You've a point about Windows finally dealing with longer extensions, although it still doesn't always do it well. But if we're going to go for better names, how about ".firehose"?
The earlier poster here suggested switching Bittorrent like protocols to UDP: I'm merely pointing out where it's going to have issues.
Well, yes. ISP's pulling that level of filtering are, as you imply, doing so where choices for cheap and freely usable bandwidth are limited. This can be because of the expense of bandwidth or a desire for casual monitoring (such as a campus network, where the student with the 3 Terabytes of MP3's and DVD's sharing them to the world is both a bandwidth and a legal problem). It's relatively common in small, insular markets, where a power user or systems dabbler such as many Slashdot posters would be regarded as a problem, not a good client base.
Some folks do find the proxy filtering an issue in corporate networks: I've certainly found it to be pesky for rsyncing or Bittorrenting freeware CD images, and had talks with upstream network managers who wondered why I was pulling so much data through the firewall (which I throttled, and did off-hours, but they noticed).
And, guess what? Some ISP's, and especially some corporate network providers, block all outbound traffic except through their proxies. It's part of their particular ISP provided "software bundle": you find that kind of setup out in the boonies, or for corporate or university services.
It needs to have a recognizable TLA at the end, to keep the Windows users and file identifiers happy. There aren't many leftp: p2p is at least descriptive.
A lot of ISP's bloock UDP outside their network, except for Port 53 for their DNS servers, and force you to use or slave from their DNS servers. Like directly reaching out on Port 25 for SMTP, it's straightforward to block the service except for those hosts you wish to permit.
And toy manufacturers are regularly sued for making dangerous toys, gun manufacturers have been sued for selling assault weapons that were not labeled as such, etc. End User License Agreements (EULA) try to avoid all responsibility for manufacturers, but they do wind up with whatever an attorney can manage to convince a judge to allow, whether a law is constituional or not.
They may not actually be paid for by Microsoft, that's merely the obvious suspicion. This may be more of a "People's Front of Judea" vs. "People's Judean Front" sort of splinter group, who've discovered that their particular prized functionality will not be done at the expense of the stability of core format. A casual look at CDF, which they seem to be supporting now, shows it focused cell phones, which is a nightmare to support due to the small screen real estate and modest system resources.
Excuse me, sir or madam. But that "English speaker's arrogance" has existed in every empire since the dawn of time. Mandarin Chines in Asia, Portuguese, French, Arabic in the current Muslim states, Latin for the Roman Empire and the later Christian states, even German and French for the werten scienes have all had their own speakers insist on using the "master language" of their literature, commerce, religion, and engineering.
Such "arrogance" is hardly unique, and given its small userbase, Esperanto stands little to no chance of gaining acceptance over *any* of those languages.
And after seeing what Microsoft did to Kerberos and to HTML, Sun was very wise to prevent Microsoft from doing their classic "embrace and extend" procedure to proprietize it in their own favor.
While that bit of logic is similar, Esperanto never had any country that actually used it. ODF does, in the OpenOffice and StarOffice suites of software. Thus, it never had a base of native users to support or to eveolve it.
No, the name is clearly designed to create an impression of a relationship with the Open Document Format. It's a common tactic for gathering webhits and credibility, much as the "Open Source Foundation" pretended to be about open source, which it never really was.
Excuse me, but there is no such thing as ".doc" format. There at least half a dozen, if not more, mutually distinct formats labeled ".doc". Each of them has features and capabilities not available in all the others, and transformations among them are non-reversible: translating a document from an old Word 95 format to Word 2003, to Word for Macintosh version whatever, will not reproduce your original document. It's even worse for spreadsheets, which are also part of the format.
The denominators for it are not "common", they're nearly fractal in their complexity.
Sadly, Microsoft often doesn't have to pay shills like this. They can sell their services in "promoting compatibility" to third parties who don't know any better.
Witness the career of Meng Weng Wong, who naively cooperated with Microsoft in accepting SenderID into his SPF standard and watched Microsoft's proprietary, patented XML lunacy effectively destroy further SPF deployment, while allowing Microsoft and SenderID to take credit for all the good SPF had already done.
It's like dealing with Wal-mart: you may be forced into doing so in the short term by the need for expansion, but in the long term, it's usually death for you company or your project.
It doesn't have to work in the long run. It needs to work for the next few years, until the next major technology alters the network landscape yet again. Building in policies now for use in 10 years is more political than it is economic or "business oriented".
In fact, such filtering of outbound port 25 traffic both inbound and outbound except to the ISP mail servers, outbound and inbound SMB, outbound and inbound NFS, outbound FTP or HTTP except through proxies, etc., are all quite helpful in managing network traffic and reducing bandwidth costs with only slight cost in tech support or customer subscriptions.
Technologically, this is merely an expansion of the same policy. Politically and socially, it's much worse, because it's not actually blocking: it's faking the traffic and causing it nto fail, which makes it a whole different ball of wax.
You've got a good point. But I have to admit that when I babysit for my friends, it's very handy to pull up something meaningful from the cable TV and watch it with the kids, explaining things and playing games with it, than to go digging through their CD's or DVD's or mine for some rainy day entertainment.
The weather channels and news channels are also amazingly useful for farmers and other low income people living closer to nature than most of us Slashdot posters.
As important as GUI is, most computing is not fundamentally GUI. That's why "Hello, World" is such a classic example of the issues. The massive and usually unnecessary overhead of many programming languages in doing such elementary functions as text handling, disk IO, mathematics, pointer handling, building arrays, memory allocation, etc. often severely punish systems that are written for GUI or other more abstracted environments.
And if all the children in the world clap 3 times, the fairy Tinkerbell will come back to life.
More seriously, porting everything Microsoft based to be usable under Linux is an amazingly herculean task, and burdened by massive issues of copyright, patent, and trade secret. It's also burdened by Microsoft's ability and demonstrated willingness to violate their own specifications: take a good look at the legal craziness Samba goes through to avoid such infringements, and at the problems we experience with Windows Media players. Many of those technologies are patented: they *cannot* be used, at least by people in the US, without serious legal repercussions from the patent holders.
This is aggravated by the confusing patent issues around OOXML in particular. Doing such porting successfully, especially with Microsoft's crown jewels of MS Office applications, is begging to get attacked economically, legally, and in public forums.
4) Getting Microsoft compatibility is so important that we should imperil our open source and open standards to get their involvement. Kind of like "invading Iraq to stop terrorism".
We've seen Microsoft involved in "adapting" open standards before, which they quite deliberately broke to their own advantage to promote incompatibility with the existing standard. When they own the standard, it's far easier for them to do, and we've seen that historically with their Word and Excel document formats, and their oddnesses in the MS Exchange calendar system.
Microsoft does not use open standards by any set of rules except "take ownership".
This is part of why I love Tom's Hardware. They do pan hardware, they detail how they tested it, and they do their best to avoid getting special shipments rather than retail versions. It's sometimes difficult to avoid a special shipment when you're reviewing something that hasn't been released yet. They admit, in clear print, that it's a pre-release.
It's the nature of the beast. Even some motherboard vendors swap chipsets without changing the model number of the board. I've had to take a jeweler's loupe to a motherboard, very carefully scrape off the obscuring hot glue, read the chip numbers off, and then hand it to my colleage with better photography tools to prove the difference. (Changing it was a violation of a particular contract with that vendor, since the chipsets did not use the same drivers and forced a painful hand-written kernel upgrade.)
There's a huge legal problem there. A lot of the tools for full NVidia use are copyrighted, others are trade secrets, and others are patented. Writing drivers that use the full capability could run afoul of any of those, and if you don't sign the non-disclousre agreements, you can't even properly say "this code is intellectual property safe". Moreover, to use those capabilities, you have to re-write the Mesa libraries, which means integrating that support into a good open source codebase, which makes stuffing these "intellectual property" protected even more awkward.
So NVidia needs to change their practices for this. Instead, announce full support for ATI over NVidia, work with the open source, and help ATI get a better market. That's the best leverage we can apply to NVidia.
Usuaally you can image the disk as a *.iso image and use tools like "Daemontools" to mount the CD image when you want to play. It works well for me for several games I like.
Slashdot Mirror
You've a point about Windows finally dealing with longer extensions, although it still doesn't always do it well. But if we're going to go for better names, how about ".firehose"?
The earlier poster here suggested switching Bittorrent like protocols to UDP: I'm merely pointing out where it's going to have issues.
Well, yes. ISP's pulling that level of filtering are, as you imply, doing so where choices for cheap and freely usable bandwidth are limited. This can be because of the expense of bandwidth or a desire for casual monitoring (such as a campus network, where the student with the 3 Terabytes of MP3's and DVD's sharing them to the world is both a bandwidth and a legal problem). It's relatively common in small, insular markets, where a power user or systems dabbler such as many Slashdot posters would be regarded as a problem, not a good client base.
Some folks do find the proxy filtering an issue in corporate networks: I've certainly found it to be pesky for rsyncing or Bittorrenting freeware CD images, and had talks with upstream network managers who wondered why I was pulling so much data through the firewall (which I throttled, and did off-hours, but they noticed).
And, guess what? Some ISP's, and especially some corporate network providers, block all outbound traffic except through their proxies. It's part of their particular ISP provided "software bundle": you find that kind of setup out in the boonies, or for corporate or university services.
I accept that it's not as common as I thought.
It needs to have a recognizable TLA at the end, to keep the Windows users and file identifiers happy. There aren't many leftp: p2p is at least descriptive.
A lot of ISP's bloock UDP outside their network, except for Port 53 for their DNS servers, and force you to use or slave from their DNS servers. Like directly reaching out on Port 25 for SMTP, it's straightforward to block the service except for those hosts you wish to permit.
And toy manufacturers are regularly sued for making dangerous toys, gun manufacturers have been sued for selling assault weapons that were not labeled as such, etc. End User License Agreements (EULA) try to avoid all responsibility for manufacturers, but they do wind up with whatever an attorney can manage to convince a judge to allow, whether a law is constituional or not.
They may not actually be paid for by Microsoft, that's merely the obvious suspicion. This may be more of a "People's Front of Judea" vs. "People's Judean Front" sort of splinter group, who've discovered that their particular prized functionality will not be done at the expense of the stability of core format. A casual look at CDF, which they seem to be supporting now, shows it focused cell phones, which is a nightmare to support due to the small screen real estate and modest system resources.
Excuse me, sir or madam. But that "English speaker's arrogance" has existed in every empire since the dawn of time. Mandarin Chines in Asia, Portuguese, French, Arabic in the current Muslim states, Latin for the Roman Empire and the later Christian states, even German and French for the werten scienes have all had their own speakers insist on using the "master language" of their literature, commerce, religion, and engineering.
Such "arrogance" is hardly unique, and given its small userbase, Esperanto stands little to no chance of gaining acceptance over *any* of those languages.
Ah. The fact that they don't care is why no one ever wrote GAIM, Jabber, IRC, the old BBS's, Usenet, or mailing lists.
Getting them to agree on format is admittedly impossible, but it's obvious that they do, in fact, care.
And after seeing what Microsoft did to Kerberos and to HTML, Sun was very wise to prevent Microsoft from doing their classic "embrace and extend" procedure to proprietize it in their own favor.
While that bit of logic is similar, Esperanto never had any country that actually used it. ODF does, in the OpenOffice and StarOffice suites of software. Thus, it never had a base of native users to support or to eveolve it.
Next strawman?
No, the name is clearly designed to create an impression of a relationship with the Open Document Format. It's a common tactic for gathering webhits and credibility, much as the "Open Source Foundation" pretended to be about open source, which it never really was.
Excuse me, but there is no such thing as ".doc" format. There at least half a dozen, if not more, mutually distinct formats labeled ".doc". Each of them has features and capabilities not available in all the others, and transformations among them are non-reversible: translating a document from an old Word 95 format to Word 2003, to Word for Macintosh version whatever, will not reproduce your original document. It's even worse for spreadsheets, which are also part of the format.
The denominators for it are not "common", they're nearly fractal in their complexity.
Sadly, Microsoft often doesn't have to pay shills like this. They can sell their services in "promoting compatibility" to third parties who don't know any better.
Witness the career of Meng Weng Wong, who naively cooperated with Microsoft in accepting SenderID into his SPF standard and watched Microsoft's proprietary, patented XML lunacy effectively destroy further SPF deployment, while allowing Microsoft and SenderID to take credit for all the good SPF had already done.
It's like dealing with Wal-mart: you may be forced into doing so in the short term by the need for expansion, but in the long term, it's usually death for you company or your project.
It doesn't have to work in the long run. It needs to work for the next few years, until the next major technology alters the network landscape yet again. Building in policies now for use in 10 years is more political than it is economic or "business oriented".
In fact, such filtering of outbound port 25 traffic both inbound and outbound except to the ISP mail servers, outbound and inbound SMB, outbound and inbound NFS, outbound FTP or HTTP except through proxies, etc., are all quite helpful in managing network traffic and reducing bandwidth costs with only slight cost in tech support or customer subscriptions.
Technologically, this is merely an expansion of the same policy. Politically and socially, it's much worse, because it's not actually blocking: it's faking the traffic and causing it nto fail, which makes it a whole different ball of wax.
You've got a good point. But I have to admit that when I babysit for my friends, it's very handy to pull up something meaningful from the cable TV and watch it with the kids, explaining things and playing games with it, than to go digging through their CD's or DVD's or mine for some rainy day entertainment.
The weather channels and news channels are also amazingly useful for farmers and other low income people living closer to nature than most of us Slashdot posters.
As important as GUI is, most computing is not fundamentally GUI. That's why "Hello, World" is such a classic example of the issues. The massive and usually unnecessary overhead of many programming languages in doing such elementary functions as text handling, disk IO, mathematics, pointer handling, building arrays, memory allocation, etc. often severely punish systems that are written for GUI or other more abstracted environments.
And if all the children in the world clap 3 times, the fairy Tinkerbell will come back to life.
More seriously, porting everything Microsoft based to be usable under Linux is an amazingly herculean task, and burdened by massive issues of copyright, patent, and trade secret. It's also burdened by Microsoft's ability and demonstrated willingness to violate their own specifications: take a good look at the legal craziness Samba goes through to avoid such infringements, and at the problems we experience with Windows Media players. Many of those technologies are patented: they *cannot* be used, at least by people in the US, without serious legal repercussions from the patent holders.
This is aggravated by the confusing patent issues around OOXML in particular. Doing such porting successfully, especially with Microsoft's crown jewels of MS Office applications, is begging to get attacked economically, legally, and in public forums.
Try this:
4) Getting Microsoft compatibility is so important that we should imperil our open source and open standards to get their involvement. Kind of like "invading Iraq to stop terrorism".
We've seen Microsoft involved in "adapting" open standards before, which they quite deliberately broke to their own advantage to promote incompatibility with the existing standard. When they own the standard, it's far easier for them to do, and we've seen that historically with their Word and Excel document formats, and their oddnesses in the MS Exchange calendar system.
Microsoft does not use open standards by any set of rules except "take ownership".
OK, write "Hello, world." in all 6 languages and compare.
This is part of why I love Tom's Hardware. They do pan hardware, they detail how they tested it, and they do their best to avoid getting special shipments rather than retail versions. It's sometimes difficult to avoid a special shipment when you're reviewing something that hasn't been released yet. They admit, in clear print, that it's a pre-release.
It's the nature of the beast. Even some motherboard vendors swap chipsets without changing the model number of the board. I've had to take a jeweler's loupe to a motherboard, very carefully scrape off the obscuring hot glue, read the chip numbers off, and then hand it to my colleage with better photography tools to prove the difference. (Changing it was a violation of a particular contract with that vendor, since the chipsets did not use the same drivers and forced a painful hand-written kernel upgrade.)
There's a huge legal problem there. A lot of the tools for full NVidia use are copyrighted, others are trade secrets, and others are patented. Writing drivers that use the full capability could run afoul of any of those, and if you don't sign the non-disclousre agreements, you can't even properly say "this code is intellectual property safe". Moreover, to use those capabilities, you have to re-write the Mesa libraries, which means integrating that support into a good open source codebase, which makes stuffing these "intellectual property" protected even more awkward.
So NVidia needs to change their practices for this. Instead, announce full support for ATI over NVidia, work with the open source, and help ATI get a better market. That's the best leverage we can apply to NVidia.
Usuaally you can image the disk as a *.iso image and use tools like "Daemontools" to mount the CD image when you want to play. It works well for me for several games I like.