The rumors are flying about a Google Linux Desktop. My take? It ain't gonna happen. Not a general-OS type release… not any time soon. Now, it's no secret that many engineers inside Google use Linux (BTW, they use Linux on a couple servers too). Last time I was at the Googleplex is was a Red Hat derivative (Gred Hat or something similar) and now it's Ubuntu based. So why don't I think Google will release this as a general purpose desktop OS? Simple – it isn't ready for that. If they released this as a PC, then every random $9 game from Walmart, AOL and all kinds of other things that don't work would be expected to. Is this a knock against Linux? No! For what it's meant to do, it works phenomenally. I've used it as my sole desktop for years, and I can't imagine going back to Windows. But, I also figured out that I needed a kernel recompile because my iPod wouldn't work, so I'm not a good case for a basic consumer study on this. Wine would basically have to work as well as Rosetta for this to work out – the lack of iTunes alone would enrage millions of people. What I do think could happen here is a very nice and very function appliance. One based on Linux. One also based on the web. You see, that hits a completely different market segment… one that may have never even owned a computer. The expectation that anything and everything needs to be compatible goes away. It's cheap, it's fast, it's slick and it's reliable. It's immune to spyware, you don't have to worry about viruses and things just work. For $100-200 you get a device that allows you to browse the web (they employ Firefox developers), chat both online and via VOIP (they employ the main Gaim developer and have GoogleTalk), check your email (you may have heard of this Gmail thing) and perform a myriad of other online tasks. It may even have basic office functionality, be it via standard OOo or more likely some OOo web derivative. That would be step one. That would lay the seed. That would build a base of loyal GoOs fans. As Linux gets more and more ready for desktop prime time, Google can plan its general release. Hardware manufacturers would now have a reason they had to support Linux. The power of Google branding should be readily apparent now. One interesting thing here is, this should worry Novell (and to a lesser extent Red Hat) at least as much as it worries Microsoft. The desktop market is a big piece of the potential OS pie. A move by Google could all but lock other Linux-based options out. The fact that it could potentially be a non-RPM distro makes it all the more interesting…
–jeremy
Google, Linux, Ubuntu, Goobuntu, Red Hat, Novell
Archive for January, 2006
Voting is now open for the 2005 LinuxQuestions.org Members Choice Awards. This is the 5th year we've done the awards, and turnout has gotten better every time. The response from both the community and the award winners has been outstanding. We've revised categories a bit based on feedback and improved the logic a bit. This is your chance to be heard, so go vote!
–jeremy
LQ, Linux, Open Source
Just booked a trip to LA for the upcoming SCALE 4x, which LQ is a sponsor of. Made this one a short trip – I'll be in LA less than 48 hours. I missed SCALE last year and it looked like a great show, so I'm glad I'm going to catch this one. From the looks of it, I'll be attending quite a few conferences this year, which is alway a great experiance. If you'll be at SCALE 4X and would like to connect, let me know. For once I'm actually staying at the conference hotel.
–jeremy
SCALE4X, Linux, Open Source
A quick follow up to this post. It looks like Linus has responded to the question of whether Linux will be released under the updated version of the GPL. The answer:
No. You think “v2 or later” is the default. It's not. The _default_ is to not allow conversion.
Conversion isn't going to happen.
Seems like s pretty definitive answer. I find two interesting things about the LKML thread. First, it seems Alan Cox doesn't necessarily agree with Linus on this and second, the thread starter is one Jeff V. Merkey – the same person who offered $50,000 for a BSD copy of the Linux kernel. You have to wonder what he's up to. One also has to wonder though, how close the final version of the GPLv3 will be to the currently released draft. Some of the new clauses do seem a bit restrictive, especially from a business perspective. It almost seems a bit too political. While I dislike DRM as much as the next person (actually, I'd guess I dislike it much more than the next person), I'm not yet convinced that a “Free Software” license is the place for that battle. It's a slippery slope. Where's the line? Will other items the FSF doesn't like creep into the next version? What happens when something the FSF doesn't like isn't something that people hate as much as DRM? Will the precedent already have been set? We'll have to wait for the final version is released to know what impact v3 will really have, but with it being incompatible with v2 (via it's additional restrictions) we could have quite a headache with some projects sticking with v2 and some going with v3.
–jeremy
GPL, GPLv3, FSF, Linus, Linux, DRM, Open Source
An extremely interesting read, Slashdot has sent questions over to the Microsoft Security VP Mike Nash, who agreed to answer them with no PR scrubbing. A ton of information is in the response, which is a fairly long read. As you may have guessed, so of the information is good news and others is not. It's clear that the original Microsoft security push in 2002 was nothing but lip service. It's also clear that, while security clearly has a much higher priority now, some within Microsoft are stilling ignoring that and just don't get it. An anonymous Microsoftie points out a specific example in one of the questions. The article also gives you a little appreciation for just how large Microsoft is. The shear number of initiatives, acronyms, procedures and policies in place is astounding. It's no wonder it's taking so long to turn the security train around. Old habits die hard and business pressures are currently clearly very high at Microsoft. Given the option of not getting a product shipped on time or shipping it on time with security flaws, I think most Microsoft teams are still choosing the latter. It is fantastic to see someone this high up at Microsoft speaking directly to users though and there is a ton from the Q/A that I didn't cover here….so I recommend you read the full article.
–jeremy
Microsoft, security, Trustworthy Computing Initiative, Mike Nash
So says this article in Forbes. Now, that's fantastic, but that's not what the EU wanted and is mostly useless. From the original EU press release:
As regards interoperability, Microsoft is required, within 120 days, to disclose complete and accurate interface documentation which would allow non-Microsoft work group servers to achieve full interoperability with Windows PCs and servers. This will enable rival vendors to develop products that can compete on a level playing field in the work group server operating system market. The disclosed information will have to be updated each time Microsoft brings to the market new versions of its relevant products.
To the extent that any of this interface information might be protected by intellectual property in the European Economic Area(6), Microsoft would be entitled to reasonable remuneration. The disclosure order concerns the interface documentation only, and not the Windows source code, as this is not necessary to achieve the development of interoperable products.
That's right, they specifically said they didn't need source code. What they want is accurate interface documentation. Why? Because in this context, the source code is not of use due to patents. If one of the Samba developers so much as looked at the code and then added a related fix to Samba, both the code and the developer would be legally tainted. What does that mean? That they wouldn't look at the code, of course! What would help, and what was asked for, was something that would be both useful and not legally encumbering. Full and accurate interface documentation. No more undocumented API's, hidden hooks and other measures that don't allow fair competition. What Microsoft has done here (and brilliantly so, from a PR standpoint) is appear to the average person to be offering more than what was asked for, while at the same time offering much less. There are still more details to come on this, so who knows – maybe the only code delivery methods will be braille and/or microfeesh. I've seen no official response from the EU yet and I notice a court date of April 24-28 mentioned, but I hope this move isn't allowed to satisfy the Remedy section of the case.
–jeremy
Microsoft, Windows, EU, Open Source
A quick update, since there seems to be more interest than I had anticipated asking for our experience with the upgraded LQ infrastructure. Apache 2.2.x has been absolutely rock solid as well as slightly faster (anecdotally) than the old 1.3.x. Nothing too substantial, but based on the CHANGELOG and the last month running it for LQ, I'd suggest the 2.2.x branch to anyone that asks. PHP 5.1.x has also been solid and some of the new features when compared to 4.4.x are fairly handy. If you're not tied to 4.x for BC reasons, I'd say go for that upgrade also. One HUGE caveat is that APC has been extremely unstable for us. I'm not sure if it's PHP5+APC or just that we are running different code, but we have seen consistent APC-related segfaults. There are many bug reports about it in the PECL bug tracker, so hopefully it'll be addressed soon. In the meantime, if you need to use APC for performance reasons, you'll definitely want some kind of watchdog on Apache to make sure things are running smoothly.
–jeremy
Apache, PHP, PECL, APC, LQ, Open Source, PHP5
With the number of registered members that LQ has (over 200,000), I though it made sense to add a little social networking to the site. Some of the bare basics were there, but the data was hard to get at. You could add someone to your buddy list, for example, but you couldn't easily view the network. As of a couple minutes ago, that changed. The feature is still in beta and is really in its early concept phase, but you can view the network starting at any arbitrary point now. Here's my network as an example (Yes, I really should add some members to my friends list). I can see some other items that would make this more interesting and it's not fully integrated into the site yet, since it's in beta. I'd also envision some other changes on the horizon to help you find other members with similar interests for example. Like I said before, the redesign was only the beginning – we have some cool stuff on the horizon (and we'll continue to fix bugs and make those small usability improvements along the way too).
–jeremy
Profiles, Social Networking, web 2.0, LQ, Linux
We all know that one of the benefits of OSS is the ability to avoid vendor lockin. One thing that often gets overlooked, however, is data lockin. As “Web Services” get more and more popular, the importance of data lockin will become more clear. One project that aims to help in this regard is FOAF, or Friend of a Friend. Based on RDF, FOAF allows the expression of personal information and relationships in a standardized way. I'm happy to report that as of today, you can get FOAF info for any LQ member via their member profile page. I'm still polishing things up a bit, but everything is working as of now. We probably should have supported this a while ago, but better let than never…
–jeremy
FOAF, semantic web, RDF, LQ, Linux, Open Source
As I announced first on the LQ Podcast, LQ is a sponsor of the upcoming OSBC. The Open Source Business Conference is brought to you by IDG, which also puts on the LinuxWorld Expo. The OSBC is in San Fransisco on Fed 14-15. I haven't bought my plane ticket yet, but I do hope to attend the conference. If any of my blog readers will be going and want to touch base, do drop me a line.
OSBC, Linux, Open Source, Business
–jeremy
Subscribe