IP Vision in the UK are in a similar boat. They ship near-identical hardware (branded as a Technika 8320HD) to the Telstra T-Box, with the only difference being DVB-T2 twin tuners instead of DVB-T. They too run Linux and have a "legal information" dialogue box in the user interface, but I haven't seen any source code for the GPL'ed items they use either. More 8320HD info in my blog...
Cricket Darts Crown Green Bowling Curling Snooker (Or Pool or Billiards - all painful) Oval Car Racing (not just NASCAR, but IndyCar also guilty) Marathon Running Weightlifting Dressage Baseball (sorry, US folks, but it's interminable if watched live) Archery Fishing
I thought most people in the UK read the BBC News Website daily and rarely go to newspaper Web sites? I know I do and if I want to read other angles on a story, I go to a UK news aggregator like UK Google News or News Now. I never used The Times Website before it went paywall and I'm sure that's the case for almost everyone else out there too. It'll be interesting to see what happens when The Sun goes paywall - that demographic might not be smart enough to realise where the BBC/Google/NewsNow sites are:-)
Whilst in beta for Linux, you'd have thought Google would have reconsidered their frankly daft decision to build Chrome on a bleeding edge Fedora distro (F11 or F12) and not providing any sort of static build either. This limits running Chrome only on recent distros and, rather astonishingly, *excludes* the world's most popular commercial Linux (RHEL 5). This looks even worse when you realise that Mozilla build their products on CentOS 5 as their base and therefore can run on a lot more distros (e.g. anything from Fedora 6 upwards, as well as CentOS 5/RHEL 5 of course). If Mozilla can build Firefox on CentOS 5, I see little excuse why Google can't do the same with Chrome.
Sadly, the open source Chromium has the same issues - you can't build it on, say, CentOS 5 because of its use of later libraries. Basically, Google have shot themselves in the foot with the Linux Chrome releases...
I'm running CentOS 5 on work desktops and I do something similar to install the latest OO.org, although my procedure is as follows:
Go to http://download.openoffice.org/other.html and *untick* the "Include the JRE" option, otherwise you end up downloading a JRE you don't want (since I keep the full Sun JDK updated via downloads from java.sun.com).
Click on the "English (US)" 32-bit RPM Download link (yes, we run 64-bit CentOS, but the lack of an official 64-bit Linux Firefox (if distros and other 3rd parties can build it, why can't Mozilla?) causes a cascade of 32-bit dependencies including Java and hence OO.org).
This downloads a (147MB!).tar.gz which I unpack into an OOO dir using tar (and no, the dir name doesn't match the root of the.tar.gz file, ho hum).
I cd into the unpacked dir and then cd again into the RPMS sub-dir. I then *remove*.rpm files I don't want to install. Typically this would be openoffice.org3-dict-es, openoffice.org3-dict-fr and ooobasis3.2-testtool, but your mileage may vary.
I also "mv desktop-integration/openoffice.org3*-redhat-menus*." so that the packages appear in my users' start menu when I finally do an "rpm -Uvh *.rpm" as root. And, yes, there's a ludicrous 47 RPMs at this point to install - there really should be something like 5 or 6 (one core [aka "common"] RPM and one RPM for each app).
One final - often later - step is to download the en-GB language pack from the other.html page I mentioned at the start. Annoyingly for 3.2.0, they haven't released an en-GB pack for it yet, which is ridiculous considering far more diverse (compared to en-US) and much less popular languages have their 3.2.0 packs (e.g. Danish, Polish, Serbian and Slovenian).
I'm surprised Mundie didn't use Wikipedia (Not Invented Here?) before he spouted off - there's already a driving licence for computers at least (though it strangely doesn't include Internet 'skills') called the European Computer Driving Licence or ECDL for short. Although it started in Europe, it's spread worldwide ("European" becomes "International", so it's then known as ICDL) to 148 countries, including the US.
I seem to remember that it's very Microsoft-oriented though - the courses typically involve MS applications and probably don't consider alternatives (OpenOffice.org etc.) at all. I guess that makes it even more ironic that Mundle didn't refer to it. And, no, I've never taken the ECDL/ICDL so does that mean I'm guilty of "computing without a licence"?
I must say I was bemused to see the sort of items that are allowed into the cargo hold (i.e. via checked-in luggage, though I'm guessing you have to declare most or all of these):
Live ammunition, firearms, stun guns, cattle prods, martial arts weapons (yes, including numchucks), axes, meat cleavers, knives of *any* length, bows and arrows, swords and throwing stars.
Enough weapons to start a small war there - I wonder if there's a limit (other than total baggage weight) on how many of each of these items you can put in your luggage?
I will admit that major OEMs may be more likely to pre-install Chrome OS than a typical Linux distro simply because of the Google brand name, but other than that, what advantages will Chrome OS have over any decently configured Linux distro?
My strong suspicion is that the critical/shiny bits of Chrome OS may be closed source only and designed *not* to run on non-Chrome OS Linux distros to provide "added value". For example, will the Google OS kernel be fully open source or will have a magic closed source module that provides, say, a/proc/google special file that cats out some crypto string that can only be decoded by again closed source code embedded in the desktop/Google browser etc.? Without a proper crypto string, the app then simply refuses to run (yes, it could be binary hacked to bypass this, but it's messy doing that on all binaries and future updates).
If Google don't close source part of their OS (BTW, close sourcing the Chrome browser only may not be enough, since that's runnable on other Linux distros), then surely Linux can run (or imitate) the best bits of Chrome OS themselves anyway?
It's really bizarre that the article author included Paint.Net in a list of "best free open source software for Windows", because the source code - as the author himself even admits - is *not* available for free download for any of the recent versions of Paint.Net.
If that wasn't enough, there's been no new release of Paint.net for almost a year and I'd have thought GIMP (or GIMPShop) was a clearly superior (and fully open source) graphics package on Windows anyway.
Whilst I'm a bit surprised they won't be using X (effectively eliminating any chance of running local existing X applications that won't have a good Web equivalent for a long time), one thing that might be quite nice is if the OS did all its system administration (and user configuration) via a Web interface. This would mean shipping a Web server (possibly with an embedded scripting language - python, PHP, whatever) - though that could be fired up on demand I guess - and has the added benefit that remote admin becomes a doddle too.
Personally, I think Google have to push this to OEMs to bundle with netbooks - Linux has never had a "fully blown" OEM push (Dell and HP horribly hide their Linux offerings) before and pre-installation is the only way Windows users will ever migrate to another OS, since 95% of Windows users have never installed an OS before.
Google's brand awareness is massive and if the OS is slick enough for Joe Public and bundled virtually for free with netbooks from big OEMs, then this might just take off. The inability to run any local pre-existing graphical apps (Windows, Linux or Mac) might be a bit of downer though.
The article seemed to be confused about the size of the HD in the gaming rig. Initially, it states that they found a 320GB drive for $43, but the final table says it's actually a 250GB drive. Either way, isn't that quite a small drive - you suspect that installing 15-20 games on that rig could potentially fill the 200GB or so that would be available after the OS install. Newegg have the excellent 1TB Samsung Spinpoint F1 for $84.99 - four times the capacity (and I bet faster, cooler and quieter too) for twice the price.
One thing I didn't see covered by the article was how noisy the rig was - it always worries me that when you get beefy power supplies, CPUs and graphics cards, the thing can sound like a jumbo jet. Now I know that game sound effects can drown most of that out, but what about when you're not playing games? Can you sleep in the same room as the gaming rig is if you leave it turned on overnight?
I downloaded the 64-bit desktop ISO for 9.04 RC and I have an ATI 2600 XT card (which is pretty low-end nowadays). The graphical installer dismally fails (blank screen though you do get the Ubuntu desktop theme sound) in "normal" graphics mode, so I tried "safe graphics mode" instead. Although this did indeed fire up a viewable X, the last few hundred horizontal pixels were off-screen, making it impossible to confirm OK/cancel on each screen (yes, I could guess how many tabs to activate and then press Return, but this varies too).
Annoyingly, the desktop install CD doesn't include the (far superior) text-mode installer for no real good reason (this is a major weakness of Ubuntu compared to Fedora) and since I didn't want to download another 700MB of alternate installer CD, I did eventually work out how to get the X up and running OK.
Basically, I took an xorg.conf from another partition (Fedora 8) that used the "radeon" driver and copied it over the one in Ubuntu (Shift-Ctrl-F2 got me a text prompt to let me do this) and then killed X, which restarted correctly. A shame Ubuntu 9.04 RC is broken and the lack of a text installer in the desktop ISO is unforgivable because the only easy fallback is missing!
On a slightly more serious note, CentOS 5.2 users that are upgrading to 5.3 should do "yum update glibc" first before a "yum update" because of a yum locking issue. For us at work, CentOS 5.2 broke nss_ldap (and it's still broken with 5.3) and 5.3 has broken sound support in the kernel (about 1 in every 4 boots, the sound modules crash and fail to load properly). Easily fixed by keeping older RPMs on the system, but a bit annoying for a clone of an enterprise distro (I suspect RHEL 5 would have identical issues).
Ah yes, tvcatchup.com, the 100% mis-named service because you actually *can't* use it to catch up on TV that you may have missed when it was transmitted. Unless it's a live sporting event or breaking live news of major proportions, I record *everything* I watch to a hard disk recorder for later viewing (unless it wasn't transmitted in the UK in the first place, in which case it'll be a "download").
This allows me to watch it virtually the same quality as it was broadcast, any time I want, any number of times I want and dump it to DVD (the recorder has a DVD+RW too) to share with any number of friends. *None* of this possible with tvcatchup.com, so it's epic fail for that service I'm afraid.
In Red Dwarf's case, its transmission on Dave is virtually forcing a chunk of the UK population to magic P2P-land, especially since in its prime, it was a massive show on BBC 2.
"Midlands" covers several primary transmitters and even more relays. You'll find that a chunk of Freeview users of the relays/transmitters in the Midlands do not see all the multiplexes - some relays/transmitters have to run on reduced power to avoid interference and others are interefered with by geographical conditions (hills etc.).
I put my postcode/house number (which is in Merseyside) into the DTG checker and The Wrekin (Midlands) and Moel-Y-Parc (Wales) are my two transmitters - neither has the full complement of channels available at my postcode address (i.e. several multiplexes are missing). Also note that in my case, Granada is unreceivable because a big hill (no, not Winter Hill!) cuts off the signal for me.
The one glimmer of hope is that the full digital switchover (Oct/Nov 2009 for Wales, 2011 in the MIdlands) will allow full power transmission and finally bring all the multiplexes online for the first time ever for me. It's quite surprising how this lack of multiplexes over parts of the country ever since Freeview started is virtually unheard of in the press, although the "missing" channels already generally have low viewership nationwide anyway (Dave included).
Sadly, the BBC decided not to get involved in these new episodes, so it's ended up on the "tiny" channel Dave and won't get the multi-million viewing figures it might have had on BBC 1 or 2.
Unfortunately, the digital multplexes where I am (yes, I have 2 aerials to pick up Wales and Midlands) both don't have Dave, plus although I have a Sky Digital dish, I don't subscribe to anything on it, so Dave isn't available there either (it's encrypted on Sky for absolutely no good reason, especially when it's in the clear on Freeview if your multiplexes carry it).
So the irony is that despite living in the UK, having access to two Freeview digital regions *and* having a Sky Digital dish, I still can't see these new episodes. So I guess it's off to "other" avenues begining with the letter "B" to find the episodes then...sigh....
The reason for the rapid rise of Safari 4 Beta recently is that Apple have taken the frankly appalling decision to push a beta version of the browser as the default one to download when you go to http://www.apple.com/safari/ (and even deviously put the word "BETA" in a fainter, smaller font so you won't notice it).
Now Apple may claim that it's stable enough for any Mac end-user to download and use, but if that's the case, why is it still labelled "BETA", which implies that the general public shouldn't download this unless they're willing to be burned by unfixed bugs/missing features?
Contrast this with Firefox, which has never offered a beta release as the default download for users who visit the primary download site. In fact, if anything, pre-release versions of Firefox are arguably a little too hard to find, but I'd rather they erred on the side of caution - one bad pre-release that the masses get hold of and it can cause some damage to their reputation.
1. Windows Update needs to be catered for, so Windows will have to ship with the bare minimum to use that (possibly including the core of IE's rendering code).
2. Many apps use IE's rendering engine - again, the core of that will probably still need to be shipped.
3. On first boot, Windows should contact a server (who runs that? The EU? MS? A consortium?) and download a list of browsers, exact download URLs (which may vary if there's localised versions) and a short description of each browser.
4. The browser list is displayed (the order of browsers is important - I'd say in most popular order, but this one is up to debate) and the user can pick *one or more* (or even none!) browsers to download and install. IE should *not* be compulsory at this point.
5. Which browsers are downloaded should be recorded centrally (i.e. from the same place the browser list was downloaded from) so that the most popular browsers can be computed for ordering in future runs.
6. The browser(s) are downloaded and installed - I think this should be from each browser vendor's site and not from a central location or from MS.
7. Updates are probably the responsibility of each browser, mainly because MS doesn't seem to allow non-MS products to be updated via Windows Update.
8. To change the set of browsers, the first-boot tool should be available from the Start menu somewhere, so that the user can re-jig the browsers as appropriate (the tool should perhaps offer to uninstall any installed browsers too if it detects any from its list are already on the system).
I don't think browser vendors should have to pay to get on the browser list - they just have to show that they release regular updates (abandoned or rarely updated browsers have no place on the list) and have a reasonable chunk of the market (e.g. 0.25% or higher). You can't present the end-user with more than about 5 or 6 browsers to install - they'll just panic otherwise.
It's quite dubious that the only beta browser tested was Chrome, especially when most of the others have publicly available beta versions available for testing. Yes, I understand that the *only* release of Chrome is a beta, but then either Chrome should be disqualified from testing since it's not a final release or other browsers' beta releases should be allowed into the test (why not include both a final and beta release of those in that case, so we can see if there are improvements in the beta?).
I'd also like to see tests on non-Windows platforms as well, although Chrome scores as badly as IE here - it's *only* available on Windows at the moment and there's been a vague promise of ports to Mac and Linux, but these seem to be predictably dragging on and on.
If you run one of the longer term Linux distros (SUSE Enterprise, RHEL, CentOS primarily), then the current release of those probably won't ever see OO.org 3.X as an update. We primarily run CentOS 5 at work (both for servers and desktops) and I've hand-upgraded every OO.org release via manual downloads from the official site since we moved to CentOS 5.
I just wish there weren't 47 (!!) RPMs to install - what's wrong with 5 or 6 [core plus each app]? - plus where's the 64-bit version of OO.org?
However with blu-ray disks, i cannot picture the average consumer, or even the less common nerder consumer giving a damn over the inability to copy 40gig movies to their computer or to where ever.
However, they will care if they've got an non-HDCP monitor (e.g. analogue CRT, non-HDMI LCD) attached to their PC or TV and the software they've just bought at $199 (WinDVD 9 Plus Blu-Ray) or the standalone player they've spent a load of money on will only play it at 25% of the HD resolution!
My Philips analogue CRT can do 1920x1440, which is better than virtually all LCD TV/monitors out there and yet apparently it's "not allowed" - thanks to DRM - to display HD content which only goes up to 1080 vertical anyway! This is why DRM is the biggest load of old crock - it limits your use of the product simply because I have a "better" monitor than the movie studios would like...
I multi-boot with several 64-bit Linux distros and 64-bit Ultimate Vista on a Dell Vostro 400 I bought back in February (does this have the TPM stuff?). Grub is installed on the MBR and I don't have BitLocker enabled in Vista (why would I - can't read the disks in Linux if I did!). I installed Vista SP1 when it came out and had absolutely no problems (I may have had to re-install GRUB on the MBR, but I do that so often that I consider it no big deal). So am I the odd one out?
If you're in the UK, one of the cheapest RapidSSL resellers I've found is Servertastic, who do individual certs at 7.12 pounds + VAT and if you buy them in batches of 10 like we've started doing, the price comes down to an impressively cheap 5.50 pounds + VAT each.
With Firefox 3 now scaring the living daylights out of the end user when it comes across self-signed secure certs, spending less than 10 pounds a year isn't a hardship for anyone, even for a non-profit org that the original poster mentioned.
One snag though - Verisign bought Geotrust who provide the RapidSSL certs that Servertastic use, so I'm afraid that you just can't avoid giving money at least indirectly to Verisign (albeit a lot less than buying one of their hugely overpriced secure certs).
Remember the good-old days when Mozilla (and Firefox) release notes actually talked about bugs fixed, features introduced, and interesting things? When each version actually informed you about what had changed? http://www.mozilla.com/en-US/firefox/3.0rc2/releasenotes/ seems fairly reasonable to me. Granted, the differences between RC1 and RC2 aren't flagged (because virtually nothing but some blocker bugs were the changes between the two), but they *did* flag "Improved in Beta 5" in the equivalent Beta 5 release notes.
Going to mozilla.org (or.com) and trying to find betas is now impossible. No, really... there are no links to non-release versions. Oh come on! How hard did you bother reading the home page? What's New on the right hand side has a "Firefox 3 Sneak Peak" link for goodness' sake! And even if you drifted to mozilla.org's home page instead, guess what? Developer News on the right hand side announces the RC2 release as I speak. You sir, are either one lazy so-and-so or just a total troll!
I miss the time when Mozilla was a user-friendly organization, when everything was public and *easy to find*. I miss the time when people actually made the effort to check the current state of Web sites before slagging them off. Everything related to Mozilla (bar a few closed security bugs, which are opened once the fix is published) is very public and trivially easy to find. It's a shame that some people just don't think before they post.
Lots of bar graphs in the article, but no mention anywhere of the prices of the various Core 2 CPUs, which is surely important (especially if the chip has the word "Extreme" in its name)? It made the conclusion page fairly lame - most people would want to know which CPU offers the best bang for your buck, but we're given no clue about this metric and therefore are just as confused as to which Core 2 CPU to buy!
Slashdot Mirror
IP Vision in the UK are in a similar boat. They ship near-identical hardware (branded as a Technika 8320HD) to the Telstra T-Box, with the only difference being DVB-T2 twin tuners instead of DVB-T. They too run Linux and have a "legal information" dialogue box in the user interface, but I haven't seen any source code for the GPL'ed items they use either. More 8320HD info in my blog...
Cricket
Darts
Crown Green Bowling
Curling
Snooker (Or Pool or Billiards - all painful)
Oval Car Racing (not just NASCAR, but IndyCar also guilty)
Marathon Running
Weightlifting
Dressage
Baseball (sorry, US folks, but it's interminable if watched live)
Archery
Fishing
I thought most people in the UK read the BBC News Website daily and rarely go to newspaper Web sites? I know I do and if I want to read other angles on a story, I go to a UK news aggregator like UK Google News or News Now. I never used The Times Website before it went paywall and I'm sure that's the case for almost everyone else out there too. It'll be interesting to see what happens when The Sun goes paywall - that demographic might not be smart enough to realise where the BBC/Google/NewsNow sites are :-)
Whilst in beta for Linux, you'd have thought Google would have reconsidered their frankly daft decision to build Chrome on a bleeding edge Fedora distro (F11 or F12) and not providing any sort of static build either. This limits running Chrome only on recent distros and, rather astonishingly, *excludes* the world's most popular commercial Linux (RHEL 5). This looks even worse when you realise that Mozilla build their products on CentOS 5 as their base and therefore can run on a lot more distros (e.g. anything from Fedora 6 upwards, as well as CentOS 5/RHEL 5 of course). If Mozilla can build Firefox on CentOS 5, I see little excuse why Google can't do the same with Chrome.
Sadly, the open source Chromium has the same issues - you can't build it on, say, CentOS 5 because of its use of later libraries. Basically, Google have shot themselves in the foot with the Linux Chrome releases...
I'm running CentOS 5 on work desktops and I do something similar to install the latest OO.org, although my procedure is as follows:
Go to http://download.openoffice.org/other.html and *untick* the "Include the JRE" option, otherwise you end up downloading a JRE you don't want (since I keep the full Sun JDK updated via downloads from java.sun.com).
Click on the "English (US)" 32-bit RPM Download link (yes, we run 64-bit CentOS, but the lack of an official 64-bit Linux Firefox (if distros and other 3rd parties can build it, why can't Mozilla?) causes a cascade of 32-bit dependencies including Java and hence OO.org).
This downloads a (147MB!) .tar.gz which I unpack into an OOO dir using tar (and no, the dir name doesn't match the root of the .tar.gz file, ho hum).
I cd into the unpacked dir and then cd again into the RPMS sub-dir. I then *remove* .rpm files I don't want to install. Typically this would be openoffice.org3-dict-es, openoffice.org3-dict-fr and ooobasis3.2-testtool, but your mileage may vary.
I also "mv desktop-integration/openoffice.org3*-redhat-menus* ." so that the packages appear in my users' start menu when I finally do an "rpm -Uvh *.rpm" as root. And, yes, there's a ludicrous 47 RPMs at this point to install - there really should be something like 5 or 6 (one core [aka "common"] RPM and one RPM for each app).
One final - often later - step is to download the en-GB language pack from the other.html page I mentioned at the start. Annoyingly for 3.2.0, they haven't released an en-GB pack for it yet, which is ridiculous considering far more diverse (compared to en-US) and much less popular languages have their 3.2.0 packs (e.g. Danish, Polish, Serbian and Slovenian).
I'm surprised Mundie didn't use Wikipedia (Not Invented Here?) before he spouted off - there's already a driving licence for computers at least (though it strangely doesn't include Internet 'skills') called the
European Computer Driving Licence or ECDL for short. Although it started in Europe, it's spread worldwide ("European" becomes "International", so it's then known as ICDL) to 148 countries, including the US.
I seem to remember that it's very Microsoft-oriented though - the courses typically involve MS applications and probably don't consider alternatives (OpenOffice.org etc.) at all. I guess that makes it even more ironic that Mundle didn't refer to it. And, no, I've never taken the ECDL/ICDL so does that mean I'm guilty of "computing without a licence"?
I must say I was bemused to see the sort of items that are allowed into the cargo hold (i.e. via checked-in luggage, though I'm guessing you have to declare most or all of these):
Live ammunition, firearms, stun guns, cattle prods, martial arts weapons (yes, including numchucks), axes, meat cleavers, knives of *any* length, bows and arrows, swords and throwing stars.
Enough weapons to start a small war there - I wonder if there's a limit (other than total baggage weight) on how many of each of these items you can put in your luggage?
I will admit that major OEMs may be more likely to pre-install Chrome OS than a typical Linux distro simply because of the Google brand name, but other than that, what advantages will Chrome OS have over any decently configured Linux distro?
My strong suspicion is that the critical/shiny bits of Chrome OS may be closed source only and designed *not* to run on non-Chrome OS Linux distros to provide "added value". For example, will the Google OS kernel be fully open source or will have a magic closed source module that provides, say, a /proc/google special file that cats out some crypto string that can only be decoded by again closed source code embedded in the desktop/Google browser etc.? Without a proper crypto string, the app then simply refuses to run (yes, it could be binary hacked to bypass this, but it's messy doing that on all binaries and future updates).
If Google don't close source part of their OS (BTW, close sourcing the Chrome browser only may not be enough, since that's runnable on other Linux distros), then surely Linux can run (or imitate) the best bits of Chrome OS themselves anyway?
It's really bizarre that the article author included Paint.Net in a list of "best free open source software for Windows", because the source code - as the author himself even admits - is *not* available for free download for any of the recent versions of Paint.Net.
If that wasn't enough, there's been no new release of Paint.net for almost a year and I'd have thought GIMP (or GIMPShop) was a clearly superior (and fully open source) graphics package on Windows anyway.
Whilst I'm a bit surprised they won't be using X (effectively eliminating any chance of running local existing X applications that won't have a good Web equivalent for a long time), one thing that might be quite nice is if the OS did all its system administration (and user configuration) via a Web interface. This would mean shipping a Web server (possibly with an embedded scripting language - python, PHP, whatever) - though that could be fired up on demand I guess - and has the added benefit that remote admin becomes a doddle too.
Personally, I think Google have to push this to OEMs to bundle with netbooks - Linux has never had a "fully blown" OEM push (Dell and HP horribly hide their Linux offerings) before and pre-installation is the only way Windows users will ever migrate to another OS, since 95% of Windows users have never installed an OS before.
Google's brand awareness is massive and if the OS is slick enough for Joe Public and bundled virtually for free with netbooks from big OEMs, then this might just take off. The inability to run any local pre-existing graphical apps (Windows, Linux or Mac) might be a bit of downer though.
The article seemed to be confused about the size of the HD in the gaming rig. Initially, it states that they found a 320GB drive for $43, but the final table says it's actually a 250GB drive. Either way, isn't that quite a small drive - you suspect that installing 15-20 games on that rig could potentially fill the 200GB or so that would be available after the OS install. Newegg have the excellent 1TB Samsung Spinpoint F1 for $84.99 - four times the capacity (and I bet faster, cooler and quieter too) for twice the price.
One thing I didn't see covered by the article was how noisy the rig was - it always worries me that when you get beefy power supplies, CPUs and graphics cards, the thing can sound like a jumbo jet. Now I know that game sound effects can drown most of that out, but what about when you're not playing games? Can you sleep in the same room as the gaming rig is if you leave it turned on overnight?
I downloaded the 64-bit desktop ISO for 9.04 RC and I have an ATI 2600 XT card (which is pretty low-end nowadays). The graphical installer dismally fails (blank screen though you do get the Ubuntu desktop theme sound) in "normal" graphics mode, so I tried "safe graphics mode" instead. Although this did indeed fire up a viewable X, the last few hundred horizontal pixels were off-screen, making it impossible to confirm OK/cancel on each screen (yes, I could guess how many tabs to activate and then press Return, but this varies too).
Annoyingly, the desktop install CD doesn't include the (far superior) text-mode installer for no real good reason (this is a major weakness of Ubuntu compared to Fedora) and since I didn't want to download another 700MB of alternate installer CD, I did eventually work out how to get the X up and running OK.
Basically, I took an xorg.conf from another partition (Fedora 8) that used the "radeon" driver and copied it over the one in Ubuntu (Shift-Ctrl-F2 got me a text prompt to let me do this) and then killed X, which restarted correctly. A shame Ubuntu 9.04 RC is broken and the lack of a text installer in the desktop ISO is unforgivable because the only easy fallback is missing!
April 1st has called and wants its release announcement back :-)
On a slightly more serious note, CentOS 5.2 users that are upgrading to 5.3 should do "yum update glibc" first before a "yum update" because of a yum locking issue. For us at work, CentOS 5.2 broke nss_ldap (and it's still broken with 5.3) and 5.3 has broken sound support in the kernel (about 1 in every 4 boots, the sound modules crash and fail to load properly). Easily fixed by keeping older RPMs on the system, but a bit annoying for a clone of an enterprise distro (I suspect RHEL 5 would have identical issues).
Ah yes, tvcatchup.com, the 100% mis-named service because you actually *can't* use it to catch up on TV that you may have missed when it was transmitted. Unless it's a live sporting event or breaking live news of major proportions, I record *everything* I watch to a hard disk recorder for later viewing (unless it wasn't transmitted in the UK in the first place, in which case it'll be a "download").
This allows me to watch it virtually the same quality as it was broadcast, any time I want, any number of times I want and dump it to DVD (the recorder has a DVD+RW too) to share with any number of friends. *None* of this possible with tvcatchup.com, so it's epic fail for that service I'm afraid.
In Red Dwarf's case, its transmission on Dave is virtually forcing a chunk of the UK population to magic P2P-land, especially since in its prime, it was a massive show on BBC 2.
I put my postcode/house number (which is in Merseyside) into the DTG checker and The Wrekin (Midlands) and Moel-Y-Parc (Wales) are my two transmitters - neither has the full complement of channels available at my postcode address (i.e. several multiplexes are missing). Also note that in my case, Granada is unreceivable because a big hill (no, not Winter Hill!) cuts off the signal for me.
The one glimmer of hope is that the full digital switchover (Oct/Nov 2009 for Wales, 2011 in the MIdlands) will allow full power transmission and finally bring all the multiplexes online for the first time ever for me. It's quite surprising how this lack of multiplexes over parts of the country ever since Freeview started is virtually unheard of in the press, although the "missing" channels already generally have low viewership nationwide anyway (Dave included).
Sadly, the BBC decided not to get involved in these new episodes, so it's ended up on the "tiny" channel Dave and won't get the multi-million viewing figures it might have had on BBC 1 or 2.
Unfortunately, the digital multplexes where I am (yes, I have 2 aerials to pick up Wales and Midlands) both don't have Dave, plus although I have a Sky Digital dish, I don't subscribe to anything on it, so Dave isn't available there either (it's encrypted on Sky for absolutely no good reason, especially when it's in the clear on Freeview if your multiplexes carry it).
So the irony is that despite living in the UK, having access to two Freeview digital regions *and* having a Sky Digital dish, I still can't see these new episodes. So I guess it's off to "other" avenues begining with the letter "B" to find the episodes then...sigh....
The reason for the rapid rise of Safari 4 Beta recently is that Apple have taken the frankly appalling decision to push a beta version of the browser as the default one to download when you go to http://www.apple.com/safari/ (and even deviously put the word "BETA" in a fainter, smaller font so you won't notice it).
Now Apple may claim that it's stable enough for any Mac end-user to download and use, but if that's the case, why is it still labelled "BETA", which implies that the general public shouldn't download this unless they're willing to be burned by unfixed bugs/missing features?
Contrast this with Firefox, which has never offered a beta release as the default download for users who visit the primary download site. In fact, if anything, pre-release versions of Firefox are arguably a little too hard to find, but I'd rather they erred on the side of caution - one bad pre-release that the masses get hold of and it can cause some damage to their reputation.
1. Windows Update needs to be catered for, so Windows will have to ship with the bare minimum to use that (possibly including the core of IE's rendering code).
2. Many apps use IE's rendering engine - again, the core of that will probably still need to be shipped.
3. On first boot, Windows should contact a server (who runs that? The EU? MS? A consortium?) and download a list of browsers, exact download URLs (which may vary if there's localised versions) and a short description of each browser.
4. The browser list is displayed (the order of browsers is important - I'd say in most popular order, but this one is up to debate) and the user can pick *one or more* (or even none!) browsers to download and install. IE should *not* be compulsory at this point.
5. Which browsers are downloaded should be recorded centrally (i.e. from the same place the browser list was downloaded from) so that the most popular browsers can be computed for ordering in future runs.
6. The browser(s) are downloaded and installed - I think this should be from each browser vendor's site and not from a central location or from MS.
7. Updates are probably the responsibility of each browser, mainly because MS doesn't seem to allow non-MS products to be updated via Windows Update.
8. To change the set of browsers, the first-boot tool should be available from the Start menu somewhere, so that the user can re-jig the browsers as appropriate (the tool should perhaps offer to uninstall any installed browsers too if it detects any from its list are already on the system).
I don't think browser vendors should have to pay to get on the browser list - they just have to show that they release regular updates (abandoned or rarely updated browsers have no place on the list) and have a reasonable chunk of the market (e.g. 0.25% or higher). You can't present the end-user with more than about 5 or 6 browsers to install - they'll just panic otherwise.
It's quite dubious that the only beta browser tested was Chrome, especially when most of the others have publicly available beta versions available for testing. Yes, I understand that the *only* release of Chrome is a beta, but then either Chrome should be disqualified from testing since it's not a final release or other browsers' beta releases should be allowed into the test (why not include both a final and beta release of those in that case, so we can see if there are improvements in the beta?).
I'd also like to see tests on non-Windows platforms as well, although Chrome scores as badly as IE here - it's *only* available on Windows at the moment and there's been a vague promise of ports to Mac and Linux, but these seem to be predictably dragging on and on.
If you run one of the longer term Linux distros (SUSE Enterprise, RHEL, CentOS primarily), then the current release of those probably won't ever see OO.org 3.X as an update. We primarily run CentOS 5 at work (both for servers and desktops) and I've hand-upgraded every OO.org release via manual downloads from the official site since we moved to CentOS 5. I just wish there weren't 47 (!!) RPMs to install - what's wrong with 5 or 6 [core plus each app]? - plus where's the 64-bit version of OO.org?
However with blu-ray disks, i cannot picture the average consumer, or even the less common nerder consumer giving a damn over the inability to copy 40gig movies to their computer or to where ever.
However, they will care if they've got an non-HDCP monitor (e.g. analogue CRT, non-HDMI LCD) attached to their PC or TV and the software they've just bought at $199 (WinDVD 9 Plus Blu-Ray) or the standalone player they've spent a load of money on will only play it at 25% of the HD resolution!
My Philips analogue CRT can do 1920x1440, which is better than virtually all LCD TV/monitors out there and yet apparently it's "not allowed" - thanks to DRM - to display HD content which only goes up to 1080 vertical anyway! This is why DRM is the biggest load of old crock - it limits your use of the product simply because I have a "better" monitor than the movie studios would like...
I multi-boot with several 64-bit Linux distros and 64-bit Ultimate Vista on a Dell Vostro 400 I bought back in February (does this have the TPM stuff?). Grub is installed on the MBR and I don't have BitLocker enabled in Vista (why would I - can't read the disks in Linux if I did!). I installed Vista SP1 when it came out and had absolutely no problems (I may have had to re-install GRUB on the MBR, but I do that so often that I consider it no big deal). So am I the odd one out?
If you're in the UK, one of the cheapest RapidSSL resellers I've found is Servertastic, who do individual certs at 7.12 pounds + VAT and if you buy them in batches of 10 like we've started doing, the price comes down to an impressively cheap 5.50 pounds + VAT each.
With Firefox 3 now scaring the living daylights out of the end user when it comes across self-signed secure certs, spending less than 10 pounds a year isn't a hardship for anyone, even for a non-profit org that the original poster mentioned.
One snag though - Verisign bought Geotrust who provide the RapidSSL certs that Servertastic use, so I'm afraid that you just can't avoid giving money at least indirectly to Verisign (albeit a lot less than buying one of their hugely overpriced secure certs).
Lots of bar graphs in the article, but no mention anywhere of the prices of the various Core 2 CPUs, which is surely important (especially if the chip has the word "Extreme" in its name)? It made the conclusion page fairly lame - most people would want to know which CPU offers the best bang for your buck, but we're given no clue about this metric and therefore are just as confused as to which Core 2 CPU to buy!