I have the native 64 bit Flash plugin I can tell you that it is not THAT smooth and accelerated on many Intel GPUs. You may wish to qualify your statement with some links (the PenguinSWF blog talks about the issues). On the same machine that struggles with certain Youtube Flash videos on Linux, the performance is fine under Windows. I believe you will need your GPU to have at certain GL features for support for significant Flash acceleration under Linux and if your GL driver returns SGI string it will typically default to not turning acceleration on to avoid problems for those with weak GL implementations.
The squashfs image that is casper/filesystem.squashfs is already compressed with LZMA. Even initrd is compressed with LZMA. I would guess compressing the whole iSO would not shrink it by much more...
Finally the argument about which style of licence is best will be settled once and for all!:)
At the minute, BSD style licences are more trendy from a business perspective and big organisations like Apple, Google and so forth see it as the best collaborative way forward. However there are GPL-esque projects have proven popular with companies (e.g. KHTML/Webikit) so it is far from clear which side will prove more popular. I'm just happy that at least there's something open source that lets me open MS Office documents in a reasonable manner - in 1999/2000 it was a lot more painful.
The problem comes when you have to switch from said fullscreen task. Let's say I have a game that is running full screen and while its playing I want to switch to another desktop. If it had an unredirected (?) window I now have to redirect the game window elsewhere which is going to take resources and introduce an unsightly flicker. I wonder if this is why OS X does its tasteful fade before certain games are run.
Perhaps what is needed is a mechanism for an individual program to say "I want to be unredirected" so that things you normally switch between that might happen to be fullscreen are left composited, whereas those that actually need the speed can request it.
It is no more sensational than what modern (lets say Vista and above) Windows faces and if that's fair game then why not OS X? Apple made a big play about not getting viruses - well this is a step on the path towards viruses so people are going to talk.
You can point out that it requires user intervention but people are conditioned to click on bouncing icons in the dock and if you ever use 3rd party apps (e.g. Skype) you become conditioned to click past "this program comes from web" dialogs. I can see this being combined with a browser exploit in the future to automate things further.
These incidents show that malware is not limited to Windows - other non-locked down platforms can face the same issue.
As another reader pointed out, you can put a Linux ISO on to a USB stick. I know Fedora and Ubuntu both come with tools to do just this. Another option that has been supported for as long as I've been using Linux distros has been net booting - you download a tiny image and then install over the network. If you can do bootp you don't even need to put the small image on the machine using removable media.
Finally something I learned is that you can quite bootstrap an install CD if you have the technical chops and a machine which has a bootloader that lets you boot arbitrary kernels. You typically unpack the ISO to a spare partition, grab the vmlinuz and initrd/initramfs and then boot that and pass the options that let it know which partition contains the rest of the install. It was initially very painful to learn how to do this but it does mean you don't need any new media on machines which have something like grub already working (or you can put grub on a USB device and then boot off that).
The reality of the "No regressions" phrase a bit more complicated. It help tremendously if the regression is noticed during the development stage of the kernel and can be narrowed down to a single patch. Further it also depends on exactly which way you regress - a "can't boot previously working and still supported system" functional regression can be treated differently to a "inserting a new disk is 0.1 of a second slower" performance regression.
Sigh. My previous comment was horribly garbled. What I meant to say is - did you test a dynamic programming solution too? It sounds like the sort of problem that such a technique would do well at (but perhaps it would run into the same problems as pre-computation).
Do you know how fast a Dynamic Programming solution would have taken (if you still have numbers to hand)? I would have though it would be somewhere between competent and your neural network solutions but perhaps it would have been brute force by another name...
...but my degree experience (CS at one university and Software Engineering at another) poor Mathematics skills just makes life more difficult. You will just have to accept that you have to work harder than others who are good at it (if only to overcome your dislike). It is also worth noting that there are different branches of Mathematics and being bad at all of them is different to being bad in only area but fine in others. Further, different courses will place different emphasis on Mathematical content (e.g. HCI style courses may emphasise statistics).
There is also a difference between programming and Computer Science even if you can argue that programming is a subset of Computer Science. Computer most certainly is a branch of Mathematics (or if you want to annoy people you can say Mathematics is a branch of Computer Science;) and there is Mathematics underlying all computation. However if you are terrible at Maths you can still create great non-mathematical programs but you have to accept that there will be certain types of programs you may not be able to write (or write well) until you conqueror the Mathematics.
Just because knowing more usually doesn't hurt doesn't mean you HAVE to learn more but whichever direction you take you simply have to accept there will be consequences.
Bufferbloat (I first came across the term bufferbloat in this blog post by Jim Gettys) is the nickname that has been given to the high latency that can occur in modern network connections due to large buffers in the network. An example could be the way that a network game on one computer starts to stutter if another computer starts to use a protocol like bittorrent to transfer files on the same network connection.
The large buffers seem to have arisen from a desire to maximise download throughput regardless of network condition. This can give rise to the situation where small urgent packets are delayed because big packets (which perhaps should not have been sent) are queued up in front of them. The system sending the big packets is not told to stop sending them so quickly because its packets are being delivered...
The linked article sounds like people have modified the Linux kernel source to allow people who know how to compile their own kernels to test ideas people have had for reducing the bufferbloat effect on their hardware and to report back their results.
I believe computers with OSX 10.4 would ship with a trial version of Office installed. I remember this being a menace because the demo version could continue to interfere with file associations even after the full version of Office was installed.
On a more pedantic note doesn't Quicktime ship with some menu items disabled (unless you upgrade?) I am also fairly sure my family's Macbook came installed with a trial version of iWork on 10.5. Perhaps you were purchasing full versions of iWork with your machines?
If you look into how modern distros control device permissions (e.g. on/dev/snd/pcmC0D0p ) you may find they make use of ACLs to allow lists of users access to things (without resorting to groups). However this gets complicated fast.
Additionally, a number of Linux security modules (SELinux, Apparmor, TOMOYO) alloow the use of common apps under a MAC model. Fedora really does run out of the box under SELinux after all.
Slashdot Mirror
I have the native 64 bit Flash plugin I can tell you that it is not THAT smooth and accelerated on many Intel GPUs. You may wish to qualify your statement with some links (the PenguinSWF blog talks about the issues). On the same machine that struggles with certain Youtube Flash videos on Linux, the performance is fine under Windows. I believe you will need your GPU to have at certain GL features for support for significant Flash acceleration under Linux and if your GL driver returns SGI string it will typically default to not turning acceleration on to avoid problems for those with weak GL implementations.
This is well known and is not a conspiracy. For a while OSX was in a similar position.
The squashfs image that is casper/filesystem.squashfs is already compressed with LZMA. Even initrd is compressed with LZMA. I would guess compressing the whole iSO would not shrink it by much more...
According to this article the Mozilla foundation started being funded by Google in 2004. Firefox was originally started while it was still in AOL back in 2002 so at most it could have only been funded by Google for at most 9 years :-). Google provide 86% of Mozilla's funding back in 2009 but those are the latest results I could find.
An interesting comment on this comes from Jeremy Allison on the blog of an Openoffice.org developer (found via Dave Neary's blog):
This is about copyleft vs. non-copyleft licensing
Finally the argument about which style of licence is best will be settled once and for all! :)
At the minute, BSD style licences are more trendy from a business perspective and big organisations like Apple, Google and so forth see it as the best collaborative way forward. However there are GPL-esque projects have proven popular with companies (e.g. KHTML/Webikit) so it is far from clear which side will prove more popular. I'm just happy that at least there's something open source that lets me open MS Office documents in a reasonable manner - in 1999/2000 it was a lot more painful.
The problem comes when you have to switch from said fullscreen task. Let's say I have a game that is running full screen and while its playing I want to switch to another desktop. If it had an unredirected (?) window I now have to redirect the game window elsewhere which is going to take resources and introduce an unsightly flicker. I wonder if this is why OS X does its tasteful fade before certain games are run.
Perhaps what is needed is a mechanism for an individual program to say "I want to be unredirected" so that things you normally switch between that might happen to be fullscreen are left composited, whereas those that actually need the speed can request it.
It is no more sensational than what modern (lets say Vista and above) Windows faces and if that's fair game then why not OS X? Apple made a big play about not getting viruses - well this is a step on the path towards viruses so people are going to talk.
You can point out that it requires user intervention but people are conditioned to click on bouncing icons in the dock and if you ever use 3rd party apps (e.g. Skype) you become conditioned to click past "this program comes from web" dialogs. I can see this being combined with a browser exploit in the future to automate things further.
These incidents show that malware is not limited to Windows - other non-locked down platforms can face the same issue.
But the systemtap devs don't seem overly concerned.
As another reader pointed out, you can put a Linux ISO on to a USB stick. I know Fedora and Ubuntu both come with tools to do just this. Another option that has been supported for as long as I've been using Linux distros has been net booting - you download a tiny image and then install over the network. If you can do bootp you don't even need to put the small image on the machine using removable media.
Finally something I learned is that you can quite bootstrap an install CD if you have the technical chops and a machine which has a bootloader that lets you boot arbitrary kernels. You typically unpack the ISO to a spare partition, grab the vmlinuz and initrd/initramfs and then boot that and pass the options that let it know which partition contains the rest of the install. It was initially very painful to learn how to do this but it does mean you don't need any new media on machines which have something like grub already working (or you can put grub on a USB device and then boot off that).
I joined in 1999 so assuming UIDs were given out sequential order...
It's definitely the right era though.
Android Webkit source Git repository? Some of the branches in that repo explicitly mention Honeycomb...
The reality of the "No regressions" phrase a bit more complicated. It help tremendously if the regression is noticed during the development stage of the kernel and can be narrowed down to a single patch. Further it also depends on exactly which way you regress - a "can't boot previously working and still supported system" functional regression can be treated differently to a "inserting a new disk is 0.1 of a second slower" performance regression.
Jargon file entry for "mu". Answering this way isn't going to work in most cases though...
Speaking of gnome-terminal, how the fuck do you set the default window size? In Terminal.app, you just resize and Shell -> Use Settings As Default.
Using gconf to set a default size in certain versions of GNOME Terminal is broken but looking at the gnome-terminal 2.33.90 shows a "Use custom default terminal size" (yeah I know that page talks about Ubuntu but the option was there in a Fedora 15 Alpha live CD too). gnome-terminal 2.30 and peering at gnome-terminal's git suggests the option would have gone in around 2.31.
But hey - Slashdot ain't a bug tracker so here might not have been the best place to ask (even if you did work at NetApp)... :)
There is a feature in my version of Chromium on the about:flags page that says:
Enable better omnibox history matching
Enables substring and multi-fragment matching within URLs from history.
This seems to let me type a piece of bookmark name and then match the bookmark in the fashion you described but its certainly not on by default.
Sigh. My previous comment was horribly garbled. What I meant to say is - did you test a dynamic programming solution too? It sounds like the sort of problem that such a technique would do well at (but perhaps it would run into the same problems as pre-computation).
Do you know how fast a Dynamic Programming solution would have taken (if you still have numbers to hand)? I would have though it would be somewhere between competent and your neural network solutions but perhaps it would have been brute force by another name...
...but my degree experience (CS at one university and Software Engineering at another) poor Mathematics skills just makes life more difficult. You will just have to accept that you have to work harder than others who are good at it (if only to overcome your dislike). It is also worth noting that there are different branches of Mathematics and being bad at all of them is different to being bad in only area but fine in others. Further, different courses will place different emphasis on Mathematical content (e.g. HCI style courses may emphasise statistics).
There is also a difference between programming and Computer Science even if you can argue that programming is a subset of Computer Science. Computer most certainly is a branch of Mathematics (or if you want to annoy people you can say Mathematics is a branch of Computer Science ;) and there is Mathematics underlying all computation. However if you are terrible at Maths you can still create great non-mathematical programs but you have to accept that there will be certain types of programs you may not be able to write (or write well) until you conqueror the Mathematics.
Just because knowing more usually doesn't hurt doesn't mean you HAVE to learn more but whichever direction you take you simply have to accept there will be consequences.
Does changing layout.css.dpi make any difference?
Your replies on Moz stories are appreciated.
Sorry Dave. A Coral cache version of the animation.
I should have also linked to a definition of bufferbloat by Jim Gettys. For the curious here's a page of links to bufferbloat resources and a 5 minute animation that shows the impact of large buffers on network communication (.avi).
My understanding may not be correct but:
Bufferbloat (I first came across the term bufferbloat in this blog post by Jim Gettys) is the nickname that has been given to the high latency that can occur in modern network connections due to large buffers in the network. An example could be the way that a network game on one computer starts to stutter if another computer starts to use a protocol like bittorrent to transfer files on the same network connection.
The large buffers seem to have arisen from a desire to maximise download throughput regardless of network condition. This can give rise to the situation where small urgent packets are delayed because big packets (which perhaps should not have been sent) are queued up in front of them. The system sending the big packets is not told to stop sending them so quickly because its packets are being delivered...
The linked article sounds like people have modified the Linux kernel source to allow people who know how to compile their own kernels to test ideas people have had for reducing the bufferbloat effect on their hardware and to report back their results.
Does this help explain things a bit?
I believe computers with OSX 10.4 would ship with a trial version of Office installed. I remember this being a menace because the demo version could continue to interfere with file associations even after the full version of Office was installed.
On a more pedantic note doesn't Quicktime ship with some menu items disabled (unless you upgrade?) I am also fairly sure my family's Macbook came installed with a trial version of iWork on 10.5. Perhaps you were purchasing full versions of iWork with your machines?
If you look into how modern distros control device permissions (e.g. on /dev/snd/pcmC0D0p ) you may find they make use of ACLs to allow lists of users access to things (without resorting to groups). However this gets complicated fast.
Additionally, a number of Linux security modules (SELinux, Apparmor, TOMOYO) alloow the use of common apps under a MAC model. Fedora really does run out of the box under SELinux after all.
However, I feel that what is needed is the ability to disclaim privileges even when running as a normal user. There is experimental user namespace work on Linux that w allow unprivileged users to create namespaces which may in the future provide such an ability.
There was an anonymous forum post earlier in the year talking about how the political situation between carriers and handset makers in the US works against major updates. Basically feature updates can cause problems for carriers and certain handset makers also charge them extra for major updates.