Slashdot Mirror
Red Hat Linux Project Merges With Fedora
An anonymous reader writes "Red Hat has announced a merger of its Red Hat Linux Project with Fedora Linux, a group that has specialized in providing high-quality RPM packages for Red Hat. According to Red Hat, '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.' From the FAQ: 'Rather than being run through product management as something that has to appear on retail shelves on a certain date, Fedora Core will be released based on schedules, set by a steering committee, that will be open and accessible to the community, as well as influenced by the community.'"
Sounds like RedHat is trying to achieve some of the advantages of Debian. I'll welcome this, although I won't switch any machines over right away.
It'll be nice to get new software packages and rpms. I think apt-rpm has illustrated the need and the market for this. RedHat also has several great advantages over Debian, notably the installation process and more up to date software, so this could really revitalize them.
With projects like Linux From Scratch and Gentoo, distribution-building has gone fomr being an arcane art of wizards to something the community can do, and I'm glad RedHat wants to partner with the community in doing this.
Secession is the right of all sentient beings.
I think it's interesting that there is what appears to be a "core" part of the Fedora team focused on artwork.
This, alone, is an excellent move by RedHat to compete with Microsoft in a space they clearly lead the market - desktop UI.
As the Fedora site says, "Making things look pretty is the name of the game."
Unfortunately what needs improvement is the GUIs of the programs, not the desktop itself. Even the best desktop is no use if 2/3 of programs have awful GUI or are commandline only.
I know this doesn't sound ideal, but you're really in the same boat with any other OS, even Windows. (Some hardware works only with NT/2000 or 9x, not both, plus old hardware often loses support.) Buying hardware without checking driver status leads to pain.
I don't think Fedora can make this better, only the hardware vendors can.
As for documentation, try checking out the RedHat manuals. That and a good introduction to the Unix command line and vi/emacs should cover you.
Yeah, but most people who are trying out Linux don't go out and purchase hardware first, nor would anyone really expect them to...
Occurs to me that RH basically bought a QA system for packages. Since in a linux distro, apart from the kernel pretty much anything is a package, it makes one wonder if they were thinking their own QA wasn't good enough.
"Release fast release often" ring a bell? Red Hat is in the business, what, 8 years, and they're heading for a double digit main release. Way too much even if you're only in the business of putting something on retail shelves.
Perhaps they were afraid of another Drake emerging from this project or saw it as an opportunity to let the community do more of the groundwork and then serve it up to businesses.
They "have a release scedule and a steering committee"? Gosh. So do the BSDs.
My bet is that Fedora will move "up" the release scale while the various Red Hat "Advanced Server" products will move down the scale. Fedora will be more like "rawhide" and AS will be more like Debian stable. Both will remain free [as in beer] but the only *easy* way to get the exact set of RPMs that constitute the Adavanced Server line will be to cough up some money. This still won't get you the support, updates, etc. It just means that Red Hat can't stop you from finding and assembling the exact same set of RPMs as constitute AS. Otherwise they violate the GPL. The RPMs will still be available and downloadable, Red Hat just don't have to provide the ISOs unless they want to and they still meet the letter of the GPL without competing with themselves by giving the product away for free.
I don't consider this at all bad. Red Hat makes more money as a *service company* selling a very stable version of Linux to companies and organizations that are willing to pay for the service. They continue to support the open source community by providing Fedora. They don't have to continue to be both on the cutting edge and providing a stable product at the same time through the same product. Linux continues to advance through Fedora with new versions getting "released" and Red Hat incorporates the results into AS when it is sufficiently shaken out. Red Hat benefits from Fedora by allowing them to steer more so than they would be able to otherwise.
One other benefit: this also takes some of the competitive pressure off of Mandrake and some of the other mainstream (not just for developers like say gentoo) consumer/desktop distros since Red Hat effectively pulls out of the "boxed set for end users" distro business.
They that can give up essential liberty to obtain a little temporary safety deserve neither safety nor liberty.
Ben
Yep great plan. Lets see I have a well trusted and for the most part well respected brand name. Perfect business plan is then:
1. Change product name to something few have head about.
2. ????
3.Profit.$$$
Repeal the 17th Amendment TODAY! Also Please Read http://www.gnu.org/philosophy/right-to-read.html
Not a dumb question at all - this is one of our major worries about RedHat at the moment too. However, looking at their site (and FAQ), it seems like this (Fedora) is going to be very similar to the RedHat we know at the moment (not the enterprise bit) - ie, major releases every six months or so. It's quite reasonable to expect support for the old releases to fall off very rapidly (it's really not economical to continue to support more than two distributions at a time), and there is certainly no indication they would do otherwise.
RedHat are, IMHO, likely to continue the push for this rapid development so that they can entice more customers into their relatively slow moving Enterprise product.
Why can't we all just get along?
Red Hat linux has also changed dramatically sine the 7.xx days; I think your best bet will be something newer (and the documentation specific to what you're running).
That was the single most frustrating thing I personally came across: documentation that didn't actually apply to what I was running (editing XFree86, for instance, when that had no effect and I needed to be editing XFree86-4).
anyways, having said that, the hardware that is supported under Linux is always improving: I was amazed that my previous sources of frustration (winmodems, multimedia keyboards, etc.) were configured and working automagically with the latest installs (I'm not claiming things are perfect, but they're not bad now, and improving at an amazing speed).
Good luck!
Pragmatically speaking. How hard would it really be to produce "legitimized" versions of protected software (particularly multimedia stuff I am thinking) for linux? I think an awful lot of people would pay a little bit a least for programs that work and are legal. I think these patents stink don't get me wrong, but what do we do in the meantime? Am I missing something here? Is this a case where peoples idealism is stopping production or are there other problems with making this work legally on Linux?
The project will produce time-based releases of Fedora Core about 2-3 times a year with a public release schedule.
.0 a .1 and a .2 in 1 year? 3 releases for the core a year sounds rediculous to me. The core is the thing you want to be stable as a rock, not being in beta forever, which is basically what a 3 time release schedule means.
.
So will RedHat release a new product everytime a core gets delivered? Will we see a
The release cycle of linux distros is what will kill them eventually if they don't slow it down. Most of them have 2 releases (not major ones, but new boxed sets anyway) a year. And they all want the users to pay for them. That's only logic, they're running a business. But the linux distro's and the software they deliver seem to be in eternal beta. People always want the latest and greatest I guess. Lots of distro's have close to 0 people running their stable release. The thing 'we' are all so proud of (stability and security) will be going down the drain real soon if we don't start focusing on them again iso getting a filemanager #311 and a desktop #24. Lets first settle down and get everything stable. And then have a look at what needs a change.
If I buy a distro version 9, it has a lifecycle of 6 months, a year at most. Then I do need to upgrade. if you want businesses to adopt your distro or joe average to use it, cut the upgrades down. It looks silly... We are so stable and secure, but you need to upgrade every 6 months to keep up. A business doesn't want to be in an eternal upgrade cycle. Neither does Joe Average. They want to get work done. Not upgrade or do a complete reinstall with the next release just a few weeks after they have their configuration just as they want.
I started using Linux in 1996 because I wanted something different, a new challenge. I loved the "if you don't need the new feature and it is not a security thing, why upgrade program X?" mentality. Now it's just the other way around. My wife is still running Windows 98 SE on the laptop. That was released what.. 5 years ago? Sure... there were upgrades for a lot of things... but did she need to upgrade the OS itsself every 6 months ? No
*sigh*... I'm getting old I guess... nevermind me.... I just want my Linux to be stable, secure, and also all the apps i'm running on it. And preferably without losing all support for it because i'm running a distro that is more than 1 yr old.
Sure, my computer doesn't crash when 1 program does. But the program shouldn't crash. I want that to be fixed, not another feature added. Microsoft won't kill Linux... It's doing just fine on its own.
What kind of lame idiots call it a GUI when a click of an icon brings up a text interface window. Also, I believe there is way to much "Burger King" GUI programming going on in Linux. To many projects doing it their own way. There needs to be lots more standards put in place, starting with the desktop itself. There needs to be a merger of the features of GNOME and the 'look and feel' of KDE into one standard desktop before Sun's MadHatter muddies the water anymore. Frankly, I'd like to see allot more organization to the whole software for Linux arena. Not anti-competition, but a standardization of package distribution and compatibility. It's almost getting to the point that you can't run 2 programs at the same time without re-writing one of them to work with the new or outdated support package that another program needs. I was completely floored when I found out that Apache even changed it's file locations when going for ver 1 to ver 2. Ok, I'm ranting, but I've finally gotten it off my chest. Let the flaming begin.
Grr... Don't feed the trolls...
...
Let's see -
JPEG - Joint Photographics Experts Group
They have standardized it, and it's royalty free, AFAIK, but they still own it.
MPEG - Moving Picture Experts Group
They have standardized it, but it IS NOT royalty free, including
MP3 - Moving Picture Experts Group (MPEG-1 Layer 3, to be exact.)
While involved with MPEG, Fraunhofer IIS-A and Thomson worked on and patented crucial parts of the MP3 format, AND THEY DO LICENCE IT.
REDHAT CANNOT LET YOU DOWNLOAD IT WITHOUT BREAKING THE LAW! What about this can't you idiots understand???
Read this...
http://www.mp3licensing.com/index.html
Grr... I won't feed the trolls, I won't feed the trolls... next time...
MS has a very consistant standard for UI
Really? When I right-click on something, nothing happens (Pre-1995 Windows). Single-clicking opens something (Post-IE4 Windows). Keyboard shortcuts vary with the application and are subject to the whim of the application developer; CTRL-N is a new email in Outlook, but a new database in Lotus Notes.
MS's devkits include standard icon sets
Icons are also subject to the whims of the developer. In the 90s, I could always tell when a new version of Visual Basic had been released, because Windows shareware would have new and inexplicable icons.
Microsoft's user interface is not consistent over time. It is not consistent between applications, except those from the same vendor (and even then it's questionable). What seems like consistency and logic in the UI is really a huge installed base and a decade of acclimatization.
This is not intended as a bash on MS; the same would be true if Macs were 90% prevalent, or if Gnome were. It is, however, intended as a bash on those who think the MS UI is the Correct Way (tm) to do things, rather than the conventional way to do things.
Your rant is nice and all, but it's largely irrelevant. This new project exists _exactly_ to cover these concerns -- well #3 and #1, at least. #2 is a matter of style.
no offense but, being good at DOS and Windows doesn't make somebody a computer expert.. it's a matter of being used to a certain system.
For instance, I'm good with Linux and Mac but I'm totally lost on Windows. I had to work with a Windows Server 2003 install and it took me a while to figure out how to set up the ethernet port. Everything is buried in "Properties" buttons, on multiple tiny dialogs (why such a small dialog for a 21" screen?) that don't make any effort to prioritize the most important settings.
Red Hat is even worst in some areas but I've gotten used to it.
So once you figure out Red Hat you'll be a "computer expert" on there too.
I can't believe they're going to allow officially sanctioned packages which conflict with core packages; this will be like the Ximian Desktop problem but worse because these Alternatives will be semi-encouraged.
That was not a useability problem you had, it was a training issue. You were expecting your ability to speak and read Klingon would help you read Narn textbooks.
How long did it take you to go from DOS, through every version of Windows, learning everything you know? More than a week, I'm sure. *NIX may not be your cup of tea, and that's fine. I'm not picking on you here. I just picked your message.
And how does Gnome or KDE compare now with their 1995 editions?
And what is the difference between correctness and convention? Point of view.
The goal of the project is to be current and up to date. That should actually make updating easier since much of the time people tracking current updates will find they basically have the next release when a release point is declared and 'official' .iso images created.
.. upgrading is no big deal. I've taken boxes from 7.1 to 9 without rebooting.
Even with current Red Hat 8, 9,
Supporting old releases is expensive and gets vastly more expensive over time. Its why nobody does it in detail for old releases except in the enterprise space, Debian included.
Various non Red Hat folks have talked about doing unofficial RH 7.3 errata, I guess it depends if enough people willing to pay them to make it cost effective.
"upgrading is no big deal."
I guess that's pretty true if you stick with a stock installations but, given enough time, I've always drifted into configurations that seem to entail some loose ends (un-official software that I've installed).
As a result, while supporting old distibutions is expensive for vendors, repeated upgrade cycles are likewise unpleasant for me.
As a result, I have to be dragged kicking and screaming from one major release to the next. I skipped RH8 altogether.
The important thing is that there seems to be a lot of work being done to deal with these issues, from/by/for both ends of the community.
--Richard
There's a difference between "prettiness" and usability.
A purely commandline OS could be extremely usable. But no linux distribution that I've seen is, mainly because it's so heterogeneous--given a new commandline tool, is it going to use short options or long options? If short options, can they be combined or not? For options that can be toggled on or off, do you indicate this by preceding them by "+" or "-", or by preceding the negative options by "no-"? Do you get the definitive documentation from man pages, info pages, "-h", "--help", or somplace else?
And the gui's used in any linux distribution often have exactly the same problems. It's not a problem when they're not pretty; what sucks is when they're inconsistent and unintuitive.
--Bruce Fields
Really? When I right-click on something, nothing happens (Pre-1995 Windows).
Jesus, that was over eight years ago.
Single-clicking opens something (Post-IE4 Windows).
No, it doesn't. Only if you turn that option on.
Keyboard shortcuts vary with the application and are subject to the whim of the application developer; CTRL-N is a new email in Outlook, but a new database in Lotus Notes.
90% of applications follow standard shortcut procedure, but there are always the exceptions, which aren't the fault of Windows consistency.
MS's devkits include standard icon sets
Icons are also subject to the whims of the developer. In the 90s, I could always tell when a new version of Visual Basic had been released, because Windows shareware would have new and inexplicable icons.
Um...huh? What does the changing of some default dev icon have to do with the interface consistency? Most apps use their own custom icons.
Microsoft's user interface is not consistent over time. It is not consistent between applications, except those from the same vendor (and even then it's questionable). What seems like consistency and logic in the UI is really a huge installed base and a decade of acclimatization.
Completely false. Windows is considered a bastion of homogenized consistency (good or bad), especially compared to the hell that is the Linux desktop attempt.
"Sufferin' succotash."
Hell? Hey, I'm willing to put up with a fast, reliable desktop, no adware or spyware or such crap, even if TWO FRIGGING CHECKBOXES DON'T LOOK IDENTICAL.
Sheesh, get some perspective. Car dashboards are different. Microwaves are different. VCRs. DVDs. Things can be intuitive and usable without having to be identical to everything else out there, you know.
What have you got to say about that, eh?
As a third party software developer, I'd like to know what will be guaranteed to remain static within a given release of Fedora and its updates. If I write software that's dependent on, say gtk2-2.2.5 and which will break down (hypothetically) with a newer version of the gtk2 package, will I be guaranteed that this won't be the case with updates to a specific version of Fedora?
One concern of software developers is guaranteeing minimum requirements for the software they develop. Look around you and you'll see developers stating their software "works with RH 9" or whatever. If Fedora becomes too much of a moving target, it will be a headache to develop software for it.
Support and fixing bugs in bugzilla are two different things. You can expect folk to be fixing bugs, scribbling in bugzilla and the like but you won't be able to pay someone to fix stuff or get guarantees anything will be fixed.
Jesus, that was over eight years ago.
Fair enough.
No, it doesn't. Only if you turn that option on.
No. It does *unless you turn that option off*. Very large difference. The vast majority of desktops use the default.
Windows is considered a bastion of homogenized consistency (good or bad)
Wow. I'm not sure exactly who've you been talking to, but they either aren't HCI or were buzzed at the time. Windows is infamous for being used as Microsoft's testing grounds for the latest version of their widgets (which go first into Office, then into IE and Windows). MS has masses of odd little don't-quite-fit controls in their apps. Witness the big-Motif-looking-button above the mailbox list in Outlook Express, or the Start Menu -- the button that acts like a menu (but a menu that acts differently from all other menus on the system). Windows is up to three user-visible layers of filesystem (8.3, long filenames, Explorer-only features like shortcuts), which are hell from a consistency point of view.
That's just with core MS software. The really atrocious UIs come from third party VB apps. Say what you will about Linux, most volunteer efforts have a far more consistent interface than their Windows shareware equivalents.
May we never see th
I'm too stupid to understand that circular dependencies can be resolved by specifying BOTH .rpms together on the command line, and that problems hardly ever occur if one uses proper Red Hat packages instead of mixing SuSE, Mandrake and Joe's Linux packages together (which the system wasn't designed for).
Hmmm, actually, now I'm glad I read that. I've never seen "dependency hell," and now I know why. I've only recently started occasionally pulling rpms off of rpmfind, and I always do it for my exact version of RedHat. And I've always known you could plunk everything down on the rpm commandline and have it resolve it. (I once typed something like rpm -i *-dev.rpm so I'd quit having to install prerequisites to compile stuff.)
Secession is the right of all sentient beings.