Some of the complainers are just whining, greedy fools, yes. But some of them did a significant fraction of that "1000s of man-hours of work", notably the Debian guys, who did (and do) much of the work needed to get XFree86 working on non-x86 architectures, work which benefits things like porting XFree86 to AMD64 (Debian developers discovered and fixed many 32 bit/64 bit issues on other 64 bit platforms). Those people have the right to complain, and if they chose to use and maintain a fork, more power to them.
One of my Linux boxes has no documentation. It lists some other third party software devs on bootup, and this is the only place where it ever lists third party software developers. Where else should I put the credit?
The version of dhcpcd I use displays "Copyright ISC" during bootup, thus crediting a third party; NET4 credits Swansea University, another third party. My Bluetooth drivers credit Qualcomm, yet another third party. Nothing in the licence stops me removing these messages, and if I were developing a "home user" distro, I'd probably hide them, but under the new XFree86 licence, since I already get other credits on bootup, I need to ensure that there's an X credit on bootup.
Let's say you're writing a system where date is significant, and is worked with. In your OO framework, a function that tries to construct a DateTime for 2002-02-29 will cause a run-time exception. In dependently typed systems, the compiler can detect that your function can attempt to create a date of 29th February 2002, and stop you; the only need for runtime error checking is thus when you take input from the outside world.
I noticed that she touched on strong typing as an aid to avoiding bugs; would a really strong type system help avoid bugs, or would it just introduce bugs into people's data types?
I ask because I'm currently looking into dependent type systems, which aren't currently practical. However, their claim to fame is that the type system is much more expressive; it is possible to define types like "date" or "mp3" in them, and ensure that wrong data cannot be supplied to functions. As I play though, I get the feeling that if the type system is too powerful, people will just create bugs in types, and we'll not improve by as much as we could do.
Further, whenever you have a game that keeps score and has a possibility of cheating, you get players who accuse others of cheating just for being different to them.
Case in point: Agents, a web based game that I play has someone posting regularly on the message boards accusing huge lists of people of using a script to play the game. I happen to be in one of those groups, and to know quite a few of the people he's accused; we're all playing it just for fun. Yet he's not happy being at the top levels of the game; he's got to accuse people who are just playing of cheating to make himself happy...
I notice that qoagle.com is unregistered; the handwriting recognition software on my Motorola A920 sometimes confuses 'q' and 'g', and 'a' and 'o', so handwriting http://www.google.com/ could become http://www.qoagle.com/ - would Verisign count as infringing trademarks then?
I intended http://www.google.com/ and Verisign gave me Sitefinder. This isn't much different from registering (say) ploybay.com for a porn site, playing on the ease of miswriting letters, and I believe that that would count as trademark infringement, even if the site was clearly not PlayBoy's. Would Google have a case against Verisign if I went looking for a search engine, did the handwriting equivalent of typos, and then used Sitefinder instead?
I guess what I'm saying is that there are a whole class of user errors that Verisign may not have forseen; since the only way to ensure that I do not get Sitefinder is to avoid making those errors or for the trademark holder to pay Verisign, I would assume that what they are doing is legally grey.
Anyway, I'm not a lawyer, but it seems fishy to me.
Slightly off topic, but still relevant; what immunizes Verisign to trademark lawsuits if they do this? I've just checked, and my employer has only registered one of three obvious variants on their trademark (which includes the word "and"). If Verisign's service redirects traffic from one obvious variant to a competitor's site, surely they are in breach of my employer's registered trademark?
Since Verisign receive a payment for each registered.com domain, they can't argue that my employer should register all variants on its trademark, without opening themselves up to a charge of extortion, and I would be surprised if the system automatically ensured that it didn't risk abusing someone else's trademark.
And a fourth vote! They are sufficiently competent at the customer service side of things that they'll even e-mail you to say that they've got your problem report and it has them stumped for now. No bullshit, no silly arguments, just a simple, "We don't get it either; we'll have time to do more research on Friday."
Plus they've got IPv6 support already, if you're looking to the future.
Price fixing in this context is anti-competitive regional discrimination. To take your Ferrari example, if Ferrari charged USD15,000 for a Modena, and EUR500,000 for the same car in Europe, and had an effective monopoly on sports cars, and was acting in an anti-competitive manner, then the EU might consider limiting the price of a Modena in the EU.
In this case, the EU has found that Microsoft has an effective monopoly on preinstalled desktop computer operating systems (nearly everyone uses MS), and that they've abused that effective monopoly to try and block competitors in the media player market. All that's left is to determine a suitable punishment; given that US customers will be upset if they are paying much more than their British counterparts, and that OEM Windows is a major moneyspinner for them, MS cannot risk the EU drawing the conclusion that MS could afford to run its OEM OS business on smaller margins, and that that would be suitable punishment.
The trouble with that answer is that the EU has already determined that Microsoft has acted anti-competitively; it now wants to know what penalties it should apply.
If MS chooses to sell software considerably cheaper in Thailand (taking into account that costs are lower out there), it massacres any argument they may have that suggests that Windows XP is fairly priced in the EU, thus opening up the concept of the EU fixing prices for MS in the realm of operating systems.
Oddly enough, in my little area of the world, there's three x86 boxes, all of which will boot off a USB key (and make it writable from DOS). Now, if I had older x86 hardware that wouldn't boot from a USB key, I'd need floppies; as it is, I can get by fine with a bootable USB key.
There's an EU anti-trust investigation ongoing into unfair practices by Microsoft. If MS can sell Windows and Office cheaply in Thailand, one of the EU's questions is likely to be "Why can't you do that here?"; this crippled version aims to do an end run around such ideas by giving an obvious answer.
Have you ever attempted to write assembly language for the ARM chips? Things like condition codes on every instruction make it much easier than on any other chip I've used. And machines are still being built (even if they are expensive) - see the Iyonix
Agencies are bad news anyway over here. They're not good at pre-sorting you, so that the potential employer doesn't get anything of value from them, but they do pull stunts like putting a "c/o Recruitment Agency" sticker over each copy of your home address; if your address is given in the wrong font or wrong size, their sticker looks odd when the fax is recieved, and they don't like it.
Re:IP6s problem is the numeric addresses r so comp
on
The State of IPv6
·
· Score: 1
Fair enough; if it's an experimental/at-risk service, then a new name makes sense. It's just that the nice IPv6 people went to all the effort of ensuring that it played nicely in DNS with IPv4 only clients, and I'm used to PHBs who believe that because it's new, it can't share a DNS name with an IPv4 node.
Re:IP6s problem is the numeric addresses r so comp
on
The State of IPv6
·
· Score: 1
This is a nitpick/rant, so feel free to ignore me.
What's the point in having two different DNS names for the same service? If your website is identical over IPv6 and IPv4 (and it should be), simply give www.alioth.net two different address records (AAAA and A), and allow the client to choose its preferred protocol automatically. Don't force me to enter a different URL.
Re:I despise IPv6 and I think it's horrid...
on
The State of IPv6
·
· Score: 1
However, you have more subnets than ever before (unless you're already working at a place with a/16 or larger). I've not yet worked anywhere where you'd have more than 65,000 logical subnets per site; autoconfiguration should deal with the issue of tracking IPs in use in a subnet, since you only manage the/64 network number by hand.
For a large corporate network, you just firewall off connections to each network, then require that services you wish to expose have an extra address, which is opened up, and manually assigned, one address per service.
It's standard HTML tags; find any decent HTML reference, and it should explain how to work the available tags.
To make a link, enter it as something like:
<a href="http://slashdot.org/">Slashdot</a>
which appears as:
Slashdot Note that in the href, the URL in quotes needs the http:// bit.
Re:Where's the source?
on
Xandros version 2
·
· Score: 2, Informative
You have to give the source, or a written offer, valid for at least three years, to provide the source at no more than the cost of media, shipping and handling, to anyone you give a binary to.
Therefore, if the retail version comes with source, they're clear. If it comes with an appropriate written offer (e.g. source available for $10 S&H from this address, $30 if shipping outside the USA), they're clear. If it comes without source, and without an offer, they're not clear.
Some mobiles do actually redirect all your calls to a customer service centre for registration; I recently purchased a prepay SIM for my phone, and as it wasn't registered in store, it diverted all my calls to the registrations team. Once I'd registered, it then acted as a normal phone.
What about the hardware to run it on? What about the OS? Is an eBook that only runs on 48K ZX Spectrums with Microdrives now good enough? Can we even read the media?
The advantage of an open specification for the format (unencrypted PDF would work, for example) is that provided I can access the data, and provided I have a copy of the specification, I can read the books. If I don't have the specification in an alternative format, I'm screwed. If the reader requires (say) a PC without PCI to work, and I don't have a spec, I'm screwed.
The second is more likely than the first, so I'd rather have a format with an open spec.
Slashdot Mirror
Nope. I have no pets at all, so I've not registered anything.
One more request; I always like biking around in GTA:VC. Police bikes (maybe just take the PCJ 600 and add a siren and flashing light) would be cool.
Some of the complainers are just whining, greedy fools, yes. But some of them did a significant fraction of that "1000s of man-hours of work", notably the Debian guys, who did (and do) much of the work needed to get XFree86 working on non-x86 architectures, work which benefits things like porting XFree86 to AMD64 (Debian developers discovered and fixed many 32 bit/64 bit issues on other 64 bit platforms). Those people have the right to complain, and if they chose to use and maintain a fork, more power to them.
One of my Linux boxes has no documentation. It lists some other third party software devs on bootup, and this is the only place where it ever lists third party software developers. Where else should I put the credit?
The version of dhcpcd I use displays "Copyright ISC" during bootup, thus crediting a third party; NET4 credits Swansea University, another third party. My Bluetooth drivers credit Qualcomm, yet another third party. Nothing in the licence stops me removing these messages, and if I were developing a "home user" distro, I'd probably hide them, but under the new XFree86 licence, since I already get other credits on bootup, I need to ensure that there's an X credit on bootup.
Let's say you're writing a system where date is significant, and is worked with. In your OO framework, a function that tries to construct a DateTime for 2002-02-29 will cause a run-time exception. In dependently typed systems, the compiler can detect that your function can attempt to create a date of 29th February 2002, and stop you; the only need for runtime error checking is thus when you take input from the outside world.
I ask because I'm currently looking into dependent type systems, which aren't currently practical. However, their claim to fame is that the type system is much more expressive; it is possible to define types like "date" or "mp3" in them, and ensure that wrong data cannot be supplied to functions. As I play though, I get the feeling that if the type system is too powerful, people will just create bugs in types, and we'll not improve by as much as we could do.
Case in point: Agents, a web based game that I play has someone posting regularly on the message boards accusing huge lists of people of using a script to play the game. I happen to be in one of those groups, and to know quite a few of the people he's accused; we're all playing it just for fun. Yet he's not happy being at the top levels of the game; he's got to accuse people who are just playing of cheating to make himself happy...
I intended http://www.google.com/ and Verisign gave me Sitefinder. This isn't much different from registering (say) ploybay.com for a porn site, playing on the ease of miswriting letters, and I believe that that would count as trademark infringement, even if the site was clearly not PlayBoy's. Would Google have a case against Verisign if I went looking for a search engine, did the handwriting equivalent of typos, and then used Sitefinder instead?
I guess what I'm saying is that there are a whole class of user errors that Verisign may not have forseen; since the only way to ensure that I do not get Sitefinder is to avoid making those errors or for the trademark holder to pay Verisign, I would assume that what they are doing is legally grey.
Anyway, I'm not a lawyer, but it seems fishy to me.
Since Verisign receive a payment for each registered .com domain, they can't argue that my employer should register all variants on its trademark, without opening themselves up to a charge of extortion, and I would be surprised if the system automatically ensured that it didn't risk abusing someone else's trademark.
Plus they've got IPv6 support already, if you're looking to the future.
In this case, the EU has found that Microsoft has an effective monopoly on preinstalled desktop computer operating systems (nearly everyone uses MS), and that they've abused that effective monopoly to try and block competitors in the media player market. All that's left is to determine a suitable punishment; given that US customers will be upset if they are paying much more than their British counterparts, and that OEM Windows is a major moneyspinner for them, MS cannot risk the EU drawing the conclusion that MS could afford to run its OEM OS business on smaller margins, and that that would be suitable punishment.
If MS chooses to sell software considerably cheaper in Thailand (taking into account that costs are lower out there), it massacres any argument they may have that suggests that Windows XP is fairly priced in the EU, thus opening up the concept of the EU fixing prices for MS in the realm of operating systems.
Oddly enough, in my little area of the world, there's three x86 boxes, all of which will boot off a USB key (and make it writable from DOS). Now, if I had older x86 hardware that wouldn't boot from a USB key, I'd need floppies; as it is, I can get by fine with a bootable USB key.
There's an EU anti-trust investigation ongoing into unfair practices by Microsoft. If MS can sell Windows and Office cheaply in Thailand, one of the EU's questions is likely to be "Why can't you do that here?"; this crippled version aims to do an end run around such ideas by giving an obvious answer.
Have you ever attempted to write assembly language for the ARM chips? Things like condition codes on every instruction make it much easier than on any other chip I've used. And machines are still being built (even if they are expensive) - see the Iyonix
Or indeed Kile for KDE. Much more up to date than KLyX, and quite effective.
Agencies are bad news anyway over here. They're not good at pre-sorting you, so that the potential employer doesn't get anything of value from them, but they do pull stunts like putting a "c/o Recruitment Agency" sticker over each copy of your home address; if your address is given in the wrong font or wrong size, their sticker looks odd when the fax is recieved, and they don't like it.
Fair enough; if it's an experimental/at-risk service, then a new name makes sense. It's just that the nice IPv6 people went to all the effort of ensuring that it played nicely in DNS with IPv4 only clients, and I'm used to PHBs who believe that because it's new, it can't share a DNS name with an IPv4 node.
What's the point in having two different DNS names for the same service? If your website is identical over IPv6 and IPv4 (and it should be), simply give www.alioth.net two different address records (AAAA and A), and allow the client to choose its preferred protocol automatically. Don't force me to enter a different URL.
For a large corporate network, you just firewall off connections to each network, then require that services you wish to expose have an extra address, which is opened up, and manually assigned, one address per service.
To make a link, enter it as something like:
<a href="http://slashdot.org/">Slashdot</a>
which appears as:
Slashdot
Note that in the href, the URL in quotes needs the http:// bit.
Therefore, if the retail version comes with source, they're clear. If it comes with an appropriate written offer (e.g. source available for $10 S&H from this address, $30 if shipping outside the USA), they're clear. If it comes without source, and without an offer, they're not clear.
Some mobiles do actually redirect all your calls to a customer service centre for registration; I recently purchased a prepay SIM for my phone, and as it wasn't registered in store, it diverted all my calls to the registrations team. Once I'd registered, it then acted as a normal phone.
The advantage of an open specification for the format (unencrypted PDF would work, for example) is that provided I can access the data, and provided I have a copy of the specification, I can read the books. If I don't have the specification in an alternative format, I'm screwed. If the reader requires (say) a PC without PCI to work, and I don't have a spec, I'm screwed.
The second is more likely than the first, so I'd rather have a format with an open spec.