While some developers undoubtably want to yield no quarter to Microsoft, I doubt that's the real problem.
Have you ever looked at code that truly supports Windows platforms? Microsoft may claim that it's a common platform, but the extra work required to support different Unix platforms (Linux, BSD, Solaris, HP/UX, AIX) is trivial compared to the extra work required to support W95, W98, WinME, NT4, W2K and now WXP.
The only reason most shops can get anything out the door is the fact that there are tools designed to hide this inconsistency. Few people program in Xlib directly, but it's accessible to those who need to do something Athena/Motif/KDE/Qt/et al don't do. But the last I heard, nobody (except maybe some games developers) gets within three or four layers of the Windows API. That makes the cost of cross-platform development extremely high, since the abstraction layers are so different.
Of course, low level programming still interacts with the APIs directly. But I remember shocking a former boss speechless when I gave him a copy of the Linux parallel printer driver. It was about 5 pages, and everything was done once. He had written similar device drivers for Windows and basically had to write the same code four times.
On a related note, this is why I continue to insist that Windows is a toy OS. The most fundamental requirement of an OS is to hide hardware and system details. I should not have to rewrite code so it works with Zip disks in addition to floppies, or SCSI drives in addition to IDE drives. Yet programs can't access NTFS disks unless the programmer recodes them. They can't migrate from Windows API to another unless extremely thick abstraction layers are used.
In contrast, with the "toy" Linux I have routinely migrated work between Solaris and HP/UX systems at work to Linux boxes, and back, creating an extremely flexible development environment. The required source code changes, if any, can be localized into #ifdef blocks. With autoconf, I don't even need to worry about different Makefiles.
If you're paid for your work, the significant extra work required to support Windows makes sense. Or if you're a major project, like Apache. But for somebody who is doing this work in their spare time, supporting Windows means that a lot of other things won't be done.
Discordianism might have been a joke, but it has a lot in common (and perhaps was even based on) real religions.
Ever hear of Coyote?
Or Raven?
Or Erdu?
Or Loki?
Or Lucifer, in his early days?
I don't think it's an exaggeration to say that all polytheistic religions contain a Trickster. Despite a few thousand years of Christian propaganda, they aren't "evil" or "destructive," just embracing of the unpredictable, of the wild. This is only a problem if you're a control freak (*cough* Leviticus *cough*).
I believe this is also why Discordianism before, and Trickster religions today (e.g., "Coyote Zen") are so popular among techies, even if we can't agree on definitions. Our working lives have to be extraordinarily controlled, so it's only natural that we're attracted to something that introduces some healthy unpredictability into our lives. Contrawise, people who have extremely chaotic lives are attracted to the highly regimented "everything has a rule" religions and cults.
On the larger issue, in the US there's the concept of the "Jeffersonian Cult of One." A religion does not require recognition from the government, does not require a minimum number of followers, etc., all it takes is ONE person who honestly believes in its tenets. Historically, like the "free exercise of religion" clause the religion still had to based on Christianity, but recently pretty much anything goes, provided you don't break (most) civil laws. But some civil laws can be broken, e.g., a recent case involved a local city that restricted the number of cars that could be parked on a city street for any household. A small church gathered in a private home, and routinely violated this law and was ticketed. The church sued, claiming the law had the de facto effect of unduely restricting First Amendment rights of assembly and religion since they did not block the street or other residences, only briefly took up most parking spaces once a week. They won, the law was thrown out as unconstitutional.
So in the US Wicca and pagan groups can get full First Amendment protection (although, in practice, there are still plenty of bigoted judges who feel no shame in proclaiming that a woman is unfit as a child's custodian because she's a Wiccan), and a "Jedi religion" would almost certainly qualify as well.
In the UK, the situation is much murkier since there's an official state religion.
This is a false analogy since Costco targets businesses in addition to individual purchasers. To be sure, Joe Smith, employee with a bene, can be tracked as an individual. But they don't know the breakdown of my purchases as a small business owner. Is that 2-line phone for home or office? What about the set of wire shelving? Even food items are problematic - is that pile of soda and candy bars for me, for clients, for employees, or for vending machines in the office?
Masers were the predecessors to lasers, producing microwave wavelength radiation instead of visible light. And saying that the research was done years ago is putting it mildly - IIRC masers were largely developed in the 50's, gas lasers in the 60's. They have absolutely nothing to do with this recent research.
That said, it's possible that some reporter with absolutely no technical background abbreviated "matter laser" to "maser," but that would be a mistake since it causes immense confusion to anyone who remembers the original definition. If you meant "matter laser," then say so.
Albertsons has publicly stated that it has no plans to introduce consumer cards. Management could change, of course, but they are undoubtably aware that the people who care enough about these cards to change grocery stores are also going to be among their most loyal customers as long as the other chains still use those cards.
Kroger and Safeway, on the other hand, seem to have these programs in place nationwide.
If you're gonna bitch, hit the right targets. Albertsons is a one of the good guys here, unless you know of specific counter-examples. If so, that would definitely be interesting since it's contrary to their stated policy.
The Challenger violently exploded because of the liquid oxygen. Fuel is important, but oxidizer is usually the limiting factor.
Dave Barry (iirc) even covered this in a column on the world's ultimate barbeque grill. Charcoal brikets, hardly an explosive, a tank of liquid oxygen, a lit cigarette in the charcoal as an ignition source, and a long rope. One tug, *boom*, and the charcoal burned fast enough to vaporize most of the cheap grill. In the Challenger explosion, you had that oxidizer dumped into the middle of gaseous hydrogen so there was an even quicker burning than you have with charcoal and its relatively low surface area/fuel ratio.
I normally hate that new age nonsense, but this is a case where it makes sense.
Don't worry about how many other sites support this, just worry about whether *you* support it. If you're sending sensitive material, you need to use an end-to-end protocol (e.g., PGP) regardless. If you're just trying to do opportunistic encryption of the channel (something which is still worthwhile to minimize the damage caused by casual sniffers), the limiting factor will always be the other side if you make sure that you're ready. If it's a site you often trade mail with, you can always encourage them to enable encryption themselves.
It makes no sense to wait for some magic threshold to be reached since that's the way (if everyone did it) to ensure that nobody acts.
For what it's worth, my outbound mail is qmail with the TLS patch. I hope stuff is encrypted, but if I'm worried I still use PGP. (My inbound mail is handled by my ISP, so I can't control encryption there. I grab mail from it via a SSH tunnel.)
This sounds a lot like the recently discovered slrn bug (see Bugtraq, LWN, Debian) that automatically executed all scripts encountered, apparently assuming they were self-extracting archive files.
However, I'm not sure Microsoft should be let off the hook for the equivalent behavior on the Mac. The Unix code was there for a very, very long time... when it was added it was a reasonable assumption that people would not send nasties because it was too easy to complain to their employer or grad department (the only way to get online) and cause the sender significant personal pain. (This is also a painful reminder that just because code is available doesn't mean that the right people are reviewing it.) In contrast, by the time somebody added that code to the Mac version of MSIE, the possibility of untraceable, hostile scripts should have been obvious.
Don't sell CVS short, it may be quicker and cheaper to expand your existing CVS system than to replace it with a whole new system.
For instance, what precisely does management want to see? Can you harvest that information with simple scripts, or via calling programs at the various places CVS already provides hooks?
If the problem is poor documentation of the changes... that's a management issue, not a tool issue. Management needs to make sure people take this seriously - it should be a line-item in the employee review process.
As for the "problem" of multiple users simultaneously checking out the same files, that's a feature, not a bug. You would hope that adults would know how to work together, but in the real world almost every shop I've worked at with an "exclusive lock" source control system has had at least one jerk who would sit on locked files for days (or weeks!) at a time, then submit monster changes. CVS strongly encourages people to fix one thing at a time, and if someone makes a huge change then it's their responsibility to merge the differences.
Finally, I'm curious about the "work has been lost" statement. Is this because of poor practices (resulting in frequent change collisions and someone discarding code), or something specific to CVS?
P.S., unless you're in a completely homogeneous environment, the multitude of CVS clients for different platforms is a benefit that should not be overlooked. I've worked in several shops with CVS being used on Win9x, NT, and various Unix boxen without a problem. Most other solutions are either specific to a single platform, or very expensive.
Shouldn't someone point out to this university that intercepting and displaying email you are not a party to is still a federal offense (ECPA - Electronic Communications Privacy Act)?
Your boss can do it because, technically, you're acting as an agent for the company and all email sent to/from your work computer should be done on behalf of the company.
Your ISP and university can block spam, strip executable attachments, etc., because the filtering can be done because 1) it serves an important public need and 2) it can be done in a mechnical manner that does not require human intervention.
But students are not "agents" of a university, they are customers. Universities often impose rules that skirt (or outright break) the law, especially for students living in the university-provided housing, but I'm not sure that they can make any blanket assertion of the right to intercept all email sent through their system. E.g., many non-traditional students will attend class with personal or company-provided laptops which may attempt to send previously queued, but unsent, confidential material that will be transmitted once a network connection can be re-established. If the university doesn't want to allow such communications, they can block outgoing SMTP ports. While it's technically possible to configure a system to only send mail when connected to some networks, it's non-trivial and rarely done in practice.
I don't recall if ECPA covers "instant messages" explicitly, but seems more likely than not to be considered protected than not since they are not broadcast.
(IANAL, but familiarity with the ECPA should be considered required knowledge for anyone with system administration duties.)
I recently had to make a frantic trip to my bank to request a credit card - I've been using debit cards exclusively for a while, but when booking a trip I learned that most (not all) car rental companies require a credit card, not a debit card, to rent a car.
Even more bizarrely, I was told that it doesn't matter if the debit card is backed by a $10k balance, while the credit card has a $1k limit (although I ended up getting a much higher limit). The "logic" was that debit cards usually have a daily limit, vs. credit cards to not. Again, this logic is rather odd since that debit card daily limit may still be higher than some credit card limits.
Don't be too cheap here. Power supplies do go out (or catch fire because of dust or pet hair that blew into them) and even a "good" unit might have problem maintaining its rated capacity after a number of years.
Keep the chassis, but replace the power supply. Besides, since this is a server this sounds like a good place to use one of those units with a built-in UPS, even if box is hooked up to an external UPS. (I've had them fail because I unwittingly overloaded them, because of poor designs that allowed me to accidently turn off the UPS protection but not the power, etc.)
When push comes to shove, I'll package it myself and install the package to/opt instead of/usr. Too bad that makes deblint complain, unless they fixed it. (Official Debian packages shouldn't install to/opt, but by the same measure my own unofficial packages shouldn't install to/usr.)
I usually update when they go into a release freeze, but 'unstable' really is too unstable for me to use. I've been burned in the past, and my spare systems (P-166) are too limited to do some of the necessary work.
To be fair (since many non-developers read this list), this release cycle has been particularly nasty because some key Debian tools now depend on Perl 5.5, and that forces you to fully convert to 'unstable' instead of installing a handful of selected packages. (Some maintainers archived the last pre-5.5 version in 'pool', but many did not.) But by that same measure the release manager should have identified introduction of Perl 5.5 and the tools as the trigger for the next release since it's a *huge* roadblock to this type of "stable + one or two updated packages" approach that many of us prefer.
I'm currently working on a certificate authority written with servlets (and JNI calls to libopenssl for the gory work of actually creating and signing the X.509 certs), and everything is using EJB beans. The goal is to have the CA entity beans handle the actual CA and X.509 tasks, another set of beans and JSP to handle the web and java client interfaces, and yet another set of beans to handle the business rules regarding content and issuance of the certs, and tying it all together with J2EE or something similar.
The only problem is that I seem to be missing a piece of the puzzle. For now, I'm creating and initializing the beans explicitly, but shouldn't this be handled automatically somewhere/somehow? I'm sure I'm just missing some small piece of information in this huge pile. Does this release address this problem, or is it an entirely different set of code?
(As a related aside, I'm gonna stop using Debian if it continues to have such long release cycles. I eventually got suitable openssl (0.9.6), postgres (7.0) and java (1.3) installed, but it took days and a lot of pain because of the length of the "to do A you must first do B, to do B you must first do C, to do... chain.)
I like my Linksys hub/router, but the support has been downright hostile once I mentioned I run Linux. Like it matters - it's an entirely separate device configured through web pages. But like many of us, I usually run with javascript disabled and their pages provide no indication of why the router can be nonresponsive.
As for the suggestion that you run an old box, please, give it up. If it works for you, great, but I switched from a box to a hub because of power consumption, noise, floor space, etc. Except for those hassles with javascript, I haven't regretted this decision.
and set up CVS to use RSA Authentication (/etc/ssh/sshd_config)
RhostsAuthentication no
RhostsRSAAuthentication yes
RSAAuthentication yes
PasswordAuthentication no
plus similar for SSH2. This requires that each user and each host have its public SSH key on the CVS server, in their home directory and/etc/ssh/ssh_known_hosts respectively. In practice, you might find it an acceptable tradeoff between security and convenience to turn off the RhostsRSAAuthentication flag.
Finally, there's some work on implementing SSL/TLS directly into the CVS server, to eliminate the need to provide local user accounts on the server. This should dramatically increase the security of the repositories since it allows them to be turned into closed systems without user shell access. In the most likely scenario, CVS will be able to function much like SSH - you can operate in anonymous mode, or you can require PKI authentication of either or both parties.
I believe that Canada has freely shared firefighters with us when fighting wildfires, and I seem to recall other crews from as far away as Australia.
The author has some valid points, but the US usually doesn't require much outside help because few disasters affect us nationally. When we do, Canada and our other friends have always come to our aid.
The funniest thing (as in, I almost hope it passes so I can be the person to personally kick your sorry ass back into the dark ages) is the stiff fine for any "computer" that can be connected to the internet that doesn't provide for DRM.
Let's see, that's your office telephone/PBX, your office hub, your cable/DSL modem, your ISP's routers, the POP, etc. It includes almost every mainframe and large server for years - the law may require all new computer hardware to include DRM, but how often are million-dollar-plus systems replaced? For that matter, what about all of the legacy mainframes which aren't manufactured today?
Even if the Senator harrumps and says that I should stop being dense because I should know that "computer" refers to "PC-class computer" (even though countless other recent laws have repeatedly driven home the axiom that you should ignore the stated intent of the law and focus on the wording in the law itself), it will criminalize those projects to build beowulf clusters out of discarded PCs, amateur scientist projects which hook up instruments to the net with old PCs, etc.
You don't have to rewrite the entire source control software, just put in a filter. CVS has a number of hooks for this - at a minimum you should run an XML validator to ensure that only valid XML is checked in.
It might be an old joke, but if so it's based in fact. Knowledgeable people I know and trust have told me the same thing - that early XML support (at least) was nothing but a thin wrapper around the existing file format.
The question is if Office XP has peeled back a few layers of the onion.
But what does that XML actually look like? Early MS XML support, at least, replaced
blob of binary data
with
<ms-office type="word">
blob of binary data
</ms-office>
Technically a well-formed XML file (assuming that the DTD shows the ms-office tag as having CDATA content), but it's useless as a shared document format.
DocBook, and any decent document XML DTD, gives you the ability to tag your text with some description of what it means. It might be "chapter" or "list," or it might be domain specific like "files," "bugs" and "see also" (for man pages). The presentation details are left to the processing software to handle.
MS-Word, in contrast, is nothing but a paint tool for words. You can certainly give your styles names that have some domain meaning to you, but it's still ultimately nothing but a set of style instructions.
For a single document, this isn't a big issue. But if you have a lot of documents and you want to reuse content, it's impossible with the MS-Word approach. With DocBook, in contrast, it's easy to set up your documents so that the same file can be reused in multiple places, but only selected content will be reused.
IMHO, if your technical writers can't make the shift to meaningful tags, you're better off without them. (The writers, not the tags.) If they can't handle this level of structure, their writing is undoubtably muddled and confused no matter how pretty it looks.
<ms-office type="word">
blob of binary data
</ms-office>
Or something like that. Just like that joke about the guy lost in the helicopter, they provided a format which is technically correct, but totally misses the point.
Personally, I think the child porn laws have gone way too far, but there's a huge difference between the arguments that a technology "might" be used in traffic kiddie porn and the ongoing prosecution of a massive (million dollars/month) ring.
Slashdot Mirror
While some developers undoubtably want to yield no quarter to Microsoft, I doubt that's the real problem.
Have you ever looked at code that truly supports Windows platforms? Microsoft may claim that it's a common platform, but the extra work required to support different Unix platforms (Linux, BSD, Solaris, HP/UX, AIX) is trivial compared to the extra work required to support W95, W98, WinME, NT4, W2K and now WXP.
The only reason most shops can get anything out the door is the fact that there are tools designed to hide this inconsistency. Few people program in Xlib directly, but it's accessible to those who need to do something Athena/Motif/KDE/Qt/et al don't do. But the last I heard, nobody (except maybe some games developers) gets within three or four layers of the Windows API. That makes the cost of cross-platform development extremely high, since the abstraction layers are so different.
Of course, low level programming still interacts with the APIs directly. But I remember shocking a former boss speechless when I gave him a copy of the Linux parallel printer driver. It was about 5 pages, and everything was done once. He had written similar device drivers for Windows and basically had to write the same code four times.
On a related note, this is why I continue to insist that Windows is a toy OS. The most fundamental requirement of an OS is to hide hardware and system details. I should not have to rewrite code so it works with Zip disks in addition to floppies, or SCSI drives in addition to IDE drives. Yet programs can't access NTFS disks unless the programmer recodes them. They can't migrate from Windows API to another unless extremely thick abstraction layers are used.
In contrast, with the "toy" Linux I have routinely migrated work between Solaris and HP/UX systems at work to Linux boxes, and back, creating an extremely flexible development environment. The required source code changes, if any, can be localized into #ifdef blocks. With autoconf, I don't even need to worry about different Makefiles.
If you're paid for your work, the significant extra work required to support Windows makes sense. Or if you're a major project, like Apache. But for somebody who is doing this work in their spare time, supporting Windows means that a lot of other things won't be done.
Discordianism might have been a joke, but it has a lot in common (and perhaps was even based on) real religions.
Ever hear of Coyote?
Or Raven?
Or Erdu?
Or Loki?
Or Lucifer, in his early days?
I don't think it's an exaggeration to say that all polytheistic religions contain a Trickster. Despite a few thousand years of Christian propaganda, they aren't "evil" or "destructive," just embracing of the unpredictable, of the wild. This is only a problem if you're a control freak (*cough* Leviticus *cough*).
I believe this is also why Discordianism before, and Trickster religions today (e.g., "Coyote Zen") are so popular among techies, even if we can't agree on definitions. Our working lives have to be extraordinarily controlled, so it's only natural that we're attracted to something that introduces some healthy unpredictability into our lives. Contrawise, people who have extremely chaotic lives are attracted to the highly regimented "everything has a rule" religions and cults.
On the larger issue, in the US there's the concept of the "Jeffersonian Cult of One." A religion does not require recognition from the government, does not require a minimum number of followers, etc., all it takes is ONE person who honestly believes in its tenets. Historically, like the "free exercise of religion" clause the religion still had to based on Christianity, but recently pretty much anything goes, provided you don't break (most) civil laws. But some civil laws can be broken, e.g., a recent case involved a local city that restricted the number of cars that could be parked on a city street for any household. A small church gathered in a private home, and routinely violated this law and was ticketed. The church sued, claiming the law had the de facto effect of unduely restricting First Amendment rights of assembly and religion since they did not block the street or other residences, only briefly took up most parking spaces once a week. They won, the law was thrown out as unconstitutional.
So in the US Wicca and pagan groups can get full First Amendment protection (although, in practice, there are still plenty of bigoted judges who feel no shame in proclaiming that a woman is unfit as a child's custodian because she's a Wiccan), and a "Jedi religion" would almost certainly qualify as well.
In the UK, the situation is much murkier since there's an official state religion.
This is a false analogy since Costco targets businesses in addition to individual purchasers. To be sure, Joe Smith, employee with a bene, can be tracked as an individual. But they don't know the breakdown of my purchases as a small business owner. Is that 2-line phone for home or office? What about the set of wire shelving? Even food items are problematic - is that pile of soda and candy bars for me, for clients, for employees, or for vending machines in the office?
Masers were the predecessors to lasers, producing microwave wavelength radiation instead of visible light. And saying that the research was done years ago is putting it mildly - IIRC masers were largely developed in the 50's, gas lasers in the 60's. They have absolutely nothing to do with this recent research.
That said, it's possible that some reporter with absolutely no technical background abbreviated "matter laser" to "maser," but that would be a mistake since it causes immense confusion to anyone who remembers the original definition. If you meant "matter laser," then say so.
One line really pisses me off....
Albertsons has publicly stated that it has no plans to introduce consumer cards. Management could change, of course, but they are undoubtably aware that the people who care enough about these cards to change grocery stores are also going to be among their most loyal customers as long as the other chains still use those cards.
Kroger and Safeway, on the other hand, seem to have these programs in place nationwide.
If you're gonna bitch, hit the right targets. Albertsons is a one of the good guys here, unless you know of specific counter-examples. If so, that would definitely be interesting since it's contrary to their stated policy.
The Challenger violently exploded because of the liquid oxygen. Fuel is important, but oxidizer is usually the limiting factor.
Dave Barry (iirc) even covered this in a column on the world's ultimate barbeque grill. Charcoal brikets, hardly an explosive, a tank of liquid oxygen, a lit cigarette in the charcoal as an ignition source, and a long rope. One tug, *boom*, and the charcoal burned fast enough to vaporize most of the cheap grill. In the Challenger explosion, you had that oxidizer dumped into the middle of gaseous hydrogen so there was an even quicker burning than you have with charcoal and its relatively low surface area/fuel ratio.
I normally hate that new age nonsense, but this is a case where it makes sense.
Don't worry about how many other sites support this, just worry about whether *you* support it. If you're sending sensitive material, you need to use an end-to-end protocol (e.g., PGP) regardless. If you're just trying to do opportunistic encryption of the channel (something which is still worthwhile to minimize the damage caused by casual sniffers), the limiting factor will always be the other side if you make sure that you're ready. If it's a site you often trade mail with, you can always encourage them to enable encryption themselves.
It makes no sense to wait for some magic threshold to be reached since that's the way (if everyone did it) to ensure that nobody acts.
For what it's worth, my outbound mail is qmail with the TLS patch. I hope stuff is encrypted, but if I'm worried I still use PGP. (My inbound mail is handled by my ISP, so I can't control encryption there. I grab mail from it via a SSH tunnel.)
This sounds a lot like the recently discovered slrn bug (see Bugtraq, LWN, Debian) that automatically executed all scripts encountered, apparently assuming they were self-extracting archive files.
However, I'm not sure Microsoft should be let off the hook for the equivalent behavior on the Mac. The Unix code was there for a very, very long time... when it was added it was a reasonable assumption that people would not send nasties because it was too easy to complain to their employer or grad department (the only way to get online) and cause the sender significant personal pain. (This is also a painful reminder that just because code is available doesn't mean that the right people are reviewing it.) In contrast, by the time somebody added that code to the Mac version of MSIE, the possibility of untraceable, hostile scripts should have been obvious.
Don't sell CVS short, it may be quicker and cheaper to expand your existing CVS system than to replace it with a whole new system.
For instance, what precisely does management want to see? Can you harvest that information with simple scripts, or via calling programs at the various places CVS already provides hooks?
If the problem is poor documentation of the changes... that's a management issue, not a tool issue. Management needs to make sure people take this seriously - it should be a line-item in the employee review process.
As for the "problem" of multiple users simultaneously checking out the same files, that's a feature, not a bug. You would hope that adults would know how to work together, but in the real world almost every shop I've worked at with an "exclusive lock" source control system has had at least one jerk who would sit on locked files for days (or weeks!) at a time, then submit monster changes. CVS strongly encourages people to fix one thing at a time, and if someone makes a huge change then it's their responsibility to merge the differences.
Finally, I'm curious about the "work has been lost" statement. Is this because of poor practices (resulting in frequent change collisions and someone discarding code), or something specific to CVS?
P.S., unless you're in a completely homogeneous environment, the multitude of CVS clients for different platforms is a benefit that should not be overlooked. I've worked in several shops with CVS being used on Win9x, NT, and various Unix boxen without a problem. Most other solutions are either specific to a single platform, or very expensive.
Shouldn't someone point out to this university that intercepting and displaying email you are not a party to is still a federal offense (ECPA - Electronic Communications Privacy Act)?
Your boss can do it because, technically, you're acting as an agent for the company and all email sent to/from your work computer should be done on behalf of the company.
Your ISP and university can block spam, strip executable attachments, etc., because the filtering can be done because 1) it serves an important public need and 2) it can be done in a mechnical manner that does not require human intervention.
But students are not "agents" of a university, they are customers. Universities often impose rules that skirt (or outright break) the law, especially for students living in the university-provided housing, but I'm not sure that they can make any blanket assertion of the right to intercept all email sent through their system. E.g., many non-traditional students will attend class with personal or company-provided laptops which may attempt to send previously queued, but unsent, confidential material that will be transmitted once a network connection can be re-established. If the university doesn't want to allow such communications, they can block outgoing SMTP ports. While it's technically possible to configure a system to only send mail when connected to some networks, it's non-trivial and rarely done in practice.
I don't recall if ECPA covers "instant messages" explicitly, but seems more likely than not to be considered protected than not since they are not broadcast.
(IANAL, but familiarity with the ECPA should be considered required knowledge for anyone with system administration duties.)
I recently had to make a frantic trip to my bank to request a credit card - I've been using debit cards exclusively for a while, but when booking a trip I learned that most (not all) car rental companies require a credit card, not a debit card, to rent a car.
Even more bizarrely, I was told that it doesn't matter if the debit card is backed by a $10k balance, while the credit card has a $1k limit (although I ended up getting a much higher limit). The "logic" was that debit cards usually have a daily limit, vs. credit cards to not. Again, this logic is rather odd since that debit card daily limit may still be higher than some credit card limits.
Don't be too cheap here. Power supplies do go out (or catch fire because of dust or pet hair that blew into them) and even a "good" unit might have problem maintaining its rated capacity after a number of years.
Keep the chassis, but replace the power supply. Besides, since this is a server this sounds like a good place to use one of those units with a built-in UPS, even if box is hooked up to an external UPS. (I've had them fail because I unwittingly overloaded them, because of poor designs that allowed me to accidently turn off the UPS protection but not the power, etc.)
When push comes to shove, I'll package it myself and install the package to /opt instead of /usr. Too bad that makes deblint complain, unless they fixed it. (Official Debian packages shouldn't install to /opt, but by the same measure my own unofficial packages shouldn't install to /usr.)
I usually update when they go into a release freeze, but 'unstable' really is too unstable for me to use. I've been burned in the past, and my spare systems (P-166) are too limited to do some of the necessary work.
To be fair (since many non-developers read this list), this release cycle has been particularly nasty because some key Debian tools now depend on Perl 5.5, and that forces you to fully convert to 'unstable' instead of installing a handful of selected packages. (Some maintainers archived the last pre-5.5 version in 'pool', but many did not.) But by that same measure the release manager should have identified introduction of Perl 5.5 and the tools as the trigger for the next release since it's a *huge* roadblock to this type of "stable + one or two updated packages" approach that many of us prefer.
I'm currently working on a certificate authority written with servlets (and JNI calls to libopenssl for the gory work of actually creating and signing the X.509 certs), and everything is using EJB beans. The goal is to have the CA entity beans handle the actual CA and X.509 tasks, another set of beans and JSP to handle the web and java client interfaces, and yet another set of beans to handle the business rules regarding content and issuance of the certs, and tying it all together with J2EE or something similar.
The only problem is that I seem to be missing a piece of the puzzle. For now, I'm creating and initializing the beans explicitly, but shouldn't this be handled automatically somewhere/somehow? I'm sure I'm just missing some small piece of information in this huge pile. Does this release address this problem, or is it an entirely different set of code?
(As a related aside, I'm gonna stop using Debian if it continues to have such long release cycles. I eventually got suitable openssl (0.9.6), postgres (7.0) and java (1.3) installed, but it took days and a lot of pain because of the length of the "to do A you must first do B, to do B you must first do C, to do... chain.)
I like my Linksys hub/router, but the support has been downright hostile once I mentioned I run Linux. Like it matters - it's an entirely separate device configured through web pages. But like many of us, I usually run with javascript disabled and their pages provide no indication of why the router can be nonresponsive.
As for the suggestion that you run an old box, please, give it up. If it works for you, great, but I switched from a box to a hub because of power consumption, noise, floor space, etc. Except for those hassles with javascript, I haven't regretted this decision.
To set up CVS over SSH, use the following environment variables:
and set up CVS to use RSA Authentication (/etc/ssh/sshd_config)
plus similar for SSH2. This requires that each user and each host have its public SSH key on the CVS server, in their home directory and
Finally, there's some work on implementing SSL/TLS directly into the CVS server, to eliminate the need to provide local user accounts on the server. This should dramatically increase the security of the repositories since it allows them to be turned into closed systems without user shell access. In the most likely scenario, CVS will be able to function much like SSH - you can operate in anonymous mode, or you can require PKI authentication of either or both parties.
I believe that Canada has freely shared firefighters with us when fighting wildfires, and I seem to recall other crews from as far away as Australia.
The author has some valid points, but the US usually doesn't require much outside help because few disasters affect us nationally. When we do, Canada and our other friends have always come to our aid.
The funniest thing (as in, I almost hope it passes so I can be the person to personally kick your sorry ass back into the dark ages) is the stiff fine for any "computer" that can be connected to the internet that doesn't provide for DRM.
Let's see, that's your office telephone/PBX, your office hub, your cable/DSL modem, your ISP's routers, the POP, etc. It includes almost every mainframe and large server for years - the law may require all new computer hardware to include DRM, but how often are million-dollar-plus systems replaced? For that matter, what about all of the legacy mainframes which aren't manufactured today?
Even if the Senator harrumps and says that I should stop being dense because I should know that "computer" refers to "PC-class computer" (even though countless other recent laws have repeatedly driven home the axiom that you should ignore the stated intent of the law and focus on the wording in the law itself), it will criminalize those projects to build beowulf clusters out of discarded PCs, amateur scientist projects which hook up instruments to the net with old PCs, etc.
You don't have to rewrite the entire source control software, just put in a filter. CVS has a number of hooks for this - at a minimum you should run an XML validator to ensure that only valid XML is checked in.
It might be an old joke, but if so it's based in fact. Knowledgeable people I know and trust have told me the same thing - that early XML support (at least) was nothing but a thin wrapper around the existing file format.
The question is if Office XP has peeled back a few layers of the onion.
But what does that XML actually look like? Early MS XML support, at least, replaced
blob of binary data
with
<ms-office type="word">
blob of binary data
</ms-office>
Technically a well-formed XML file (assuming that the DTD shows the ms-office tag as having CDATA content), but it's useless as a shared document format.
You're dealing with different problems here.
DocBook, and any decent document XML DTD, gives you the ability to tag your text with some description of what it means. It might be "chapter" or "list," or it might be domain specific like "files," "bugs" and "see also" (for man pages). The presentation details are left to the processing software to handle.
MS-Word, in contrast, is nothing but a paint tool for words. You can certainly give your styles names that have some domain meaning to you, but it's still ultimately nothing but a set of style instructions.
For a single document, this isn't a big issue. But if you have a lot of documents and you want to reuse content, it's impossible with the MS-Word approach. With DocBook, in contrast, it's easy to set up your documents so that the same file can be reused in multiple places, but only selected content will be reused.
IMHO, if your technical writers can't make the shift to meaningful tags, you're better off without them. (The writers, not the tags.) If they can't handle this level of structure, their writing is undoubtably muddled and confused no matter how pretty it looks.
Last I heard, the MS XML format replaced
blob of binary data
with
<ms-office type="word">
blob of binary data
</ms-office>
Or something like that. Just like that joke about the guy lost in the helicopter, they provided a format which is technically correct, but totally misses the point.
Personally, I think the child porn laws have gone way too far, but there's a huge difference between the arguments that a technology "might" be used in traffic kiddie porn and the ongoing prosecution of a massive (million dollars/month) ring.