Since there is a cache on each node that discards data based on a LRU algoritm, is there a way to set aside a "static" piece of that cache that the user can fill in, to act as a publishing source. So if people want something to stay alive, they just transfer it from the LRU cache to the static storage, and that's that.
When other nodes request something, they still wouldn't know if it comes from the static part or the cache. But this way unpopular would never HAVE to disseappear from the network, purely a matter for the node operator.
I wonder how many of the people moaning about Netscape being released too soon actually took the time to install it. I'm typing this in NS 6.0 on win2k, and i have to say i was expecting way worse. I'm quite satisfied with the speed. I haven't crashed yet. And i'm pleasantly surprised with some little details (smooth install - checks all the settings from IE, didn't even have to fill in my password in slashdot). Yes, there are bugs. But nothing that annoys me yet.
All in all, i get the feeling that there's NO way that Netscape can "do good". As pointed out before: or people complain they shipped too late (people were saying they wanted an interim version that saved them from 2 years of netscape 4), or they complain that they ship too soon (not all the(important) bugs closed)
I know what an important argument is: if they ship a non-compliant browser, it's another exception that webdesigners have to watch out for.
However, 2 things:
a) webdesigners have to code for exceptions in any case
b) this netscape version atleast provides a very decent subset of the official spec. if you code for it now, it won't crash and burn completely when viewed in the next version.
Basically I'm working a lot wih zips here, and event though me, my friends and collueges have really physically abused them at times I never even had as much as a bad cluster or ANY lost data, in the years i've been using them...
For me they are proven, not too slow (that is if you take ide or scsi: here most of the (beige) macs and pcs have scsi internal zips) and probably cheap by now. I don't know about the 250 variant, but for me, it's not worth it. 100 MB is plent for what i need, if we need more we write a CD.
One thing though: stay away from the JAZ... i can't begin to tell you how much horror stories i heard with these, esp. wrt reliability.
I don't know about things like orb and such, but if you go down that lane, you'd probably be willing to shell out for a CDRW.
go with zip. (you can also boot from them - recent motherboard bios support it for the ide, and with scsi it was never really a problem)
Like some other poste mentioned, ideally packages/programs should be self contained - check out MacOSX's solution for things like shared libraries and the like (I think there's an Arstechnica article about it that is pretty good)
But I'm afraid _that_ will never happen...
The second best thing would be a system that:
- has list of dependencies (like virtual packages in debian) for every distro, with the corresponding packages for that distro.
- describes the distro fs differences in a flexible way (do i dare mention xml?) - the Redhat, Suse, Debian, whacked-out-beyond-recognition structure...
if we go even more crazy:
- that includes the source, a standard way of making it, and repackaging it immediately to the native package format of your choice (distro)
- that can include source patches, that can be optionally applied (by user choice/distro description whatever)
- (really crazy) can even (if the appropriate build tools are installed, and the source supports it of course) automatically build packages for other platforms (MacOS, BeOS, Win32)
Can you imagine, one day saying "and now something completely different...",- and just making a new "distro description", downloading a bunch of next-gen source packages, and just rebuilding a fresh new distro overnight -> blammo
It would definately make the lives easier for the poor slobs that have to maintain binary packages for different platforms/distros.
I'm positive i'm oversimplifying things, but universal source packages would be VERY helpful, esp for the open source community. Universal binary packages would be a good start too though.
What _isn't_ gonna happen is that distros converge to a specific package manager... But we could try to put a layer in between.
Oh well... in the meantime i'll just pretend my good ole debian setup is the only distro... They're actually doing good stuff, with apt and debconf, and a bunch of tools to make it easier for debian-maintainers. Too bad it's of little use with respect to interoperability with other distributions.
Anyone else notice just how good IE 5 for mac is? (as opposed to Netscape 4 and IE 5 for PC)
It's doing pretty good with standards, and is small and light-weight (takes like 4mb of memory - 7 mb install without java)
Surprisingly enough it even trounced the win IE in more than one way, on a platform they don't control; What's even more surprising were the reports i heard later on that the mac IE development team was dissolved - who knows what the reasoning behind that was. (check out http://www.mozillazine.org/talkback.html?article=1 403) (Making the mac a better browsing platform than windows?! what were they thinking:) )
IIRC I thought what was really revolutionary about Berlin was that the _server_ had more or less control over where and how to put things. Programs would just say what to display and the server decides on the best way to display it, basically allowing thesame programs to be run on a variety/several kinds of displays (Palmtop, lowres LCD, highres CRT, etc etc) Don't like the way programs look and are arranged? get another server(plugin?)
Basically you could use thesame program on Pixelscreen(tm), or in vectorspace (postscript) or even in 3D (opengl), and each time it would probably use features available to the best of its abilities, no lowest common denominator stuff. (i guess they could even make a text based server or something:)
This isn't another XServeralike with the windowmanager layered on a toolkit layered on X etc. It's conceptually different.
Of course i might be totally wrong cause it's a real long time ago that i actually looked into it.
This is probably more a GTK question, but here goes:
There are currently a couple of projects that intend to bring Display Postscript to Linux (DPS/X and gnusteps display ghostscript). IIRC Gtk+ 1.4 will be more easily portable to other platforms.
The question is: would it be possible to extend GTK+ and Gnome to base it on the vector model of DPS, a la Quartz on MacOSX? Maybe it would be useful in the areas where Gnome could use some more work? (printing model, fonts, anti-aliasing)
Also, a bit related, do you see Gnome being a X-only thing, or do you imagine it being ported to for instance embedded applications, experimental environments (Berlin anyone?) or even (gasp) windows?
Right now, i see developers struggle trying to find "reasonable" settings, and package maintainers have to always play catch-up.
what if:
- there were a couple of files (using xml of course) with mappings for the current and other distros/platforms (a debian mapping, a redhat mapping, BSD, Solaris,...)
- a standard way of asking configuration questions, independent of the installer program ala debconf, with resonable standard settings of course.
- some way to handle dependencies
... so you can install your own wacky distro in for instance this way:
* download all the universal source packages you want/need.
* make a new "mapping file" with all the settings (fs hierarchy, whatever...) you want for your new distro
* let the universal source installer rip, on all your downloaded packages - pointing to some freshly formatted and mounted drive.
* let it simmer overnight, while it compiles, configures and installs all the source packages on the drive.
* blammo, your own super-optimized/personalized distro in the convenience of your own home.
On a side note... i really like the way NextStep bundles relevant files together in a ".app" dir. It hides the files, but keeps them accessible and together.
maybe you could have something similar for source ".src" bundles. and compile them with one command to an ".app" bundle or something.
What's sure is that anything is better than what we have now. If we had a good way of handling source packages, the whole problem of standardizing linux would become a moot point, because you could just write a little mapping for whatever distro you want. Once you have that mapping you hardly need to maintain seperate packages for the distro of your choice.
I'm currently looking into setting up a rather complex network. This is going to be implemented in a month or two-three.
Here's the situation:
We have a number of people/offices, that all have cable access or adsl access (6 cable 1 adsl) The cable provider blocks all incoming ports 1000, and forces proxy use on port 80. A dynamic IP is gotten through dhcp. The ADSL provider is pretty much thesame.
In at least 3 of these places, wireless networking is going to be used. Since we're in a mixed environment (Windows, Mac, Linux) we have no option but to use Lucent's cards. (do we?)
In each of these 7 places there's a lowly 486 box doing ip masquerading.
What I want to do:
- have the ability to surf the net in each of those places. - be able to "see" the pc's on the lans of the distributed locations, as though they were on thesame lan. (VPN) - use crypto to secure the VPN. - use a transparent proxy setup to mask the forced use of a proxy for webtraffic. - have the ability to use the laptops in each of the locations transparently. (3 locations with wireless) - i'm hesitating to use dhcp on the lan/vpn: to avoid being stranded if a node goes down, all the routers must deal out ip's to their own lans. They have to watch it not to step on each other's toes in the vpn though.
My questions: - you think all this masquerading, VPN forwarding, and encrypting is too much for lowly 486 boxes? - what about latencies? Does crypto + wireless + vpn increase the latency alot? - Apple's airport base station - how configurable is this from linux? - what about the limitations of the providers: how will the routers be able to find each other if they all have dynamic ip's. What if a node/router gets disconnected - how will it be reintegrated on the vpn?
- and finally: how can you maintain a bunch of routers like that? I'm using Debian potato for the masq'ed routers now, and i already have to spend too much time looking/maintaining them. Anyone has some good ideas/insights on how to manage all of them?
I think it must be possible, but it's not really a run of the mill configuration. We have hosting where we put our email/databases and stuff, but we don't have access to a name server - best we can get is a CNAME in our domain.
Once this is up and running, it will be fun though.
The only way i can think of to avoid "duplicates" is to gradually expand your brain with a digital/networked one... The moment you die you'd feel it (like being amputated a part of your brain) but by then your "consciousness" would have expanded over more than just your physical brain.
Kindof like making your brain bigger first, and then later on losing a (hopefully small part) of the brain. Hopefully by then your memories will have migrated to the digital part.
I agree with you totally. I like gnome and the fact that gtk+ is becoming so prevalent. Let me rehash what i said on gnome-news.
I've been thinking about something similar to Aqua, ie using some vector-based layer to do all the drawing. I think Apple did a great job.
I am a total newbie on gnome/gtk internals, but IIRC gtk+ is not tied to X (esp. with the upcoming version), seeing as there are X-less BeOS/win32 ports.
Would it be possible to write a backend to gtk that writes to postscript (or like mac, pdf). This could sit on a layer that displays pdf to opengl, pdf to ggi (speedier?), or pdf to X. Since you're working with postscript, I guess printing subsystems might be easier too.
I don't know if it's worth it. It would be easier to create vector based programs, since they could make use of the extra layer. Things like scaling and anti-aliasing would be smoother too. OTOH, you're adding another layer, which means really lowend hardware would have a harder time.
I have no idea how feasible this is. The ideal would be that all apps except the apps depending on the extra layer would keep working transparantly.
On another note: since Miguel is aiming for people that never used a computer, it might be interesting to hide files (i know a lot of users won't agree with me on this one, but i think that files you don't use just clutter screen real estate - as long as i can quickly reach them if i really want to it's ok by me) and also to use metadata (again i kinda like macosX's implementation with respect to metadata)
Could someone enlighten me if there already are similar things being done?
I see a lot of people saying something to the effect of "When will RDS be available in the states" (I thought they had it already a long time but hey).
RDS is pretty cool because you get extra digital info. But wait! DAB (Digital Audio Broadcasting) is already around the corner. DAB is already broadcasting in pilot projects in Europe (and probably in other places too) and has interference free "CD quality" (mpeg layer 2 200kbit/s) audio. Of course, there will be a host of other goodies (and not so goodies) like the ability to incorporate "pay-radio" etc.
The fun part is this: (taken from the dab faq)
What are 'additional services'? DAB was developed with the aim of improving radio reception. This is why audio transmission was at the forefront of the development process. But DAB, as a digital transmission system, can transmit other data as well as audio. In principle any type of information can be transmitted by DAB, provided simply that it is available in digital form and does not exceed the maximum available DAB data rate (approx. 1.7 Mbit/s). Examples of such additional services are still pictures accompanying radio programmes, digitalised traffic messages (Traffic Message Channel), electronic newspapers, software updates and even animated video. This process leads to 'multimedia broadcasting' in which all forms of information can be conveyed via the common transmission medium DAB. In this context DAB could be described as the 'cordless information highway'. In contrast to multimedia applications via TV/cable, DAB additional services can also be received in-car and with portable equipment.
I wonder how long it will take for pirates to tap into this flexible resource, full of possibilities:)
* the videocard: If you're only using 2d it's adequate but overkill. A G200 or G400 would likely give better (cleaner) 2D at high resolutions. If however you need opengl accelartion things are ALOT different. Nvidia's TNT driver support is horrible (not performant). I wouldn't count on them (don't want to be an open source zealot, but if the specs are kept closed, you have to always rely on them for drivers) 3DFX actually has the best support for 'other' OS's like BeOS, Mac, Linux,... And the drivers are all open. If you want better peformance/features you might want to go with matrox' g400 - crisp 2D and GLX is pretty performant already. The moment XFree 4.0 comes out there should be G400 drivers VERY soon after. bottom line: get a G400. - or a Voodoo4 when(/if:) they come out.
* cpu/chipset: i see you went for an smp chipset. that's ok. i really like the 840. Just a couple of things to think about: - test one out before you buy a bunch of them. You can't afford mobo flakiness. The 840 is very young and untested. Same goes for the (dual channel) rambus. Maybe you should consider a BX/GX board - you'll also save a bunch of money on memory (1 gig of rambus memory - money must be of no concern). Are these workstations mission critical? Think about that too. If you need uniprocessor boards i'd consider athlons (depending again what you think is most important: performance/cost or maturity) * scsi: I'm not sure if you need this. If you only use one or two harddrives you won't notice much performance gain. If you absolutely need to blow some money, by all means do. OTOH, if you intend to often use more than 3 harddrives/devices on the bus at thesame time, like in raid configurations or heavy duty workstation stuff, or you need to attach plenty of devices to your chain (scsi zip/cd/writer/dvd/scanner/harddrives/...) you might get good use out of scsi.
btw: are you sure the Adaptec ultra 160/m is already supported under linux??
I think the one thing one should consider is the long term effect on the genepool...
If 1) everybody is gonna go for thesame genes, the genepool will inevitably get smaller. Bad thing - variation is essential. Note that just eradicating common diseases would probably have thesame effect.
But if 2) scientists introduce mutations and variations that are as of yet uncommon, that would actually increase the genepool. Of course environmentalists will probably be screaming we're "playing god" again.
Fact is that's how the system works: mutation introduces variety, evolution makes/takes the best of it. Evolution on its own will not (always) result in a perfect solution.
(I heard there's this grasshopper that has nerves running down it's lower body and back up to it's upper body to control it's wings. That just makes it more prone to defects/less resistant to trauma. Apparently it had evolved from a species where the wings were lower or something - the ones with the higher wings were just more successful over generations.)
OTOH one might argue if the rate of mutation is increased too much with respect to evolution, you might get mutation unfiltered by evolution (evolution filters out the more successful properties)
What I'm saying is that both should still be there: We can't just pretend were "evolution" by scrapping a bunch of genes we don't like, without increasing mutation (-> poorer genepool, less resilient) OTOH we can't just be "mutation", cause the bad stuff might not get filtered out so quickly.
As always there should be a balance between the two. On top of that we'll probably do a better job at mutation (since natural mutation is mostly random), but we'll also probably do a worse job at evolution; (we might scrap some genes that would have proven to be benificial over generations)
What would be cool would be to have a bootable CD that starts up Suse, auto probes for hardware, and runs Q3. I know that auto probes for hardware may be a bit tricky, but people can always install it themselves if they want to play it from harddrive/have wacky hardware configurations. I thought about this before. You put games on CD's in any case. You can boot from CD's. Linux is free. Hardware support for Linux is excellent. Why not just boot and play from CD.
Slashdot Mirror
Since there is a cache on each node that discards data based on a LRU algoritm, is there a way to set aside a "static" piece of that cache that the user can fill in, to act as a publishing source. So if people want something to stay alive, they just transfer it from the LRU cache to the static storage, and that's that.
When other nodes request something, they still wouldn't know if it comes from the static part or the cache. But this way unpopular would never HAVE to disseappear from the network, purely a matter for the node operator.
Okay,
I wonder how many of the people moaning about Netscape being released too soon actually took the time to install it. I'm typing this in NS 6.0 on win2k, and i have to say i was expecting way worse. I'm quite satisfied with the speed. I haven't crashed yet. And i'm pleasantly surprised with some little details (smooth install - checks all the settings from IE, didn't even have to fill in my password in slashdot). Yes, there are bugs. But nothing that annoys me yet.
All in all, i get the feeling that there's NO way that Netscape can "do good". As pointed out before: or people complain they shipped too late (people were saying they wanted an interim version that saved them from 2 years of netscape 4), or they complain that they ship too soon (not all the(important) bugs closed)
I know what an important argument is: if they ship a non-compliant browser, it's another exception that webdesigners have to watch out for.
However, 2 things:
a) webdesigners have to code for exceptions in any case
b) this netscape version atleast provides a very decent subset of the official spec. if you code for it now, it won't crash and burn completely when viewed in the next version.
Basically I'm working a lot wih zips here, and event though me, my friends and collueges have really physically abused them at times I never even had as much as a bad cluster or ANY lost data, in the years i've been using them...
For me they are proven, not too slow (that is if you take ide or scsi: here most of the (beige) macs and pcs have scsi internal zips) and probably cheap by now. I don't know about the 250 variant, but for me, it's not worth it. 100 MB is plent for what i need, if we need more we write a CD.
One thing though: stay away from the JAZ... i can't begin to tell you how much horror stories i heard with these, esp. wrt reliability.
I don't know about things like orb and such, but if you go down that lane, you'd probably be willing to shell out for a CDRW.
go with zip. (you can also boot from them - recent motherboard bios support it for the ide, and with scsi it was never really a problem)
Ugh. Packages. Hell.
Like some other poste mentioned, ideally packages/programs should be self contained - check out MacOSX's solution for things like shared libraries and the like (I think there's an Arstechnica article about it that is pretty good)
But I'm afraid _that_ will never happen...
The second best thing would be a system that:
- has list of dependencies (like virtual packages in debian) for every distro, with the corresponding packages for that distro.
- describes the distro fs differences in a flexible way (do i dare mention xml?) - the Redhat, Suse, Debian, whacked-out-beyond-recognition structure...
if we go even more crazy:
- that includes the source, a standard way of making it, and repackaging it immediately to the native package format of your choice (distro)
- that can include source patches, that can be optionally applied (by user choice/distro description whatever)
- (really crazy) can even (if the appropriate build tools are installed, and the source supports it of course) automatically build packages for other platforms (MacOS, BeOS, Win32)
Can you imagine, one day saying "and now something completely different...",- and just making a new "distro description", downloading a bunch of next-gen source packages, and just rebuilding a fresh new distro overnight -> blammo
It would definately make the lives easier for the poor slobs that have to maintain binary packages for different platforms/distros.
I'm positive i'm oversimplifying things, but universal source packages would be VERY helpful, esp for the open source community. Universal binary packages would be a good start too though.
What _isn't_ gonna happen is that distros converge to a specific package manager... But we could try to put a layer in between.
Oh well... in the meantime i'll just pretend my good ole debian setup is the only distro... They're actually doing good stuff, with apt and debconf, and a bunch of tools to make it easier for debian-maintainers. Too bad it's of little use with respect to interoperability with other distributions.
Anyone else notice just how good IE 5 for mac is? (as opposed to Netscape 4 and IE 5 for PC)
1 403) :) )
It's doing pretty good with standards, and is small and light-weight (takes like 4mb of memory - 7 mb install without java)
Surprisingly enough it even trounced the win IE in more than one way, on a platform they don't control; What's even more surprising were the reports i heard later on that the mac IE development team was dissolved - who knows what the reasoning behind that was. (check out http://www.mozillazine.org/talkback.html?article=
(Making the mac a better browsing platform than windows?! what were they thinking
... or maybe i don't
IIRC I thought what was really revolutionary about Berlin was that the _server_ had more or less control over where and how to put things. Programs would just say what to display and the server decides on the best way to display it, basically allowing thesame programs to be run on a variety/several kinds of displays (Palmtop, lowres LCD, highres CRT, etc etc) Don't like the way programs look and are arranged? get another server(plugin?)
Basically you could use thesame program on Pixelscreen(tm), or in vectorspace (postscript) or even in 3D (opengl), and each time it would probably use features available to the best of its abilities, no lowest common denominator stuff. (i guess they could even make a text based server or something
This isn't another XServeralike with the windowmanager layered on a toolkit layered on X etc. It's conceptually different.
Of course i might be totally wrong cause it's a real long time ago that i actually looked into it.
This is probably more a GTK question, but here goes:
There are currently a couple of projects that intend to bring Display Postscript to Linux (DPS/X and gnusteps display ghostscript). IIRC Gtk+ 1.4 will be more easily portable to other platforms.
The question is: would it be possible to extend GTK+ and Gnome to base it on the vector model of DPS, a la Quartz on MacOSX? Maybe it would be useful in the areas where Gnome could use some more work? (printing model, fonts, anti-aliasing)
Also, a bit related, do you see Gnome being a X-only thing, or do you imagine it being ported to for instance embedded applications, experimental environments (Berlin anyone?) or even (gasp) windows?
I think this is the right thing to do...
...)
Right now, i see developers struggle trying to find "reasonable" settings, and package maintainers have to always play catch-up.
what if:
- there were a couple of files (using xml of course) with mappings for the current and other distros/platforms (a debian mapping, a redhat mapping, BSD, Solaris,
- a standard way of asking configuration questions, independent of the installer program ala debconf, with resonable standard settings of course.
- some way to handle dependencies
... so you can install your own wacky distro in for instance this way:
* download all the universal source packages you want/need.
* make a new "mapping file" with all the settings (fs hierarchy, whatever...) you want for your new distro
* let the universal source installer rip, on all your downloaded packages - pointing to some freshly formatted and mounted drive.
* let it simmer overnight, while it compiles, configures and installs all the source packages on the drive.
* blammo, your own super-optimized/personalized distro in the convenience of your own home.
On a side note... i really like the way NextStep bundles relevant files together in a ".app" dir. It hides the files, but keeps them accessible and together.
maybe you could have something similar for source ".src" bundles. and compile them with one command to an ".app" bundle or something.
What's sure is that anything is better than what we have now. If we had a good way of handling source packages, the whole problem of standardizing linux would become a moot point, because you could just write a little mapping for whatever distro you want. Once you have that mapping you hardly need to maintain seperate packages for the distro of your choice.
I'm currently looking into setting up a rather complex network. This is going to be implemented in a month or two-three.
Here's the situation:
We have a number of people/offices, that all have cable access or adsl access (6 cable 1 adsl) The cable provider blocks all incoming ports 1000, and forces proxy use on port 80. A dynamic IP is gotten through dhcp. The ADSL provider is pretty much thesame.
In at least 3 of these places, wireless networking is going to be used. Since we're in a mixed environment (Windows, Mac, Linux) we have no option but to use Lucent's cards. (do we?)
In each of these 7 places there's a lowly 486 box doing ip masquerading.
What I want to do:
- have the ability to surf the net in each of those places.
- be able to "see" the pc's on the lans of the distributed locations, as though they were on thesame lan. (VPN)
- use crypto to secure the VPN.
- use a transparent proxy setup to mask the forced use of a proxy for webtraffic.
- have the ability to use the laptops in each of the locations transparently. (3 locations with wireless)
- i'm hesitating to use dhcp on the lan/vpn: to avoid being stranded if a node goes down, all the routers must deal out ip's to their own lans. They have to watch it not to step on each other's toes in the vpn though.
My questions:
- you think all this masquerading, VPN forwarding, and encrypting is too much for lowly 486 boxes?
- what about latencies? Does crypto + wireless + vpn increase the latency alot?
- Apple's airport base station - how configurable is this from linux?
- what about the limitations of the providers: how will the routers be able to find each other if they all have dynamic ip's. What if a node/router gets disconnected - how will it be reintegrated on the vpn?
- and finally: how can you maintain a bunch of routers like that? I'm using Debian potato for the masq'ed routers now, and i already have to spend too much time looking/maintaining them. Anyone has some good ideas/insights on how to manage all of them?
I think it must be possible, but it's not really a run of the mill configuration. We have hosting where we put our email/databases and stuff, but we don't have access to a name server - best we can get is a CNAME in our domain.
Once this is up and running, it will be fun though.
The only way i can think of to avoid "duplicates" is to gradually expand your brain with a digital/networked one... The moment you die you'd feel it (like being amputated a part of your brain) but by then your "consciousness" would have expanded over more than just your physical brain.
Kindof like making your brain bigger first, and then later on losing a (hopefully small part) of the brain. Hopefully by then your memories will have migrated to the digital part.
I agree with you totally. I like gnome and the fact that gtk+ is becoming so prevalent. Let me rehash what i said on gnome-news.
I've been thinking about something similar to Aqua, ie using some vector-based layer to do all the drawing. I think Apple did a great job.
I am a total newbie on gnome/gtk internals, but IIRC gtk+ is not tied to X (esp. with the upcoming version), seeing as there are X-less BeOS/win32 ports.
Would it be possible to write a backend to gtk that writes to postscript (or like mac, pdf). This could sit on a layer that displays pdf to opengl, pdf to ggi (speedier?), or pdf to X. Since you're working with postscript, I guess printing subsystems might be easier too.
I don't know if it's worth it. It would be easier to create vector based programs, since they could make use of the extra layer. Things like scaling and anti-aliasing would be smoother too. OTOH, you're adding another layer, which means really lowend hardware would have a harder time.
I have no idea how feasible this is. The ideal would be that all apps except the apps depending on the extra layer would keep working transparantly.
On another note: since Miguel is aiming for people that never used a computer, it might be interesting to hide files (i know a lot of users won't agree with me on this one, but i think that files you don't use just clutter screen real estate - as long as i can quickly reach them if i really want to it's ok by me) and also to use metadata (again i kinda like macosX's implementation with respect to metadata)
Could someone enlighten me if there already are similar things being done?
I see a lot of people saying something to the effect of "When will RDS be available in the states" (I thought they had it already a long time but hey).
:)
RDS is pretty cool because you get extra digital info. But wait! DAB (Digital Audio Broadcasting) is already around the corner. DAB is already broadcasting in pilot projects in Europe (and probably in other places too) and has interference free "CD quality" (mpeg layer 2 200kbit/s) audio. Of course, there will be a host of other goodies (and not so goodies) like the ability to incorporate "pay-radio" etc.
The fun part is this: (taken from the dab faq)
What are 'additional services'?
DAB was developed with the aim of improving radio reception. This is why audio transmission was at the forefront of the development process. But DAB, as a digital transmission system, can transmit other data as well as audio. In principle any type of information can be transmitted by DAB, provided simply that it is available in digital form and does not exceed the maximum available DAB data rate (approx. 1.7 Mbit/s). Examples of such additional services are still pictures accompanying radio programmes, digitalised traffic messages (Traffic Message Channel), electronic newspapers, software updates and even animated video. This process leads to 'multimedia broadcasting' in which all forms of information can be conveyed via the common transmission medium DAB. In this context DAB could be described as the 'cordless information highway'. In contrast to multimedia applications via TV/cable, DAB additional services can also be received in-car and with portable equipment.
I wonder how long it will take for pirates to tap into this flexible resource, full of possibilities
look at http://www.worlddab.org for more
Hi. I have a couple of remarks. Beginning with ...
... And the drivers are all open. If you want better peformance/features you might want to go with matrox' g400 - crisp 2D and GLX is pretty performant already. The moment XFree 4.0 comes out there should be G400 drivers VERY soon after.
* the videocard: If you're only using 2d it's adequate but overkill. A G200 or G400 would likely give better (cleaner) 2D at high resolutions. If however you need opengl accelartion things are ALOT different. Nvidia's TNT driver support is horrible (not performant). I wouldn't count on them (don't want to be an open source zealot, but if the specs are kept closed, you have to always rely on them for drivers) 3DFX actually has the best support for 'other' OS's like BeOS, Mac, Linux,
bottom line: get a G400. - or a Voodoo4 when(/if:) they come out.
* cpu/chipset: i see you went for an smp chipset. that's ok. i really like the 840. Just a couple of things to think about:
- test one out before you buy a bunch of them. You can't afford mobo flakiness. The 840 is very young and untested. Same goes for the (dual channel) rambus. Maybe you should consider a BX/GX board - you'll also save a bunch of money on memory (1 gig of rambus memory - money must be of no concern). Are these workstations mission critical? Think about that too. If you need uniprocessor boards i'd consider athlons (depending again what you think is most important: performance/cost or maturity) * scsi: I'm not sure if you need this. If you only use one or two harddrives you won't notice much performance gain. If you absolutely need to blow some money, by all means do. OTOH, if you intend to often use more than 3 harddrives/devices on the bus at thesame time, like in raid configurations or heavy duty workstation stuff, or you need to attach plenty of devices to your chain (scsi zip/cd/writer/dvd/scanner/harddrives/...) you might get good use out of scsi.
btw: are you sure the Adaptec ultra 160/m is already supported under linux??
I think the one thing one should consider is the long term effect on the genepool...
If 1) everybody is gonna go for thesame genes, the genepool will inevitably get smaller. Bad thing - variation is essential. Note that just eradicating common diseases would probably have thesame effect.
But if 2) scientists introduce mutations and variations that are as of yet uncommon, that would actually increase the genepool. Of course environmentalists will probably be screaming we're "playing god" again.
Fact is that's how the system works: mutation introduces variety, evolution makes/takes the best of it. Evolution on its own will not (always) result in a perfect solution.
(I heard there's this grasshopper that has nerves running down it's lower body and back up to it's upper body to control it's wings. That just makes it more prone to defects/less resistant to trauma. Apparently it had evolved from a species where the wings were lower or something - the ones with the higher wings were just more successful over generations.)
OTOH one might argue if the rate of mutation is increased too much with respect to evolution, you might get mutation unfiltered by evolution (evolution filters out the more successful properties)
What I'm saying is that both should still be there: We can't just pretend were "evolution" by scrapping a bunch of genes we don't like, without increasing mutation (-> poorer genepool, less resilient)
OTOH we can't just be "mutation", cause the bad stuff might not get filtered out so quickly.
As always there should be a balance between the two. On top of that we'll probably do a better job at mutation (since natural mutation is mostly random), but we'll also probably do a worse job at evolution; (we might scrap some genes that would have proven to be benificial over generations)
Oh well, bit of a long rant.
What would be cool would be to have a bootable CD that starts up Suse, auto probes for hardware, and runs Q3. I know that auto probes for hardware may be a bit tricky, but people can always install it themselves if they want to play it from harddrive/have wacky hardware configurations. I thought about this before. You put games on CD's in any case. You can boot from CD's. Linux is free. Hardware support for Linux is excellent. Why not just boot and play from CD.