Freespire Lives, Goes Back To Debian 104
nerdyH writes "Following Xandros's acquisition of Linspire, some feared for the future of Freespire, the free version of Linspire. However, Xandros today announced a new version of Freespire that will return the popular free Linux distro to its Debian-based roots."
Re:Popular? (Score:3, Interesting)
There's a small company in Chile that sells custom-built boxes with it installed, and they sell quite a lot of them to lower-middle income families in the capital.
Don't ask me why they picked it, I have no idea. Having seen one of these systems up close, they're really crappy (hardware-wise), but I guess they work well enough. They also provide tech support for a nominal fee.
Re:Xandros and Linspire (Score:3, Interesting)
The only thing they really "Have" is the application store. It's the only place I know of that is like the app store on the iPhone (the first of that type I ever saw actually) where it combines free and commercial apps, has a single install/remove point, is trivial to use automatically adds it to your menus, ...
The thing is, Ubuntu's is at least as good now, so I'm guessing that the only reason they have to stick around is so that some current users can avoid change.
As I've been told when trying to update the family's apps: "Nobody likes change"
Re:Xandros and Linspire (Score:3, Interesting)
The only way I can see something like LSB working is to have a distro that is ONLY LSB... as in, nothing else. But seriously... does the spec even encompass a whole working OS?
Re:Xandros and Linspire (Score:3, Interesting)
Oh, come on now. I prefer Debian myself, but I put my non-geek friends and family on Ubuntu. If I put them on Debian I'll have to explain so much crap their eyes will glaze over and they'll ask me to put Windows back on. It may be trivial to you and me when something goes wrong (an apt-get dist-upgrade breaks the menu system, or optical discs suddenly stop mounting automagically, or the wireless network card no longer shows up in the Gnetwork box), but to a non-geek any one of those things is a total deal killer. Ubuntu may just be Debian slowed down by a bit of idiot-proofing, but it's exactly that idiot-proofing that makes it usable for so many people. When those same people see my ion3 Debian desktop they're like, "WTF? Why don't you just use the easy one, like you gave me?". And of course I explain it's because a more CLI-centric install is easier for me.
Re:Xandros and Linspire (Score:5, Interesting)
I still haven't u what we need Ubuntu for. We have Debian. Granted, not that trendy but it works.
Except that
On the topic of working out of the box:
They often avoided a difficult (political or technical) decision and left it to the user. Who was supposed to "know better what to do", or to read and study in order to take any decision. Increasing the dedication necessary to run the system.
In short, Debian //never// went own to produce a system for someone who wasn't, at least, a hobbyist UNIX sysadmin.
Sure, many of these points are probably much better now, but this was surely the context that made Ubuntu a welcomed offering.
The greatest plague of modern computing is complexity. Debian tackles a whole world of it through its dependency work, testing, and dpkg/APT. But they still (leave?) left way too much unnecessary complexity into the system.
Why did Linspire's CNR fail? (Score:5, Interesting)
I'm honestly not sure why CNR [cnr.com] hasn't done better (which is to say, generate any noticeable use). It's free-as-in-beer, supports several major distros in a central location, offers social features such as reviews and ratings, allows grouping of apps into "aisles" for easy one-click installation and sharing, handles commercial software sales as well as free software installation compatibly and rather efficiently, and generally provides a rather nice experience.
Why has it wilted like a Friendster? Because it's not free-as-in-speech? Is Applications -> Add / Remove or Synaptic simply "good enough"? Do enough Linux users really object to their Microsoft deal and abstain on moral grounds?
Of course, I don't use it personally. And I'm not sure why. Would a FOSS version by a more credible member of the community generate more interest and enjoy some success?
Re:Why doesn't Debian adopt Ubuntu? (Score:3, Interesting)
Specifically, starting a bunch of unnecessary modules at boot just in case you may have that hardware may be OK for a system that "just works" but some of us prefer to optimize for faster bootup time and the reliability that comes from running fewer unknowns in kernel space.
(I'm not saying that Debian necessarily boots faster, as it will let you add all manner of services if you tell it to install the kitchen sink, but, well, you get the picture...)
Re:Xandros and Linspire (Score:3, Interesting)
I'm just going to respond to you, though it seems several people were suggesting the same thing.
Ubuntu is a bad example largely because their fork features major changes. Mint basically just includes codecs. PCLinuxOS was originally largely just changing the defaults of the desktop. Then are hundreds of active distros, many of which offer minor changes at best, yet pull away tons of developer time to maintain different repos and such.
Re:Xandros and Linspire (Score:3, Interesting)
Portage allows you to install commercial software like crossover-office. You still need a license, but portage will pull in the installer through the standard install process, and keep track that it is installed for dependency purposes.
Re:Xandros and Linspire (Score:5, Interesting)
My point still stands. The Ubuntu devs could have focused their efforts on Debian. Their distro today still is binary compatible with Debian.
Actually, they have focused their attention on Debian. You would have to examine Ubuntu and Debian source packages to really see just how much effort Ubuntu is putting into Debian.
In every distro there is a list of stuff that nobody wants to do. For instance, getting 'bashisms' out of init and other shell scripts so that a fully POSIX compatible shell (such as dash) can parse them correctly. Ubuntu tackled a lot of that list.
If you look at the Ubuntu source packages, you will see a ton of patches in debian/ , Ubuntu has structured their patches so that Debian can cherry pick from their improvements easily. Debian has and will continue to do this. For instance, if Debian just wants the patch that takes bashisms out of a given script, they can just take that and leave the rest.
Similarly, Debian security updates and other things are easily cherry picked by Ubuntu. Managing patches like this is very time consuming, Ubuntu could have said 'screw that' but they didn't.
Its a rather interesting symbiosis. While the projects are going in separate directions, devs from both camps continue to ensure that improvements remain isolated and rather portable.
My desktop is a mix of Ubuntu and Debian packages, for instance. Most things I use begin with Debian source packages, then I grab the Ubuntu source packages and get the patches that I want ... then make my own thing. Granted, this isn't typical use but it illustrates the benefits of a larger cooperative effort.
Re:Why did Linspire's CNR fail? (Score:1, Interesting)
It's not done better because it's a commercial project.
Rightly or wrongly, Linux people HATE anything with money attached. Unless that money is being given to them, with no strings attached, in which case they'll sit on it forever more like demented spinsters.
Re:Xandros and Linspire (Score:3, Interesting)
Why should Debian be like Ubuntu? New users could perfectly well use Suse, Mandrake or Red Hat before shit brown became fashionable.
While I don't actively dislike Ubuntu, I'm not a big fan because it feels like they equate usability with condescension. Many people, not limited to them, seem to think that it's impossible to create something that's both easy and powerful. I don't agree.
On the other hand, this argument also justifies its existence. The people who don't agree with me get to use what they prefer. Mostly I'm worried that a lot of new users with the potential to understand more won't realize that there are other options, since it gets so much publicity.
Re:Why doesn't Debian adopt Ubuntu? (Score:3, Interesting)
If I'm sure I won't ever be needing a specific service and I disable it manually, I'm not exactly worried that it won't get upgraded, I'm probably just trying it temporarily and will soon get by to uninstalling it. But if I were worried about that and I disabled it due to a false dependency, then I would restore things to the way the installer left it, and then upgrade as normal.
There are various ways to do that. The quickest way is probably to keep a separate directory for initscripts you don't want, and move the junk to there. During an upgrade, you can move them all back (maybe keep a list), do the upgrade, and then promptly put them back in their holding area. It's a terrible hack - but it's easy and quick regardless of whether its debian or something else. Another way is to check the scripts into a source code revision system, edit them to comment out the start section and then backtrack the changes during an upgrade. Yet another way is to write a shell script that creates a K link for each S link for each init level and removes the S links, with a mirror script that does the reverse. I may be overlooking an even simpler way of doing this with unionfs, but I don't think of it as anything more than a convenient hack which isn't going to cause any problems I can't fix.
If all of the above sounds like much ado about nothing because it only represents a tiny fraction of system resources, you may understand my motivation a bit better by opening a command terminal in Ubuntu, sudo bash, then lsmod. If the last time you did this was before ACPI was fully implemented, this will open your eyes because that list used to fit on one screen. After ACPI and bluetooth, that list grew by leaps and bounds, and I see drivers loaded in memory for hardware I *know* I don't have... I also won't accept the fact that you can't simply rmmod ipv6 if you know you're not using it. Once you load that module it sinks its teeth into the kernel like some sort of memory leech and won't come out. [Insert some George Carlin-like expletives here]! Unused code sitting in memory is a problem waiting to happen, and the larger it is the bigger the potential security hole.
Each running service is also a potential problem. Don't use the "at" daemon? Nix it. Don't use NFS? Maybe you can do without RPC... Don't use samba? No need for samba daemon either. The dictionary program wants to run a server by default... *rolls eyes*. Some game wants its own sound server to start before anything else. Gone. A desktop system wants to start a sound server each time it utters a sound.... There's all this stuff that wants to be running all of the time that could just have been linked to a library or something. So, excuse me if I sometimes use a hack to route the trash elsewhere...
Now, if we are to keep Linux viable for running on embedded stuff like digital cameras and the like, it's going to have to become more systematic to get rid of this sort of excess. Maybe someday this sort of optimization will get easier. Case in point: Did you know that it's technically possible to boot Linux from flash in as little as .5 seconds? On a 200Mhz arm processor?
If someday my Linux systems boot in a sweet fraction of a second, I figure it will be because debian or something similar will be installed, not Ubuntu, and certainly not some hardware-treadmilled adware-laden commercial product. Ubuntu just doesn't seem headed in that philosophical direction, and if modules are going to be resisting their removal by the superuser then that's not a good omen for the way things could become. I mean, if a software monopoly wants to bloat its OS to the point of nonfunctionality then by all means, but they should keep the brain damage to themselves and not require all Linux users to standardize on the same philosophy just to keep hardware companies happy.