These are the dual format units, capable of writing +R/RW and -R/RW media. Strangely I've seen them available in CompUsa and OfficeDepot but almost all of the online retailers don't have them or show them backordered.
I got one from a reseller at work, and the + media is nearly worthless. +RW isn't readable on anything but the Sony drive. -R had the greatest readability (two DVD players and all models of DVD-R drives I could find).
The last time I looked at spamass-milter it wasn't quite ready for prime time. It hung on long messages and was generally a LOT less reliable than running vanilla spamassassin out of.procmailrc or something.
I've gotten spam at an email address at a private ISP that is used exclusively to send/recieve test emails to myself.
My understanding is that spammers will brute force some ISPs via either dictionaries of the left hand side of email addresses or through random generation (which given the user base of AOL or MSN or Hotmail might work).
8 guage shotguns haven't been usable for any hunting in the US since 1938 and probably haven't been produced in any quantity since the 19th century.
The largest you can buy today is the 10 guage, and you'd probably have to work pretty hard to find a load bigger than double-aught buck. I don't think anyone makes a slug for 10 guage since its primarily a waterfowl gun.
What would a mercury-filled steel slug do, anyway, besides wobble in flight and do long-term nerve damage (which would be mitigated by the short-term fatal hydrostatic shock)? A lead slug would be more dense. Depleted uranium would be denser yet, if you needed something exotic.
If you need to throw slugs like that you'd be better off with a.50 cal machine gun.
The scenerio you outline is kind of true, but there are still reasons for a GUI front-end. A modern GUI at 1024x768 can display a lot more information in a single screen than a 24x80 character display (I've never seen a mainframe-type app use 132 x yy displays, although I'm sure some have, but it'd be hard to see).
This alone is valuable; it requires fewer screen switches, fewer DB transactions, and the user can possible draw different conclusions based on more data.
It can lead to better/faster navigation; the user may be able to view ancillary data without losing the screen they're on, or they can go to an arbitrary application stage without navigating a bunch of unneeded screens.
It can even be possible to have a GUI client and the old screen-based method at the same time if the DB host can accessed client/server style, either directly or through a middle tier.
The most important thing, though, is to not just translate screen-for-screen a character-mode app to a GUI; it's important to re-think the application's flow and interface when converting it to a GUI mode, and to do so in a way that preserves whatever investments in hardware and technology you have.
If you have problems with the host hardware that prevent it from running the firmware, how do you diagnose those problems? Get on an airplane? Fly, drive to god knows where on short notice? Pay somebody else closer to do it? A free PSU replacement is suddenly a $10k expense because you don't have the ability to 'see' the machine anymore because your host CPU won't run.
We have these management functions in our servers and they have saved us thousands of dollars in travel by being able to diagnose what's broken remotely when the host CPU won't boot. It's typically been something trivial to replace that non-technical staff can replace if given the parts.
Console on serial (like with FreeBSD or Linux) in software doesn't cut it if the functionality you need involves the physical console. And firmware doesn't give you power on/off functionality either ("the building called and they will be shutting off power in 20 minutes for 4 hours. What do we do with the computers?").
I'm all for a better x86 BIOS, but having a seperate management computer adds a lot of value that even the best firmware can't give you because it relies on the very hardware you want to monitor.
Here, they are reinventing the wheel as well: HTTP/HTML is arguably the best GUI for these kinds of functions. But if something else is needed, VNC would seem like the right choice: a simple and open protocol.
You're missing the point. The idea is that the management card has the ability to redirect the keyboard and mouse of the console independant of the host environment *and* perform what amounts to real-time streaming of console display. You're not doing VNC/PC ANywhere/Terminal Services, you're able to get whatever's on the console's physical display, in the same way that a KVM-over-IP box works.
being IN the semiconductor test industry, it's really interesting how rarely does people really consider the necessity, and challenges, let alone costs, in testing.
Given the low quality and low reliability of so many devices, I didn't think they were testing anything!
I say this mostly tongue-in-cheek, but I have bought products that simple would not work. It wasn't a case of the devices being defective (ie, faulty single unit), but devices that just don't work *at all*.
I just returned a Hauppage WinTV PvR-350. For $200, it promised hardware MPEG2 encode and decode and the usual suite of video in/outs. The first system I tried it on it didn't work at all. Tech support sent me a suite of 'beta' drivers and some convoluted instructions wildly different than the documentation. The card started actually working (ie, video input was displayed and captured), but system performance (on a ~900Mhz PIII system) was so abysmal even when not writing streams to disk that the computer was unusable. I moved it to another system and it didn't work *at all* with any driver suite or graphics card I could find.
This isn't the first time I've run into products like this that aren't just somewhat disappointing but actually totally fail to function. If you run into this often enough, you start to ask yourself if these designs were ever tested at all.
Don't buy the Hauppage card. Ick -- even if it had worked, you couldn't capture from a third party application and the Hauppage application was pretty ugly (bad GUI, etc) and it was a pretty big hodgepodge of software from different vendors. Worst, I don't think the card does hardware MPEG2 decode to screen, I think its software decode, based on the low system performance.
In the Intel server world it seems that trend is going away from better BIOSes/OpenFirmware/Etc, and towards smarter and more functional management boards which are computers in their own right, not bound to the host CPU, RAM, and in some cases NIC. You can get full console redirection as well, including BIOS control and other hardware-console specific configuration, as well as diagnostics, environmental information and power off capability/
GUI OSes are kind of crippling for console redirection, though, and most vendors kludge together something using PC Anywhere or other dependent tools. I was told by a Compaq rep that eventually we might see a management controller with the ability to do what some of the KVM-Over-IP people are doing, except natively in the PC management board, providing a remote keyboard/mouse/display functionality. A side benefit of this is that you can ditch your KVMs, too, since you could just use a laptop with a NIC as the display and switch in software.
Personally I'm interested in this and it'd be interesting to see it take off and benefit from all the usual mass-production efficiencies, as well as seing it develop more OS-like features (such as perhaps running user-supplied modules).
Forget compression, what about transcoding of the files between various formats or bitrates? Forget about the aural impact of transcoding for a second, but the datastream impact. My rusty ol' ears won't hear anything different, but the data stream will have a completely different signature.
If its watermarking, would transcoding it destroy the watermark?
Unfortunately, it's being supplanted by infighting & organised crime[...]
From what I understand this has become a far bigger problem in both Republican and Loyalist areas for far longer than anyone wants to admit.
My understanding is that it got its start with IRA "containment" of heroin trafficking, which became a significant revenue stream. The same is true for Loyalist forces, and they both make good money on extortion and protection rackets.
My wife does business with some Northern Ireland companies and one of them had two pricipals killed in a bombing by Loyalists angry over hiring of Catholic labor.
In my scenerio, the item was published, but in a limited fashion due to cost and low demand, then let go unpublished for the expiry time. At that point its public domain, but after it goes public domain a new technology enables the trivial redistribution of it, cutting the artist out of profits.
Another version of this could involve a filmmaker who shoots a film on 70mm stock, makes a dozen complicated widescreen transfers to LD in the mid 80s. Won't allow VHS versions of any kind. LD version goes out of print due to low demand. Voila -- a 16:9 480p DVD version is now possible, and anyone can then legally (if I read your rules right) make and sell a DVD made from the LD transfer.
Films are even more complicated because what constitutes "in print"? Public performances? Say I have a graymarket print of the film (not an unauthorized copy, but one acquired from a movie house being torn down or something), does no public showings count? LD or VHS transfers to DVD? My own transfers from film to DVD?
The gaming industry may have good reason to complain; a given game cart may not be economically viable due to the obsolence of the platform, and its not technologically viable to emulate the platform *yet*, but they hang onto them based on the value of the character or that decent emulation may allow them to get a few sales for a new platform in the future.
I'm not saying I agree with this logic, but its kind of easy to see why people might want to hang onto copyrights for even 20 years after the original is no longer economically viable. I guess I'd make personal copying legal, but sales illegal just to protect the owner against economic exploitation.
The best spam control isn't new laws about who can send mail or newer and fancier statistical analysis filters, it's vigorous and severe punishment for the fraudulent schemes/products pushed through email.
Even if you passed a law that said "No spamming" the people running spam servers or doing bulk emailing would still be nearly impossible to catch since they can be hijacking open relays or using false-name disposable mail servers.
But ALL of the spam ultimately has a 'reachable' person there to collect the money, and that provides the perfect means to catch the PEOPLE invovled.
If the fraud merchants behind these schemes began to get indicted and convicted with a visible public frequency, it might start to have a real impact on spam. It wouldn't be a 100% perfect solution, but it would stave off needless government regulation of email which 'spam' laws would need to be effective and it might scare off a big chunk of the amateurs and part-timers trying to run their own solo operations.
I'd love to see that spam bitch from Florida that was on Slashdot a few months ago lose that house, the SUV and all the other goodies she has as she tries to ward off a federal wire fraud investigation. I know she was just acting as an advertising service, but isn't aiding and abetting illegal too?
I like this, although I wonder if there's maybe a valid argument against it. Say for example I, as an artist, produce a recording in glorious quadrophonic sound. The recording in fact doesn't sound good in stereophonic sound, and as a small artist I find the costs of recording my work in quadrophonic vinyl too expensive to sell in the market. So I let it go out of print.
By these terms, I lose my copyright after 10 years. But let's say that a new medium comes into existance that trivially can encode all four channels for no more cost than two channels. I've lost my copyright, and anyone can sell my work, it becomes semi-popular but I don't get any royalties.
How can you deal fairly with this (admittedly contrived) scenerio?
Ohters in this thread suggest dictionary-style spam campaigns aimed at high-profile domains, which is a good explanation.
However, I wonder if this isn't the beginning of an attempt to corner the market on (euphamism mode on) "permissioned, targeted email marketing" to hotmail users.
Step 1: chase away spammers who don't buy protection from MS Step 2: sell access to spammers (high quality lists, demographics, etc) Step 3: $$$$$
It's got to frost MS a little that there are all those "consumers" at hotmail and somebody else is sending them marketing info that MS doesn't get a big slice of.
I think that's reasonable, but I also think it's difficult to enshrine into law. The minute we say it's ok to make three copies and give them to friends, someone will make 10 copies and then make a copy of each copy three times and give them away, for a total of *30* and argue that they followed the law -- no more than three copies of an original may be given away..
Google's not a public corporation, but that doesn't meant they're the FSF either.
I can assure you that their private corporation status means that the financial stakeholders is a smaller, more coherent and more invovled group of people and that the pressure around money will be larger, and more focused than it would be if it was a public company, where those forces can be more easily diffused among many shareholders whose interests are often in conflict.
I'd be up for changing the copyright law on this one. I'd like to see copyright holders be required to prove that they have existing plans to re-release something currently out of print within a reasonable timeframe before they can invoke their copyrights.
I realize this might hinder the "business model" of some copyright holders that manufacture scarcity by letting popular items go out of print and stay out of print for a while so they get a bubble of sales when they re-release them.
I also realize this might hurt performers whose material has gone out of print because their sales fall below the million-a-day required to be considered businessworthy by large labels, but at least their material would be available.
Compared to many businesses, Google seems pretty good -- geek friendly, low ad content, good service, cool technology, et al.
But its also important to not lose sight that Google is a corporation, with investors, debtors and other people who are solely (or primarily) motivated by MAKING MONEY. They're not motivated by some pro-geek/anti-corporate ethos.
So as long as you keep in mind that they might turn around and do something that protects profit first and makes privacy or other goals take a back seat then you'll be OK.
Take a look at the amount of abandoned train tracks throughout america, it's extremely sad.
In high school a friend and I thought it would be fun to get a map of train tracks and then build a kind of hand-car that we could mount our bikes on and use pedal power for locomotion. We figured we could go farther and faster on the rails due to their relatively straight and flat paths and the decreased rolling resistance. It'd probably be safer than riding on rural highways.
I'm pretty sure it would have been a big problem if we would have been caught riding something like this without authoriztion on *active* train tracks, even though I don't think it'd be much of a safety risk, although crossing a long bridge might still be a little iffy.
It would have been a lot of fun, especially since trains often take paths different from roadways and you'd get to see scenery from a different persective. The light weight and bicycle nature of it would have meant we could have yanked it off the tracks and camped at night or ridden into towns for supplies, etc.
Slashdot Mirror
These are the dual format units, capable of writing +R/RW and -R/RW media. Strangely I've seen them available in CompUsa and OfficeDepot but almost all of the online retailers don't have them or show them backordered.
I got one from a reseller at work, and the + media is nearly worthless. +RW isn't readable on anything but the Sony drive. -R had the greatest readability (two DVD players and all models of DVD-R drives I could find).
You're full of it. Hotmail's outgoing mail is still handled by customized qmail code.
Isn't that kind of like saying that Tony Stewart won the Nascar championship using a customized Chevy?
I mean, if you customize, it just naturally follows that it can do anything.
The last time I looked at spamass-milter it wasn't quite ready for prime time. It hung on long messages and was generally a LOT less reliable than running vanilla spamassassin out of .procmailrc or something.
I've gotten spam at an email address at a private ISP that is used exclusively to send/recieve test emails to myself.
My understanding is that spammers will brute force some ISPs via either dictionaries of the left hand side of email addresses or through random generation (which given the user base of AOL or MSN or Hotmail might work).
8 guage shotguns haven't been usable for any hunting in the US since 1938 and probably haven't been produced in any quantity since the 19th century.
.50 cal machine gun.
The largest you can buy today is the 10 guage, and you'd probably have to work pretty hard to find a load bigger than double-aught buck. I don't think anyone makes a slug for 10 guage since its primarily a waterfowl gun.
What would a mercury-filled steel slug do, anyway, besides wobble in flight and do long-term nerve damage (which would be mitigated by the short-term fatal hydrostatic shock)? A lead slug would be more dense. Depleted uranium would be denser yet, if you needed something exotic.
If you need to throw slugs like that you'd be better off with a
Here's the definition of an EJB from the http://java.sun.com site.
If the abbreviation is EJB, doesn't that make it three words? Just because you jammed two together doesn't make them one word...
The scenerio you outline is kind of true, but there are still reasons for a GUI front-end. A modern GUI at 1024x768 can display a lot more information in a single screen than a 24x80 character display (I've never seen a mainframe-type app use 132 x yy displays, although I'm sure some have, but it'd be hard to see).
This alone is valuable; it requires fewer screen switches, fewer DB transactions, and the user can possible draw different conclusions based on more data.
It can lead to better/faster navigation; the user may be able to view ancillary data without losing the screen they're on, or they can go to an arbitrary application stage without navigating a bunch of unneeded screens.
It can even be possible to have a GUI client and the old screen-based method at the same time if the DB host can accessed client/server style, either directly or through a middle tier.
The most important thing, though, is to not just translate screen-for-screen a character-mode app to a GUI; it's important to re-think the application's flow and interface when converting it to a GUI mode, and to do so in a way that preserves whatever investments in hardware and technology you have.
If you have problems with the host hardware that prevent it from running the firmware, how do you diagnose those problems? Get on an airplane? Fly, drive to god knows where on short notice? Pay somebody else closer to do it? A free PSU replacement is suddenly a $10k expense because you don't have the ability to 'see' the machine anymore because your host CPU won't run.
We have these management functions in our servers and they have saved us thousands of dollars in travel by being able to diagnose what's broken remotely when the host CPU won't boot. It's typically been something trivial to replace that non-technical staff can replace if given the parts.
Console on serial (like with FreeBSD or Linux) in software doesn't cut it if the functionality you need involves the physical console. And firmware doesn't give you power on/off functionality either ("the building called and they will be shutting off power in 20 minutes for 4 hours. What do we do with the computers?").
I'm all for a better x86 BIOS, but having a seperate management computer adds a lot of value that even the best firmware can't give you because it relies on the very hardware you want to monitor.
Here, they are reinventing the wheel as well: HTTP/HTML is arguably the best GUI for these kinds of functions. But if something else is needed, VNC would seem like the right choice: a simple and open protocol.
You're missing the point. The idea is that the management card has the ability to redirect the keyboard and mouse of the console independant of the host environment *and* perform what amounts to real-time streaming of console display. You're not doing VNC/PC ANywhere/Terminal Services, you're able to get whatever's on the console's physical display, in the same way that a KVM-over-IP box works.
being IN the semiconductor test industry, it's really interesting how rarely does people really consider the necessity, and challenges, let alone costs, in testing.
Given the low quality and low reliability of so many devices, I didn't think they were testing anything!
I say this mostly tongue-in-cheek, but I have bought products that simple would not work. It wasn't a case of the devices being defective (ie, faulty single unit), but devices that just don't work *at all*.
I just returned a Hauppage WinTV PvR-350. For $200, it promised hardware MPEG2 encode and decode and the usual suite of video in/outs. The first system I tried it on it didn't work at all. Tech support sent me a suite of 'beta' drivers and some convoluted instructions wildly different than the documentation. The card started actually working (ie, video input was displayed and captured), but system performance (on a ~900Mhz PIII system) was so abysmal even when not writing streams to disk that the computer was unusable. I moved it to another system and it didn't work *at all* with any driver suite or graphics card I could find.
This isn't the first time I've run into products like this that aren't just somewhat disappointing but actually totally fail to function. If you run into this often enough, you start to ask yourself if these designs were ever tested at all.
Don't buy the Hauppage card. Ick -- even if it had worked, you couldn't capture from a third party application and the Hauppage application was pretty ugly (bad GUI, etc) and it was a pretty big hodgepodge of software from different vendors. Worst, I don't think the card does hardware MPEG2 decode to screen, I think its software decode, based on the low system performance.
In the Intel server world it seems that trend is going away from better BIOSes/OpenFirmware/Etc, and towards smarter and more functional management boards which are computers in their own right, not bound to the host CPU, RAM, and in some cases NIC. You can get full console redirection as well, including BIOS control and other hardware-console specific configuration, as well as diagnostics, environmental information and power off capability/
GUI OSes are kind of crippling for console redirection, though, and most vendors kludge together something using PC Anywhere or other dependent tools. I was told by a Compaq rep that eventually we might see a management controller with the ability to do what some of the KVM-Over-IP people are doing, except natively in the PC management board, providing a remote keyboard/mouse/display functionality. A side benefit of this is that you can ditch your KVMs, too, since you could just use a laptop with a NIC as the display and switch in software.
Personally I'm interested in this and it'd be interesting to see it take off and benefit from all the usual mass-production efficiencies, as well as seing it develop more OS-like features (such as perhaps running user-supplied modules).
Forget compression, what about transcoding of the files between various formats or bitrates? Forget about the aural impact of transcoding for a second, but the datastream impact. My rusty ol' ears won't hear anything different, but the data stream will have a completely different signature.
If its watermarking, would transcoding it destroy the watermark?
Heh, I was thinking the same thing. I wouldn't mind having XP on something that size, especially if you could get 512MB of RAM.
Unfortunately, it's being supplanted by infighting & organised crime[...]
From what I understand this has become a far bigger problem in both Republican and Loyalist areas for far longer than anyone wants to admit.
My understanding is that it got its start with IRA "containment" of heroin trafficking, which became a significant revenue stream. The same is true for Loyalist forces, and they both make good money on extortion and protection rackets.
My wife does business with some Northern Ireland companies and one of them had two pricipals killed in a bombing by Loyalists angry over hiring of Catholic labor.
In my scenerio, the item was published, but in a limited fashion due to cost and low demand, then let go unpublished for the expiry time. At that point its public domain, but after it goes public domain a new technology enables the trivial redistribution of it, cutting the artist out of profits.
Another version of this could involve a filmmaker who shoots a film on 70mm stock, makes a dozen complicated widescreen transfers to LD in the mid 80s. Won't allow VHS versions of any kind. LD version goes out of print due to low demand. Voila -- a 16:9 480p DVD version is now possible, and anyone can then legally (if I read your rules right) make and sell a DVD made from the LD transfer.
Films are even more complicated because what constitutes "in print"? Public performances? Say I have a graymarket print of the film (not an unauthorized copy, but one acquired from a movie house being torn down or something), does no public showings count? LD or VHS transfers to DVD? My own transfers from film to DVD?
The gaming industry may have good reason to complain; a given game cart may not be economically viable due to the obsolence of the platform, and its not technologically viable to emulate the platform *yet*, but they hang onto them based on the value of the character or that decent emulation may allow them to get a few sales for a new platform in the future.
I'm not saying I agree with this logic, but its kind of easy to see why people might want to hang onto copyrights for even 20 years after the original is no longer economically viable. I guess I'd make personal copying legal, but sales illegal just to protect the owner against economic exploitation.
The best spam control isn't new laws about who can send mail or newer and fancier statistical analysis filters, it's vigorous and severe punishment for the fraudulent schemes/products pushed through email.
Even if you passed a law that said "No spamming" the people running spam servers or doing bulk emailing would still be nearly impossible to catch since they can be hijacking open relays or using false-name disposable mail servers.
But ALL of the spam ultimately has a 'reachable' person there to collect the money, and that provides the perfect means to catch the PEOPLE invovled.
If the fraud merchants behind these schemes began to get indicted and convicted with a visible public frequency, it might start to have a real impact on spam. It wouldn't be a 100% perfect solution, but it would stave off needless government regulation of email which 'spam' laws would need to be effective and it might scare off a big chunk of the amateurs and part-timers trying to run their own solo operations.
I'd love to see that spam bitch from Florida that was on Slashdot a few months ago lose that house, the SUV and all the other goodies she has as she tries to ward off a federal wire fraud investigation. I know she was just acting as an advertising service, but isn't aiding and abetting illegal too?
I like this, although I wonder if there's maybe a valid argument against it. Say for example I, as an artist, produce a recording in glorious quadrophonic sound. The recording in fact doesn't sound good in stereophonic sound, and as a small artist I find the costs of recording my work in quadrophonic vinyl too expensive to sell in the market. So I let it go out of print.
By these terms, I lose my copyright after 10 years. But let's say that a new medium comes into existance that trivially can encode all four channels for no more cost than two channels. I've lost my copyright, and anyone can sell my work, it becomes semi-popular but I don't get any royalties.
How can you deal fairly with this (admittedly contrived) scenerio?
Ohters in this thread suggest dictionary-style spam campaigns aimed at high-profile domains, which is a good explanation.
However, I wonder if this isn't the beginning of an attempt to corner the market on (euphamism mode on) "permissioned, targeted email marketing" to hotmail users.
Step 1: chase away spammers who don't buy protection from MS
Step 2: sell access to spammers (high quality lists, demographics, etc)
Step 3: $$$$$
It's got to frost MS a little that there are all those "consumers" at hotmail and somebody else is sending them marketing info that MS doesn't get a big slice of.
I think that's reasonable, but I also think it's difficult to enshrine into law. The minute we say it's ok to make three copies and give them to friends, someone will make 10 copies and then make a copy of each copy three times and give them away, for a total of *30* and argue that they followed the law -- no more than three copies of an original may be given away..
Google's not a public corporation, but that doesn't meant they're the FSF either.
I can assure you that their private corporation status means that the financial stakeholders is a smaller, more coherent and more invovled group of people and that the pressure around money will be larger, and more focused than it would be if it was a public company, where those forces can be more easily diffused among many shareholders whose interests are often in conflict.
I'd be up for changing the copyright law on this one. I'd like to see copyright holders be required to prove that they have existing plans to re-release something currently out of print within a reasonable timeframe before they can invoke their copyrights.
I realize this might hinder the "business model" of some copyright holders that manufacture scarcity by letting popular items go out of print and stay out of print for a while so they get a bubble of sales when they re-release them.
I also realize this might hurt performers whose material has gone out of print because their sales fall below the million-a-day required to be considered businessworthy by large labels, but at least their material would be available.
I thought it was supposed to be faster too, but the machine I have it installed on (a Dual PIII650) is noticably slower than the same box with Win2k.
The running performance is more sluggish, too. Maybe I need to upgrade.
Compared to many businesses, Google seems pretty good -- geek friendly, low ad content, good service, cool technology, et al.
But its also important to not lose sight that Google is a corporation, with investors, debtors and other people who are solely (or primarily) motivated by MAKING MONEY. They're not motivated by some pro-geek/anti-corporate ethos.
So as long as you keep in mind that they might turn around and do something that protects profit first and makes privacy or other goals take a back seat then you'll be OK.
The other day I had occasion to boot a Win95 machine (complete Netware 4.11 networking, IP, printers, etc) and was blown away by how fast it booted.
It was on a PII450, but still, even high end PIII systems take double that amount of time to boot 2K and XP.
Take a look at the amount of abandoned train tracks throughout america, it's extremely sad.
In high school a friend and I thought it would be fun to get a map of train tracks and then build a kind of hand-car that we could mount our bikes on and use pedal power for locomotion. We figured we could go farther and faster on the rails due to their relatively straight and flat paths and the decreased rolling resistance. It'd probably be safer than riding on rural highways.
I'm pretty sure it would have been a big problem if we would have been caught riding something like this without authoriztion on *active* train tracks, even though I don't think it'd be much of a safety risk, although crossing a long bridge might still be a little iffy.
It would have been a lot of fun, especially since trains often take paths different from roadways and you'd get to see scenery from a different persective. The light weight and bicycle nature of it would have meant we could have yanked it off the tracks and camped at night or ridden into towns for supplies, etc.