About a year ago I designed a new email system. It was pretty kickass.
It was kind of a cross between usenet and standard email. When you "sent" an email, it was in reality uploaded to your message store (the idea of the inbox was removed). Then notifications were sent to each person that a message was in the To field. That meant that for instance you could edit messages after they were sent, you could bring people in on threaded conversations half way through preserving the threading and so on. It also meant the attachment limit was decided by the senders account, not the receivers. Want to send a 200mb video to your hotmail using friend? No problem.
One of the features of this system was that key signing was built in from the start. That meant, you could opt to trust certain "roots", probably international ISPs. If you wanted to setup a newMail server, you'd have to get your hosting ISP to sign it for you, probably requiring a contract to be signed saying you'd shut down any abusive accounts etc.
Mailing lists were dealt with specially, I've never been happy with the way they currently work.
Combined with send limits (how often do you email >100 people?), that meant that spam could be cut down quite significantly. In particular, because it could be shut off at the source, if a spammer did somehow manage to spam lots of people at once, all it'd take is one report and the email would magically disappear from peoples message stores, before they'd even seen it in some cases. If the spammers were running their own servers, revoking their certs would do a similar trick.
It wouldn't eliminate spam of course, that's not possible. Smart enough people will figure out ways around it. However, having accountability built in from the start would help curb the situation a lot.
Originally I was going to write the client as a commercial app, but make the protocols open (with a non-commercial free license available). However, I ended up working on autopackage instead, so I never got around to it. If somebody thinks it'd be cool, contact me and I'll fill you in.
Come on now, you guys didn't really think Microsoft was going to sit back and let someone else build a free implementation of their golden goose before it was even in wide acceptance, did you?
This doesn't fly. Wine has been around for over a decade, and short of trying to claim header files were copyrighted (yeah, right) they have not done anything to it. Why? Because there's nothing innovative there, and you can't patent or even copyright interfaces.
.NET is simply yet another set of APIs, just like Win32. We've been cloning Win32 for a very long time now, and believe me, if Microsoft could kill Wine they would. I've spoken to MS execs, and they are scared of Wine.
So, I don't understand all this Mono bashing. We need mono for two reasons:
1) Compatability. There will be.NET apps soon enough. We'll need to be able to run them.
2) Good tools for us.
Microsoft has nothing to gain by researching, developing, and standardizing a platform that could conceivably allow network AND desktop applications to run on non-Microsoft platforms.
Tough on them. They can't stop people replicating the platform, that is not legally possible.
Did Miguel et al just think Microsoft had learned the error of their ways?
No, but they do have a better understanding of the law than you do.
Probably an easy to use library that provides a simple interface to it. Oh, and some momentum wouldn't go amiss. Basically when somebody is convinced it's cool enough to sit down and integrate.
1) A way to write Windows/desktop apps sanely and
2) Competition to J2EE.
Number 1 is imo more important than 2 to MS. You may not know, but writing apps to the raw Win32 APIs is a recipe for insanity. I've done a bit of work on Wine (and hope to do more in future) and the braindamages of that API is incredible. So,.NET is a chance for them to clean things up a lot, for the large number of Windows only apps out there. All this web services garbage is just hype they're churning out.
People here on/. seem to miss this one. with Sun, MS, or another standard server OS based on *NIX you have to pay per-seat lincensing out the wazoo! UNLIMITED clients for an OS which is SUPPORTED is a phenominal deal.
You do?
Funny. I could have sworn Redhat or Debian didn't have per-seat licensing, and in fact Redhat is also commercially supported - they have a long track record in that business.
So, remind me again, why is this a good deal? What possible reason is there to go with an expensive solution that gives you sod all (except integration with Mac clients I guess) that Linux on a cheap Intel box does not?
I'm yet to be convinced people who bought Apple servers unless they have a lot of Apple clients also are doing so more because they think it looks good than any sound technical or financial reasons.
I haven't seen much discussion here about alternatiev economic systems. Most of the talk is about whether we have cash or electronic cards. Why not talk about more interesting stuff, like altering the way money itself works. Ever heard of demurrage? Neither had I until I read some stuff over at transaction.net - the concept of negative currency is a fascinating one.
If you've ever read the Mars trilogy, you'll probably have some idea of the kinds of things I'm thinking about: limitations on the size of corporations and so on. Basically Capitalism version 2 (or 3).
I've been wondering the same thing myself. Aren't there any better ideas in the past 3 decades?
Yes, of course, but not as many as you might think. Lots of people seem to miss this point, even Ritchie - when Stallman set out on the GNU project, his aim was not to build a gee-whiz cutting edge computer system, it was to produce free software that would be very useful to people. At the time, everybody used UNIX, so that's what it made sense to "make free". Also, there was a lot of experience with UNIX systems, and communications wasn't as good back then as it is now, so the modularity of UNIX meant the work could easily be split into various teams.
Even though the driving force was a desire for software freedom rather than cool features, Linux and the rest of the GNU system today have all kinds of stuff that wasn't in the old UNIX systems. In fact, stuff like/proc was stolen directly from Plan9.
Note that some of the ideas that might sound good at first, have been tried, and basically don't work, or don't work as well as you might expect. The microkernel for instance. The Hurd is of course a microkernel based system, yet we all use Linux. Why? Because it was there, and it was pretty good. And now highly modular monolithic kernels have many of the advantages of microkernels, and microkernels have steadily increased in size as performance issues weighed in.
Ditto for a lot of other ideas that seemed good at the time, but actually perhaps weren't. AppFolders for instance (my pet one):)
Because, in the real world, people use computers to get things done. They're not used to make a political statement or fight for human rights in Burma. They're tools, not toys.
I guess you'd think that the boycott of Nestle baby milk, or not buying cosmetics products tested on animals, or not buying CDs because there the RIAA do things you don't like - I guess you'd think all those things are stupid yes?
Because really people will make political statements about things all the time. Most stuff is political, even if you don't realise it. Politics is just one facet of the interactions of humans. So, if people wish to take the piss out of Microsoft because they've done bad things, let them.
use a desktop computer and desktop software that actually works - to be productive instead of to feel technically and morally "superior" (whatever that means).
I'm under the distinct impression that believing you are the paragon of rationality who would never even conceive of using anything but the best tool for the job (in your opinion) actually makes you feel morally and technically superior.
Just because some people consider more than one factor when choosing a product, doesn't mean they're wrong, it just means they have different priorities to you.
Who says it has to work with other companies browsers?
Uh, we do. As in, society does, in much the same way that we demand our kitchen appliances use the type of electricty that is most common in the country. In fact, the govt enforce such standards, partly to stop an electricty company trying to force its way into the market for cookers by giving you free "special" electricity that only works with its products.
Why can't I refuse to hire somebody because they are black, or because they drive a Fiat? Because that'd be unfair discrimination, and it'd be illegal. I don't see why it should be different for products. Clearly it's very easy to make it work OK in Opera, just remove the browser sniffing code.
When big companies pull tricks like this, everybody loses. The web becomes more fragmented, and some idiots might look at such behaviour and think that it's actually ok to do something similar.
I've seen this with MSDN I think, or maybe it was TechNet. Anyway, it's not just MSN that do this, I've seen content shifted too far to the left in Gecko based browsers in deeply buried pages. Very irritating. I never bothered looking to see if it was deliberate sabotage, but this sounds too close to be coincidence.
Are you lying? Have you last used KDE in 1997?
You are wrong.
The (configurable) throbber will quit no matter what happens after 30 seconds.
Not really, I'm just inaccurate. I realised about a minute after I posted it that my wording was sloppy there.
I have yet to find an application that does not "SUPPORT" KDE startup notification.
That's because it gave you startup notification regardless of whether the app "supported" it or not. Hence for many apps, the throbber would hang there until it was killed by a timeout (yes, 30 seconds, i remember now).
Support in this sense means that the app signals the startup notifier when it's done loading (the notification is also cancelled when the first window opens, but that's unreliable iirc). You add the line StartupNotify=true to the.desktop file to signal support.
Are you telling me that the GNOME startup notification will work only with GNOME apps and that it is more "robust and useful" because of this? I can't believe it.
No. Again, sloppy wording, the startup notification is a freedesktop.org spec that both KDE and GNOME are using (or will be using), as well as 3rd parties. That means that eventually hopefully all apps will support it. In fact, any app linked with GTK2.2 gets supports for it semi-automatically I think.
Bit of trivia: so far there have been > 250k hits in the screenshots dir!
They should have taken this month to implement a better file dialog.
Well, a new file dialog is planned for GTK2.4, and is developing in libegg CVS.
Note a few things:
1) GTK is a separate project from GNOME. That means, it's not simply something the gnome team have forgotten to get around to, it has to fit into the gtk feature cycle, it has to go through usability reviews, it has to be tested etc. New APIs will be needed probably. Then it has to be integrated with standardised mime typing and so on. There's a lot more work than you might think.
2) For GNOME2.2, gtk2.2 was released in time, great, fab. That probably won't be the case for gtk2.4, for reasons I won't go into, but basically the release schedules just don't mesh. However, GTK2.4 is backwards compatable with 2.2, so when it does come out, you just drop it in - all the apps get the new dialog.
Re:Woohoo! More Gnome than you can shake a stick a
on
Gnome 2.2 Released
·
· Score: 2, Informative
Does anyone know whether the KDE equivalent is a past attempt they are talking about?
Yes. The KDE system was a good first attempt, but suffered from not having toolkit/desktop integration as well as it could have. If an app didn't support it, or crashed on startup, the throbber would hang indefinately. This system has toolkit/X integration, and is only activated for programs that flag themselves as supporting it, so it's more robust and useful.
Most users don't give a damn about the philosophy behind the software.
They don't now, but that can be changed.
All Linux developers or contributors were once users of something else, mostly Windows. Yet, many get extremely passionate about the philosophy of the software. Are some people hardwired to be programmers, hardwired to care about software philosophy?
I don't think so, I think the ideas RMS had operate at a more human level. They are, essentially, ideas about sharing and co-operation. That's a recurring theme in Stallmans writings and can be found in the GNU manifesto document, and it trancends sofware.
Most users don't give a damn about software licensing today, because most don't know about it. It sounds boring, and mostly is. But, so far I have yet to meet somebody who isn't fascinated by the idea of hundreds of thousands, possibly millions of people from all over the world working together on such an enormous project, sharing their work with others, and ensuring it can always be shared.
That's why RMS opposes the open source movement. It's not because he's a fantatic (though i guess it depends on your definition of the term). It's because it reduces his core ideas of sharing and long term benefit for society to ones of "you can get free software" - he sees this as totally missing the point. RMS was a programmer, so today the GPL is for software, but it could have been different.
So basically, I think it's quite possible that in future, nearly all computer users will understand the importance of the freedom part of free software, just as nearly all people today understand the importance of a democratically elected government. Of course, it's up to us to make sure people don't think of it as "freeware" (i have met people who think that). When people do get confused, I normally point out how the Windows freeware community never built anything much more than a few freebie utilities, certainly nothing on the scale of Linux. For managers, I point out that the free as in cost is just a side effect of the real benefits - the flexibility, and the lack of vendor lockin - the whole freedom thang. We all should do that.
In general, selling Linux because it's free or cheap is ulimately long term harmful, it just plays into the hands of Microsoft, and makes us forget the true motivations behind the whole thing in the first place.
If you look at the desktop experiences of one advanced user who isn't a developer [glowingplate.com], I think it's safe to say that this is an ongoing problem.
I just read this. I assume you wrote it? You're quite right in that it'd get modded down on Slashdot, and that's because it's mostly redundant. Let's take a few of the points shall we?
To be able to get the foot in the door, it should at least install easily on whatever piece of dogshit machine you throw it at. There are distros which run on a 386SX with 2 megs of RAM (http://www.superant.com/smalllinux/). Let's see that as the baseline to get a running kernel.
So you want this OS, that can run office suites capable of doing polynomial regression to run on anything? Not going to happen. Nobody cares what kind of dogshit hardware you might throw at it, if you want a desktop, most people know they should use a desktop machine.
In Red Hat 7.1 - not that old - there's no support for my mouse's scroll wheel by default. I don't care the reason, scroll wheel mice have been popular since 1998. Four years is a lifetime in Internet time, even with a recession.
If you insist on pointlessly bitching because some wierd mouse doesn't use one of the 2 or 3 mainstream mousewheel protocols, then you should be doing it, with helpful technical facts, in bugzilla. Using a distro that is in fact quite old won't help, update it and try again. Or, sit back and know that while you are writing rants, somebody else is probably trying to fix it.
Xine is arguably the best multimedia player for *nix, but it doesn't have something as simple as a repeat button, from what I can tell.
Oh me oh my. And QuickTime for Windows doesn't have a fullscreen mode, despite being the only program that can now display most trailors out there. I guess Windows is a failure on the desktop.
Again, these kind of odd feature requests belong in a bug tracker, not in a rant.
Who cares. I am the end user, and that's what the end user wants
No, it's what you, an individual wants. Clearly, not enough people actually want or need such a feature to make it worthwhile implementing. Please don't choose ridiculous features with practically no use cases (you can't even give any yourself) and then state that it doesn't matter, because you are an End User.
instead, we should be looking to emulate the styles of popular GUIs - MacOS X and Windows. Clean, conservative, efficient and neutral.
First point, who is this "we" white man? You have said yourself you don't actually code. Being bad at it is a pretty poor excuse, you could learn. Or you could write docs. Or do bug triage. Or artwork. You could do something, but as you've failed to mention it if you do, I can't see where this 'we' business comes from.
Second point, quite why you think MacOS or Windows are "conserative" I don't know. Glowing buttons? Teletubbies window decorations? Compared to this GNOME is a haven of simplicity and understatement.
Microsoft's Outlook has it, so it's a feature that end users are used to. Qualcomm's Eudora has it. I don't care about the technical reasons why it has not been implemented, or why KMail's spell checker sucks as much as it does.
I was going to give you a link to the KDE bugzilla, but realised that the last thing the KDE developers need is you spamming bugs with "I don't care why it's not done, I just demand it" style comments. Grow up. This whole thing reads like one big troll.
It's hard to imagine pure computer geeks being as cliquish and superficial as 14-year-old girls in a schoolyard, yet I know when I copy this to a comment form in Slashdot, I'll be modded down.
Yes, quite right, that's the moderation system working. That whole page has practically nothing of value or interest on it, it's just somebody spouting off because things are not exactly as you want them. It's full of ridiculous insults, baseless assertion and pointless flames. In short, it's classic flamebait material, and I'm only replying to this because the above post is basically just more of the same, yet it's been modded up.
Additionally, humans have one thing robots cannot: Imagination.
So? For most stuff we do in space imagination is a curse, not a blessing. It's extremely boring, and there are millions of possible threats, that could kill you at any moment. Imagining them just makes it worse.
They have the ability to, in a pinch, come up with solutions to problems that no machine technically can. When they had to build a CO2 scrubber from spare parts on Apollo 13, do you think a robot with the same computational power available in those days could have done the same? Of course not.
Ignoring the fact that robots are a) expendable and b) much simpler to keep alive than humans, so such a scrubber would have been unnecessary, I think this is more an argument for improving AI than sending humans into space.
Obviously in "those days" robots couldn't have done it, but these days they probably could have done. And in terms of applications here on Earth, advances in robotics would probably pay for themselves.
Additionally, humans seeing an anomalous phenomena would be immediately intrigued by it
You're assuming they would see it and know it's anomalous. Humans can't process anywhere near the amount of raw data machines can - an anomalous nebulae would probably be more obvious to the computers than the pilots.
A robot would see known gasses, shrug because it's known, and ignore it, going on its way
If it's known, then presumably it's not anomalous.
And one other critical factor: Humans have a survival instinct. Robots do not. Humans, when threatened, can respond almost immediately. Robots cannot.
How does that work? Robots are expendable anyway, so it's not like they need highly tuned survival instincts, but the latencies for machines to react are so much better than humans that all things where human life may be at risk, especially in flight, are nowadays computer controlled (autopilots etc).
Sawmill is a very old name for Sawfish, which is probably one of the more configurable window managers around.
GNOME is a project to produce a cohesive desktop environment. What is such a beast?
Desktop manager programs, typically in kde/gnome a panel implementation, a program to control the desktop, panel applets (start menu etc), and a window manager. In Windows and MacOS, the window management is not performed by an user pluggable process, on Windows the app performs its own window management to some extent, on MacOS it's now a part of the display server.
A bunch of related tools and utilities you'd expect to find on a desktop. Text editors, calculators, character maps, audio players, file managers and so on
A developer platform: a set of APIs that the developer can use to take advantage of advanced desktop services like a VFS, integrated help, icon management, "recently used items", multimedia and so on. In GNOME, there are many such APIs, usually fairly discrete. Some of them are separate projects from GNOME, like GTK+ (which in turn provides the GLib, GObject, and Pango APIs). GTK+ as you are aware provides widget services to applications.
The projects do various other things needed for a nice desktop environment, like spec out usability guidelines and so on.
GNOME-Compliance is a thing of the past, and has been for years. I don't know where you're getting this information, but it's seriously out of date. In the early years, both desktops extended the x protocol to support window icons and such in incompatible ways. Eventually they standardised on the NET_WM extensions, and today there are only "standards compliant" apps and window managers.
Volunteer developers don't have a rats chance in hell of actually replicated all the software any given may want or need to the same quality. The GIMP has been around for years and is a damn fine program, but Photoshop still has it beat. If somebody has bought Photoshop, why do we have to say "sorry, that investment is dead meat, use the gimp instead"?
Lots of people have investments in software which is either higher quality or there simply is no native equivalent on Linux. So we need to be able to run Windows apps.
Actually, I suspect that the reason Red Hat aren't part of it is that they don't really care that much about Linux on the desktop.
I think it's more to do with not being in a hurry, and being realistic.
Linux on the desktop just isn't ready yet. I wish it were. That'd be sweet. But it's not. There are still fundamental issues with things like hardware support, Windows compatability (ntfs/wine), app packaging, dvd/multimedia support, consistancy, standardisation and so on. Not to mention that the desktop platforms (kde/gnome) themselves are still maturing.
In short, there's an absolute buttload of work to do first. The corporate desktop is a realistic short-medium term target. It's not ready yet, but I'd guess by the start of next year it will be. A lot of the current issues don't affect the corporate desktop - DVD playback? So what. Packaging? The administrator can install the apps. Hardware support will come once business customers start phoning up and asking for drivers. Redhat are busy working on app consistancy and app compatability issues.
So the corp desktop will happen first, and it'll still take years to gain acceptance, just like Linux on the server. Then maybe by the time we've gone through that, desktop Linux will be smoother than OS X, Wine will be perfect, hardware support will be 99% out of the box and we'll stand a chance of getting other kinds of desktop converted.
Redhat know this. Hence their general lack of enthusiasm for home desktop at the moment.
Slashdot Mirror
It was kind of a cross between usenet and standard email. When you "sent" an email, it was in reality uploaded to your message store (the idea of the inbox was removed). Then notifications were sent to each person that a message was in the To field. That meant that for instance you could edit messages after they were sent, you could bring people in on threaded conversations half way through preserving the threading and so on. It also meant the attachment limit was decided by the senders account, not the receivers. Want to send a 200mb video to your hotmail using friend? No problem.
One of the features of this system was that key signing was built in from the start. That meant, you could opt to trust certain "roots", probably international ISPs. If you wanted to setup a newMail server, you'd have to get your hosting ISP to sign it for you, probably requiring a contract to be signed saying you'd shut down any abusive accounts etc.
Mailing lists were dealt with specially, I've never been happy with the way they currently work.
Combined with send limits (how often do you email >100 people?), that meant that spam could be cut down quite significantly. In particular, because it could be shut off at the source, if a spammer did somehow manage to spam lots of people at once, all it'd take is one report and the email would magically disappear from peoples message stores, before they'd even seen it in some cases. If the spammers were running their own servers, revoking their certs would do a similar trick.
It wouldn't eliminate spam of course, that's not possible. Smart enough people will figure out ways around it. However, having accountability built in from the start would help curb the situation a lot.
Originally I was going to write the client as a commercial app, but make the protocols open (with a non-commercial free license available). However, I ended up working on autopackage instead, so I never got around to it. If somebody thinks it'd be cool, contact me and I'll fill you in.
I meant a programming interface, an API. So, you go ahead and enjoy the irony, while I do as well.
This doesn't fly. Wine has been around for over a decade, and short of trying to claim header files were copyrighted (yeah, right) they have not done anything to it. Why? Because there's nothing innovative there, and you can't patent or even copyright interfaces.
So, I don't understand all this Mono bashing. We need mono for two reasons:
1) Compatability. There will be .NET apps soon enough. We'll need to be able to run them.
2) Good tools for us.
Tough on them. They can't stop people replicating the platform, that is not legally possible.
No, but they do have a better understanding of the law than you do.
Probably an easy to use library that provides a simple interface to it. Oh, and some momentum wouldn't go amiss. Basically when somebody is convinced it's cool enough to sit down and integrate.
1) A way to write Windows/desktop apps sanely and
2) Competition to J2EE.
Number 1 is imo more important than 2 to MS. You may not know, but writing apps to the raw Win32 APIs is a recipe for insanity. I've done a bit of work on Wine (and hope to do more in future) and the braindamages of that API is incredible. So, .NET is a chance for them to clean things up a lot, for the large number of Windows only apps out there. All this web services garbage is just hype they're churning out.
Why do you think they haven't changed Windows to break Wine? Because they can't. That's not the way it works.
You do?
Funny. I could have sworn Redhat or Debian didn't have per-seat licensing, and in fact Redhat is also commercially supported - they have a long track record in that business.
So, remind me again, why is this a good deal? What possible reason is there to go with an expensive solution that gives you sod all (except integration with Mac clients I guess) that Linux on a cheap Intel box does not?
I'm yet to be convinced people who bought Apple servers unless they have a lot of Apple clients also are doing so more because they think it looks good than any sound technical or financial reasons.
I haven't seen much discussion here about alternatiev economic systems. Most of the talk is about whether we have cash or electronic cards. Why not talk about more interesting stuff, like altering the way money itself works. Ever heard of demurrage? Neither had I until I read some stuff over at transaction.net - the concept of negative currency is a fascinating one.
If you've ever read the Mars trilogy, you'll probably have some idea of the kinds of things I'm thinking about: limitations on the size of corporations and so on. Basically Capitalism version 2 (or 3).
Actually I could have sworn that FastTrack was the first P2P app with encryption, or even FreeNet. The summary was way off base there.
Yes, of course, but not as many as you might think. Lots of people seem to miss this point, even Ritchie - when Stallman set out on the GNU project, his aim was not to build a gee-whiz cutting edge computer system, it was to produce free software that would be very useful to people. At the time, everybody used UNIX, so that's what it made sense to "make free". Also, there was a lot of experience with UNIX systems, and communications wasn't as good back then as it is now, so the modularity of UNIX meant the work could easily be split into various teams.
Even though the driving force was a desire for software freedom rather than cool features, Linux and the rest of the GNU system today have all kinds of stuff that wasn't in the old UNIX systems. In fact, stuff like /proc was stolen directly from Plan9.
Note that some of the ideas that might sound good at first, have been tried, and basically don't work, or don't work as well as you might expect. The microkernel for instance. The Hurd is of course a microkernel based system, yet we all use Linux. Why? Because it was there, and it was pretty good. And now highly modular monolithic kernels have many of the advantages of microkernels, and microkernels have steadily increased in size as performance issues weighed in.
Ditto for a lot of other ideas that seemed good at the time, but actually perhaps weren't. AppFolders for instance (my pet one) :)
I guess you'd think that the boycott of Nestle baby milk, or not buying cosmetics products tested on animals, or not buying CDs because there the RIAA do things you don't like - I guess you'd think all those things are stupid yes?
Because really people will make political statements about things all the time. Most stuff is political, even if you don't realise it. Politics is just one facet of the interactions of humans. So, if people wish to take the piss out of Microsoft because they've done bad things, let them.
use a desktop computer and desktop software that actually works - to be productive instead of to feel technically and morally "superior" (whatever that means).
I'm under the distinct impression that believing you are the paragon of rationality who would never even conceive of using anything but the best tool for the job (in your opinion) actually makes you feel morally and technically superior.
Just because some people consider more than one factor when choosing a product, doesn't mean they're wrong, it just means they have different priorities to you.
Uh, we do. As in, society does, in much the same way that we demand our kitchen appliances use the type of electricty that is most common in the country. In fact, the govt enforce such standards, partly to stop an electricty company trying to force its way into the market for cookers by giving you free "special" electricity that only works with its products.
Why can't I refuse to hire somebody because they are black, or because they drive a Fiat? Because that'd be unfair discrimination, and it'd be illegal. I don't see why it should be different for products. Clearly it's very easy to make it work OK in Opera, just remove the browser sniffing code.
When big companies pull tricks like this, everybody loses. The web becomes more fragmented, and some idiots might look at such behaviour and think that it's actually ok to do something similar.
They do? I thought it made massive, massive losses even with all kinds of funny accounting tricks.
I've seen this with MSDN I think, or maybe it was TechNet. Anyway, it's not just MSN that do this, I've seen content shifted too far to the left in Gecko based browsers in deeply buried pages. Very irritating. I never bothered looking to see if it was deliberate sabotage, but this sounds too close to be coincidence.
Not really, I'm just inaccurate. I realised about a minute after I posted it that my wording was sloppy there.
I have yet to find an application that does not "SUPPORT" KDE startup notification.
That's because it gave you startup notification regardless of whether the app "supported" it or not. Hence for many apps, the throbber would hang there until it was killed by a timeout (yes, 30 seconds, i remember now).
Support in this sense means that the app signals the startup notifier when it's done loading (the notification is also cancelled when the first window opens, but that's unreliable iirc). You add the line StartupNotify=true to the .desktop file to signal support.
Are you telling me that the GNOME startup notification will work only with GNOME apps and that it is more "robust and useful" because of this? I can't believe it.
No. Again, sloppy wording, the startup notification is a freedesktop.org spec that both KDE and GNOME are using (or will be using), as well as 3rd parties. That means that eventually hopefully all apps will support it. In fact, any app linked with GTK2.2 gets supports for it semi-automatically I think.
Bit of trivia: so far there have been > 250k hits in the screenshots dir!
Well, a new file dialog is planned for GTK2.4, and is developing in libegg CVS.
Note a few things:
1) GTK is a separate project from GNOME. That means, it's not simply something the gnome team have forgotten to get around to, it has to fit into the gtk feature cycle, it has to go through usability reviews, it has to be tested etc. New APIs will be needed probably. Then it has to be integrated with standardised mime typing and so on. There's a lot more work than you might think.
2) For GNOME2.2, gtk2.2 was released in time, great, fab. That probably won't be the case for gtk2.4, for reasons I won't go into, but basically the release schedules just don't mesh. However, GTK2.4 is backwards compatable with 2.2, so when it does come out, you just drop it in - all the apps get the new dialog.
Here, from Havoc
Here, from jdub
There is a list for GTK2.4 on the website, but unfortunately I think GTK2.4 won't be ready in time for GNOME2.4, so 2.4 will still be based on GTK2.2
Yes. The KDE system was a good first attempt, but suffered from not having toolkit/desktop integration as well as it could have. If an app didn't support it, or crashed on startup, the throbber would hang indefinately. This system has toolkit/X integration, and is only activated for programs that flag themselves as supporting it, so it's more robust and useful.
Oh yes. Sums up the whole of free software, that does. I like it lots.
I guess I should put something relatively on topic. Oooh, I know, this'll make fellow slashbots happy:
YOU CAN NOW HAVE TRANSPARENT PANELS!
Yes, 'tis true. Transparent panels rock. There are a few minor glitches, not all applets support it yet, but everybody loves transparency.
They don't now, but that can be changed.
All Linux developers or contributors were once users of something else, mostly Windows. Yet, many get extremely passionate about the philosophy of the software. Are some people hardwired to be programmers, hardwired to care about software philosophy?
I don't think so, I think the ideas RMS had operate at a more human level. They are, essentially, ideas about sharing and co-operation. That's a recurring theme in Stallmans writings and can be found in the GNU manifesto document, and it trancends sofware.
Most users don't give a damn about software licensing today, because most don't know about it. It sounds boring, and mostly is. But, so far I have yet to meet somebody who isn't fascinated by the idea of hundreds of thousands, possibly millions of people from all over the world working together on such an enormous project, sharing their work with others, and ensuring it can always be shared.
That's why RMS opposes the open source movement. It's not because he's a fantatic (though i guess it depends on your definition of the term). It's because it reduces his core ideas of sharing and long term benefit for society to ones of "you can get free software" - he sees this as totally missing the point. RMS was a programmer, so today the GPL is for software, but it could have been different.
So basically, I think it's quite possible that in future, nearly all computer users will understand the importance of the freedom part of free software, just as nearly all people today understand the importance of a democratically elected government. Of course, it's up to us to make sure people don't think of it as "freeware" (i have met people who think that). When people do get confused, I normally point out how the Windows freeware community never built anything much more than a few freebie utilities, certainly nothing on the scale of Linux. For managers, I point out that the free as in cost is just a side effect of the real benefits - the flexibility, and the lack of vendor lockin - the whole freedom thang. We all should do that.
In general, selling Linux because it's free or cheap is ulimately long term harmful, it just plays into the hands of Microsoft, and makes us forget the true motivations behind the whole thing in the first place.
I just read this. I assume you wrote it? You're quite right in that it'd get modded down on Slashdot, and that's because it's mostly redundant. Let's take a few of the points shall we?
So you want this OS, that can run office suites capable of doing polynomial regression to run on anything? Not going to happen. Nobody cares what kind of dogshit hardware you might throw at it, if you want a desktop, most people know they should use a desktop machine.
If you insist on pointlessly bitching because some wierd mouse doesn't use one of the 2 or 3 mainstream mousewheel protocols, then you should be doing it, with helpful technical facts, in bugzilla. Using a distro that is in fact quite old won't help, update it and try again. Or, sit back and know that while you are writing rants, somebody else is probably trying to fix it.
Oh me oh my. And QuickTime for Windows doesn't have a fullscreen mode, despite being the only program that can now display most trailors out there. I guess Windows is a failure on the desktop.
Again, these kind of odd feature requests belong in a bug tracker, not in a rant.
No, it's what you, an individual wants. Clearly, not enough people actually want or need such a feature to make it worthwhile implementing. Please don't choose ridiculous features with practically no use cases (you can't even give any yourself) and then state that it doesn't matter, because you are an End User.
First point, who is this "we" white man? You have said yourself you don't actually code. Being bad at it is a pretty poor excuse, you could learn. Or you could write docs. Or do bug triage. Or artwork. You could do something, but as you've failed to mention it if you do, I can't see where this 'we' business comes from.
Second point, quite why you think MacOS or Windows are "conserative" I don't know. Glowing buttons? Teletubbies window decorations? Compared to this GNOME is a haven of simplicity and understatement.
I was going to give you a link to the KDE bugzilla, but realised that the last thing the KDE developers need is you spamming bugs with "I don't care why it's not done, I just demand it" style comments. Grow up. This whole thing reads like one big troll.
Yes, quite right, that's the moderation system working. That whole page has practically nothing of value or interest on it, it's just somebody spouting off because things are not exactly as you want them. It's full of ridiculous insults, baseless assertion and pointless flames. In short, it's classic flamebait material, and I'm only replying to this because the above post is basically just more of the same, yet it's been modded up.
So? For most stuff we do in space imagination is a curse, not a blessing. It's extremely boring, and there are millions of possible threats, that could kill you at any moment. Imagining them just makes it worse.
They have the ability to, in a pinch, come up with solutions to problems that no machine technically can. When they had to build a CO2 scrubber from spare parts on Apollo 13, do you think a robot with the same computational power available in those days could have done the same? Of course not.
Ignoring the fact that robots are a) expendable and b) much simpler to keep alive than humans, so such a scrubber would have been unnecessary, I think this is more an argument for improving AI than sending humans into space.
Obviously in "those days" robots couldn't have done it, but these days they probably could have done. And in terms of applications here on Earth, advances in robotics would probably pay for themselves.
Additionally, humans seeing an anomalous phenomena would be immediately intrigued by it
You're assuming they would see it and know it's anomalous. Humans can't process anywhere near the amount of raw data machines can - an anomalous nebulae would probably be more obvious to the computers than the pilots.
A robot would see known gasses, shrug because it's known, and ignore it, going on its way
If it's known, then presumably it's not anomalous.
And one other critical factor: Humans have a survival instinct. Robots do not. Humans, when threatened, can respond almost immediately. Robots cannot.
How does that work? Robots are expendable anyway, so it's not like they need highly tuned survival instincts, but the latencies for machines to react are so much better than humans that all things where human life may be at risk, especially in flight, are nowadays computer controlled (autopilots etc).
GNOME is a project to produce a cohesive desktop environment. What is such a beast?
The projects do various other things needed for a nice desktop environment, like spec out usability guidelines and so on.
GNOME-Compliance is a thing of the past, and has been for years. I don't know where you're getting this information, but it's seriously out of date. In the early years, both desktops extended the x protocol to support window icons and such in incompatible ways. Eventually they standardised on the NET_WM extensions, and today there are only "standards compliant" apps and window managers.
Lots of people have investments in software which is either higher quality or there simply is no native equivalent on Linux. So we need to be able to run Windows apps.
I think it's more to do with not being in a hurry, and being realistic.
Linux on the desktop just isn't ready yet. I wish it were. That'd be sweet. But it's not. There are still fundamental issues with things like hardware support, Windows compatability (ntfs/wine), app packaging, dvd/multimedia support, consistancy, standardisation and so on. Not to mention that the desktop platforms (kde/gnome) themselves are still maturing.
In short, there's an absolute buttload of work to do first. The corporate desktop is a realistic short-medium term target. It's not ready yet, but I'd guess by the start of next year it will be. A lot of the current issues don't affect the corporate desktop - DVD playback? So what. Packaging? The administrator can install the apps. Hardware support will come once business customers start phoning up and asking for drivers. Redhat are busy working on app consistancy and app compatability issues.
So the corp desktop will happen first, and it'll still take years to gain acceptance, just like Linux on the server. Then maybe by the time we've gone through that, desktop Linux will be smoother than OS X, Wine will be perfect, hardware support will be 99% out of the box and we'll stand a chance of getting other kinds of desktop converted.
Redhat know this. Hence their general lack of enthusiasm for home desktop at the moment.