I found it very hard to track what was going on, since despite it allegedly being about open development most of the work went on behind closed doors and then suddenly hit projects' CVS repositories a couple of weeks ago. In most cases I couldn't even find out what the projects were until they were checked in.
I'm sure some cool stuff has come out of this, but I wish it had been run better. The stated purpose of all this was to get the young 'uns into open source. For me, a major part of open source development is working with others, and yet I've seen little of that here. Instead, the students were essentially doing "work for hire" and releasing the results at the end. There's nothing there to keep the students interested now it's over; they didn't become part of the community and enjoy the social aspects of open source development, and so I would expect that many of them won't be seen again for some time.
I find that these days, having used the command prompt far too much, I'm always pressing TAB every few characters, even at completely inappropriate moments, and even when it doesn't work I still seem to bash that poor key several times before my concious mind realises that my shell can't auto-complete whatever I was in the middle of typing.
LCC is a C cross-compiler written by some people at some university somewhere. It's used to compile the game source to the engine's virtual machine. If its licence offends you there's always the option of linking your game code directly or compiling a native code shared library.
According to the Google Talk developer page, Google is only planning pre-arranged peering with a set of providers. Their goal, it appears, is to reduce spam and other abuses by ensuring that all clients are connecting through trusted services.
While I see their point, it does seem like a bit of a cop out. "Service choice" doesn't really mean much unless I can choose to use my own service and still inter-operate. A truly open system should allow anyone to play, not just the big boys.
Personally, I hate webmail and other such applications because they treat my expensive PC as a dumb terminal. My options are limited by what the service choses to provide. By running a real mail client, I get the raw data and can process it in any way I choose, including grepping the raw maildir files if I so choose. Sure, it doesn't come up that often in practice, but it's nice to have the ability when it's useful.
Of course, I won't deny that webmail is useful when you're away from your main computer. For that reason, I have a webmail interface to my IMAP server so that I can get the best of both worlds. Since the webmail is an interface to my own local mailstore, I can even just SSH into my server and grep the maildirs (or more likely just run pine) if that becomes useful.
Everyone sends HTML in my workplace too. It seems the only purpose of this is to attach a bloated corporate signature containing a company logo in JPEG format. The body of the message is generally devoid of any special markup.
I have my email client configured to ignore the HTML part of incoming mail, so I always see an amusingly butchered version of this signature. I don't actually include the signature in my outgoing mail, but no-one's really called me on it since I'm not in a customer-facing position. (They don't let the developers talk to the customers.:) )
I do keep meaning to make a text-only rendition of the signature, though. Possibly featuring the logo in ASCII Art, though probably just eschewing the logo completely.
But anybody can listen to anything they want, and can make antennas for receiving of any sort. In fact, as long as you're only receiving and not transmitting, I don't see how wardriving could ever be illegal.
Here in the UK at least there is a crime called something like "Interception of a transmission not intended for you", which was originally intended to handle the fact that it is illegal to listen to the police radio signals. It covers all sorts of things, though. Reportedly the associated law actually makes to technically illegal to answer the phone if the caller dialled a wrong number. I suspect that "wardriving" could actually be considered illegal under UK law.
My argument was simply that it'd be better to fix it at the level of having machine-readable directory indexes, not by adding yet another verb. Directory indexes are just another resource that you can GET; they don't need any special treatment.
Note that most "webmasters" go out of their way to disable directory listings, so this proposed new standard (whether it be a new verb or simply a new response format) would be mainly applicable to FTP-like sites where having the root of the site being a directory listing wouldn't be such a wacky thing. Alternatively, of course, sites could start providing machine-readable links in their HTML using the HTML link element, though of course that would get quite bloaty quite quickly as they'd have to essentially include the entire navigation set twice.
HTTP has no intrinsic concept of a "directory", it just happens that most webservers map the URL onto a filesystem by default. There's not really anything to stop you configuring your server to serve up directory listings in some XML format, and a more proper way to activate it would be to have the Accept: header contain the MIME type for your new XML format, or you could even just serve up the XML with an XSLT sheet referenced and not worry about having several versions.
This way a directory listing is just a resource like any other, which is true to the spirit of HTTP.
The episode where they put the Amiga against the Atari ST is brilliant. It's funny watching all of the predictions about what would happen in what became the war between Amiga and Atari for a while.
Most VCRs have in them their own tuner. This is why you are able to record one channel while watching another: the VCR's applying its own tuner to the incoming signal and then passing it through to another socket which is then connected to your TV which also has a tuner.
It sounds, though, like either your VCR was defective or you just didn't understand how to tune it! The first few VCRs I owned had mechanical tuning dials on the top under a cover. Later on as technology got more sophisticated they started having an integrated computer which controlled the tuner, meaning the tuning could be managed from an on-screen display. The last few had the ability to "auto-tune", though this meant that the channels ended up in broadcast order and not in the conventional numbering order.
I think the main thing I like about a VCR is its simplicity. There's very little that can go wrong with it. I once tried to do a fancy visual display at an event I was running using a modded XBox hooked up to a video projector with videos playing from a DVD and remote control over bluetooth from the other side of the room. Needless to say, on the night the DVD drive started to struggle with the DVD for whatever reason and the video began to stutter, until eventually it just froze up altogether. The next time I kept it simple and just took a VCR and a video cassette containing what I wanted to show. I lost the real-time control over the progression, but it actually worked flawlessly and I was able to stop and resume playback in response to unexpected events without going through a major ordeal.
I have similar issues with DVDs. If I'm watching a video and I'm interrupted, I can stop playback and remove the cassette, place it back in its box and forget about it for a while. When I come back later and put it back in it'll be exactly how I left it. With a DVD player, I have to sit through a few minutes of intros that my DVD player won't let me skip and then try to find the closest chapter point to where I was in the ever-useless menus. Once there, I inevitably end up re-watching several minutes of the programme because the chapter point didn't coincide with where I halted playback.
The sysadmin at my office is almost constantly making life difficult by reconfiguring permissions on directories I'm trying to serve web documents from, or just moving my home directory altogether, or removing my ability to restart my development webserver. It drives me crazy.
I'm glad he wastes an hour or two every day browsing random websites and chatting to his friends on MSN Messenger. It means I can actually get some work done! (Of course, when he's broken something for the nth time it does give me a handy downtime opportunity, which is useful if I need a break without it counting against me. I do actually prefer to get work done at work, though; maybe I'm just weird.)
That's a little off-topic, though. I think you were a little hard on your receptionist. You were paying her to be available to answer the phone, and that is what she was doing. What did you expect her to do when there were no phone calls to deal with?
There are a few things you can do in Java to make your app look more like a real OS X app, though- use a screen ( instead of window ) menu bar, have Apple UI-compliant menus, support the Application menu commands ( about, preferences, quit )... that takes a small amount of actual code, but there are plenty of apps out there that just don't take that extra step.
Having to take these extra steps defeats the object of a cross-platform UI library. The library should be handling this stuff. If it can't, then it shows that the design is flawed. Such things as screen vs. window menu bars should just happen, and there should be higher-level objects representing things like a collection of documents, which might be an MDI interface on one platform but a bunch of windows which live together in the z-order on another.
If Swing requires what you say it does, then you're back to writing special code for each platform, at which point you're not gaining much vs. just using the native Cocoa bindings for the MacOS version of your interface.
I don't see any reason why you can't use it as part of authentication if you want. Just have a list of OpenID identities which are allowed access and let OpenID handle the identity checking step. If the user successfully logs in with an approved identity, you then let them have all of the associated priviledges.
In fact, from what I understand LiveJournal is already doing something like this: users can add OpenID identities to a list which then allows those entities to post comments with a greater level of trust than the default, which is to treat them the same as anonymous commenters.
Two things in the transaction could be percieved as "permission". Firstly, the access point is (presumably) periodically advertising itself to the world, inviting any nearby computers to connect. Some computers will do this automatically without prompting, as mine did when I turned it on in my new office the other day and it discovered the access point in the office next door. Secondly, once the computer had associated with the access point it sent a DHCP request onto the network. Think of this as walking up to someone's open door and yelling "Can I come in?". The DHCP server then responded "Sure, you can come in and sit in this seat!" (you can use this IP address). This is also often done unattended by a computer once it has completed the previous step.
Not only, then, is the wireless network sending out periodical invitations to everyone, but when they respond it is helping them to get connected. This guy might be able to claim "hacking" if neither of these were true, but I think in this case it's clear to me that the owner of the wireless network has the liability for sharing his Internet connection in breach of his ISP contract.
As a side note, I was taught in school that in the UK you can legally access any system which doesn't make attempts to stop you. Of course, if you then go ahead and break it or cause disruption you can be charged with damage to property and other such crimes, but just "seeing what's out there" and making use of what you find is legal, assuming what I was taught in school was correct. If this wasn't true, it would be illegal to connect to amazon.com on port 80 without prior permission; the fact that it isn't restricted implies permission to use it. If it required a password and I brute-forced the password to gain access, I would be breaking the law.
Actually, it appears that you can. If I'm not mistaken, logging in there gets you a normal login session at which point you can do anything a normal user can do. The only thing that seems to be disabled is for OpenID users to keep their own journals.
Perhaps they'll lock it down more in the future, though. I've also heard that GreatestJournal gives full access to their photo hosting service to anyone who logs in with OpenID.
Actually, having downloaded Qt4 for Windows I've found that the GPL version doesn't "support" VC++. Only the makefiles for mingw are included. The configure "script" (which is actually an executable) just seems to display the GPL (which you can edit in an accompanying text file so you don't have to hit enter repeatedly), ask if you agree and then bail out. Perhaps that's because I don't have mingw installed; I'm not sure, because it didn't give any kind of error.
It's quite annoying that it doesn't come with the makefiles for VC++. I don't use the Visual Studio IDE, but I always use Microsoft's compiler (CL.EXE) and have a big collection of libraries already built for it which I don't fancy having to rebuild to switch compilers just to appease Qt. No doubt I could, if I had enough knowledge about Qt and the time to do it, create my own makefiles for CL based on the ones supplied for mingw, but that TrollTech didn't include them when they clearly have them for the commercial version is a little annoying. Why are they mandating which compiler open source developers must use? It's not like I'm suddenly going to plonk down a $4000 licence fee just because I want to use a different toolchain. What difference would it have made?
Perhaps "blind viewers" was an unfortunate turn of phrase, but many blind people do enjoy television and movies. Some DVDs even have a special audio track with descriptions accompanying parts of the movie that are very visual. Also, many "blind" people aren't completely without sight, so they may be able to discern scene changes and perhaps people moving around but not see with enough clarity to read and navigate the DVD menus.
My point, though, was more that the menus could be tailored to the needs and preferences of the viewer, regardless of what those needs and preferences are.
DVD menus should be completely declarative, and the exact layout should be decided and implemented by the player. This would allow for players which can attempt to read the menu captions out for blind viewers, among other benefits. It would also mean that they would by necessity be less flashy and annoying, and they'd work the same for every DVD. It wouldn't work the same on every DVD player, but then people might start shopping for DVD players based on who has the best UI, which would be fine by me.
Has there been posted a list of all of the accepted projects for all organisations yet? I was trying to find it on the mailing list but it doesn't seem that it has been announced yet. I'm curious to see what kinds of things people are going to be working on.
I think it's better to look at it from the point of view of the public's desired outcome. The public doesn't want an end to copyright, the public wants access to creative works at reasonable prices and the ability to share them with friends and family legally. The answer here, then, is to do away with the media cartels to force competition, and to make it illegal to infringe on the fair use provisions of copyright law. Doing away with copyright completely would certainly have negative effects which I'd bet most of the general public hasn't considered, or don't care about in the short term.
The public's view on law isn't really the best thing to go on, as the government is the one that'll get it in the neck when it bowed to public wishes and it didn't turn out how the public imagined. We elect supposedly-educated people into office because they are supposedly better at figuring out how to achieve the public's desired outcome, even if it isn't exactly in line with how the public would like it implemented at this point.
I've heard good things about Kivio. I've never actually used it, because I've never felt the need to make diagrams. The screenshots seem sensible, though! ;)
Slashdot Mirror
I found it very hard to track what was going on, since despite it allegedly being about open development most of the work went on behind closed doors and then suddenly hit projects' CVS repositories a couple of weeks ago. In most cases I couldn't even find out what the projects were until they were checked in.
I'm sure some cool stuff has come out of this, but I wish it had been run better. The stated purpose of all this was to get the young 'uns into open source. For me, a major part of open source development is working with others, and yet I've seen little of that here. Instead, the students were essentially doing "work for hire" and releasing the results at the end. There's nothing there to keep the students interested now it's over; they didn't become part of the community and enjoy the social aspects of open source development, and so I would expect that many of them won't be seen again for some time.
I find that these days, having used the command prompt far too much, I'm always pressing TAB every few characters, even at completely inappropriate moments, and even when it doesn't work I still seem to bash that poor key several times before my concious mind realises that my shell can't auto-complete whatever I was in the middle of typing.
LCC is a C cross-compiler written by some people at some university somewhere. It's used to compile the game source to the engine's virtual machine. If its licence offends you there's always the option of linking your game code directly or compiling a native code shared library.
According to the Google Talk developer page, Google is only planning pre-arranged peering with a set of providers. Their goal, it appears, is to reduce spam and other abuses by ensuring that all clients are connecting through trusted services.
While I see their point, it does seem like a bit of a cop out. "Service choice" doesn't really mean much unless I can choose to use my own service and still inter-operate. A truly open system should allow anyone to play, not just the big boys.
Personally, I hate webmail and other such applications because they treat my expensive PC as a dumb terminal. My options are limited by what the service choses to provide. By running a real mail client, I get the raw data and can process it in any way I choose, including grepping the raw maildir files if I so choose. Sure, it doesn't come up that often in practice, but it's nice to have the ability when it's useful.
Of course, I won't deny that webmail is useful when you're away from your main computer. For that reason, I have a webmail interface to my IMAP server so that I can get the best of both worlds. Since the webmail is an interface to my own local mailstore, I can even just SSH into my server and grep the maildirs (or more likely just run pine) if that becomes useful.
Everyone sends HTML in my workplace too. It seems the only purpose of this is to attach a bloated corporate signature containing a company logo in JPEG format. The body of the message is generally devoid of any special markup.
I have my email client configured to ignore the HTML part of incoming mail, so I always see an amusingly butchered version of this signature. I don't actually include the signature in my outgoing mail, but no-one's really called me on it since I'm not in a customer-facing position. (They don't let the developers talk to the customers. :) )
I do keep meaning to make a text-only rendition of the signature, though. Possibly featuring the logo in ASCII Art, though probably just eschewing the logo completely.
Here in the UK at least there is a crime called something like "Interception of a transmission not intended for you", which was originally intended to handle the fact that it is illegal to listen to the police radio signals. It covers all sorts of things, though. Reportedly the associated law actually makes to technically illegal to answer the phone if the caller dialled a wrong number. I suspect that "wardriving" could actually be considered illegal under UK law.
My argument was simply that it'd be better to fix it at the level of having machine-readable directory indexes, not by adding yet another verb. Directory indexes are just another resource that you can GET; they don't need any special treatment.
Note that most "webmasters" go out of their way to disable directory listings, so this proposed new standard (whether it be a new verb or simply a new response format) would be mainly applicable to FTP-like sites where having the root of the site being a directory listing wouldn't be such a wacky thing. Alternatively, of course, sites could start providing machine-readable links in their HTML using the HTML link element, though of course that would get quite bloaty quite quickly as they'd have to essentially include the entire navigation set twice.
HTTP has no intrinsic concept of a "directory", it just happens that most webservers map the URL onto a filesystem by default. There's not really anything to stop you configuring your server to serve up directory listings in some XML format, and a more proper way to activate it would be to have the Accept: header contain the MIME type for your new XML format, or you could even just serve up the XML with an XSLT sheet referenced and not worry about having several versions.
This way a directory listing is just a resource like any other, which is true to the spirit of HTTP.
The episode where they put the Amiga against the Atari ST is brilliant. It's funny watching all of the predictions about what would happen in what became the war between Amiga and Atari for a while.
I test my smoke detector daily by cooking dinner. As of yesterday, it's still working.
Most VCRs have in them their own tuner. This is why you are able to record one channel while watching another: the VCR's applying its own tuner to the incoming signal and then passing it through to another socket which is then connected to your TV which also has a tuner.
It sounds, though, like either your VCR was defective or you just didn't understand how to tune it! The first few VCRs I owned had mechanical tuning dials on the top under a cover. Later on as technology got more sophisticated they started having an integrated computer which controlled the tuner, meaning the tuning could be managed from an on-screen display. The last few had the ability to "auto-tune", though this meant that the channels ended up in broadcast order and not in the conventional numbering order.
I think the main thing I like about a VCR is its simplicity. There's very little that can go wrong with it. I once tried to do a fancy visual display at an event I was running using a modded XBox hooked up to a video projector with videos playing from a DVD and remote control over bluetooth from the other side of the room. Needless to say, on the night the DVD drive started to struggle with the DVD for whatever reason and the video began to stutter, until eventually it just froze up altogether. The next time I kept it simple and just took a VCR and a video cassette containing what I wanted to show. I lost the real-time control over the progression, but it actually worked flawlessly and I was able to stop and resume playback in response to unexpected events without going through a major ordeal.
I have similar issues with DVDs. If I'm watching a video and I'm interrupted, I can stop playback and remove the cassette, place it back in its box and forget about it for a while. When I come back later and put it back in it'll be exactly how I left it. With a DVD player, I have to sit through a few minutes of intros that my DVD player won't let me skip and then try to find the closest chapter point to where I was in the ever-useless menus. Once there, I inevitably end up re-watching several minutes of the programme because the chapter point didn't coincide with where I halted playback.
(...and yes, I'm British.)
No matter how many root servers you have, there always must be some authority to decide what goes in the root zone.
As a kid, I'd always say it as "I before E, except after C, unless the word is weird." I used to think I was so clever. I still do, actually. :)
The sysadmin at my office is almost constantly making life difficult by reconfiguring permissions on directories I'm trying to serve web documents from, or just moving my home directory altogether, or removing my ability to restart my development webserver. It drives me crazy.
I'm glad he wastes an hour or two every day browsing random websites and chatting to his friends on MSN Messenger. It means I can actually get some work done! (Of course, when he's broken something for the nth time it does give me a handy downtime opportunity, which is useful if I need a break without it counting against me. I do actually prefer to get work done at work, though; maybe I'm just weird.)
That's a little off-topic, though. I think you were a little hard on your receptionist. You were paying her to be available to answer the phone, and that is what she was doing. What did you expect her to do when there were no phone calls to deal with?
Having to take these extra steps defeats the object of a cross-platform UI library. The library should be handling this stuff. If it can't, then it shows that the design is flawed. Such things as screen vs. window menu bars should just happen, and there should be higher-level objects representing things like a collection of documents, which might be an MDI interface on one platform but a bunch of windows which live together in the z-order on another.
If Swing requires what you say it does, then you're back to writing special code for each platform, at which point you're not gaining much vs. just using the native Cocoa bindings for the MacOS version of your interface.
I don't see any reason why you can't use it as part of authentication if you want. Just have a list of OpenID identities which are allowed access and let OpenID handle the identity checking step. If the user successfully logs in with an approved identity, you then let them have all of the associated priviledges.
In fact, from what I understand LiveJournal is already doing something like this: users can add OpenID identities to a list which then allows those entities to post comments with a greater level of trust than the default, which is to treat them the same as anonymous commenters.
Two things in the transaction could be percieved as "permission". Firstly, the access point is (presumably) periodically advertising itself to the world, inviting any nearby computers to connect. Some computers will do this automatically without prompting, as mine did when I turned it on in my new office the other day and it discovered the access point in the office next door. Secondly, once the computer had associated with the access point it sent a DHCP request onto the network. Think of this as walking up to someone's open door and yelling "Can I come in?". The DHCP server then responded "Sure, you can come in and sit in this seat!" (you can use this IP address). This is also often done unattended by a computer once it has completed the previous step.
Not only, then, is the wireless network sending out periodical invitations to everyone, but when they respond it is helping them to get connected. This guy might be able to claim "hacking" if neither of these were true, but I think in this case it's clear to me that the owner of the wireless network has the liability for sharing his Internet connection in breach of his ISP contract.
As a side note, I was taught in school that in the UK you can legally access any system which doesn't make attempts to stop you. Of course, if you then go ahead and break it or cause disruption you can be charged with damage to property and other such crimes, but just "seeing what's out there" and making use of what you find is legal, assuming what I was taught in school was correct. If this wasn't true, it would be illegal to connect to amazon.com on port 80 without prior permission; the fact that it isn't restricted implies permission to use it. If it required a password and I brute-forced the password to gain access, I would be breaking the law.
Actually, it appears that you can. If I'm not mistaken, logging in there gets you a normal login session at which point you can do anything a normal user can do. The only thing that seems to be disabled is for OpenID users to keep their own journals.
Perhaps they'll lock it down more in the future, though. I've also heard that GreatestJournal gives full access to their photo hosting service to anyone who logs in with OpenID.
Actually, having downloaded Qt4 for Windows I've found that the GPL version doesn't "support" VC++. Only the makefiles for mingw are included. The configure "script" (which is actually an executable) just seems to display the GPL (which you can edit in an accompanying text file so you don't have to hit enter repeatedly), ask if you agree and then bail out. Perhaps that's because I don't have mingw installed; I'm not sure, because it didn't give any kind of error.
It's quite annoying that it doesn't come with the makefiles for VC++. I don't use the Visual Studio IDE, but I always use Microsoft's compiler (CL.EXE) and have a big collection of libraries already built for it which I don't fancy having to rebuild to switch compilers just to appease Qt. No doubt I could, if I had enough knowledge about Qt and the time to do it, create my own makefiles for CL based on the ones supplied for mingw, but that TrollTech didn't include them when they clearly have them for the commercial version is a little annoying. Why are they mandating which compiler open source developers must use? It's not like I'm suddenly going to plonk down a $4000 licence fee just because I want to use a different toolchain. What difference would it have made?
Perhaps "blind viewers" was an unfortunate turn of phrase, but many blind people do enjoy television and movies. Some DVDs even have a special audio track with descriptions accompanying parts of the movie that are very visual. Also, many "blind" people aren't completely without sight, so they may be able to discern scene changes and perhaps people moving around but not see with enough clarity to read and navigate the DVD menus.
My point, though, was more that the menus could be tailored to the needs and preferences of the viewer, regardless of what those needs and preferences are.
DVD menus should be completely declarative, and the exact layout should be decided and implemented by the player. This would allow for players which can attempt to read the menu captions out for blind viewers, among other benefits. It would also mean that they would by necessity be less flashy and annoying, and they'd work the same for every DVD. It wouldn't work the same on every DVD player, but then people might start shopping for DVD players based on who has the best UI, which would be fine by me.
Has there been posted a list of all of the accepted projects for all organisations yet? I was trying to find it on the mailing list but it doesn't seem that it has been announced yet. I'm curious to see what kinds of things people are going to be working on.
I think it's better to look at it from the point of view of the public's desired outcome. The public doesn't want an end to copyright, the public wants access to creative works at reasonable prices and the ability to share them with friends and family legally. The answer here, then, is to do away with the media cartels to force competition, and to make it illegal to infringe on the fair use provisions of copyright law. Doing away with copyright completely would certainly have negative effects which I'd bet most of the general public hasn't considered, or don't care about in the short term.
The public's view on law isn't really the best thing to go on, as the government is the one that'll get it in the neck when it bowed to public wishes and it didn't turn out how the public imagined. We elect supposedly-educated people into office because they are supposedly better at figuring out how to achieve the public's desired outcome, even if it isn't exactly in line with how the public would like it implemented at this point.
I've heard good things about Kivio. I've never actually used it, because I've never felt the need to make diagrams. The screenshots seem sensible, though! ;)