TreoQuestions.org Officially Launched

I am happy to announce that TreoQuestions.org is now out of beta and officially launched. I'm looking forward to the challenge of growing a site from scratch, while also continuing to build and grow LQ (which is just about to hit 250,000 registered members). While this blog will remain mostly Linux and Open Source related, I'll likely now throw a Treo tidbit out there from time to time, especially as Linux and the Treo converge.
–jeremy
, , ,

Why the light has gone out on LAMP

I have no idea how things like this make news sites like slashdot (or maybe I do I guess…). It's a couple paragraph rant that contains no real facts, huge generalizations and not much more. The fact is, MySQL and PHP are just fine if you use them correctly. Ask Yahoo! if they scale. The thing is both will allow you to be lazy, but that's your fault – not theirs. The fact is, with PHP5 and MySQL 4.1/5.x both offer feature rich solutions that are highly scalable. One thing I notice about PHP and MySQL bashers, is that they are stuck in MySQL 3.23 and PHP 4 land. Things like:
For years the MySQL developers were quite vocal that things like referential integrity, transactions, subselects, etc were little more than baggage that could be done better another way. Let me clue you in: you need these things or you need to let someone else handle your database work for you.
Let me clue you in – MySQL has supported these things for a long time now. Now, don't get be wrong – PHP and MySQL historically had major problems (very real problems – like magic quotes and register_globals being on by default) and still do have some problems. If you are going to bash them, at least know what you are bashing, as opposed to comparing them to BASIC..which is really a bit silly. PostgreSQL and Python are both great, but as always the right tool for the right job (and knowing what each tool actually does is a huge part of that).
–jeremy
, , ,

The 25 Worst Tech Products of All Time

I found it interesting that 3 of the top 10 in the 25 Worst Tech Products were of Microsoft origin. The one that really sticks out to me is Microsoft Internet Explorer 6 (2001), coming in at number 8. The fact that the most current browser from a company that has a huge majority share of the browser market is over 5 years old is almost unintelligible. Think of how much has changed on the web since 2001. That's redefining stagnation. Ask yourself if they'd be working so hard on IE 7 if Firefox hadn't seen its meteoric rise in popularity. Monopolies are bad – and this is just a single example of why.
–jeremy
, ,,

Google Follows Through

Following up on a topic I've talked about before, Google has released Picasa for Linux. As speculated though, it's not a native port, but a Wine rollup. It's great to see Google follow up on the promise they made to support Linux apps, but it is a bit unfortunate they went the Wine route IMHO. There are a lot of downsides. First, it *looks* a bit out of place after you run it, ie it looks like a Windows app stuck in Linux and has absolutely no Window Manager integration. The package includes a bundled version of Wine, so is a hefty 25M. If other companies follow, will we now need to have 30 different wine installs (since this does not use a standard win install, even if you have one). I'm hoping this isn't a dangerous precedent where Linux support equals Wine. Now, unlike some, I'm not a Wine basher and think it is a useful piece of software, but if Linux support starts to equal “runs in Wine” we will forever be intrinsically tied to Windows and that's a very bad thing. I'm hoping it's only used as a sort of temporary migration platform while they gauge demand and get more comfortable with Linux as an end user platform, but that remains to be seen. All in all though, kudos to Google for staying true to what they said and also for potentially furthering the exposure Linux gets. One truly great thing here is that Google contributed over 200 patches back to Wine, so even if you never touch Picasa it's very possible you may benefit from this. More info is available here and here.
FWIW I did download and install this and it does run very nicely for a Wine app. Google makes 3 downloads available – RPM, DEB and a Loki-based installer. They even contributed some patches back to Loki. It should be interesting to see if they continue this three release paradigm as they support more of their applications on the Linux desktop. This is certainly a sticking point area for Linux and how Google handles it could set a precedent for many.
–jeremy
, , , ,

Mac OS X Kernel Source Now Closed

That (or something close to it) is the title of many articles that seems to be circulating. The odd part is, OS X was *never* Open Source code in the first place. Nor did Apple pitch it as such. Carbon, Quartz, Cocoa – you name it, no code. Does sound like a good headline though, doesn't it? What the articles are actually referring to is that Apple is no longer releasing the code for XNU, the OS X kernel, for the x86 platform. I don't follow Darwin too closely, but it seems to have been dead for some time now. The reality is that Apple just isn't an Open Source company. They claim to support OSS in their marketing, and do in fact release a fair amount of code, but it's just not their bread and butter…nor is it really their culture. They were, however, able to leverage OSS to make a nice platform.
Much of the speculation going around centers around why Apple chose to do this now. One of the most popular theories is that it's to prevent piracy. I'd hope they have a better reason then that, because it's clear that making something closed source in no way prevents piracy. Other theories abound though. Could it be because of an embarrassing disparity between the PPC and x86 trees? Or does it go deeper than that. Is Apple working on a non-MACH version of the kernel and trying to slowly work some items in without people knowing? Are they working on implementing the full Windows API to enable you to run Windows apps right in OS X? With the next release of OS X coming up fairly soon, I'd guess we'll know more soon. In the meantime, if this is a subject that interests you, I suggest reading these two extremely informative articles.
–jeremy
, , ,

Microsoft considers taking admin rights from employees

From the ZDnet article: As Microsoft moves its internal desktop systems to Windows Vista, the company is contemplating whether to change a long running tradition and take away admin rights from its employees in order to improve security.
This explain quite a bit I think. While some job functions absolutely need admin privileges, every employee at a company as large as Microsoft is a bit much. I think, though, it shows the mentality of the company when it comes to security. That mentality comes though in the end product, which has been historically ridden with security issues. This may also explain why so many applications that quite simply should not require admin privileges do – if Microsoft runs everyone as admin, that's the example that's being set. I've seen many companies give up and give people admin for the sole reason that it's the only way they can get things to work. Enter a decade of viruses, malware and fun. Good to see this mentality may finally be changing, although it's a bit late. Maybe in 5-10 years most Windows software will understand the proper security paradigm and the security issues will calm down. The Vista successor should be out by then ;)
–jeremy
, , , , ,

Red Hat Not Satisfied with Sun's New Java License

This seems like an odd move by Red Hat. Now, I know they don't necessarily get along with Sun, but calling them out on the Distro License after Sun has announced its intention to Open Source Java and giving incompatibility with OLPC as the main reason?? Come on guys, you're better than that. I can only hope the article is a poor summation by the journalist, which is certainly possible these days. While OLPC is an amiable goal (one I should discuss on this blog, and will at some point), it's not a reason for any company to base this kind of decision on. Looking at the OLPC specs, the Sun implementation wouldn't even run in a usable manner anyway. Additionally, Red Hat contributes to classpath and could include that implementation. The one valid reason given in the article is that Red Hat would like to make some specific improvement to Java, centering around real-time capabilities, and I'd guess we'll see a 64-bit Java plugin available soon after it's officially Open Sourced. It will be interesting to see what Red Hat's reaction to the Open Sourcing of Java is if/when it's not GPL compatible. It constantly seems like the biggest proponents of Open Source Java are direct Sun competitors, which along with what Microsoft tried to pull is probably why Sun has been so skiddish on the issue. It should also be noted that, from what I can tell Debian does not approve of the license. If we reach a point that Debian approves of a license that Red Hat doesn't, we're probably in trouble :)
–jeremy
, , , ,

Novell Partner Linux Driver Process

Novell recently did a press release entitled “Novell Delivers Device Driver Breakthrough to Accelerate Linux Adoption. With device drivers being a sometimes maligned aspect of Linux, I was interested in exactly what Novell was going to do. After reading the press release, I really had no idea. After poking around the Novell site a little, I have a slightly better idea, *I think*, but there's still not a lot of technical detail available (or I am missing it). It appears that if you join the program and develop your driver in accordance to the Suse Kernel Module Packagers Manual, then your driver will be available via YaST and you'll also be notified in advance if an ABI change will break your driver. If the driver is certified, you can also get some level of support for it from Novell. At a certain level of certification, you can even ensure that your driver is available at the same time a kernel security update is released. Once nice thing here, is that Novell appears to be making an attempt to encourage development to take place in the mainline kernel, while allowing this as a way to obtain a driver either 1) before the driver is accepted upstream or 2) without having to upgrade your kernel. Both of those are nice, but hardly a “Breakthrough”, so I think I must be missing some other aspect. I'll have to keep reading. Once thing I don't really see mentioned is binary drivers. Will they be allowed as part of the process? If so, will a binary driver be allowed to certify somehow (say a Novell NDA)? I'm also waiting to see if Greg K-H comments on this, so I can see the nonmarketing side of things.
–jeremy
, , , , ,

Sun promises to open source Java

In a session at the JavaOne conference, it looks like the new Sun CEO has promised to Open Source Java.
“It's not a question of whether we'll open source Java, the question is how, ” Schwartz told delegates in his opening keynote at the tradeshow.
Interestingly, it hasn't created as much buzz as I'd have thought. The one thing I keep seeing that's incorrect is that the source to Java isn't currently available. It is and has been for a long time now, it just isn't available under an Open Source license. It will be interesting to see how this progresses. No timeline has been given and Sun certainly still seems a little apprehensive. There are a lot of other questions here also. What will the current Open Source implementations do? How much of this was was caused by C# uptake. How will Microsoft and IBM react? I'd guess the answer to most of those questions will depend on how the Open Sourcing objective plays out. No license has been officially announced, but one has to think that the CDDL is a likely candidate. No matter which license it is, it almost certainly won't be GPL-compatible. Also, Sun seem very concerned with forking. With the right to fork being one of the main Open Source ideals, they'll have to be careful how they go about trying to prevent forks. It will probably be a trademark enforcement, but if they try to work something into the actual license, it could get messy. On the Linux side, you have to wonder if this is going to get more distros to ship with Java preinstalled. I'd guess it once again will likely depend how Sun proceeds. So now I see why the issue may not be sparking the buzz I had expected. While it's great that Sun is continuing their commitment to Open Source, with so many aspects of this still up in the air, many people may just be taking a wait and see approach.
–jeremy
, , , , ,

Can the ordinary computer user ditch Windows for Linux?

That's the questions Mark Golden recently asked in the Wall Street Journal. His conclusion?
For me, though, using the Linux systems didn't make sense. I often send documents and spreadsheets between my home PC and the one at work, which uses Microsoft Office. And the files are sometimes complex. Meanwhile, for both personal and professional computer use, I want access to all multimedia functions.
While solutions may exist to almost every problem I encountered, I was willing to invest only a limited amount of time as a system administrator. Claims by some Linux publishers that anybody can easily switch to Linux from Windows seem totally oversold.

Despite a few minor errors (Linux was really written as a Minix replacement, I'd hardly call Usenet in 1993 a “Web bulletin board”, Linus really is not in charge of maintaining central Linux standards) I think Mark gives Linux a fair assessment (although I think he would have faired slightly better with newer distribution versions). Linux isn't ready for the desktop. But, things really are changing. During the first round of Linux on the desktop talk about 4-5 years ago, Linux wasn't ready for technical reasons. To a large degree, that has changed. It's not ready now mostly due to legal encumbrances, software patents and other non-technical stumbling blocks. For many reasons though, I think these will be much harder to overcome than the technical obstacles were. You see – the Linux community is filled with some really smart people that excel at overcoming tough technical issues. We were bound to fix the technical issues. The Linux community however is not necessarily filled with people interested in marketing and it's especially not filled with people who are willing to concede freedom in the name of marketing. Look at the drubbing Linspire took when they released what from one angle can be seen as a legitimate attempt to fix the problems Mark had.
As you know, I've covered this topic a lot recently. I'm coming to see this situation as sort of the Ying and the Yang or the fire and the water. On one side we have the people who will fight for freedom, do what's right and solve the technical issues. On the other side we have the people that are interested in making things easy and catering to the masses. The intersection is that they both want to help Linux adoption (albeit for much different reasons). I'm slowly coming to an understanding that both groups need each other more than they may think at first. You see, the first group – they want open specs, open solutions and Open Source. But, with the current marketshare, it's not hard to understand why some companies aren't listening. Very few companies are interested in doing the “right thing”. They are interested in maximizing profits. So in come the people who are attempting to market Linux to the so called unwashed masses (ie. the ones who don't see there computer as a tool or something to tinker with, but merely as a way to check email). So, the question becomes can both sides meet at the intersecting point and agree to work to the same end via different means? Can we work towards to end goal of group 1 (which I think in the right one, personally) via group 2 gaining marketshare and mindshare in the main stream while avoiding the binary doomsday scenario? I'm still thinking about that one, but I'm interested in what others have to say.
So, does the fire need water?
–jeremy
, , , ,