Slashdot was the turning point for technology journalism. On Slashdot, the geeks picked the news, and the geeks revelled in the news. It's hard to remember how bad things were before Slashdot. It was ugly out there in tech news land, the news was stale, shallow, irrelevant and patronising, and it was embarassing to be a geek. Now we have so many ways to connect and to dream. Thanks Rob, for leading the pack. Geek pride forever!
An openly documented office file format is a step in the right direction.
But I think the bigger problem with MS Office is that the file format itself is horrid. From what I can gather, it's a cross between a raw object dump and a virtual memory cache.
This would explain why even MS Office can't reliably read its own documents, and why reverse engineering has been so slow and difficult.
In Microsoft Word 2.0, if you were editing a file on a floppy disk, and you swapped the floppy for another one (e.g. to copy text out of another file) without closing Word, then Word crashed and your doc file was trashed. I reckon the reason the file was trashed is that Word was using it as a swap file.
The same bug exists in Word 2002, released 10 years later. It happened to me when a network drive disconnected.
The fact that so many people rely on such a fragile file format is holding a lot of things back. How can anyone build on shifting sand? The sooner its replaced with something decent the better.
Imagine if digital cameras all used a closed image file format. A lossy format like JPEG, except that instead of discarding details humans don't notice, the format loses information that humans really care about. And instead of compressing data, it expands it.
I think a lot fewer people would enjoy using digital cameras if they worked that way.
Of course the file format is only the tip of the iceberg. There are substantial features in Word that are so broken that users quickly learn it's not worth using them. Other features are so unpredictable that they effectively discourage editing and experimentation. Expert users subconciously avoid features that cause problems.
I reckon the overall productive uptime with Word is about 90% - 95%, depending on the job at hand and the user. That's not too bad until you multiply 5% downtime by several hundred million daily users.
The quandry is that people won't be motivated to switch until something comes along that is both compatible and yet significantly improved. It's hard to be compatible with a dodgy file format, and projects like OpenOffice.org seem to be more of the same but open. It reminds me of the ponderous Mozilla browser. Mozilla and OOo are excellent projects, but what's needed is something like Firefox that strips off the barnacles, is open source so people can build on it, works roughly the way that people are used to, and yet simpler, faster and more reliable.
There are 5 billion literate people in the world. Isn't it time we had a writing tool that doesn't suck?
Unless of course people boot Knoppix to do their banking, then stay in Knoppix to do their webmail, their online shopping, and pretty soon they're never using Windows because it's too inconvenient to have to reboot.
I just thought this through a bit more. My scenario above means that Microsoft might end up buying Red Hat! Not to squash it, but to buy into Linux as their OS layer.
Of course the GPL would give them serious indigestion, so more likely they would go with BSD like Apple did. We know they already use a little BSD in the Windows networking code after all.
We'll know Microsoft understands what's happening when they dump the Windows OS layer and start selling a closed source window manager and office application suite that runs on top of an open source Unix. They are in a good situation to compete with Wine, allowing people to reliably run Win32 apps on top of Unix, and it's the kind of backwards compatibility hack they have always managed to pull off.
If they're really smart, their slimmed stack will include a closed source graphics layer (with lots of drivers) before XFree86 gets user friendly enough to compete. Setting up XFree86 with full 3D support is still one of the most fiddly and uncertain things about Linux.
Microsoft's next version of Windows, Longhorn, is trudging along because they are burdened with maintaining all levels of the stack, while Apple is starting to fly along now that they have specialised on the top of the software stack.
Mac OS X isn't available on Intel chips yet, so Microsoft has a breather until some other open or closed source project starts gaining critical mass in the window manager and application space.
If Microsoft tries to hold on to the whole stack, they may feel like they are succeeding for longer, only to fall harder by losing more of the stack when they do inevitably lose market dominance of the OS layer for both desktops and servers.
There'll still be plenty of competition for the top of the stack, and closed and open source will both be in the game. But I think Tim's genius is in spotting the way that free software is assimilating its way up the software stack from the network protocols up the through the OS and beyond.
I agree there's often a tradeoff between money and performance.
However,.org.au is for non-profit organisations. There are hundreds of non-profits in Sydney who are completely volunteer, and run on budgets of $1000 a year or less.
$200 a year for a domain is a lot of money for them.
Yet because.org.au is free, and because http://www.cat.org.au is grungy and volunteer itself, we've been able to host many.org.au websites.
http://www.green.net.au is in a similar situation.
So, if some large non-profits want to pay $200 for an instant response, good for them. But please let there be a cheaper option (with slower response times?) for smaller non-profits, or they will not be able to afford domains at all.
There's nothing I could find on auDA's website about how they plan to manage.org.au when they get their hands on it. Indeed, it's still not clear how long it will be before the switch.
I imagine that they will introduce charges for.org.au domains.
Asking for.org.au domains has seemed like a bit of a lucky dip with Elz, and my experience agrees with earlier comments that sometimes he's prompt (2 weeks), and sometimes your request goes into a black hole.
But anyone in a real hurry could get.org or.asn.au or something if they had the cash.
Robert Elz handled.org.au for free for all these years, and many non-profits were able to get their own domain name without the cost and hassle of startup and annual fees.
They're in damage control mode now. Saying they will only monitor traffic after a complaint from a copyright holder.
However, it's still not clear if such a complaint can only lead to investigation of an individual user, or if it can lead to blanket monitoring of the optus network for that complainers' content.
Re: low RAM is still an important test case
on
Linux: Browser Wars
·
· Score: 1
anyone using computers more than 3 years old will know how hard to find and expensive the old RAM is. so you're often effectively stuck with the 32 MB of RAM the box came with.
me, i'm using a pentium 166 with about 50 MB RAM. with browsers like opera and konqueror it just hangs in there as far as being usably fast. but for lots of other things, including web programming, it's quite useful, although emacs is a bit of a stretch.
some people can't afford to spend a thousand dollars a year on their computer.
this is especially important for projects seeking to recycle old computers and put linux on them.
yes the codec problem needs to be solved, but what about simple outgoing bandwidth costs?
witness slashdot, it's not even streaming media, and it needs advertising to support its bandwidth costs (i think?). streaming media chews huge amounts of bandwidth.
and slashdot is in the US, where outgoing bandwidth is relatively cheap, especially now with DSL. here in australia, outgoing bandwidth can be 10 - 20 times as expensive.
what it all means, is that the internet is great for anyone to publish any of their ideas as stories, as long as they don't get too popular. if they get popular, currently they might find they need to get ads, which some ideas and stories may not be compatible with.
definitely a hero, definitely unsung. he's been working on xfree86 for eight years. we all use X. although you'd really want to give the gong to all four founders of the project.
the world health organisation says that depression will soon be one of the world's biggest diseases. and i'd guess that young people, especially geeks being picked on at school, would be particularly vulnerable to depression.
now different people might react to depression differently. some of the more extreme reactions could be:
* going overboard on drugs
* growing up and oppressing the hell out of other people
* committing suicicde
* shooting people
now the last is the most extreme reaction. perhaps a reason why mainly white men seem to do this, could be that white men get depressed like all humans can, but they believe the hype about white men being superior, and that drives them that extra step into insanity where they think it would be a good idea take their rightful dominant place and murder people.
i dunno, just a theory.
people getting depressed suck. people overdosing on drugs or attempting suicide sucks. it'd be good if we could figure out why this is escalating. i have my own theories about centralised media and suppressed creativity.
people murdering other people sucks and i have no sympathy for it. but if we can figure out how people get that way, and maybe reduce it, then that would be good.
m.
com+ apps a big threat to samba and linux?
on
Microsoft's COOL
·
· Score: 1
one of the reasons linux has become very popular as a server is because samba makes it such a good replacement for winNTserver, even if people are running MSoffice apps on the desktop.
the way i understand it, com+ apps may depend on win2000 server - will samba be able to serve com+ objects to com+ apps like MSoffice 2000?
otherwise, this could be the excuse pointy haired bosses are looking for to install win2k server instead of samba/linux. i.e. upgrade all desktops to MSoffice 20000, which will then force all file servers to be running win2000 server.
MSoffice's market share is not currently threatened by open source apps much, so this could be a way for MS to use its app monopoly to bolster it's weakening server business.
OK, Microsoft is trying to invent a market... crashing home appliances, it's gotta try to keep those shareholders happy by going for enourmous growth rates.
About the only way I rekcon they could succeed is by bundling a DNA probe in your toaster (for our safety of course) then doing a licensing deal with Monsanto so that the toaster GPFs on anything but Monsanto wheat bread. Or more importantly, the fridge complains of a virus and blue screens if you've got things in there without Monsanto genetic copyright signals.
Kind of off-topic, but meanwhile, I think I figured out how Microscoft might be planning to lock in NT. Even if they don't de-commoditize the internet, they still have a monopoly on Office. And Office 2000 hype is already coming out.
Part of Office is COM/DCOM stuff, which means bits of the app don't get installed on your drive until you click particular menus. And I reckon the only way this will work is if you have a DCOM server running on NT.
So who cares if samba runs rings around NT? Corporates are going to want Office 2000 because MS have lock-in on the deliberate file format mess that is Word. And if Office 2000's fancy new features depend on an NT server, maybe Linux won't get a look in.
Slashdot Mirror
Slashdot was the turning point for technology journalism. On Slashdot, the geeks picked the news, and the geeks revelled in the news. It's hard to remember how bad things were before Slashdot. It was ugly out there in tech news land, the news was stale, shallow, irrelevant and patronising, and it was embarassing to be a geek. Now we have so many ways to connect and to dream. Thanks Rob, for leading the pack. Geek pride forever!
An openly documented office file format is a step in the right direction.
But I think the bigger problem with MS Office is that the file format itself is horrid. From what I can gather, it's a cross between a raw object dump and a virtual memory cache.
This would explain why even MS Office can't reliably read its own documents, and why reverse engineering has been so slow and difficult.
In Microsoft Word 2.0, if you were editing a file on a floppy disk, and you swapped the floppy for another one (e.g. to copy text out of another file) without closing Word, then Word crashed and your doc file was trashed. I reckon the reason the file was trashed is that Word was using it as a swap file.
The same bug exists in Word 2002, released 10 years later. It happened to me when a network drive disconnected.
The fact that so many people rely on such a fragile file format is holding a lot of things back. How can anyone build on shifting sand? The sooner its replaced with something decent the better.
Imagine if digital cameras all used a closed image file format. A lossy format like JPEG, except that instead of discarding details humans don't notice, the format loses information that humans really care about. And instead of compressing data, it expands it.
I think a lot fewer people would enjoy using digital cameras if they worked that way.
Of course the file format is only the tip of the iceberg. There are substantial features in Word that are so broken that users quickly learn it's not worth using them. Other features are so unpredictable that they effectively discourage editing and experimentation. Expert users subconciously avoid features that cause problems.
I reckon the overall productive uptime with Word is about 90% - 95%, depending on the job at hand and the user. That's not too bad until you multiply 5% downtime by several hundred million daily users.
The quandry is that people won't be motivated to switch until something comes along that is both compatible and yet significantly improved. It's hard to be compatible with a dodgy file format, and projects like OpenOffice.org seem to be more of the same but open. It reminds me of the ponderous Mozilla browser. Mozilla and OOo are excellent projects, but what's needed is something like Firefox that strips off the barnacles, is open source so people can build on it, works roughly the way that people are used to, and yet simpler, faster and more reliable.
There are 5 billion literate people in the world. Isn't it time we had a writing tool that doesn't suck?
Unless of course people boot Knoppix to do their banking, then stay in Knoppix to do their webmail, their online shopping, and pretty soon they're never using Windows because it's too inconvenient to have to reboot.
I just thought this through a bit more. My scenario above means that Microsoft might end up buying Red Hat! Not to squash it, but to buy into Linux as their OS layer.
Of course the GPL would give them serious indigestion, so more likely they would go with BSD like Apple did. We know they already use a little BSD in the Windows networking code after all.
We'll know Microsoft understands what's happening when they dump the Windows OS layer and start selling a closed source window manager and office application suite that runs on top of an open source Unix. They are in a good situation to compete with Wine, allowing people to reliably run Win32 apps on top of Unix, and it's the kind of backwards compatibility hack they have always managed to pull off.
If they're really smart, their slimmed stack will include a closed source graphics layer (with lots of drivers) before XFree86 gets user friendly enough to compete. Setting up XFree86 with full 3D support is still one of the most fiddly and uncertain things about Linux.
Microsoft's next version of Windows, Longhorn, is trudging along because they are burdened with maintaining all levels of the stack, while Apple is starting to fly along now that they have specialised on the top of the software stack.
Mac OS X isn't available on Intel chips yet, so Microsoft has a breather until some other open or closed source project starts gaining critical mass in the window manager and application space.
If Microsoft tries to hold on to the whole stack, they may feel like they are succeeding for longer, only to fall harder by losing more of the stack when they do inevitably lose market dominance of the OS layer for both desktops and servers.
There'll still be plenty of competition for the top of the stack, and closed and open source will both be in the game. But I think Tim's genius is in spotting the way that free software is assimilating its way up the software stack from the network protocols up the through the OS and beyond.
I agree there's often a tradeoff between money and performance.
.org.au is for non-profit organisations. There are hundreds of non-profits in Sydney who are completely volunteer, and run on budgets of $1000 a year or less.
.org.au is free, and because http://www.cat.org.au is grungy and volunteer itself, we've been able to host many .org.au websites.
However,
$200 a year for a domain is a lot of money for them.
Yet because
http://www.green.net.au is in a similar situation.
So, if some large non-profits want to pay $200 for an instant response, good for them. But please let there be a cheaper option (with slower response times?) for smaller non-profits, or they will not be able to afford domains at all.
Asking for .org.au domains has seemed like a bit of a lucky dip with Elz, and my experience agrees with earlier comments that sometimes he's prompt (2 weeks), and sometimes your request goes into a black hole.
But anyone in a real hurry could get .org or .asn.au or something if they had the cash.
Robert Elz handled .org.au for free for all these years, and many non-profits were able to get their own domain name without the cost and hassle of startup and annual fees.
Thanks Mr. Elz!
They're in damage control mode now. Saying they will only monitor traffic after a complaint from a copyright holder. However, it's still not clear if such a complaint can only lead to investigation of an individual user, or if it can lead to blanket monitoring of the optus network for that complainers' content.
anyone using computers more than 3 years old will know how hard to find and expensive the old RAM is. so you're often effectively stuck with the 32 MB of RAM the box came with.
me, i'm using a pentium 166 with about 50 MB RAM. with browsers like opera and konqueror it just hangs in there as far as being usably fast. but for lots of other things, including web programming, it's quite useful, although emacs is a bit of a stretch.
some people can't afford to spend a thousand dollars a year on their computer.
this is especially important for projects seeking to recycle old computers and put linux on them.
yes the codec problem needs to be solved, but what about simple outgoing bandwidth costs?
witness slashdot, it's not even streaming media, and it needs advertising to support its bandwidth costs (i think?). streaming media chews huge amounts of bandwidth.
and slashdot is in the US, where outgoing bandwidth is relatively cheap, especially now with DSL. here in australia, outgoing bandwidth can be 10 - 20 times as expensive.
what it all means, is that the internet is great for anyone to publish any of their ideas as stories, as long as they don't get too popular. if they get popular, currently they might find they need to get ads, which some ideas and stories may not be compatible with.
definitely a hero, definitely unsung. he's been working on xfree86 for eight years. we all use X. although you'd really want to give the gong to all four founders of the project.
the world health organisation says that depression will soon be one of the world's biggest diseases. and i'd guess that young people, especially geeks being picked on at school, would be particularly vulnerable to depression.
now different people might react to depression differently. some of the more extreme reactions could be:
* going overboard on drugs
* growing up and oppressing the hell out of other people
* committing suicicde
* shooting people
now the last is the most extreme reaction. perhaps a reason why mainly white men seem to do this, could be that white men get depressed like all humans can, but they believe the hype about white men being superior, and that drives them that extra step into insanity where they think it would be a good idea take their rightful dominant place and murder people.
i dunno, just a theory.
people getting depressed suck. people overdosing on drugs or attempting suicide sucks. it'd be good if we could figure out why this is escalating. i have my own theories about centralised media and suppressed creativity.
people murdering other people sucks and i have no sympathy for it. but if we can figure out how people get that way, and maybe reduce it, then that would be good.
m.
one of the reasons linux has become very popular as a server is because samba makes it such a good replacement for winNTserver, even if people are running MSoffice apps on the desktop.
the way i understand it, com+ apps may depend on win2000 server - will samba be able to serve com+ objects to com+ apps like MSoffice 2000?
otherwise, this could be the excuse pointy haired bosses are looking for to install win2k server instead of samba/linux. i.e. upgrade all desktops to MSoffice 20000, which will then force all file servers to be running win2000 server.
MSoffice's market share is not currently threatened by open source apps much, so this could be a way for MS to use its app monopoly to bolster it's weakening server business.
bye bye exponential linux server growth...?
OK, Microsoft is trying to invent a market... crashing home appliances, it's gotta try to keep those shareholders happy by going for enourmous growth rates.
About the only way I rekcon they could succeed is by bundling a DNA probe in your toaster (for our safety of course) then doing a licensing deal with Monsanto so that the toaster GPFs on anything but Monsanto wheat bread. Or more importantly, the fridge complains of a virus and blue screens if you've got things in there without Monsanto genetic copyright signals.
Kind of off-topic, but meanwhile, I think I figured out how Microscoft might be planning to lock in NT. Even if they don't de-commoditize the internet, they still have a monopoly on Office. And Office 2000 hype is already coming out.
Part of Office is COM/DCOM stuff, which means bits of the app don't get installed on your drive until you click particular menus. And I reckon the only way this will work is if you have a DCOM server running on NT.
So who cares if samba runs rings around NT? Corporates are going to want Office 2000 because MS have lock-in on the deliberate file format mess that is Word. And if Office 2000's fancy new features depend on an NT server, maybe Linux won't get a look in.
Just something to watch out for I reckon.