes establishing individual trust webs more cumbersome. Some would argue this is a good thing. But, ultimately, who's signature you trust on a certificate is up to you, and no, it should not be blind.
As for vouching for the repeatability of reaching a given endpoint, even without CA signatures, this is only true when the endpoint's private key has not been compromised. So, you have to trust the care your correspondent places in their servers. This is also an argument for accepting self-signed certs as siging certs from known conterparties, but not as endpoint identification-serving certs: if the endpoint is compromised, but the CA cert isn't (managed by the same entity), CRLs and OCSP servers can still be trusted to say "certificate X has been compromised". If the signing cert is compromised, ALL certs signed by it can no longer be trusted, nor any related CRLs or OCSP servers.
Always kmow where your signing cert is.
A bit more than that -- in some places, telephone charges (analog modem), or internet access via other means (DSL, cable), might be prohibitively expensive.
All is fine if you originally distributed binaries that way (since then source was available "with" binaries), but not if you didn't.
I had one vendor argue with me that "FTP was a medium for data exchange" per the GPL, but of course the GPL was addressing physical media (we needed their mods to the Tulip ethernet driver in the kernel so we could apply subsequent patches -- and more importantly ALL their modified source to their binaries that we were redistributing -- which was impractical to FTP from them since their server died frequently). We did eventually get a source CD but only after a hostile email exchange.
Remind your bean counters that simply distributing the source with the binaries is, by far, the best way to avoid costly litigation like in TFA.
That may be the case today, but in the 90s, the likelyhood of GPL enforcement was largely regarded as a joke by many. Often it was those of us who respected it that went the extra mile to ensure that our employers complied by investing our time to make source distros, etc.
I was responsible for GPL complience at a former employer, and had RMS come and give a lecture on the subject to our devs and set up a build process where source RPMs would be packaged for all the GPL-tagged RPMS in our software install CDs automaticaly (so long as the SPEC files had the right license tags). It did not guarantee complience, of course (since it didn't check linking GPL and non-GPL code), but it helped avoid mistakes of omission.
But, for many products, like routers that sell for $20, including a $0.50 CD IS an exorbitant expense in a commodity product market of razor-thin margins. So, complience does have costs associated with it that some might try to illegally avoid.
Actually, yes.
The software may not be in a form for easy building and consumption. Think about dependency hell and configuration issues. A lot of shops have a golden "build machine" that does the right thing. This poor practice is less common today, but was du rigeur in a lot of shops where I worked in the 90s -- scrambling to get product out. Often, there is a desire to not publicize the "lack of polish".
Furthermore, there might not be a budget to support an internet-facing server if you don't already have one. Often, the "web presense" is managed by a different department than the engineering group. Again, less of an issue than it used to be, but still a problem. One place I worked (which was not exactly known for it's support of free software, and a large presence in Redmond, WA) would have to "internally lease" a server from the department that did this, and get billed $10k/month for the privilege.
Then, there is the issue of "mail requests". If you don't distribute source with binaries, and are not an academic institution, you must be willing to provide an offer, good for three years, to provide source to anyone who asks! -- not just those you distributed binaries to. A lot of businesses are not set up to do this. [This is for GPLv2. I have not checked if GPLv3 is as onerous.]
It actually is a lot easier to ship source with product if you don't already have the infrastructure to distribute to any and all comers. If you're not an agile shop with a dynamic internet presence, you don't have the infrastructure. Even if you do, the department that makes stuff and the department that related "publicly" often do not communicate efficiently enough.
But, then the bean counters that try to shave every penny on the cost of an item will balk at including a CD that does nothing for functionality -- arguing that a web server is cheaper in the long run. But, you lose the market window in the time it takes you to set one up in a manner consistent with corporate policies, and might not have the budget in the short term.
Finally, you might have customers who do not want to receive source, and balk at the CDs you send them (because getting rid of them is now an expense for them). Can you treat them in the "offer good for three years" manner? No, because you are not set up for the transferrable clause in that offer.
This is a case of "what is easy for the individual is hard for the corporation" because of political and scalability issues.
Er, what about an electrostatic solution: charge the damn panels and use electrostatic repulsion to drive the dust off?
I can see a problem if the charge also attracts charged particles, but I should think that could be mitigated,
But, if you prepare your own, yes, it does reek to high heaven as it's fermenting.
You can purchase it in Korean groceries, and the smell is nowhere near as rank as the homemade stuff.
My father used to like upenceller cheese. It made limburger smell like roses, by comparison. He swore it was delicious, but, to my nose, it had such a fecal odor as to be wretched beyond belief. He'd smear it quite thickly on rye bread -- "shitloaf" I think I called the result.
I once built a SBC based on a 6809 CPU with a 6883 (IIRC) Video/DRAM refresh controller and 64 KB RAM, but programmed a PAL wrong, and had no refresh. Would stay stable for about 30 seconds after reboot. Easy to fix, though.
4 hours is fine.
Er, meet my BIGGER friends, "Fat Man" and "Little Boy". And go tell that blowhard "Tsar Bomba" to get lost.
Er, no. Meet my friends: Smith and Wesson. :-)