You need to use the current generation of DVCS in anger and then go back to the old stuff and see how much you miss it.
While the feature list isn't neccessarily that much longer than old stuff, a couple of things about DVCS don't just change the speed at which things happen, they change the way you work.
Firstly, the speed of common operations is so much greater that things that were previously something you considered carefully before doing just become routine. When "status" takes less than a second to return, you don't think twice about doing it. But it's not just status that gets so much faster.
Secondly, the most important feature of DVCS is branching. Branching is something that you avoid like the plague in older VCS ; you think carefully about it and avoid doing it. Why? Because the merging is so slow and painful. Merging in a DVCS isn't just fast, it's properly done too. Renamed a file in one branch and changed it in the other? DVCS copes. Merged a revision already? DVCS copes. It's not the 10 minute long process working out which revisions you need to merge, which ones have been merged already, and then unpicking the mess that happens after you do it. The merge takes a few seconds, and afterwards you have very high confidence that you only have to review files marked as conflicted. Couple this with a good set of unit tests and you can almost wave your merge blues goodbye.
Because merging works so much better, branches become a useful tool instead of the "reponse to dysfunctional personal relationships" that one guy above thinks they are.
Fix a bug? Make a branch for the fix. Test a feature idea? Make a branch. Have dysfunctional relationships?:-) Merge their changes in a staging branch, and test it before merging to trunk.
Once you overcome your (justified) branch phobia from using older VCS, they make
Aside from this ;
You intrinsically get a degree of distributed backup
Road warriors and commuters love being able to do everything they want to offline
Did I mention how much faster it is?
You mention VSS. I have so many painful memories of using VSS I would fight tooth and nail to avoid the current incarnation, even if it was rewritten from scratch by a crew of top-notch VCS geniuses. VSS 6 was so painful to have to use that it could put people off VCS for life ; in what other VCS can you commit files in the future by setting your clock forward, and have it believe you - the file won't show up for other users until that future date (unless they set their clock forward too).
You're neglecting the fact that SVN keeps a full, pristine copy of the base revision locally in each working copy. This can actually be larger than the repository storage on the server in some cases.
It depends on variables like your tree size and how many revisions are in it, and how many branches you keep going at once, but somewhere the lines cross over and DVCS systems start to win in terms of local storage as well, even though they replicate the entire project history onto your local filesystem.
It's particularly telling if you have multiple branches going at once. On SVN, the performance of "switch" leaves much to be desired and it has some problems with renames and such, so it's more practical to just check out another working copy for the branch ; and another pristine set of base revisions.
The metadata in SVN working copies is also spread across every folder in the tree and more than doubles the file count in that tree, which slows down recursive operations. It's also a PITA to avoid recursing through when you're scripting or searching.
I'd try Bazaar because it's easier to learn than git and can also be used as if it were SVN (central server, update/commit cycle) in tandem with a more advanced DVCS workflow.
This thread is prompting me to have another go with git on Windows ; I've been using Bazaar for my project (both the source for the project, and the version control for the actual content of the project).
The performance on Windows has come along in leaps and bounds since v 0.9 - it started off just "a lot faster than Subversion", it's got to the point where it's "wheee, that's fast".
Of course, you try out the same file tree with Bazaar on Linux and the speed makes you a little green with envy...
Bazaar isn't as "industrial" as git - in both ways. It may not perform quite as well, but it's also less rough around the edges and supports more workflows. And it has the tremendous advantage of being mostly written in Python, which means that hacking on it is a lot easier. If you have an itch, it's a lot easier to scratch it.
I would have hated to try to teach the default porcelain in git to my (non techie) users.
If you go to a movie theater and didn't like the movie, are you entitled to a refund?
It's unlikely you'd get one, but I'd interpret UK consumer law as allowing it ; goods have to be "fit for purpose". If the purpose is to be entertaining, and the goods suck at it, then yes, you should be entitled to a refund.
I feel the same ; I bought Mass Effect (only when it came down to half it's published price, but still, I didn't like myself).
It's a great game, but I felt ashamed to be supporting EA. I also felt uncomfortable with the activation scheme - it's definitely dampening my ardour for a new GPU, because that will require me to burn an activation.
Games are the only things keeping a "real" Windows install on my disk now. I need Windows for work, but I'm prepared to run it in a virtual machine if necessary. The software I write is cross-platform (and runs much faster on Linux, given the same hardware). I'm much less interested in gaming in general. The few games that actually get Linux ports would probably be enough to satisfy me (and be something to look forward to, instead of the almost limitless choice on Windows).
Maybe I should just grow up and kick the gaming habit ; the big publishers seem to want that, they are taking all the joy out of it.
That is correct. The Regulation of Investigatory Powers Act demands that you hand over your encryption key, and also stipulates that the maximum penalty for refusal is five years imprisonment.
So far, there have been 8 notices served under the act. 2 complied, 2 are being prosecuted, 2 may yet be prosecuted, and 2 of them are not being spoken about. So far none of the cases have come to court yet.
Tivoization is a technical restriction, and was specificically concocted to get around the "no restrictions" clauses in the original license ; by ensuring that your hardware will only run, for example, signed binaries.
Tivoization thus _does_ restrict what you can do with the code - you can't change it and run the changed build output on original hardware. It's therefore debateable just who started this particular fight.. It's inevitable the GPL 3 would include a reaction to this. Tivoizers gain all the benefits of the vast collective efforts put into GPLed code whilst dealing themselves out of the game that put that code in their hands in the first place - not cricket.
If people want to make Tivoized devices they should stick to (new variant) BSD licensed code, which is entirely happy to let them do whatever they like and not contribute. Of course, this limits their choices. Que sera.
This is one end of the spectrum of definition of "open source".
The more commonly understood definition is a license that meets the Open Source Definition, which MS-LPL obviously does not (contravenes point 10 at least).
It would not be outlandish to suggest that MS are trying to dilute the expectations of those hearing the words "open source" to include their more, well, useless definitions like MS-RSL which allows you to refer to the source for the purpose of increasing your understanding of the library (so you can make your stuff work around their "quirks", presumably), but doesn't actually grant you the right to build it, fix it, distribute it, or even use the binaries.
As long as the author of the driver from that site is the copyright holder of the original source, it doesn't matter if it was originally GPL ; the author may choose to relicense the code any way he sees fit.
I'm not saying that this is the case ; just presenting a potential means of this driver being both based on GPL code and legitimately distributed in a closed-source manner simultaneously.
No, "open source" means no license restrictions on what you do with the code. If the license forbids you to run the code on another OS, or another pieces of hardware, it is not open. Neither is it open if you prevent the use of the code for a particular purpose. If you want to use the code to tabulate a list of people who you intend to round up, incarcerate, and incinerate, people will deplore your morals, but the OSS movement in principle defend your right to use open source code to do it (but does not allow you to create a license that says that same group of people cannot use your software).
Hmm... this is probably why the Volt and the EV1 make such a focus out of drag ; the EV1 had a drag coefficient of 0.19
Here's hoping those shiny new battery/ultracapacitor technologies take the combustion engine out of the equation for smaller vehicles... ditch the engine and fuel tanks, pack batteries into the spare space.
I wonder if wheel hub motors would go some way to allowing a bit more room for the power train.
It is indeed a shame ; software is an enormously influential part of the lives of most civilised nations now. It's also astonishingly powerful ; you just have to look at the financial losses chalked up to a successful virus to realise that.
Putting so much power in the hands of purely selfish entities is not a good idea. Take the trojan that encrypts your files and demands that you pay money to regain access to them. This is illegal because the user concerned did not agree to it. But their actions allowed it to happen because they are not in that small group that understands what their computer is doing.
Microsoft would love you to participate in the legal version - Trusted Computing, where you not only agree to pay to have your files encrypted in the first place, you have to pay again to be able to access them each time you buy a new computer (because you need an MS operating system and software, as anything else will not have the requisite encryption keys).
A coal power station gets much better efficiency than a gasoline powered combustion engine, and an electric motor is >90% efficient.
You're not going to see many people fitting a carbon sequestration system to their car either. Or switching their gas engine to a windmill powered one, or a fusion reactor.
You don't need such large ones, of course, because the car is smaller in the first place.
Certain parts will be of a fixed size. Certain parts will scale with vehicle size. I'm actually finding it tough to think of parts that are necessarily fixed size, to be honest.
And don't forget ; you're not taking a standard combustion engine and bolting on a generator and electric drivetrain. You're ripping it out completely, and replacing it with a much smaller combustion engine.
If the car is twice as efficient, you only need half the engine output. And since the bulk of engine power in modern vehicles is usually only reserve for acceleration, you can get away with an even smaller combustion engine.
You want to anyway - most ICE have their top efficiency at 60% of their highest power output. When the engine is running, you want it running at that speed all the time, charging the battery at max efficiency... but you can run it at 100% if you really need to.
So you want an ICE that can only _just_ kick out enough power to maintain the maximum cruising speed of the car. You don't need spare capacity for acceleration - capacitors will take care of that.
Of course, both our opinions are worthless without the numbers.:-)
Well, this would be a nice assumption ; but in the ultra-paranoid world of intelligence you can't do that.
Keyboard? Might have a keystroke logger with tens of gigabytes of recall these days.
Anything on a USB connection could conceal a flash drive. Anything with BlueTooth could have data pushed to it via OBEX.
Monitor? Gigabytes of data pass down that cable every hour, you could easily secrete a storage device in it and hide data in the output stream.
You can't even assume that a cable doesn't have a concealed data storage device built into it's length.
If it was my responsibility to secure these resources, I would have them crushed, and then destroyed with thermite. All IT hardware would be bought in, and never leave (whole). Any hardware entering the building would have to stay in the building, so workers should leave personal hardware at the door.
Maybe the data isn't stored on the RFID in the passport. But there's the headshot ; such an obvious biometric that people forget about it.
What many people noticed on applying for a UK passport recently was the leaflet that came with the form telling you exactly how to pose for your photograph... you were only allowed certain margins, certain backgrounds, you had to face forward, you had to take off your glasses. It was pretty clear to those with a technical bent that the photograph was intended for consumption by a computer, so I'd suggest that anyone with a recent UK passport is already in a large database of facial geometry metrics somewhere in the Home Office (and maybe on your passport chip too). This would mean that you are ripe for rapid recognition from any sufficiently detailed CCTV footage ; and as we know, the UK has more CCTV cameras than anywhere else in the world. Nice.
Now, people don't habitually carry their passport in the UK, partly because it's a valuable document, partly because you don't need it for everyday usage, and partly because of the form factor - a little red book that doesn't conveniently fit into your pocket without the risk of being bent. A credit card sized ID on the other hand, is VERY easy to slip into your wallet and forget about.
If I were the UK government wanting to promote the routine carrying of an RFID enabled ID, I'd make the UK passport modular - a red book for the visa stamps, with a pocket in the back to carry the wallet-sized photo / RFID card when you're travelling. A lot of people would take to carrying their "passport card" routinely because suddenly, it's convenient.
Many is the time I've turned up at a place and found I needed a photo-ID or my passport and not had one, buying foreign currency, for example. It would probably work on me (after I put the tinfoil weave in my wallet, of course).
Heh, I mailed Weta to ask them what their position is... as a non NZ-national, I think lobbying Kiwi politicos would be a bit much.
I'd imagine that Weta would be grateful for the software foundation of their enormous 3200 processor render farm, but they also might have a few innovations they'd rather keep to themselves. I shall be interested to see what they say, if they respond. Maybe it'll kick up some dust.
"Don't dare expose it to the Internet"? Not true. Research, please.
It might be as secure as a nuns corset, but my IT dept still don't dare to expose it to the internet based on their past experience. Never mind actual objective proof. I've personally seem OWA 2003 pwned heavily, so I'd feel nervous too. It's how it IS configured that is annoying me, not how it SHOULD be configured.
"Outlook Anywhere" is NOT OWA.
Didn't say it was ; what I meant was that the RPC calls that the OWA client makes are probably the same calls that Anywhere makes ; it makes sense from a code-reuse point of view to be using the same APIs.
Again, this was about my personal frustration with how my IT services dept have configured it, not how it should be optimally configured. The documents they have sent around detailing the restrictions are very plain about spelling out how inconvenient it's going to be for me personally. It so happens that the way they are configuring things does suffer from the retriction that I detailed ; you have to use OWA lite, or pay MS something.
Disliking a particular implementation of an MS solution does not make one a Linux fanboi any more than it makes you an Apple fanboi or a OS/2 fanboi. For the record, I think the state of calendaring apps that are not Exchange is lamentable.
Weta workshop being the only major example of a software developer in New Zealand that I can think of (not being a Kiwi myself), I wonder what it is that they think on the matter?
Getting influential players to forswear the Dark Side is probably the best move you can make.
Our email is being moved over to Exchange.. after being moved off Exchange, to something else.
Previously, the admins dared not place Exchange on the internet, lest it be hacked. So the only way to get your mail was via VPN. Since they configure the concentrator to only allow Windows clients with the firewalling on, you can't access anything on your local network, and yea verily, this did sucketh.
Presently, there is a public IMAP server (running some variety of not-Exhange). And it's nice to be able to get your email without crippling your network connection, and from the IMAP client of your choice (ie, Thunderbird), installed on the device of your choice.
Soon, they intend to move us back onto Exchange. Because they still dare not place Exchange onto the internet, it will be secured behind something called Intelligent Application Gateway, which appears to be some kind of SSL proxy server.
So our options are....
Use an IAG client, an MS only payware product, to tunnel IMAP.
Use Outlook 2007 which conveniently has the "Outlook Anywhere" feature, which seems to combine an IAG client and use XMLRPC calls, and i probably the same client implementation as....
Outlook Web Access, which comes in "functional version for IE" and "crap version for dirty smelly hippyware browsers"
Given that the current solution works fine, I'm none too happy ; reading the announcement the first question that arose was "Are they idiots?", closely followed by "How fat was the wad of sweaty Billbucks they were given?"
Your options are ; give money to MS, or use a client that sucks (OWA lite). All the other clients suck LESS than OWA Lite, but to access any of them you must give some money to MS. Minimum spend being "a copy of a MS operating system", for IE, and maximum being Outlook. I'm not sure what the license cost of an IAG tunnel client is, but since you have to run it on Windows, it's a guaranteed winner for MS.
Postscript can contain function calls and as such, is often marked as a potential scripting threat. Google, for example, refuses to send raw eps files as attachments.
A similar principle to Windows MetaFile, which is essentially a list of calls to the Windows graphics library ; several Windows exploits owe their birth to WMF calling unchecked functions in the graphics library.
Note that just because a file format doesn't contain function calls or scripting does not make it secure. A poor implementation of any file reader can be vulnerable to a well crafted file. But active content makes things much easier, because it's much harder to check for security.
Slashdot Mirror
I disagree immensely.
You need to use the current generation of DVCS in anger and then go back to the old stuff and see how much you miss it.
While the feature list isn't neccessarily that much longer than old stuff, a couple of things about DVCS don't just change the speed at which things happen, they change the way you work.
Firstly, the speed of common operations is so much greater that things that were previously something you considered carefully before doing just become routine. When "status" takes less than a second to return, you don't think twice about doing it. But it's not just status that gets so much faster.
Secondly, the most important feature of DVCS is branching. Branching is something that you avoid like the plague in older VCS ; you think carefully about it and avoid doing it. Why? Because the merging is so slow and painful. Merging in a DVCS isn't just fast, it's properly done too. Renamed a file in one branch and changed it in the other? DVCS copes. Merged a revision already? DVCS copes. It's not the 10 minute long process working out which revisions you need to merge, which ones have been merged already, and then unpicking the mess that happens after you do it. The merge takes a few seconds, and afterwards you have very high confidence that you only have to review files marked as conflicted. Couple this with a good set of unit tests and you can almost wave your merge blues goodbye.
Because merging works so much better, branches become a useful tool instead of the "reponse to dysfunctional personal relationships" that one guy above thinks they are.
Fix a bug? Make a branch for the fix. :-) Merge their changes in a staging branch, and test it before merging to trunk.
Test a feature idea? Make a branch.
Have dysfunctional relationships?
Once you overcome your (justified) branch phobia from using older VCS, they make
Aside from this ;
You mention VSS. I have so many painful memories of using VSS I would fight tooth and nail to avoid the current incarnation, even if it was rewritten from scratch by a crew of top-notch VCS geniuses. VSS 6 was so painful to have to use that it could put people off VCS for life ; in what other VCS can you commit files in the future by setting your clock forward, and have it believe you - the file won't show up for other users until that future date (unless they set their clock forward too).
You're neglecting the fact that SVN keeps a full, pristine copy of the base revision locally in each working copy. This can actually be larger than the repository storage on the server in some cases.
It depends on variables like your tree size and how many revisions are in it, and how many branches you keep going at once, but somewhere the lines cross over and DVCS systems start to win in terms of local storage as well, even though they replicate the entire project history onto your local filesystem.
It's particularly telling if you have multiple branches going at once. On SVN, the performance of "switch" leaves much to be desired and it has some problems with renames and such, so it's more practical to just check out another working copy for the branch ; and another pristine set of base revisions.
The metadata in SVN working copies is also spread across every folder in the tree and more than doubles the file count in that tree, which slows down recursive operations. It's also a PITA to avoid recursing through when you're scripting or searching.
I'd try Bazaar because it's easier to learn than git and can also be used as if it were SVN (central server, update/commit cycle) in tandem with a more advanced DVCS workflow.
This thread is prompting me to have another go with git on Windows ; I've been using Bazaar for my project (both the source for the project, and the version control for the actual content of the project).
The performance on Windows has come along in leaps and bounds since v 0.9 - it started off just "a lot faster than Subversion", it's got to the point where it's "wheee, that's fast".
Of course, you try out the same file tree with Bazaar on Linux and the speed makes you a little green with envy...
Bazaar isn't as "industrial" as git - in both ways. It may not perform quite as well, but it's also less rough around the edges and supports more workflows. And it has the tremendous advantage of being mostly written in Python, which means that hacking on it is a lot easier. If you have an itch, it's a lot easier to scratch it.
I would have hated to try to teach the default porcelain in git to my (non techie) users.
If you go to a movie theater and didn't like the movie, are you entitled to a refund?
It's unlikely you'd get one, but I'd interpret UK consumer law as allowing it ; goods have to be "fit for purpose". If the purpose is to be entertaining, and the goods suck at it, then yes, you should be entitled to a refund.
I hope they make a paper bicycle like the one in Virtual Light.
Millenium Development Goals :
Yes, you're right, that is un-American.
click-click
click
<password><enter>
Damn, that was cryptic. Oh, wait.
TrueCrypt file volume. I now have a nice safe drive U:
Full disk encryption just prompts you for the password or smartcard+PIN at boot time.
I feel the same ; I bought Mass Effect (only when it came down to half it's published price, but still, I didn't like myself).
It's a great game, but I felt ashamed to be supporting EA. I also felt uncomfortable with the activation scheme - it's definitely dampening my ardour for a new GPU, because that will require me to burn an activation.
Games are the only things keeping a "real" Windows install on my disk now. I need Windows for work, but I'm prepared to run it in a virtual machine if necessary. The software I write is cross-platform (and runs much faster on Linux, given the same hardware). I'm much less interested in gaming in general. The few games that actually get Linux ports would probably be enough to satisfy me (and be something to look forward to, instead of the almost limitless choice on Windows).
Maybe I should just grow up and kick the gaming habit ; the big publishers seem to want that, they are taking all the joy out of it.
That is correct. The Regulation of Investigatory Powers Act demands that you hand over your encryption key, and also stipulates that the maximum penalty for refusal is five years imprisonment.
So far, there have been 8 notices served under the act. 2 complied, 2 are being prosecuted, 2 may yet be prosecuted, and 2 of them are not being spoken about. So far none of the cases have come to court yet.
Tivoization is a technical restriction, and was specificically concocted to get around the "no restrictions" clauses in the original license ; by ensuring that your hardware will only run, for example, signed binaries.
Tivoization thus _does_ restrict what you can do with the code - you can't change it and run the changed build output on original hardware. It's therefore debateable just who started this particular fight.. It's inevitable the GPL 3 would include a reaction to this. Tivoizers gain all the benefits of the vast collective efforts put into GPLed code whilst dealing themselves out of the game that put that code in their hands in the first place - not cricket.
If people want to make Tivoized devices they should stick to (new variant) BSD licensed code, which is entirely happy to let them do whatever they like and not contribute. Of course, this limits their choices. Que sera.
This is one end of the spectrum of definition of "open source".
The more commonly understood definition is a license that meets the Open Source Definition, which MS-LPL obviously does not (contravenes point 10 at least).
It would not be outlandish to suggest that MS are trying to dilute the expectations of those hearing the words "open source" to include their more, well, useless definitions like MS-RSL which allows you to refer to the source for the purpose of increasing your understanding of the library (so you can make your stuff work around their "quirks", presumably), but doesn't actually grant you the right to build it, fix it, distribute it, or even use the binaries.
As long as the author of the driver from that site is the copyright holder of the original source, it doesn't matter if it was originally GPL ; the author may choose to relicense the code any way he sees fit.
I'm not saying that this is the case ; just presenting a potential means of this driver being both based on GPL code and legitimately distributed in a closed-source manner simultaneously.
No, "open source" means no license restrictions on what you do with the code. If the license forbids you to run the code on another OS, or another pieces of hardware, it is not open. Neither is it open if you prevent the use of the code for a particular purpose. If you want to use the code to tabulate a list of people who you intend to round up, incarcerate, and incinerate, people will deplore your morals, but the OSS movement in principle defend your right to use open source code to do it (but does not allow you to create a license that says that same group of people cannot use your software).
Read point 10 of the Open Source Definition
Hmm... this is probably why the Volt and the EV1 make such a focus out of drag ; the EV1 had a drag coefficient of 0.19
Here's hoping those shiny new battery/ultracapacitor technologies take the combustion engine out of the equation for smaller vehicles... ditch the engine and fuel tanks, pack batteries into the spare space.
I wonder if wheel hub motors would go some way to allowing a bit more room for the power train.
It is indeed a shame ; software is an enormously influential part of the lives of most civilised nations now. It's also astonishingly powerful ; you just have to look at the financial losses chalked up to a successful virus to realise that.
Putting so much power in the hands of purely selfish entities is not a good idea. Take the trojan that encrypts your files and demands that you pay money to regain access to them. This is illegal because the user concerned did not agree to it. But their actions allowed it to happen because they are not in that small group that understands what their computer is doing.
Microsoft would love you to participate in the legal version - Trusted Computing, where you not only agree to pay to have your files encrypted in the first place, you have to pay again to be able to access them each time you buy a new computer (because you need an MS operating system and software, as anything else will not have the requisite encryption keys).
I am at a loss as to what to do about it though.
A coal power station gets much better efficiency than a gasoline powered combustion engine, and an electric motor is >90% efficient.
You're not going to see many people fitting a carbon sequestration system to their car either. Or switching their gas engine to a windmill powered one, or a fusion reactor.
You don't need such large ones, of course, because the car is smaller in the first place.
Certain parts will be of a fixed size. Certain parts will scale with vehicle size. I'm actually finding it tough to think of parts that are necessarily fixed size, to be honest.
And don't forget ; you're not taking a standard combustion engine and bolting on a generator and electric drivetrain. You're ripping it out completely, and replacing it with a much smaller combustion engine.
If the car is twice as efficient, you only need half the engine output. And since the bulk of engine power in modern vehicles is usually only reserve for acceleration, you can get away with an even smaller combustion engine.
You want to anyway - most ICE have their top efficiency at 60% of their highest power output. When the engine is running, you want it running at that speed all the time, charging the battery at max efficiency... but you can run it at 100% if you really need to.
So you want an ICE that can only _just_ kick out enough power to maintain the maximum cruising speed of the car. You don't need spare capacity for acceleration - capacitors will take care of that.
Of course, both our opinions are worthless without the numbers. :-)
Well, this would be a nice assumption ; but in the ultra-paranoid world of intelligence you can't do that.
Keyboard? Might have a keystroke logger with tens of gigabytes of recall these days.
Anything on a USB connection could conceal a flash drive. Anything with BlueTooth could have data pushed to it via OBEX.
Monitor? Gigabytes of data pass down that cable every hour, you could easily secrete a storage device in it and hide data in the output stream.
You can't even assume that a cable doesn't have a concealed data storage device built into it's length.
If it was my responsibility to secure these resources, I would have them crushed, and then destroyed with thermite. All IT hardware would be bought in, and never leave (whole). Any hardware entering the building would have to stay in the building, so workers should leave personal hardware at the door.
So what does this one make it so special?
You can etch the board yourself and make it at home from parts.
Maybe the data isn't stored on the RFID in the passport. But there's the headshot ; such an obvious biometric that people forget about it.
What many people noticed on applying for a UK passport recently was the leaflet that came with the form telling you exactly how to pose for your photograph... you were only allowed certain margins, certain backgrounds, you had to face forward, you had to take off your glasses. It was pretty clear to those with a technical bent that the photograph was intended for consumption by a computer, so I'd suggest that anyone with a recent UK passport is already in a large database of facial geometry metrics somewhere in the Home Office (and maybe on your passport chip too). This would mean that you are ripe for rapid recognition from any sufficiently detailed CCTV footage ; and as we know, the UK has more CCTV cameras than anywhere else in the world. Nice.
Now, people don't habitually carry their passport in the UK, partly because it's a valuable document, partly because you don't need it for everyday usage, and partly because of the form factor - a little red book that doesn't conveniently fit into your pocket without the risk of being bent. A credit card sized ID on the other hand, is VERY easy to slip into your wallet and forget about.
If I were the UK government wanting to promote the routine carrying of an RFID enabled ID, I'd make the UK passport modular - a red book for the visa stamps, with a pocket in the back to carry the wallet-sized photo / RFID card when you're travelling. A lot of people would take to carrying their "passport card" routinely because suddenly, it's convenient.
Many is the time I've turned up at a place and found I needed a photo-ID or my passport and not had one, buying foreign currency, for example. It would probably work on me (after I put the tinfoil weave in my wallet, of course).
Heh, I mailed Weta to ask them what their position is... as a non NZ-national, I think lobbying Kiwi politicos would be a bit much.
I'd imagine that Weta would be grateful for the software foundation of their enormous 3200 processor render farm, but they also might have a few innovations they'd rather keep to themselves. I shall be interested to see what they say, if they respond. Maybe it'll kick up some dust.
"Don't dare expose it to the Internet"? Not true. Research, please.
It might be as secure as a nuns corset, but my IT dept still don't dare to expose it to the internet based on their past experience. Never mind actual objective proof. I've personally seem OWA 2003 pwned heavily, so I'd feel nervous too. It's how it IS configured that is annoying me, not how it SHOULD be configured.
"Outlook Anywhere" is NOT OWA.
Didn't say it was ; what I meant was that the RPC calls that the OWA client makes are probably the same calls that Anywhere makes ; it makes sense from a code-reuse point of view to be using the same APIs.
Again, this was about my personal frustration with how my IT services dept have configured it, not how it should be optimally configured. The documents they have sent around detailing the restrictions are very plain about spelling out how inconvenient it's going to be for me personally. It so happens that the way they are configuring things does suffer from the retriction that I detailed ; you have to use OWA lite, or pay MS something.
Disliking a particular implementation of an MS solution does not make one a Linux fanboi any more than it makes you an Apple fanboi or a OS/2 fanboi. For the record, I think the state of calendaring apps that are not Exchange is lamentable.
Weta workshop being the only major example of a software developer in New Zealand that I can think of (not being a Kiwi myself), I wonder what it is that they think on the matter?
Getting influential players to forswear the Dark Side is probably the best move you can make.
Our email is being moved over to Exchange.. after being moved off Exchange, to something else.
Previously, the admins dared not place Exchange on the internet, lest it be hacked. So the only way to get your mail was via VPN. Since they configure the concentrator to only allow Windows clients with the firewalling on, you can't access anything on your local network, and yea verily, this did sucketh.
Presently, there is a public IMAP server (running some variety of not-Exhange). And it's nice to be able to get your email without crippling your network connection, and from the IMAP client of your choice (ie, Thunderbird), installed on the device of your choice.
Soon, they intend to move us back onto Exchange. Because they still dare not place Exchange onto the internet, it will be secured behind something called Intelligent Application Gateway, which appears to be some kind of SSL proxy server.
So our options are....
Given that the current solution works fine, I'm none too happy ; reading the announcement the first question that arose was "Are they idiots?", closely followed by "How fat was the wad of sweaty Billbucks they were given?"
Your options are ; give money to MS, or use a client that sucks (OWA lite). All the other clients suck LESS than OWA Lite, but to access any of them you must give some money to MS. Minimum spend being "a copy of a MS operating system", for IE, and maximum being Outlook. I'm not sure what the license cost of an IAG tunnel client is, but since you have to run it on Windows, it's a guaranteed winner for MS.
Postscript can contain function calls and as such, is often marked as a potential scripting threat. Google, for example, refuses to send raw eps files as attachments.
A similar principle to Windows MetaFile, which is essentially a list of calls to the Windows graphics library ; several Windows exploits owe their birth to WMF calling unchecked functions in the graphics library.
Note that just because a file format doesn't contain function calls or scripting does not make it secure. A poor implementation of any file reader can be vulnerable to a well crafted file. But active content makes things much easier, because it's much harder to check for security.