Fedora Core 1 Released 566
EvilAlien writes "The Fedora Project has released Fedora Core 1, aka Yarrow. The release was expected on November 3rd, but was briefly delayed. The release notes has quite a bit of good detail, and is worth checking out for any preliminary questions you may have. Download options include BitTorrent in addition to the traditional collection of FTP mirrors."
Still concerns about security errata (Score:3, Interesting)
Still, looks like RH's first-rate QA has been put into place (unlike in Mandrake), and hopefully they'll keep that up as the community gets more involved.
M
Sounds like... (Score:2, Interesting)
The Fedora Project is a Red-Hat-sponsored and community-supported open source project. It is also a proving ground for new technology that may eventually make its way into Red Hat products.
Hmmm, what does that remind [apple.com] me of?
best feature: up2date does apt and yum! (Score:3, Interesting)
silly, but not... (Score:4, Interesting)
VNC installation (Score:5, Interesting)
That's really cool, and more useful than it sounds... I was looking for just this feature several months ago when installing RH on a laptop whose video card was supported by XFree but for some reason wouldn't work with the graphical installer. (Tweaks were required for the configuration file.) I know there's a text-based installer as well, but it's so much easier to select packages on the GUI install. It sounds like this will be a nice successor to RH 9.
Re:Features (Score:3, Interesting)
Re:Still concerns about security errata (Score:1, Interesting)
There's still one thing I don't quite understand, though -- if Fedora will use new upstream packages for updates (which is undoubtedly useful in many cases and stops stagnation), won't it cause a nightmare for other package maintainers?
For example, say a security issue is found in Mozilla 1.4.1 (the Moz shipped with FC 1). If the policy is to release Mozilla 1.5 as an update, that'll require new Epiphany builds too (and other Gecko-dependant packages), which in turn could require new GNOME supporting packages.
In the end, this means that a tiny hole in Mozilla could lead to a huge amount of updates. Yes, this is a worst-case scenario, but I'm not sure how the no-backport policy will work here.
Re:Features (Score:3, Interesting)
This is the way BSD does it, and I like the idea a lot. Do you have a link to this ISO? I checked their download site, and no boot.iso was listed. Here's what I see:
Parent Directory 04-Nov-2003 11:49 -
MD5SUM 04-Nov-2003 12:00 1k
yarrow-SRPMS-disc1.iso 04-Nov-2003 11:39 610M
yarrow-SRPMS-disc2.iso 04-Nov-2003 11:39 610M
yarrow-SRPMS-disc3.iso 04-Nov-2003 11:38 610M
yarrow-i386-disc1.iso 04-Nov-2003 11:38 630M
yarrow-i386-disc2.iso 04-Nov-2003 11:38 637M
yarrow-i386-disc3.iso 04-Nov-2003 11:38 616M
Re:um.. (Score:4, Interesting)
A couple of other links (Score:2, Interesting)
Check out this [xades.com] for more details.
There's also a lot of extras at Freshrpms [freshrpms.net] (although not updated for Yarrow yet).
Stability? (Score:5, Interesting)
I looked around the website and they don't really explain how important a priority stability is. They do have a QA "project", but they also say they want to "Be on the leading edge of open source technology..."
I take it we're not talking OpenBSD/Debian-stable level of reliability. That's fine. But what's the goal? Will this stuff be /directly/ used by RedHat, or is there a "polishing" step?
If you are talking about this stuff ending up in RH Advanced Server, then yes, this will be heavily polished before release.
If you are talking about RH Linux 10, well, that won't be happening. This is the new world right here.
Speaking as someone who runs Mandrake Cooker (someone pick up that reader who just fainted) stability concerns really don't worry me too much as I can hack the problems as they occur :-) However, this stuff has gone through basic Redhat QA so it can be assumed that it won't eat your dog or sleep with your wife :-)
This is a distribution for the release early, release often crowd. The primary release (which this is) should be treated as being a reasonable base to build on. Once you hook up the apt-get or yum tools to the respective repositories, upgrading broken packages should be easy enough. Fedora will be making an appearance on my laptop in the next week or so - time will tell whether Fedora is stable enough. If you are nervous about being an early adopter, sit back and watch the forums, newsgroups and mailing lists for show stoppers that might hit your configuration.
I'm happy to see Fedora hit the streets. I've been running RH 8.0 on this laptop for a while now and I miss the absolute bleeding edge that the Mandrake Cooker tree gives me. Running Fedora on this laptop will allow me to track the latest stable release series. Mandrake Cooker allows me to track the latest developer releases on my desktop box.
Cheers,
Toby Haynes
Re:One important issue... (Score:5, Interesting)
All the flaming of Redhat for switching to a model that resembles a commercial Debian has been amusing. They take away boxed CD's that you could buy for XX dollars (which tons of slashdotters would flame anyways, 'who would pay for what you can get for free', as seen by SuSE 9 threads) and replace it with a leading edge distro that focuses on quick updates that would be impossible to accomplish with a boxed distrobution method (call it debian unstable). They will then take what they learn from Fedora and incorporate it into Redhat Enterprise line of software (think of it as analogous to debian stable).
Of course that asks why use Redhat instead of just Debian? For personal use it really comes down to flavor, but for Enterprise use it is an easy question.
-Eyston
Re:Why? (Score:3, Interesting)
Fedora Core is the best Redhat so far, I'm using it flawlesly on my notebook since beta2.
Being in bussiness of which important part is Linux on desktop, I'm really happy with this decision. While normal distro is cool for normal user, it's not so good for corporate user.
Reasons:
1. Don't need 5 programs fo each need, I need one, and one that works.
2. Don't need such urge on being up2date with everything
3. I want terrily tested and really working stable distro
4. Give me a clean distro and I'm filling only the gaps that don't suit me or the gaps that are not covered
As home user:
1. I want to test software to decide which one is better
2. If let's say Gnome 2.6 would be today, today would be the day I would be installing it, some softwares I keep daily with cvs
3. I want everything
4. Read 3, I have everything
DIfferences between Fedora Core and Readhat WS Professional are the same as I stated with previous descriptions.
But as here Fedora Core is just a test bed for Redhat WS (difference is as follows, al features are tested in Fedora before they are implemented in WS, Fedora Core is accessible for users to interact but WS isn't, Fedora will support outside package repository, WS won't), everything goes as I want, I get one fresh and up2date and one stable for the price of one.
Why should you run this? Because people that were bitchin over Debian being more_free, well this Redhat is practically the same but with Redhat support, because developers are still working on Fedora, hell they test there things for WS, so I doubt that Redhat would stop working on Fedeora.
Killing their desktop? No, they just extended it to likes of more people
Re:RedHat Sucks (Score:1, Interesting)
Way back when, there was really no time set on how long support for a release would last. It was just 'a long time'. About a year ago, RedHat said of their support that they had a new policy:
Redhat up to 7.3's support would end on December 31, 2003. (In the same announcement, they closed off support for some even older releases like the 6.x series, giving 6.3 support for just a little longer).
Further, any new release would be supported for a minimum of 1 year, but with no guaruntee of longer.
So, yesterday, with the Fedora release imminent, they said exactly the same thing; all the Redhat releases were getting their one year of update support, and being cut off -- exactly what they said months and months ago.
I procrastinated too and am screwed right about now - I've got a collection of 7.1 servers I'm going to have to do something about. But that's my fault, not RedHat's. They said up front what their support policy would be. If I really want the long term support, Redhat's enterprise offerings will do five years. That may well be where I'm headed next.
Re:A couple of links (Score:4, Interesting)
The King is Dead, Long Live the King (Score:3, Interesting)
Red Hat still puts resources into Fedora.
Red Hat still puts QA into Fedora (in fact they caught flack for delaying it).
Red Hat opened up the development to outsiders through Fedora.
Red Hat changed the products name into something everyone can use and sell.
We get a more open, supported, release often OS.
So far this sounds great. In the coming months we will see if this really is a win/win.
Re:A couple of links (Score:3, Interesting)
I'm quite happy with the new kernel version, since it supports DMA mode on my new SATA drive, while the kernel in RH9 would occasionally hang if you tried to turn DMA on. Not a huge thing, unless you happen to have an affected drive, in which case it makes a world of difference. There are some other kernel changes that should make life nicer for laptop users.
Some other random changes that I notice in their release notes:
They've also removed a number of packages that are either no longer necessary (e.g. QT2, LPRng) or have licensing issues (e.g. pine, some aspell packages).
One other big change is that they're moving from using ASCII to UTF8 wherever they can. That's one of those things that's kind of annoying while the switch is taking place (they mention that this could cause problems with telnet and ssh, since they don't explicitly negotiate the encoding) but will presumably make everything easier in the long run.
This looks good to me. (Score:1, Interesting)
- I always liked Red Hat, but found their release schedule aggravatingly long, and when the products came out, slow to release updates.
- I like the Red Hat systems since 8.0, on a whole. I've been using Red Hat since 4.0, and it wasn't until 8.0 that it became a sufficiently stable, worthwhile distro, IMO.
So, I think that Fedora will continue to build -- and, most importantly improve upon -- Red Hat's strengths. And if Red Hat tanks, who cares? It's a community-based system that would still keep ticking.
not too late (Score:4, Interesting)
Re:Fedora vs. RedHat, and RHCE (Score:4, Interesting)
And no, I don't expect to be sued. RedHat understands the consequences of releasing software under the GPL. Besides, they would have to be zarking mad to try sueing a public library for publishing GPL software. They would be reviled second only to SCO.
If they find something I have missed I expect they will drop an email, I'll make a corrected set of images available and that will be that. They aren't at war with us, they just made a business decision that non-enterprise customers weren't all that profitable. Personally I think it is going to cost them in the long term, but that's just my opinion and it is their call to make. They are the ones who answer to the shareholders.
Re:Still concerns about security errata (Score:3, Interesting)
(I would use a different description, maybe "with an open development system", rather than "open-source", since the are neither mutually exlusive nor mutually inclusive)
I really wanted to know though how that differed from (say) Debian, Gentoo and Mandrake (who have been had open development systems for at least a year each, especially Debian).
If a lot of people want backported security fixes, there's nobody stopping them from doing the work and putting up an apt or yum repository with those packages.
Sure, but considering it takes time and hard work to get on the early vulnerability annoucement lists, it is unlikely for this to happen any time soon, so Fedora-lagacy updates will be a few days behind other distros.
Re:um.. (Score:3, Interesting)
I was told that it was a long-term goal, 6+ months at least before anybody would be allowed to contribute.
Any idea if those plans have moved forward?
(For reference, if this is shown to anybody else who participated in the discussion at the time, my handle was "ElectricElf")
Open Source Licenses Software, Not Freeloaders (Score:3, Interesting)
Unless you're a stockholder, RedHat owes you squat. If you are a stockholder, RedHat has a moral responsibility to make a profit and pay you dividends.
Re:What the... (Score:3, Interesting)
Now, to explain why you need that much, is that the entire installer is loaded into RAM on a RAM disk. It makes testing easier, and it speeds up in the install a lot. If the installer is on disk, and memory is tight, you end up disk seeking a lot to read the program you want to execute, along with the fact, that you probably lose the disk space the installer was on, and a partition (any other way has a chicken and the egg problem). If the installer mounted loop back on the CD, it takes more room on the CD (because loopback can't be compressed that I know of), and then you'll have to seek around the disk.
You need that much RAM for the filesystem, which eats into how much RAM you can have use for the running programs. They could probably build a different installer that took a lot less read (16-32) RAM, but it would be much harder test both methods (RAM disk, loop off the CD, and network basd installs would be fundamentally different, where as right now most installs are nearly identical), so they opted for the one that has the most flexibility. If you need a very small install, do the install on a big machine and use cpio or tar to copy the files over to the dinky install machine.
You can get a RedHat machine to run on a 32-64MB machine if you run the right software. You can't run Mozilla. You can't run a full GNOME or KDE desktop. However, if you are willing to crack out fvwm or twm, it's not a that hard to get a GUI desktop. You'll need to slim down your Linux kernel, and remove stuff you don't need. You might need to rebuild your glibc to take up less space.
You'll need plenty of patience, and plenty of swap to do it on, but heck, I ran X on a 4MB RAM, 20MB of swap on a 386 25Mhz. I used fvwm2 and I couldn't run anything except lynx to browse the web. I read my e-mail in elm, and ran elvis as my text editor. Last time I checked, most of that stuff or something nearly equivilent is still on the RedHat CD (minimal vim, and links).
Look at how much more your current machine can do. I run Linux (Leaf based distro) off a 64MB flash disk in a dinkly little router configuration. (It has 512MB of RAM, but that's because that was the only stick of memory I had to spare when I built the thing, it'd work with no more then 16-32MB of RAM easily, probably 8 if I worked hard at it). It's about what I used to do with my linux installs. It's not like the resource conservative installs made great desktop machines. Sure they we're great if all you wanted to do was run a bunch of xterms, and maybe have a cute background. All the features that are on RedHat cost memory. OpenOffice has the ability to read Office Documents, and the ability to do all that WYSIWYG formatting. 6-7 years ago, you did that via a text editor an latex, easily done in 16MB of RAM. That will still work today on RedHat. You can browse the Web using Mozilla, that takes a lot more memory then Netscape 3.0 did when rendering stuff. You can't use KDE with all it's eye candy and slick integration, you used to do that with fvwm or twm (or God forbid, by switching virtual terminals, and screen). Six years ago you could run X in a lot less memory, and I'll bet you still can if you dumped all of the advanced rendering extensions used for 3D/DRM and other recent feature advancements. If you want smaller memory foot print, recompile all your apps to take out PAM support, to take out LDAP integration, to take out the X app support. Go back to running straight up inetd,
Re:A couple of links (Score:4, Interesting)
No MP3...as long as the license for the codec is what it is you'll never see anything even remotely associated with Red Hat including it.
No 2.6...well 2.6 is not ready for the parameters of this type of release yet. And 2.4.22*.nptl does moderately rock...
prelink is absolulely amazing
I'm actually looking forward to how the "extras" path will pan out. For me on my personal boxes Fedora is a no brainer
but not for my servers.
unlike many of the
For them it's the "security" factor.Easily understood in their finacial world
For me it's the oppertunity to finally pay back Red Hat for some quality production level code that I have used over the years.