If China government is trying to control what Chinese people can do on Internet...what's the point of having a high-performance router if they can't download pr0n movies? Linux distros?
There's nothing wrong with the file abstractiong but, how it's used
With soundcards, you should be able to do thigns like cp sng.mp3/dev/soundcard and it should work. Of course for that you should have a mp3 player behind that file (and NOT in the kernel, but as a userspace "service").
In Linux you have "io priorities" (check CFQ) and you will be able to give portions of CPU usage to a give group of users in the near future. I don't think it's not possible to do such things in unix, ie: its not that "Unix is wrong" is that "nobody has built it for unix". Just look at mac os x, it _is_ unix and it is far from being a "antithesis of a practical user interfaces"
Having been a windows users, I hate how this stuff is handled in windows
There's a reason why directories are separated in unix,/var for "data which changes often": you can put it in the fatest disk, you can't do the same (or it's much harder) in windows.
/etc itself is ok for me except for the different configuration formats. It allows me to do a "tar c/etc > configuration-backup.tar" . Try to do the same in Windows, where each program puts its configuration file in its directory and you don't know what file it is (or even worse, the registry..)
Seriously, the way unix handles installations is 100% OK for me. Yes, you don't know where programs are installed, but such is unix: you don't need to care (or you shouldn't) about where things are installed. Package tools are supposed to do that. If you need to know where files are saved you are probably using a operative system with a vomitive packaging tool - or without one, like Windows (until they figured out that most of the mainteinance problems you've in Windows are because people has to write their own installers and their installers are crappy, a system-side installer would have been the right thing from the start)
In plan9 you don't have a "$PATH variable", instead you have several directories (/whatever/arch-dependent-bin,/whatever/arch-independent-bin, ~/my-own-bin) and you just "join" them in a single directory:/bin (In plan9 every process can configure their filesystem namespace like they want and normal users are allowed to do things like that)
Well, breaking everything all time would be a pain indeed
My point is that we shouldn't do like microsoft who put "backwards compatibility" as top 1 on the list, creating crappy OSes like the 9x series (no multiuser etc), or the whole concept of "units" that NT has inherited. At some point you stop making your software better because you've to keep "backwards compatibility", and nobody does everything right the first time.
I like what gnome/kde does: "we guarantee you binary compatibility along the 2.X series". The kernel also has pretty good backwards compatibility, but only for syscalls, ioctls, proc stuff etc, binary drivers however are more like a "internal API" IMO
Backwards compatibility is important in the closed source world.
In the open source world if something can be fixed you fix it and fix the software which depended on the bad thing
Many people complains about the kernel breaking compatibility with binary drivers, but people knows when this happens and does it on purpose, because when they break the "binary compatibility" they fix all the open source drivers in the kernel, so at the end the didn't break the compatibility because all the in-tree drivers still works because they could be modified.
If you rely in closed software which depends on things that can be improved and your software provider refuses to improve its software by adapting to the changes...you're screwed anyway.
Todd: I'm talking Windows [Division] in general, or Microsoft in general. The Longhorn wave... we kind of took a year off. We kind of stopped the train, went back and fixed some problems in XP, and now we're gearing the momentum back up. We are getting ready to focus on Longhorn.
Now, at the same time all this has been going on, there has been a lot of complaining about the constantly slipping Longhorn release date. I haven't weighed in on that too much yet, but I think it's time to break my silence. Microsoft shifted between 80-90% of the Windows Client Team off Longhorn development and onto Windows XP SP2.
Is not that the SP2 is a bad thing. Is a great improvement, but it took so many time, it was delayed so many times...that's all what Microsoft can do? I mean, they just put all they resources in the SP2 and it took them forever to release it.
Perhaps it's just me, but the open source world evolves much faster and has more resources than Microsoft. Every 6 months I see more evolution in the OSS field than what I saw in SP2 (and again, it's not that the SP2 was bad - it was great! But just look at fedora 3 with its SELinux integrationand all the rest. We're being faster than them IMHO, and how fast can you evolute is more important than "how good are you today"
In Spain, sharing your broadband connection (via wireless or with a ethernet cable to your neightbourd) is forbidden by law.
In fact, some small villages had made a public wireless net, so everybody could use internet (we're talking of tiny villages with no access to broadband etc), and they were denounced by some stupid "teleccomunication comission"
The new gobernment told them that they shouldn't have denounced those villages since they were trying to spread internet's access but well...the point is: we have some law that forbids it:(
One of the reasons why OSS is not being as succesful as it could be is because people look at Openoffice, Firefox etc. and it looks all nice, but there's no equivalent to Outlook. So people continue using Office (and Word and the whole Office stack
Explorer is quite fast I must say. However firefox is not slow.
I guess people don't care about those benchmarks because:
o Some years ago, Internet wasn't "mainstream", people who used internet was interested in technical details, these days lots of people don't know what "render" means
o Browsers are fast these days. Some years ago it was "the start" of the "internet revolution". Browsers started to emerge and I guess there was a lot of field to build fast rendering engines. aslo computers are a lot faster so the difference between brosers is not very noticeable
o These days people cares more about things like popup blocking and tabs
Windows directly depends on 3rd party hardware manufacturers. That in fact hurts them A LOT
They've to keep old kernel interfaces and crappy compatibility layers because they don't have the source.
Linux however has the source of most of the hardware. I don't think many people realize how much advantages has linux because of that. It allows linux to evolve *anywhere* without waiting for 3rd party hardware manufacturers to catch up. It allows, in fact, to ignore binary compatibility, which is what Linus has been doing for a while in 2.6 - he don't care about binary compatibility because all the open source drivers are already fixed.
Google's desktop search is a BETA product. That means that it doesn't works always and that Google doesn't even need to fix it since you shouldn't be using it for serious purposes in first place.
They have lots of governments paying automatically lots of millions every year to upgrade their computers (plus hardware upgrade, etc etc)
When it comes to servers Windows may not be the king, but when it comes to "desktop computer with a office suite and corporate support", Microsoft Office has pretty much the 100% of the market.
Which is why Openoffice is so important. Openoffice is damaging Microsoft more than any other OSS project.
If you look at the kernel, pretty much all the core kernels are IBM/Redhat/Suse workers. GCC (which BSD users also use) have lots of redhat people in their lists. Lots of gnome hackers are paid to (project Utopia in Novell, Sun and Redhat, etc). Openoffice is not something that a couple of geeks can do in a weekend too. X.org has keith packards who did lots of work. KDE has several Suse/Mandrake/Lindows contributions. There're Freebsd hackers too, etc, etc etc..
OSs programmers probably won't see money, but instead companies hire people to work on their projects and they give you the code (lots of times those people are guys who were one of the main developers of the project and they got hired) Which is fair. We give them our work, they got money, and they hire people to work on our projects and release the source of the modifications. We get better software, they get money.
Well, my point is, "is not true that they give absolutely nothing back". NTPL, good SMP support, latest improvements in the incoming GCC 4.0, Gnome usability, Gnome accesibility from the Sun guys, openoffice...there're LOTS of things that wouldn't have happened without those companies.
I disagree. It shows you the number of hits a give word has. I use it often to check if I've spelled correctly a english word (it's not my mother tongue)
I agree however that it "distracts" you from the real search, because it shows a annoying menu. I wish the results would appear anywhere on the page, though.
(No, it doesn't use fancy html features but it's HTML. The "announcements" liked in the story are announcements for developers. Developers like pure ascii)
I also have the impression that GTK is not the fastest toolkit on earth, but it looks like they're improving!. Just look at the release notes:
Performance improvements
The chunk size for incremental transfers of big selections
has been increased, reducing the number of necessary
roundtrips. GTK+ uses sync counters to speed up window
resizing. The efficiency of GtkListStore and GtkUIManager
has been improved by changing the used algorithms and data
structures. Icon themes are cached in an mmap()able cache file
to reduce memory consumption and disk seek overhead. The cost
of intra-library function calls has been reduced by avoiding
PLT redirections.
Slashdot Mirror
If China government is trying to control what Chinese people can do on Internet...what's the point of having a high-performance router if they can't download pr0n movies? Linux distros?
IMHO the right fix is to have a good browser which don't allow phising.
There's nothing wrong with the file abstractiong but, how it's used
/dev/soundcard and it should work. Of course for that you should have a mp3 player behind that file (and NOT in the kernel, but as a userspace "service").
With soundcards, you should be able to do thigns like cp sng.mp3
In Linux you have "io priorities" (check CFQ) and you will be able to give portions of CPU usage to a give group of users in the near future. I don't think it's not possible to do such things in unix, ie: its not that "Unix is wrong" is that "nobody has built it for unix". Just look at mac os x, it _is_ unix and it is far from being a "antithesis of a practical user interfaces"
Having been a windows users, I hate how this stuff is handled in windows
/var for "data which changes often": you can put it in the fatest disk, you can't do the same (or it's much harder) in windows.
/etc > configuration-backup.tar" . Try to do the same in Windows, where each program puts its configuration file in its directory and you don't know what file it is (or even worse, the registry..)
There's a reason why directories are separated in unix,
/etc itself is ok for me except for the different configuration formats. It allows me to do a "tar c
Seriously, the way unix handles installations is 100% OK for me. Yes, you don't know where programs are installed, but such is unix: you don't need to care (or you shouldn't) about where things are installed. Package tools are supposed to do that. If you need to know where files are saved you are probably using a operative system with a vomitive packaging tool - or without one, like Windows (until they figured out that most of the mainteinance problems you've in Windows are because people has to write their own installers and their installers are crappy, a system-side installer would have been the right thing from the start)
Or what plan9 does, just kill the $PATH variable.
/whatever/arch-independent-bin, ~/my-own-bin) and you just "join" them in a single directory: /bin (In plan9 every process can configure their filesystem namespace like they want and normal users are allowed to do things like that)
In plan9 you don't have a "$PATH variable", instead you have several directories (/whatever/arch-dependent-bin,
evolve, sorry
Well, breaking everything all time would be a pain indeed
My point is that we shouldn't do like microsoft who put "backwards compatibility" as top 1 on the list, creating crappy OSes like the 9x series (no multiuser etc), or the whole concept of "units" that NT has inherited. At some point you stop making your software better because you've to keep "backwards compatibility", and nobody does everything right the first time.
I like what gnome/kde does: "we guarantee you binary compatibility along the 2.X series". The kernel also has pretty good backwards compatibility, but only for syscalls, ioctls, proc stuff etc, binary drivers however are more like a "internal API" IMO
D'oh, didn't know
Backwards compatibility is important in the closed source world.
In the open source world if something can be fixed you fix it and fix the software which depended on the bad thing
Many people complains about the kernel breaking compatibility with binary drivers, but people knows when this happens and does it on purpose, because when they break the "binary compatibility" they fix all the open source drivers in the kernel, so at the end the didn't break the compatibility because all the in-tree drivers still works because they could be modified.
If you rely in closed software which depends on things that can be improved and your software provider refuses to improve its software by adapting to the changes...you're screwed anyway.
Todd: I'm talking Windows [Division] in general, or Microsoft in general. The Longhorn wave
As I had previously read this is not a joke, just look at this quote from a Microsoft worker: http://www.longhornblogs.com/robert/archive/2004/
Now, at the same time all this has been going on, there has been a lot of complaining about the constantly slipping Longhorn release date. I haven't weighed in on that too much yet, but I think it's time to break my silence. Microsoft shifted between 80-90% of the Windows Client Team off Longhorn development and onto Windows XP SP2.
Is not that the SP2 is a bad thing. Is a great improvement, but it took so many time, it was delayed so many times...that's all what Microsoft can do? I mean, they just put all they resources in the SP2 and it took them forever to release it.
Perhaps it's just me, but the open source world evolves much faster and has more resources than Microsoft. Every 6 months I see more evolution in the OSS field than what I saw in SP2 (and again, it's not that the SP2 was bad - it was great! But just look at fedora 3 with its SELinux integrationand all the rest. We're being faster than them IMHO, and how fast can you evolute is more important than "how good are you today"
In Spain, sharing your broadband connection (via wireless or with a ethernet cable to your neightbourd) is forbidden by law.
:(
In fact, some small villages had made a public wireless net, so everybody could use internet (we're talking of tiny villages with no access to broadband etc), and they were denounced by some stupid "teleccomunication comission"
The new gobernment told them that they shouldn't have denounced those villages since they were trying to spread internet's access but well...the point is: we have some law that forbids it
Oh my God!. It's 100% now! Where is Superman when we need it?
Why not get only one and run two operative systems? (that don't forbids you from buying two comptures either)
Which is the problem - it only runs in Linux.
One of the reasons why OSS is not being as succesful as it could be is because people look at Openoffice, Firefox etc. and it looks all nice, but there's no equivalent to Outlook. So people continue using Office (and Word and the whole Office stack
Explorer is quite fast I must say. However firefox is not slow.
I guess people don't care about those benchmarks because:
o Some years ago, Internet wasn't "mainstream", people who used internet was interested in technical details, these days lots of people don't know what "render" means
o Browsers are fast these days. Some years ago it was "the start" of the "internet revolution". Browsers started to emerge and I guess there was a lot of field to build fast rendering engines. aslo computers are a lot faster so the difference between brosers is not very noticeable
o These days people cares more about things like popup blocking and tabs
That linux limit is OK for 2.4
Linux 2.6 runs on 512-cpu altix. I don't think Solaris has ever run in a computer so big
Windows directly depends on 3rd party hardware manufacturers. That in fact hurts them A LOT
They've to keep old kernel interfaces and crappy compatibility layers because they don't have the source.
Linux however has the source of most of the hardware. I don't think many people realize how much advantages has linux because of that. It allows linux to evolve *anywhere* without waiting for 3rd party hardware manufacturers to catch up. It allows, in fact, to ignore binary compatibility, which is what Linus has been doing for a while in 2.6 - he don't care about binary compatibility because all the open source drivers are already fixed.
Google's desktop search is a BETA product. That means that it doesn't works always and that Google doesn't even need to fix it since you shouldn't be using it for serious purposes in first place.
They have lots of governments paying automatically lots of millions every year to upgrade their computers (plus hardware upgrade, etc etc)
When it comes to servers Windows may not be the king, but when it comes to "desktop computer with a office suite and corporate support", Microsoft Office has pretty much the 100% of the market.
Which is why Openoffice is so important. Openoffice is damaging Microsoft more than any other OSS project.
If you look at the kernel, pretty much all the core kernels are IBM/Redhat/Suse workers. GCC (which BSD users also use) have lots of redhat people in their lists. Lots of gnome hackers are paid to (project Utopia in Novell, Sun and Redhat, etc). Openoffice is not something that a couple of geeks can do in a weekend too. X.org has keith packards who did lots of work. KDE has several Suse/Mandrake/Lindows contributions. There're Freebsd hackers too, etc, etc etc..
OSs programmers probably won't see money, but instead companies hire people to work on their projects and they give you the code (lots of times those people are guys who were one of the main developers of the project and they got hired)
Which is fair. We give them our work, they got money, and they hire people to work on our projects and release the source of the modifications. We get better software, they get money.
Well, my point is, "is not true that they give absolutely nothing back". NTPL, good SMP support, latest improvements in the incoming GCC 4.0, Gnome usability, Gnome accesibility from the Sun guys, openoffice...there're LOTS of things that wouldn't have happened without those companies.
I disagree. It shows you the number of hits a give word has. I use it often to check if I've spelled correctly a english word (it's not my mother tongue)
I agree however that it "distracts" you from the real search, because it shows a annoying menu. I wish the results would appear anywhere on the page, though.
"With their computing lives complete"
That's how I feel since I use 100% open source, strange.
Like this? http://www.gtk.org/gtk-2.6.0-notes.html
(No, it doesn't use fancy html features but it's HTML. The "announcements" liked in the story are announcements for developers. Developers like pure ascii)
I also have the impression that GTK is not the fastest toolkit on earth, but it looks like they're improving!. Just look at the release notes:
Performance improvements
The chunk size for incremental transfers of big selections has been increased, reducing the number of necessary roundtrips. GTK+ uses sync counters to speed up window resizing. The efficiency of GtkListStore and GtkUIManager has been improved by changing the used algorithms and data structures. Icon themes are cached in an mmap()able cache file to reduce memory consumption and disk seek overhead. The cost of intra-library function calls has been reduced by avoiding PLT redirections.