Yes, but why should I come on as your employee for that? Why shouldn't I come on as a short-term contractor instead? If you aren't going to offer me a stable, decent base salary, I'm better off avoiding W2 status (as a contractor I can write off a lot of stuff on my taxes that I can't as a W2 employee).
Date: Fri, 7 Apr 2006 10:09:27 -0700 (PDT)
From: Todd Knarr <xxxx@xxxxxx.xxx>
To: sale@dlink.com, customerservice@dlink.com
Subject: DLink router use of Danish NTP server
This is in reference to the open letter to DLink from Danish sysadmin Poul-Henning Kamp (http://people.freebsd.org/~phk/dlink/). Abuse of an NTP server in express violation of the service agreement in the Stratum-1 server list is, in my opinion, inexcusable. Willful refusal to correct the abuse when requested is, if anything worse. Hard-coding the server name into the firmware, so that changes are difficult or infeasible, as opposed to DLink maintaining their own DNS records so that changes are simple, is also inexcusable in any technically-competent organization.
I have been comtemplating purchase of a DLink DI-784 router/AP, a DWL-7100AP access point and a DWL-AG660 CardBus adapter. If DLink doesn't correct their error as Mr. Kamp asks, I will be taking my purchases to NetGear instead. They, at least, have demonstrated a willingness to fix their mistakes when asked. I will also be recommending to my friends that they avoid DLink products in the future.
Oh, I know all too well that e-mail spam costs money. But it doesn't cost money in a way that you can tie directly to the spam in the ledgers. Doing it requires a lot of complicated work with the techies, and the numbers are almost always ambiguous (at least to the business types). With FAXes the accountant can say "FAX consumable costs were up 180% last month. Legitimate FAX volume was down 4%, so the increase isn't from normal traffic. Junk FAXes were 52% of received FAXes, and they're the only thing that could've caused the increase in costs.". That's something even an MBA gets immediately.
The FCC's made a mis-step here. Junk e-mail is one thing, it costs time and hassle but not money. Junk faxes, though, cost money. The accountants will see the cost of consumables (paper, ink/toner) go up, and they'll be able to tie it directly to junk faxes. That's when the business groups start calling their Congressmen saying "Your FCC's decision is costing our members money. Do something, or come election time our contributions go to your opponent.". That's why the junk-fax provisions of the TCPA were put in in the first place.
Of course, there's also another catch. The FAX-sending entity probably has a FAX line too. If they're claiming an existing business relationship with you, they can't very well deny you having an existing business relationship with them, now can they? And these new rules allow you to send junk FAXes to entities you have an existing business relationship with, don't they?
Actually he has entered into an agreement. Unfortunately, in all but 2 states it's going to be the default contract of sale defined by the Uniform Commercial Code. That means by default you have the rights granted under copyright law as regards making additional copies, plus the rights granted by the UCC as regards use of the copy you bought. This is one thing the copyright cartels try fiercely to avoid dealing with, with their copyright-law and "you only have a license" rhetoric, because there's a lot of UCC case law out there and most of it contradicts the position they'd like to take.
Sorry, AT&T. I've used 1.5mbps, 3mbps and 5mbps services in the last year. There very definitely is a discernable difference in speed. You won't notice it until you actually use the higher speeds, but once you have you won't want to go back.
Now, AT&T may be rigth that their backbone is limiting their customers to slower speeds, but that's a problem with their backbone they need to fix. Other providers don't have that problem.
Because one of Microsoft's big arguments in favor of sticking with MS software is the cost and hassle of converting all those Microsoft-format documents into the other software's formats. If customers don't have to convert documents, there's not much argument in favor of MS when license renewal comes up and Finance says "Why should we spend $BIGNUM on MSOffice licenses when we can spend $BIGNUM/10 on OpenOffice instead and be able to do everything we need?".
This is the real reason Microsoft is worried about open source but deathly afraid of open formats.
If you've looked at ODF and MSXML, you'll see why it matters which one's the standard. If you look at a comparison of ODF and MSXML, you'll see the differences. You'll notice that the XHTML and ODF examples read like document mark-up: you have the recognizable text of the document and things like paragraph and italic marks occur at the obvious places. This makes it really easy to manipulate ODF via XSLT to turn it into other formats. MSXML, by contrast, reads like an XML encoding of the internal object representation within Word. Instead of text being a paragraph you have a paragraph object with several attributes, one of which is the text of the paragraph. Notice how complex the nesting gets when you've got words in italics and boldface within a paragraph. This structure is going to be a real bitch to manipulate using XSLT. In fact Microsoft themselves admit this, saying that MSXML isn't meant to be manipulated by standard tools like XSLT but by programs using Microsoft's own APIs. What's the use of XML if you can't manipulate the document using standard tools for manipulating XML?
Only access to the web sites of about half the people I know. And access to half my hardware vendors (including such minor things as case-maker Lian-Li and thermal product vendor Zalman). And access to the support site for my motherboard (made by Soyo). And a huge number of anime-related sites.
Shortly after the first FCC-approved implementation by an ISP, we'll see this:
Google announced today that, effective immediately, ISPs who wish enhanced bandwidth when using Google's services can obtain it by paying the nominal fee of 1 cent per kilobyte of transit. ISPs who do not wish to pay for enhanced access will retain normal best-effort access to bandwidth not used by those entitled to enhanced access. A Google spokesperson said "The ISPs are the ones who think this is a good idea, we're firmly committed to letting them live by their beliefs.". Elsewhere, ISPs expressed outrage at this idea, and fear that they'd be forced to either pay for enhanced access or lose users due to degraded access to the most popular search engine. Google responded by quoting back the ISPs' own statements about how providing enhanced access for a fee was not degrading service for those who didn't pay.
That depends. Number of looks isn't nearly as important as number of conversions. As an advertiser, given a choice between an advertisement that'll bring 1000 people into your store but only 1 will likely buy or an advertisement that'll only bring in 100 people but 10 of them are likely to buy, which one is a better advertisement?
A high number of views may mean that you're getting your ad out in front of a lot of the kind of people who click through an ad solely to cost the advertiser money or to give the site the ad's on money without any intention of actually buying what's advertised.
It's not a suprise that ugly-but-functional websites are successful in attracting and retaining visitors. Remember this: people visit a site looking for something. First and foremost, the site has to make it fairly easy (or at least straightforward) for visitors to get what they came there for. All else matters only after you've done that. If you meet that first requirement, visitors will forgive an awful lot of ugly. Functional and attractive is best, but if they have to choose visitors will choose functional over pretty when it comes time to actually use things. Note that your focus groups may not reflect this, because they're generally giving only the user's first impressions. First impressions fade fast when it comes time to actually do something.
It won't included in the current back. But it will still be in previous backups, and those aren't thrown out immediately because the admins may still need them. The current backup may be unreadable (hardware fault, tape's too old or damaged, any number of reasons), the data may have been corrupted so the last few backups are of bad data and you have to go back a week to a point before the corruption started to get a good restore, etc. etc..
To give you an idea of a normal backup rotation, the standard rotation I learned long ago was daily incremental backups (everything that's changed since the last backup) and weekly full backups (the whole system). Normal was 2 weeks worth of incrementals and 5 weeks worth of fulls (enough to span any full month). The first full backup of each month rotates out to the monthly backups, and monthlies got kept for a full year.
And what would you (and other users) say if a mistake by a Google sysadmin deleted every mailbox in the system and they said "Sorry, we don't have any backups, your mail is gone forever."?
This isn't a suprise. What Google's policy says is simple and obvious: "We make backups of our systems. That includes data files like your mailbox. We archive the backups on a rotating schedule that you don't know, so don't go assuming you know when any particular day's backup will be wiped. And we don't go back and alter those backups when you modify your data, so don't assume that deleting something today makes it disappear from all backups back to the beginning of time (or the inception of our service).". This subpoena is no different from a standard subpoena to a company asking for all documents including archived copies. If you wrote a memo, it got archived and then later you decided to shred your copies of the memo, the archived copies still have to be turned over in response to the subpoena. And note that GMail's not special in this regard. If you recieve your e-mail through your ISP and use their POP3/IMAP server to get it, it's probably backed up the same way and subject to the same risk of being subpoena'd
First rule: if you want control over your data and when it's destroyed, you must never allow it onto systems which you don't control.
It's taken them this long to notice this one? The cricket book discusses it, fer cryin' out loud, and had a good recommended solution: refuse recursive queries by default, then enable them only on those nameservers that'll be used by your client machines and only if the query comes from your local network. I thought everybody setting up a nameserver knew this one, BIND even comes with options specifically to make it easy to do.
Unlikely. The developers have tested all the combinations of configuration settings, which includes things like compiler options. Your distribution may not have tested it, but someone else has. It's like buying an HP computer: just because the store didn't test it for compatibility of parts doesn't mean it's untested, HP tested it for compatibility before they shipped it to the store.
Variations in libraries causing a problem is possible, but developers already think about that kind of thing. The standard autoconfiguration scripts check versions and will refuse to use libraries old enough to be known to be problems, and when making new versions of libraries developers tend to either insure backwards compatibility or arrange it so old code just won't compile with the newer library version.
It's theoretically possible that the specific version of the compiler you're using has some strange bug in it that causes it to generate bad code, but that's rare and such problems tend to get fixed very quickly for obvious reasons. You could also be using some strange compiler that nobody's heard of and the vendor doesn't support anymore, but in that case you probably can't get pre-built packages anyway and you're used to the situation.
In short, your statement's akin to claiming I don't know the sky's blue because I'm not looking at it right this exact second: theoretically the sky could have turned green since I last looked at it, but it's unlikely it could and not cause such a commotion that I'd know something was up (likewise, if it wasn't blue because it was storming or night I'd probably know too).
I don't think so. The authorization is implicit in my job.
To you, it is. Maybe to your manager, it is. However, neither of you are in IT's chain of command. The relevant question is has IT's bosses said that you have the authority to authorize IT to do what you want. If they haven't, then IT's hands are tied. And this isn't a matter of IT structuring things that way, it's a matter of the people IT answers to structuring things that way. If the CEO or CIO says "This is the way it shall be.", then that's the way it shall be whether IT wants it that way or not.
You seem to think that IT can ignore their own superiors any time they want. Sorry, it doesn't work that way. And IT doesn't get any more say in how the CEO organizes things than you do.
If you're compiling your own for performance reasons, don't bother. There's a few packages that can benefit from being compiled for specific hardware, the kernel and libc, for example, and things like math libraries that can use special instructions when compiled for specific hardware. For the most part, though, your apps aren't going to be limited by the instruction set, they'll be limited by things like graphics and disk I/O performance and available memory. IMHO if you're trying to squeeze the last 1% of performance out of a system, you probably should look at whether you need to just throw more hardware horsepower at the problem.
The big benefits and drawbacks of custom-compiled vs. pre-built packages is in the dependencies. Pre-built packages don't require you to install development packages, compilers and all the cruft that goes along with a development environment. You can slap those packages in and go with a lot less installation, and you can be sure they're built with a reasonable selection of features. On the other hand, those pre-built packages come with dependencies. When you install a pre-built package you pretty much have to have the versions of all the libraries and other software it depends on. By contrast, when you compile your own packages they'll build against the versions of other software you already have, using all the same compiler options, and they'll probably auto-configure to not use stuff you don't have installed. This leads to a more consistent set of binaries, fewer situations where you need multiple versions of other packages installed (eg. having to install Python 2.3 for package X alongside Python 2.4 for package Y) and overall a cleaner package tree.
Where the cut-off is depends on your situation. If there's only a few instances of dependency cruft, it may not be an issue. If you have a lot of dueling dependencies, it may be worth while to recompile to reduce the number of different versions of stuff you need. If you've got externally-dictated constraints (eg. only one version of OpenSSL is approved for your systems and everything must use it or not use SSL at all) then you may have no choice but to compile your own if you can't find a pre-built package that was built against the appropriate versions of libraries.
Windows is usually locked down after demonstrated problems (eg. a virus outbreak caused by people installing random software). My usual response to people wanting an unlocked desktop is "OK, but IT's responsibility for that machine's limited to giving you the installation media and license keys for corporate-provided software, and disabling the switch port it's connected to if it causes problems. Control and responsibility go hand-in-hand, either you have both or IT has both.".
Unfortunately, that usually doesn't work. IT is required by their bosses to keep the corporate systems running smoothly so everyone can do their work. If IT lets you run your own machine and it interferes with the network, IT has to explain to the CEO why they let the production networks get hosed during peak hours costing the business $X thousands (or millions) of dollars. Before they can let you have control, you have to convince upper management to hold you and not IT responsible for any problems your machine might cause. I'd place your chances of managing that somewhere in the neighborhood of "nitrocellulose dog chasing asbestos cat through the hottest furnace in Gehenna".
Myself, I personally agree with you, and I'd love to be able to run my own Linux box. It'd speed up my job a lot. But I'm not the one on the hot-seat if anything goes wrong, and until I am I'll just have to live with IT's restrictions (or convince my managers to get the appropriate authorities to approve the neccesary exceptions or changes, which is often a non-starter due to things completely unrelated to my doing my job).
My number one IT complaint is that they make rules like:
no unsupported access/mysql/etc databases, but provide no alternative.
Odds on IT doesn't have a choice about this rule. They've likely been told by upper management that no software will be installed unless it's been approved by the Powers That Be and vendor support has been arranged. Which means that if IT goes ahead and allows your unsupported database software to be installed like you want, they get fired.
So what can I do? My leaders are too clueless to understand the problem. They just demand that I collect all the data so they can later magically get 'six-sigma' data from me.
And this is IT's problem how? Your managers are just that: your managers. It's your job to get them to approve and authorize the things you need to do what they ask of you. Of course, that also means your managers will have to authorize paying for those things. That Oracle mobile application you could use may require a license the company hasn't paid for, and IT can't just start using unlicensed software whenever it wants. Ditto for Oracle Open Interfaces. So do your job first: convince your managers to approve and authorize what you want. Then take those authorizations over to IT and I suspect you'll find them much more willing to do what you ask.
And if your managers won't approve what you want and you're getting frustrated, think about how frustrated IT must be with every department in the company asking them for things and upper management refusing to let IT do it.
If the departments can charge back IT for "lost time", then let IT charge those departments for work. A request requires 1 hour of work from IT? 1 hour of the tech's time gets charged to that department's personnel budget. A department wants more storage? They get to pay for the drives out of their budget, likewise installation time and any other costs like maintenance contracts and additional UPS capacity.
Oddly, I've found that when this idea gets seriously considered, the departments raising the fuss tend to get really quiet of a sudden.:)
Additionally, they haven't been able to sell needed changes to senior management. Unacceptable server down time, maxed network storage, and no backups systems have hit the bottom line, and those on top are starting to notice.
The first and second sentences here are mutually exclusive. If IT's asking for needed changes and upper management won't buy into them, then either upper management isn't noticing problems or upper management is the problem. You'd be well advised to investigate the actual constraints and mandates IT is operating under first, else you may find that some Bastard over in IT has his ducks in a row and his written evidence in order and instead of those "incompetents" over in IT you'll be revolting against the CFO or some other extremely senior executive. Senior executives, BTW, don't like revolts by the rank-and-file, nor do they like being backed into a corner where they've no choice but to publicly admit one of their decisions was wrong. Ponder the concept of Career-Limiting Move carefully. On the other hand, if you get down off your high horse and work with IT to provide the business-side leverage needed to pry some of what IT's been asking for for years out of senior management, you may find IT solidly on your side in the future.
Not quite. Any company can get and use GPL'd software with no restrictions whatsoever. They can modify it however they want, keep it as secret as they want, put as many NDAs on their modifications as they want. It's when they redistribute it that the strings come into play. The same with other things. If Google were redistributing those books, they'd be in the wrong. But they've been very careful to not redistribute the contents of the books beyond the small snippets allowed by fair use. And they got the contents from libraries, whose sole purpose for existing is to archive and lend out copies of books. So no, I don't take a different attitude, I simply make the same distinction between use and redistribution as copyright law does. Certain copyright owners would rather not make that distinction, as it'd benefit them to control and charge for each use rather than each copy made, but that's their problem.
I think his mistake was in arguing about his authority to delete files at all. His argument should have been that all the files alleged to have been deleted were personal files, personal use of the laptop was authorized by his employment agreement (quote the relevant paragraph from the agreement, and the company has no right to demand that those files be turned over in the first place. You can't be charged under the law he was if the only things you "damaged" belonged to you.
Of course, this only works if he was scrupulous to avoid mixing his personal stuff with company data and can clearly show that all the files the company can prove were theirs are untouched. If he did delete files from an area normally or provably containing company data, he's pretty much SOL.
Slashdot Mirror
Yes, but why should I come on as your employee for that? Why shouldn't I come on as a short-term contractor instead? If you aren't going to offer me a stable, decent base salary, I'm better off avoiding W2 status (as a contractor I can write off a lot of stuff on my taxes that I can't as a W2 employee).
I sent the following:
Date: Fri, 7 Apr 2006 10:09:27 -0700 (PDT)
From: Todd Knarr <xxxx@xxxxxx.xxx>
To: sale@dlink.com, customerservice@dlink.com
Subject: DLink router use of Danish NTP server
This is in reference to the open letter to DLink from Danish sysadmin Poul-Henning Kamp (http://people.freebsd.org/~phk/dlink/). Abuse of an NTP server in express violation of the service agreement in the Stratum-1 server list is, in my opinion, inexcusable. Willful refusal to correct the abuse when requested is, if anything worse. Hard-coding the server name into the firmware, so that changes are difficult or infeasible, as opposed to DLink maintaining their own DNS records so that changes are simple, is also inexcusable in any technically-competent organization.
I have been comtemplating purchase of a DLink DI-784 router/AP, a DWL-7100AP access point and a DWL-AG660 CardBus adapter. If DLink doesn't correct their error as Mr. Kamp asks, I will be taking my purchases to NetGear instead. They, at least, have demonstrated a willingness to fix their mistakes when asked. I will also be recommending to my friends that they avoid DLink products in the future.
One customer, voting with his dollars.
We'll see what kind of response I get.
Oh, I know all too well that e-mail spam costs money. But it doesn't cost money in a way that you can tie directly to the spam in the ledgers. Doing it requires a lot of complicated work with the techies, and the numbers are almost always ambiguous (at least to the business types). With FAXes the accountant can say "FAX consumable costs were up 180% last month. Legitimate FAX volume was down 4%, so the increase isn't from normal traffic. Junk FAXes were 52% of received FAXes, and they're the only thing that could've caused the increase in costs.". That's something even an MBA gets immediately.
The FCC's made a mis-step here. Junk e-mail is one thing, it costs time and hassle but not money. Junk faxes, though, cost money. The accountants will see the cost of consumables (paper, ink/toner) go up, and they'll be able to tie it directly to junk faxes. That's when the business groups start calling their Congressmen saying "Your FCC's decision is costing our members money. Do something, or come election time our contributions go to your opponent.". That's why the junk-fax provisions of the TCPA were put in in the first place.
Of course, there's also another catch. The FAX-sending entity probably has a FAX line too. If they're claiming an existing business relationship with you, they can't very well deny you having an existing business relationship with them, now can they? And these new rules allow you to send junk FAXes to entities you have an existing business relationship with, don't they?
Actually he has entered into an agreement. Unfortunately, in all but 2 states it's going to be the default contract of sale defined by the Uniform Commercial Code. That means by default you have the rights granted under copyright law as regards making additional copies, plus the rights granted by the UCC as regards use of the copy you bought. This is one thing the copyright cartels try fiercely to avoid dealing with, with their copyright-law and "you only have a license" rhetoric, because there's a lot of UCC case law out there and most of it contradicts the position they'd like to take.
Sorry, AT&T. I've used 1.5mbps, 3mbps and 5mbps services in the last year. There very definitely is a discernable difference in speed. You won't notice it until you actually use the higher speeds, but once you have you won't want to go back.
Now, AT&T may be rigth that their backbone is limiting their customers to slower speeds, but that's a problem with their backbone they need to fix. Other providers don't have that problem.
Because one of Microsoft's big arguments in favor of sticking with MS software is the cost and hassle of converting all those Microsoft-format documents into the other software's formats. If customers don't have to convert documents, there's not much argument in favor of MS when license renewal comes up and Finance says "Why should we spend $BIGNUM on MSOffice licenses when we can spend $BIGNUM/10 on OpenOffice instead and be able to do everything we need?".
This is the real reason Microsoft is worried about open source but deathly afraid of open formats.
If you've looked at ODF and MSXML, you'll see why it matters which one's the standard. If you look at a comparison of ODF and MSXML, you'll see the differences. You'll notice that the XHTML and ODF examples read like document mark-up: you have the recognizable text of the document and things like paragraph and italic marks occur at the obvious places. This makes it really easy to manipulate ODF via XSLT to turn it into other formats. MSXML, by contrast, reads like an XML encoding of the internal object representation within Word. Instead of text being a paragraph you have a paragraph object with several attributes, one of which is the text of the paragraph. Notice how complex the nesting gets when you've got words in italics and boldface within a paragraph. This structure is going to be a real bitch to manipulate using XSLT. In fact Microsoft themselves admit this, saying that MSXML isn't meant to be manipulated by standard tools like XSLT but by programs using Microsoft's own APIs. What's the use of XML if you can't manipulate the document using standard tools for manipulating XML?
Only access to the web sites of about half the people I know. And access to half my hardware vendors (including such minor things as case-maker Lian-Li and thermal product vendor Zalman). And access to the support site for my motherboard (made by Soyo). And a huge number of anime-related sites.
Is the picture clearer now?
Shortly after the first FCC-approved implementation by an ISP, we'll see this:
Google announced today that, effective immediately, ISPs who wish enhanced bandwidth when using Google's services can obtain it by paying the nominal fee of 1 cent per kilobyte of transit. ISPs who do not wish to pay for enhanced access will retain normal best-effort access to bandwidth not used by those entitled to enhanced access. A Google spokesperson said "The ISPs are the ones who think this is a good idea, we're firmly committed to letting them live by their beliefs.". Elsewhere, ISPs expressed outrage at this idea, and fear that they'd be forced to either pay for enhanced access or lose users due to degraded access to the most popular search engine. Google responded by quoting back the ISPs' own statements about how providing enhanced access for a fee was not degrading service for those who didn't pay.
That depends. Number of looks isn't nearly as important as number of conversions. As an advertiser, given a choice between an advertisement that'll bring 1000 people into your store but only 1 will likely buy or an advertisement that'll only bring in 100 people but 10 of them are likely to buy, which one is a better advertisement?
A high number of views may mean that you're getting your ad out in front of a lot of the kind of people who click through an ad solely to cost the advertiser money or to give the site the ad's on money without any intention of actually buying what's advertised.
It's not a suprise that ugly-but-functional websites are successful in attracting and retaining visitors. Remember this: people visit a site looking for something. First and foremost, the site has to make it fairly easy (or at least straightforward) for visitors to get what they came there for. All else matters only after you've done that. If you meet that first requirement, visitors will forgive an awful lot of ugly. Functional and attractive is best, but if they have to choose visitors will choose functional over pretty when it comes time to actually use things. Note that your focus groups may not reflect this, because they're generally giving only the user's first impressions. First impressions fade fast when it comes time to actually do something.
It won't included in the current back. But it will still be in previous backups, and those aren't thrown out immediately because the admins may still need them. The current backup may be unreadable (hardware fault, tape's too old or damaged, any number of reasons), the data may have been corrupted so the last few backups are of bad data and you have to go back a week to a point before the corruption started to get a good restore, etc. etc..
To give you an idea of a normal backup rotation, the standard rotation I learned long ago was daily incremental backups (everything that's changed since the last backup) and weekly full backups (the whole system). Normal was 2 weeks worth of incrementals and 5 weeks worth of fulls (enough to span any full month). The first full backup of each month rotates out to the monthly backups, and monthlies got kept for a full year.
And what would you (and other users) say if a mistake by a Google sysadmin deleted every mailbox in the system and they said "Sorry, we don't have any backups, your mail is gone forever."?
This isn't a suprise. What Google's policy says is simple and obvious: "We make backups of our systems. That includes data files like your mailbox. We archive the backups on a rotating schedule that you don't know, so don't go assuming you know when any particular day's backup will be wiped. And we don't go back and alter those backups when you modify your data, so don't assume that deleting something today makes it disappear from all backups back to the beginning of time (or the inception of our service).". This subpoena is no different from a standard subpoena to a company asking for all documents including archived copies. If you wrote a memo, it got archived and then later you decided to shred your copies of the memo, the archived copies still have to be turned over in response to the subpoena. And note that GMail's not special in this regard. If you recieve your e-mail through your ISP and use their POP3/IMAP server to get it, it's probably backed up the same way and subject to the same risk of being subpoena'd
First rule: if you want control over your data and when it's destroyed, you must never allow it onto systems which you don't control.
It's taken them this long to notice this one? The cricket book discusses it, fer cryin' out loud, and had a good recommended solution: refuse recursive queries by default, then enable them only on those nameservers that'll be used by your client machines and only if the query comes from your local network. I thought everybody setting up a nameserver knew this one, BIND even comes with options specifically to make it easy to do.
Unlikely. The developers have tested all the combinations of configuration settings, which includes things like compiler options. Your distribution may not have tested it, but someone else has. It's like buying an HP computer: just because the store didn't test it for compatibility of parts doesn't mean it's untested, HP tested it for compatibility before they shipped it to the store.
Variations in libraries causing a problem is possible, but developers already think about that kind of thing. The standard autoconfiguration scripts check versions and will refuse to use libraries old enough to be known to be problems, and when making new versions of libraries developers tend to either insure backwards compatibility or arrange it so old code just won't compile with the newer library version.
It's theoretically possible that the specific version of the compiler you're using has some strange bug in it that causes it to generate bad code, but that's rare and such problems tend to get fixed very quickly for obvious reasons. You could also be using some strange compiler that nobody's heard of and the vendor doesn't support anymore, but in that case you probably can't get pre-built packages anyway and you're used to the situation.
In short, your statement's akin to claiming I don't know the sky's blue because I'm not looking at it right this exact second: theoretically the sky could have turned green since I last looked at it, but it's unlikely it could and not cause such a commotion that I'd know something was up (likewise, if it wasn't blue because it was storming or night I'd probably know too).
I don't think so. The authorization is implicit in my job.
To you, it is. Maybe to your manager, it is. However, neither of you are in IT's chain of command. The relevant question is has IT's bosses said that you have the authority to authorize IT to do what you want. If they haven't, then IT's hands are tied. And this isn't a matter of IT structuring things that way, it's a matter of the people IT answers to structuring things that way. If the CEO or CIO says "This is the way it shall be.", then that's the way it shall be whether IT wants it that way or not.
You seem to think that IT can ignore their own superiors any time they want. Sorry, it doesn't work that way. And IT doesn't get any more say in how the CEO organizes things than you do.
If you're compiling your own for performance reasons, don't bother. There's a few packages that can benefit from being compiled for specific hardware, the kernel and libc, for example, and things like math libraries that can use special instructions when compiled for specific hardware. For the most part, though, your apps aren't going to be limited by the instruction set, they'll be limited by things like graphics and disk I/O performance and available memory. IMHO if you're trying to squeeze the last 1% of performance out of a system, you probably should look at whether you need to just throw more hardware horsepower at the problem.
The big benefits and drawbacks of custom-compiled vs. pre-built packages is in the dependencies. Pre-built packages don't require you to install development packages, compilers and all the cruft that goes along with a development environment. You can slap those packages in and go with a lot less installation, and you can be sure they're built with a reasonable selection of features. On the other hand, those pre-built packages come with dependencies. When you install a pre-built package you pretty much have to have the versions of all the libraries and other software it depends on. By contrast, when you compile your own packages they'll build against the versions of other software you already have, using all the same compiler options, and they'll probably auto-configure to not use stuff you don't have installed. This leads to a more consistent set of binaries, fewer situations where you need multiple versions of other packages installed (eg. having to install Python 2.3 for package X alongside Python 2.4 for package Y) and overall a cleaner package tree.
Where the cut-off is depends on your situation. If there's only a few instances of dependency cruft, it may not be an issue. If you have a lot of dueling dependencies, it may be worth while to recompile to reduce the number of different versions of stuff you need. If you've got externally-dictated constraints (eg. only one version of OpenSSL is approved for your systems and everything must use it or not use SSL at all) then you may have no choice but to compile your own if you can't find a pre-built package that was built against the appropriate versions of libraries.
Windows is usually locked down after demonstrated problems (eg. a virus outbreak caused by people installing random software). My usual response to people wanting an unlocked desktop is "OK, but IT's responsibility for that machine's limited to giving you the installation media and license keys for corporate-provided software, and disabling the switch port it's connected to if it causes problems. Control and responsibility go hand-in-hand, either you have both or IT has both.".
Unfortunately, that usually doesn't work. IT is required by their bosses to keep the corporate systems running smoothly so everyone can do their work. If IT lets you run your own machine and it interferes with the network, IT has to explain to the CEO why they let the production networks get hosed during peak hours costing the business $X thousands (or millions) of dollars. Before they can let you have control, you have to convince upper management to hold you and not IT responsible for any problems your machine might cause. I'd place your chances of managing that somewhere in the neighborhood of "nitrocellulose dog chasing asbestos cat through the hottest furnace in Gehenna".
Myself, I personally agree with you, and I'd love to be able to run my own Linux box. It'd speed up my job a lot. But I'm not the one on the hot-seat if anything goes wrong, and until I am I'll just have to live with IT's restrictions (or convince my managers to get the appropriate authorities to approve the neccesary exceptions or changes, which is often a non-starter due to things completely unrelated to my doing my job).
My number one IT complaint is that they make rules like:
no unsupported access/mysql/etc databases, but provide no alternative.
Odds on IT doesn't have a choice about this rule. They've likely been told by upper management that no software will be installed unless it's been approved by the Powers That Be and vendor support has been arranged. Which means that if IT goes ahead and allows your unsupported database software to be installed like you want, they get fired.
So what can I do? My leaders are too clueless to understand the problem. They just demand that I collect all the data so they can later magically get 'six-sigma' data from me.
And this is IT's problem how? Your managers are just that: your managers. It's your job to get them to approve and authorize the things you need to do what they ask of you. Of course, that also means your managers will have to authorize paying for those things. That Oracle mobile application you could use may require a license the company hasn't paid for, and IT can't just start using unlicensed software whenever it wants. Ditto for Oracle Open Interfaces. So do your job first: convince your managers to approve and authorize what you want. Then take those authorizations over to IT and I suspect you'll find them much more willing to do what you ask.
And if your managers won't approve what you want and you're getting frustrated, think about how frustrated IT must be with every department in the company asking them for things and upper management refusing to let IT do it.
If the departments can charge back IT for "lost time", then let IT charge those departments for work. A request requires 1 hour of work from IT? 1 hour of the tech's time gets charged to that department's personnel budget. A department wants more storage? They get to pay for the drives out of their budget, likewise installation time and any other costs like maintenance contracts and additional UPS capacity.
Oddly, I've found that when this idea gets seriously considered, the departments raising the fuss tend to get really quiet of a sudden. :)
Additionally, they haven't been able to sell needed changes to senior management. Unacceptable server down time, maxed network storage, and no backups systems have hit the bottom line, and those on top are starting to notice.
The first and second sentences here are mutually exclusive. If IT's asking for needed changes and upper management won't buy into them, then either upper management isn't noticing problems or upper management is the problem. You'd be well advised to investigate the actual constraints and mandates IT is operating under first, else you may find that some Bastard over in IT has his ducks in a row and his written evidence in order and instead of those "incompetents" over in IT you'll be revolting against the CFO or some other extremely senior executive. Senior executives, BTW, don't like revolts by the rank-and-file, nor do they like being backed into a corner where they've no choice but to publicly admit one of their decisions was wrong. Ponder the concept of Career-Limiting Move carefully. On the other hand, if you get down off your high horse and work with IT to provide the business-side leverage needed to pry some of what IT's been asking for for years out of senior management, you may find IT solidly on your side in the future.
Not quite. Any company can get and use GPL'd software with no restrictions whatsoever. They can modify it however they want, keep it as secret as they want, put as many NDAs on their modifications as they want. It's when they redistribute it that the strings come into play. The same with other things. If Google were redistributing those books, they'd be in the wrong. But they've been very careful to not redistribute the contents of the books beyond the small snippets allowed by fair use. And they got the contents from libraries, whose sole purpose for existing is to archive and lend out copies of books. So no, I don't take a different attitude, I simply make the same distinction between use and redistribution as copyright law does. Certain copyright owners would rather not make that distinction, as it'd benefit them to control and charge for each use rather than each copy made, but that's their problem.
I think his mistake was in arguing about his authority to delete files at all. His argument should have been that all the files alleged to have been deleted were personal files, personal use of the laptop was authorized by his employment agreement (quote the relevant paragraph from the agreement, and the company has no right to demand that those files be turned over in the first place. You can't be charged under the law he was if the only things you "damaged" belonged to you.
Of course, this only works if he was scrupulous to avoid mixing his personal stuff with company data and can clearly show that all the files the company can prove were theirs are untouched. If he did delete files from an area normally or provably containing company data, he's pretty much SOL.