the time loop one is straight out of star trek and who knows where else...
I've not seen the SG1 take on it, but the earliest incarnation of this story that I know of is "A Little Something for Us Tempunauts" by Philip K. Dick.
The entire first season built the story. Without it, the following seasons have no foundation, they wouldn't work.
The first episode of B5 that I saw was TKO. That one *really* sucks, as far as I'm concerned. And you can miss it, and about half of the rest of series 1, without really losing out on too much of the story.
I think the series 1 episodes you need to see are:
Midnight on the Firing Line Born to the Purple Parliament of Dreams Mind War And the Sky Full of Stars Deathwalker (?) Survivors Signs and Portents Eyes (?) Legacies A Voice in the Wilderness (2 eps) Babylon Squared The Quality of Mercy Chrysalis
That's 15 out of 24, not much over half. And I'd consider a couple of those optional but highly recommended, too.
Compare that to the series 2 episodes you could skip: The Long Dark, A Spider in the Web, GROPOS (perhaps), Knives.
What have Confederates got to do with it? The sci-fi part is pretty much an excuse for having characters on the losing side of a civil war *without* them being Confederates. Look at the politics they talk about: it is _nothing like_ the same as the cause of the American Civil war.
Either I don't understand you, or you don't understand me, but I'm not sure which, so I'll try to be more explicit, and then if I'm misunderstanding you you can point out where.
My scenario is this:
I own domain example.com; my ISP is example.net. I clearly want to be able to send e-mails using "julesh@example.com" as my address, because that's what I want people to see as my address, record in their address books, etc. Because I might want to change ISP at short notice, I also want all replies to go there too. The example.com domain is hosted by a forwarding service that just forwards all e-mails it receives to my example.net address.
When I send an e-mail, I send it using example.net's servers. This is obviously the best way for the Internet in general because the traffic is local rather than crossing the 'net, and therefore costs less to transfer, so I would rather keep it this way. smtp.example.net then transfers it to the recipient domain's MX server, from where it is delivered into the recipient's mailbox. The recipient MX might lookup example.com's SPF record, see that it contains a reference to example.net's domains, looks up example.net's SPF record and sees smtp.example.net listed, so accepts it. I might have other users at different ISPs, so I have all of those ISPs listed in my SPF record. This traffic is probably not much, because most of those ISP records are probably cached by the recipient's mail server.
In your system, I send a message by uploading to a server on (e.g.) pmtp.example.net, which then sends a notification to the recipient that they have a message from julesh@example.com: but the recipient looks up the PMX record for example.com, snd one of two things happen:
1. The PMX record lists only a single server. Recipient connects to pmx.example.com, finds no messages waiting. Delivery failed.
2. The PMX record lists the servers of all 5 ISPs the different example.com users can send with, so the recipient now has to check with all of those, causing a lot of wasted connections.
That's if example.net will even allow me to send mail through their servers with an example.com address; they might get fed up of all these extra connection attempts and deny access to anyone not using an address in their own domain.
Re:Can someone explain to me?
on
Tetherless Wireless
·
· Score: 2, Informative
Any connection has a finite amount of bandwidth that must be shared between both directions of transfer.
Most home users (or rather, those that don't run servers or filesharing, which was once most of them, I don't know about now) would rather have faster download speeds and slower upload. It just works better for web browsing, e-mail reading, and most other things the average user wants to do with their Internet connection.
This explains most of the asymmetries involved. The only one *not* explained is the fact that 56K modems only have a 28.8K upstream (which is not widely reported, but true), whereas there's actually equal bandwidth in each direction on phone lines -- what you use in one direction doesn't affect the availability in the other, AFAIK. So I don't know why 56K modems do this... perhaps to keep the hardware cheaper?
If you spoof the address the recipients don't get their messages!
SPF is much more workable for verifying sender addresses, and that has serious flaws.
The problem is that most small businesses (i.e. a very large proportion of legitimate e-mail users) send e-mail with the 'from' address set to a domain that is not hosted by their ISP, but rather by a web hosting company. In SPF you can work around this by including your ISP's domains in your own domain's registration record. In your system, I can see no easy way of achieving it -- your e-mail address *must* map directly to a single IP address that you will be sending from. Doesn't work.
I think a protocol change is in order. Instead of sending the message via SMTP, only send a notification that an email is waiting to be picked up, and it's location. When your email is checked, it loads the email from the server. This has the following benefits:
1. The cost is on the sender. It would be prohibitively expensive to send millions of messages - you'd have to host all the clients!
Huh? You'd still only transfer the same amount of data, it would just become unpredictable how much you would transfer, when.
Plus, you'd be able to tell which of those e-mails were actually being retrieved, therefore you'd get a better form of address verification out of it than is currently available.
Also, it would create problems for people who check their e-mail infrequently (because the originating message might have disappeared).
Decrease bandwidth usage? No, SPF will *increase* it due to all the extra DNS checks people will make.
You don't think spammers are going to look at this and think "fuck that, I'll sell my stuff some other way" do you? They'll just try sending even more messages to counter its effects.
Aren't those the exact same settings that you can access in the javascript options section?
On my copy of firefox, only "hide the status bar" is in the prefs dialog, although I remember these options having been there in the past. Perhaps its mozilla suite only? Or maybe an extension enables them?
Well, yes, the interface is intended to allow you to change internal settings that are provided only for power users. Somewhat similar to the program, regedit.exe, that comes with Windows. Neither is a brilliant solution; in an ideal world there would be simple preference dialogs to change all of these options, but that would require to much work to maintain and would be difficult to do in a way that's not confusing for the average user.
Funny analogy, but a more accurate one would be to say that doors have a flaw because you can't see who's on the other side of them before you open them.
There's a simple solution, you can install a spy hole that lets you see through to the other side first.
If you don't like it, take it up with the standards committee and try to get the behavior redesigned for the next revision of the spec, but don't try to blame the browsers, that's just stupid
What spec? Which precise specification requires that web browsers, when requested to by web sites, should open a popup window in which their address bar is hidden? Not ECMAscript, that doesn't standardise the 'window' object.
Fair enough. I've never had occasion to search for a regexp in a web page (except while editing them!), but clearly other people have, so maybe it would be a useful feature.
Re:Is 100% compatibility really what we want?
on
KOffice 1.4 Released
·
· Score: 1
So every word processor on the planet has to not only implement, but also adaptively default to and provide an interface to select between, the line-breaking algorithms used by every other word processing product whose data can be imported?
No, that's a reducto ad absurdum argument. As long as there is a single de-facto standard for such things, everything should make as much effort as is reaosnable to support it. I'd put the threshold at about 80% market share -- once a word processor drops below that point, there's no longer a need to be layout-for-layout compatible with its document formats. Until that time, there is.
I'm certainly not convinced it's desirable just to avoid a very occasional layout change.
The changes aren't just occasional. I've been working with OO.o receiving documents from clients written with MS Word for about 2 years now, and about 10% of the documents render with noteable differences under OO.o. And these differences frequently matter to my clients.
The most jarring changes I've seen are with page-anchored floating objects, which frequently seem to appear in completely incorrect places. I'm not sure if this is an OO.o bug, or simply because it is impossible to import MS Word documents that well without duplicating MS Word's formatting options.
One thing I will note is that the simple typographical features you talk about (line breaking, kerning, etc.) should probably not have an MS-imitation mode applied. This is because there is substantial variation even between different versions of Word, and Word will even display documents differently if you change your printer driver sometimes. If Word can't standardise itself well enough for this to work, obviously nothing else can hope to duplicate it.
But other word processors should be able to emulate Word's box model, including precise details of how spacing affects paragraph positioning, etc.
Not sure about that. The big problems with P2P are:
1. Getting online on a decentralized network. The network only has to be decentralized because it's illegal, this network only trades in authorised files so it's perfectly legal, hence the problem doesn't exit.
2. Getting bad files. You can't, because they're only allowing upload/download of authorised files, so the content is under central control -- you *can't* find a bad file on this network. Unles you count those Brittney MP3s.
- None of the goodness.
Depends on what you think is good. This is probably better than most other legal download services, so if you're not willing to infringe copyright, then it's good.
- DRM on top.
According to the FAQ, they'll be providing MP3s (among other format). I don't know of any mechanism that can add DRM to an MP3 file.
This will be reaaaaal popular.
Yeah, I think it will. It's a lot better than iTMS, which is very popular.
Slashdot Mirror
the same plot as "The Core" only slightly more unrealistic
;)
Wow. That's... errr... interesting. How on _earth_ did they achieve it?
the time loop one is straight out of star trek and who knows where else...
I've not seen the SG1 take on it, but the earliest incarnation of this story that I know of is "A Little Something for Us Tempunauts" by Philip K. Dick.
The entire first season built the story. Without it, the following seasons have no foundation, they wouldn't work.
The first episode of B5 that I saw was TKO. That one *really* sucks, as far as I'm concerned. And you can miss it, and about half of the rest of series 1, without really losing out on too much of the story.
I think the series 1 episodes you need to see are:
Midnight on the Firing Line
Born to the Purple
Parliament of Dreams
Mind War
And the Sky Full of Stars
Deathwalker (?)
Survivors
Signs and Portents
Eyes (?)
Legacies
A Voice in the Wilderness (2 eps)
Babylon Squared
The Quality of Mercy
Chrysalis
That's 15 out of 24, not much over half. And I'd consider a couple of those optional but highly recommended, too.
Compare that to the series 2 episodes you could skip: The Long Dark, A Spider in the Web, GROPOS (perhaps), Knives.
I love Babylon 5 dearly, but there was a hell of a lot more than 1 episode that could have driven a new viewer away from the show.
Including about half of the first series, unfortunately.
What have Confederates got to do with it? The sci-fi part is pretty much an excuse for having characters on the losing side of a civil war *without* them being Confederates. Look at the politics they talk about: it is _nothing like_ the same as the cause of the American Civil war.
Along with the mod who sent the original post down as flamebait, I guess. What _was_ the joke?
OK. Then I guess you'd use something like SPF to verify that pmtp.example.net was an acceptable place to pick up e-mail for that address.
:)
I guess that works.
Either I don't understand you, or you don't understand me, but I'm not sure which, so I'll try to be more explicit, and then if I'm misunderstanding you you can point out where.
My scenario is this:
I own domain example.com; my ISP is example.net. I clearly want to be able to send e-mails using "julesh@example.com" as my address, because that's what I want people to see as my address, record in their address books, etc. Because I might want to change ISP at short notice, I also want all replies to go there too. The example.com domain is hosted by a forwarding service that just forwards all e-mails it receives to my example.net address.
When I send an e-mail, I send it using example.net's servers. This is obviously the best way for the Internet in general because the traffic is local rather than crossing the 'net, and therefore costs less to transfer, so I would rather keep it this way. smtp.example.net then transfers it to the recipient domain's MX server, from where it is delivered into the recipient's mailbox. The recipient MX might lookup example.com's SPF record, see that it contains a reference to example.net's domains, looks up example.net's SPF record and sees smtp.example.net listed, so accepts it. I might have other users at different ISPs, so I have all of those ISPs listed in my SPF record. This traffic is probably not much, because most of those ISP records are probably cached by the recipient's mail server.
In your system, I send a message by uploading to a server on (e.g.) pmtp.example.net, which then sends a notification to the recipient that they have a message from julesh@example.com: but the recipient looks up the PMX record for example.com, snd one of two things happen:
1. The PMX record lists only a single server. Recipient connects to pmx.example.com, finds no messages waiting. Delivery failed.
2. The PMX record lists the servers of all 5 ISPs the different example.com users can send with, so the recipient now has to check with all of those, causing a lot of wasted connections.
That's if example.net will even allow me to send mail through their servers with an example.com address; they might get fed up of all these extra connection attempts and deny access to anyone not using an address in their own domain.
So, how do you make this work?
Why would it?
Any connection has a finite amount of bandwidth that must be shared between both directions of transfer.
Most home users (or rather, those that don't run servers or filesharing, which was once most of them, I don't know about now) would rather have faster download speeds and slower upload. It just works better for web browsing, e-mail reading, and most other things the average user wants to do with their Internet connection.
This explains most of the asymmetries involved. The only one *not* explained is the fact that 56K modems only have a 28.8K upstream (which is not widely reported, but true), whereas there's actually equal bandwidth in each direction on phone lines -- what you use in one direction doesn't affect the availability in the other, AFAIK. So I don't know why 56K modems do this... perhaps to keep the hardware cheaper?
If you spoof the address the recipients don't get their messages!
SPF is much more workable for verifying sender addresses, and that has serious flaws.
The problem is that most small businesses (i.e. a very large proportion of legitimate e-mail users) send e-mail with the 'from' address set to a domain that is not hosted by their ISP, but rather by a web hosting company. In SPF you can work around this by including your ISP's domains in your own domain's registration record. In your system, I can see no easy way of achieving it -- your e-mail address *must* map directly to a single IP address that you will be sending from. Doesn't work.
I think a protocol change is in order. Instead of sending the message via SMTP, only send a notification that an email is waiting to be picked up, and it's location. When your email is checked, it loads the email from the server.
This has the following benefits:
1. The cost is on the sender. It would be prohibitively expensive to send millions of messages - you'd have to host all the clients!
Huh? You'd still only transfer the same amount of data, it would just become unpredictable how much you would transfer, when.
Plus, you'd be able to tell which of those e-mails were actually being retrieved, therefore you'd get a better form of address verification out of it than is currently available.
Also, it would create problems for people who check their e-mail infrequently (because the originating message might have disappeared).
No thanks, I think the current method is better.
Decrease bandwidth usage? No, SPF will *increase* it due to all the extra DNS checks people will make.
You don't think spammers are going to look at this and think "fuck that, I'll sell my stuff some other way" do you? They'll just try sending even more messages to counter its effects.
Yeah, but that way you gotta watch out for them Langford Visual Hacks.
Cool! Except... this flaw isn't about opening windows. It's about dialog boxes.
OK, so that's the solution to a different flaw. But an equally important one.
Aren't those the exact same settings that you can access in the javascript options section?
On my copy of firefox, only "hide the status bar" is in the prefs dialog, although I remember these options having been there in the past. Perhaps its mozilla suite only? Or maybe an extension enables them?
Well, yes, the interface is intended to allow you to change internal settings that are provided only for power users. Somewhat similar to the program, regedit.exe, that comes with Windows. Neither is a brilliant solution; in an ideal world there would be simple preference dialogs to change all of these options, but that would require to much work to maintain and would be difficult to do in a way that's not confusing for the average user.
Funny analogy, but a more accurate one would be to say that doors have a flaw because you can't see who's on the other side of them before you open them.
There's a simple solution, you can install a spy hole that lets you see through to the other side first.
If you don't like it, take it up with the standards committee and try to get the behavior redesigned for the next revision of the spec, but don't try to blame the browsers, that's just stupid
What spec? Which precise specification requires that web browsers, when requested to by web sites, should open a popup window in which their address bar is hidden? Not ECMAscript, that doesn't standardise the 'window' object.
400 AUD = 168 GBP (or $307 US) according to xe.com.
That's over 3 times as much as you get here in the UK. Why so much?
Fair enough. I've never had occasion to search for a regexp in a web page (except while editing them!), but clearly other people have, so maybe it would be a useful feature.
So every word processor on the planet has to not only implement, but also adaptively default to and provide an interface to select between, the line-breaking algorithms used by every other word processing product whose data can be imported?
No, that's a reducto ad absurdum argument. As long as there is a single de-facto standard for such things, everything should make as much effort as is reaosnable to support it. I'd put the threshold at about 80% market share -- once a word processor drops below that point, there's no longer a need to be layout-for-layout compatible with its document formats. Until that time, there is.
I'm certainly not convinced it's desirable just to avoid a very occasional layout change.
The changes aren't just occasional. I've been working with OO.o receiving documents from clients written with MS Word for about 2 years now, and about 10% of the documents render with noteable differences under OO.o. And these differences frequently matter to my clients.
The most jarring changes I've seen are with page-anchored floating objects, which frequently seem to appear in completely incorrect places. I'm not sure if this is an OO.o bug, or simply because it is impossible to import MS Word documents that well without duplicating MS Word's formatting options.
One thing I will note is that the simple typographical features you talk about (line breaking, kerning, etc.) should probably not have an MS-imitation mode applied. This is because there is substantial variation even between different versions of Word, and Word will even display documents differently if you change your printer driver sometimes. If Word can't standardise itself well enough for this to work, obviously nothing else can hope to duplicate it.
But other word processors should be able to emulate Word's box model, including precise details of how spacing affects paragraph positioning, etc.
- All of the crap from p2p.
Not sure about that. The big problems with P2P are:
1. Getting online on a decentralized network. The network only has to be decentralized because it's illegal, this network only trades in authorised files so it's perfectly legal, hence the problem doesn't exit.
2. Getting bad files. You can't, because they're only allowing upload/download of authorised files, so the content is under central control -- you *can't* find a bad file on this network. Unles you count those Brittney MP3s.
- None of the goodness.
Depends on what you think is good. This is probably better than most other legal download services, so if you're not willing to infringe copyright, then it's good.
- DRM on top.
According to the FAQ, they'll be providing MP3s (among other format). I don't know of any mechanism that can add DRM to an MP3 file.
This will be reaaaaal popular.
Yeah, I think it will. It's a lot better than iTMS, which is very popular.
Simple; they're giving away an infinite amount of it. Hey everyone! Infinite free music over at PeerImpact!
I don't get the no iPod support bit:
Peer Impact supports the following file formats:
Audio
* MP3 Audio File (.mp3)
I've played about with iPods before now, and all the ones I've seen play MP3s just fine.
Maybe the thing doesn't integrate with iTunes, but that's not the same as not working with iPods at all.