Red Hat releases Netscape Directory Server to OSS

parry writes "Red Hat has released the Netscape Directory server acquired from Netscape Security Solutions under a "GPL + Exception" license. The Fedora Directory Server is made up of a number of different pieces of software, each with their own licensing. "

Comments

Slashdot releases a dupe

[Anonymous Coward]

Red Hat Opens Netscape Directory - This is a bad omen for the start of the work week.

Re:Slashdot releases a dupe

[JamesD_UK]

Please reread both articles. The first announces that Red Hat will be opening Netscape Directory, the second actually points directly to the openned code which wasn't available at first. So if it is a dupe, it's a very useful one!

Re:Slashdot releases a dupe

[Anonymous Coward]

So what? You'll tell your boss "Sorry, I f-ed up because /. duped" ?

In that case I don't see a reason to work at all anymore :)

Re:Slashdot releases a dupe

[Anonymous Coward]

If you notice a dupe - your not doing enough work.

(I noticed it too)

A little help

[manno]

So what exactly does this do? And what closed source products does this replace?

Re:A little help

[ScribeOfTheNile]

It was originaly closed-source, so, er, it's replacing itself.

Re:A little help

[diegocgteleline.es]

Microsoft's Active Directory and propietary directory servers for $BIG UNIXES. Red Hat plays in the "Big server" market so this is a mejor milestone for them - and for linux in the enterprise. Lots of netscape directory's customers will want to switch to redhat too.

Re:A little help

[Cylix]

Having been a netscape directory customer...

I can't say I would really want it on redhat either.

Maybe it's gotten better though.

Not precisely a dupe

[The Woodworker]

The article last week was a press release type article. Lots of fluff, no content. Now we're getting the content. So it's not really a dupe. More of a late update.

Calling licensing brainiacs

[flacco]

anyone care to examine the license - "GPL with Exception" - and give us an Evil / Not Evil summary?

Re:Calling licensing brainiacs

[LiquidCoooled]

The Exception is explained here

It states:

GPL Exception License Text
From Fedora Directory Server

This is the text of the Licensed used in the Core of the Directory Server code. For more of an explaination, please see the annotated license text for a more in-depth description.

This Program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; version 2 of the License.

This Program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this Program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.

LC:here it comes

In addition, as a special exception, Red Hat, Inc. gives You the additional right to link the code of this Program with code not covered under the GNU General Public License ("Non-GPL Code") and to distribute linked combinations including the two, subject to the limitations in this paragraph. Non-GPL Code permitted under this exception must only link to the code of this Program through those well defined interfaces identified in the file named EXCEPTION found in the source code files (the "Approved Interfaces"). The files of Non-GPL Code may instantiate templates or use macros or inline functions from the Approved Interfaces without causing the resulting work to be covered by the GNU General Public License. Only Red Hat, Inc. may make changes or additions to the list of Approved Interfaces. You must obey the GNU General Public License in all respects for all of the Program code and other code used in conjunction with the Program except the Non-GPL Code covered by this exception. If you modify this file, you may extend this exception to your version of the file, but you are not obligated to do so. If you do not wish to provide this exception without modification, you must delete this exception statement from your version and license this file solely under the GPL without exception.

I will let others decide on its evilness factor.

Re:Calling licensing brainiacs

[Tooky]

Here's an annotated version of the license:

http://directory.fedora.redhat.com/wiki/Annotated_ GPL_Exception_License

Re:Calling licensing brainiacs

[palfrey]

IANAL, and also not any sort of licensing expert, but the really iffy bit looks like

Only Red Hat, Inc. may make changes or additions to the list of Approved Interfaces.

This means that only Red Hat, Inc. gets to make changes to the list of Approved Interfaces. Since the file that contains the list of Approved Interfaces is also distributed under the GNU GPL, it's perfectly fine to make changes to that file. We just wanted to make sure that if you do make changes to that file that the additional rights granted through this exception do not apply.

I think there's some unresolved issues there, namely that (for the sake of argument) Red Hat goes out of business, the software becomes significantly restricted as no-one else would be allowed to extend the Approved Interfaces list (without getting permissions from the now disappeared major copyright holder i.e. Red Hat), and so the development of new interfaces for plugins to non-GPL'ed code becomes impossible. Ok, this means we just stick to GPL plugins, but still...

Re:Calling licensing brainiacs

[hey!]

IANALB, but...

Well, the exception relaxes the GPL restriction on linking non-GPL code, provided that the non-GPL code access the GPL code exclusively through a set of defined and free APIs.

This "feels" a bit LGPL-ish. I'm guessing that RH has some particular scenario in mind that would be allowed under this exception that might not be under LGPL. Perhaps this is to enable proprietary extensions to be made (e.g. to interoperate with non-free directory servers) to be made and distributed, via something like a plug-in interface. LGPL allows this, but requires that the resulting derivative work (the original LGPL plus proprietary) must allow reverse engineering -- presumably of the proprietary part. According to LGPL: "Such a contradiction means you cannot use both them and the Library together in an executable that you distribute."

Of course, they could have started LGPL and relaxed that, so it's also possible that LGPL may be too permissive in other cases. For example, you could under LGPL conceivably create your own proprietary directory server, deep linking to various routines containd in the LGPL source code. As long as you kept the source code unmingled with the LGPL source code, released your changes to the LGPL part appropriately, could could claim your directory server was a "work that uses the Libarary" under section 5; and redistribute the combined proprietary, LGPL work under the terms of section 6.

The LGPL doesn't say you have to used the specific APIs intended for public consumption. The APIs that RH has blessed for linking may be sufficient for interoperability, without providing derivative proprietary works access to all kinds of implemenation goodness they could mooch off of.

Re:Calling licensing brainiacs

[DaveHowe]

Given the purpose of a directory service it seems a reasonable exception - you may not take the code and use it in a non-gpl package, but you *can* use the published api (this is at a similar level to being able to use SQL to access a MySQL or PostgreSQL database, but not actually use any of the code in your own non-gplled programs)

Re:Calling licensing brainiacs

[swv3752]

I'm betting there is some specific closed source Enterprise software that interoperates with the Netscape Directory. Red Hat probably intend for the Netscpae directory to continue to be delivered with this software but did not want the permissiveness of the LGPL so everyone and thier dog could leech off thier work.

Re:Calling licensing brainiacs

[Anonymous Coward]

If Red Hat goes out of business, who will be left to sue you for breaking the copyright?

Re:Calling licensing brainiacs

[petermgreen]

its tighter than the lgpl in that it doesn't let people add new approved interfaces to the code for the propietry software to call.

though i can't see any rules stopping you from modifyiing the stuff behind the appoved interfaces and therefore basterdising them in some way (for example making a previously undifined result into some feature you wan't)

imho this is less evil than what mysql and trolltech do (which is understandable, they wan't to make money but shows that free software isn't really thier goal lukilly for us trolltech were slow enough about releasing thier code that we got a decent lgpl alternative) but not as nice as say lgpl.

Re:Calling licensing brainiacs

[Lucractius]

This is an improvement for directory services. Now theres at least one corporate grade open interface... there will be a defacto standard as apps for Fedora Directory will see other back ends adopt the standard in order to become compatible in a similar fashion to SQL as pointed out above.

Re:Calling licensing brainiacs

[Chris Blizzard]

We chose to use a GPL+Exception license for a few very specific reason:

1. We wanted something that was GPL-compatible.

2. We wanted people to be able to ship it with non-free software.

Directory Server has been shipped for a long time as the backend for lots of other apps, like all the old Netscape Suite apps like mail server. We thought this was a valuable form of delivery and wanted people to be able to continue to do this.

3. We wanted people to be able to build plugins for the Directory Server under non-free licenses.

We thought that the ability to build plugins with any software was an important right to grant.

4. We wanted to make sure the core of the Directory Server was maintained as free software.

This was the most important goal. If someone extends some piece of core functionality in the Directory Server they need to give it back to the community. We didn't want forks that included proprietary software right in the core of the directory server.

Unlicensed brainiac

[MountainMan101]

Worth pointing out, the exception is an extra freedom not a less-freedom (which of course would make it not GPL).

I'm not too clear on GPL vs LGPL but the extra "you can link this from non GPL...." sounds like a cross between the two.

Re:Unlicensed brainiac

[dolmen.fr]

This licence, contrary to LGPL, explicitely lists the Approved Interfaces in a file named EXCEPTION.
So if external code is linked to code not declared in this file, it is covered by the GPL.

Where the LGPL-like freedom stops is here: "Only Red Hat, Inc. may make changes or additions to the list of Approved Interfaces." So if you (assuming you are not Red Hat) want to add interfaces to a fork, they will be covered by the pure GPL as the exception clause will not apply.

What we need is:

[nighty5]

An easy to use Directory Service that:
* childsplay to install
* hides the LDAP schema from admins that don't need to know
* a GUI / web console to add, delete, alter LDAP attributes
* easy integration into the O/S with a few file changes: PAM modules
* is easier to get going than OpenLDAP

I hope that the new Fedora project will do something like this, I saw the admin toolkit but no source is available yet, only binary packages - since I run gentoo i'll wait... Might be interesting!

Re:What we need is:

[davecb]

I hope the quality is good: I've used it in
the past and it needed some work.

--dave

Re:What we need is:

[parryFromIndia]

Netscape Directory used to have a GUI (Java Swing based, and so it used to suck but..) for performing routine management tasks (searching, setting up replication, adding/deleting/modifying the attributes etc.). So you can hope that the Fedora DS which is based on Netscape will have atleast some GUI tools to make an admin's life easier. Some one should probably write a Softerra LDAP Administrator alike tool integrated with the Fedora Directory server and it will be an ideal OSS directory server to use. And last time I checked Netscape Directory server was fairly easy to install.

Re:What we need is:

[mr_jrt]

Well, for one of your points, namely PAM modules, libnss-ldap works fine for me, and it's hardly rocket science to set up. 30 minutes, tops, including importing all the existing data.

Only thing that would be nice is if Debian was happy to have automated user operations performed on the directory instead of just passwd...but I'm sure a few scripts could sort that out if I could be bothered.

Only things that bother me about OpenLDAP:
* ACLs in a config file rather than the directory itself...yes its more secure but come on.
* No decent GUI tools. I have a dodgy (as in doesn't work correctly) Java LDAP editor and phpldapadmin, but these two are no way as quick 'n easy as MS's MMC.

Re:What we need is:

[G Money]

Novell's eDirectory satsifies all of those requirements. It installs as an rpm (everyone else will have to alien it to a deb or whatever you like). The ldap schema is completely hidden unless you want to extend it in which case you can use either a web frontent (iManager) or a java app (ConsoleOne). The same two tools will let you manage everything you would ever need to touch on it as well as manage just about every other Novell application. Both tools work fine under the browser/os of your choice as well. If you're using SUSE the PAM setup takes about 2 minutes using YaST. Otherwise, you just have to create and /etc/ldap.conf file and modify whatever services you want to authenticate via ldap under /etc/pam.d/. If you want it even eaiser you can setup the entire thing from the web interface with very advanced ACLs for who can access what server which will setup pam_nam automatically for you.
I've used OpenLDAP a number of times and while it works well in smaller environments, the ability of eDirectory to scale out to millions of objects and transparently replicate bidrectionally with no setup required makes it quite a bit more useful. It's not OSS, but they practically give it away so you usually never even have to pay for it.

Re:What we need is:

[Chris Blizzard]

We've got a set of admin tools there (as mentioned elsewhere in this thread.) We intend to open source them, we just haven't gotten that far yet.

If you want to see them download the binaries from the web site. They are included with the rpm, just not with source yet.

Re:What we need is:

[bheerssen]

Looks like you can pull it from anonymous CVS, though the build procedure looks to be quite complex, but I don't see why you can't do it on Gentoo. There may be valid reasons to choose another LDAP implentation, but this one should at least build on any Linux.

http://directory.fedora.redhat.com/wiki/Building

Re:What we need is:

[killjoe]

An admin who does not know the ldap schema he is working is downright dangerous. They should immediately be fired before they cause major harm to the company they are working for.

Re:What we need is:

[Xoth]

I have used this DS for many years and am pleased that Red Hat bought the Netscape product and didnt let it die. Sun aquired the code through a partnership called iPlanet. AOL ended up with the code too out of iPlanet and just liked the name "Netscape" for commercials. So now you got Sun and Red Hat moving forward with it.

Whats not to like? All operating systems have a decent DS now! You can stop announcing what you use, it just doesnt matter. If you run Suse edirectory, solaris Sun DS, windows AD, RHEL Red Hat DS.

GUI!? Interestingly enough in trying to get this DS to run on linux and other OS's in the past, its the console that caused problems not the actual service. Havent tried the latest Sun version for RHEL 3 so dont know the status but the darn java console always installed an old version (4) of the netscape http server. That was the culprit. One of the reasons the java console is there is to support central admin for other Sun products. Makes sense but now we dont need that anymore :) So I agree with all the posts. Red Hat and fedora crowd should develop a console that lives up to the quality of the service, or make it an option at install so people who can, can use the command-line. Put a nice front-end on this baby and the world is yours!

Re:What we need is:

[Anonymous Coward]

Ah yes.. The old "he should be fired" line.. Almost always coming from someone who ISN'T in the industry and has no idea what the reality of it is.

Re:What we need is:

[hyc]

In OpenLDAP 2.3 (which has been in beta for a few months and is due for a public release in a few days) the full configuration is editable via LDAP, so you no longer need to restart the server to tweak ACLs, add schema, or reconfigure indexing. In fact you can generate a complete server configuration via LDAP - you can add new backends and start them up and shut them down on the fly.

re: decent GUI tools - yes, this is a problem, but you have to frame it correctly - it is a problem with LDAP software in general. You shouldn't *need* a custom console to configure the service, you should be able to do everything through the basic protocol, regardless of how you access the protocol - command line, standalone GUI, or web application. Most site admins will always need a combination of these tools - command line tools for automated/batched, bulk access and for trivial incidental operations, a GUI to aid in visualizing an overview of the data, and a variety of customized web forms for the day-to-day trivia like resetting users' passwords. A custom console may be helpful, but the ones I've seen so far are overkill for day to day administration, and not flexible enough for real problem resolution. It will be nice to work toward a generic console that can address all of these needs, but I think we'll be stuck writing one-offs for some time yet.

Re:What we need is:

[mr_jrt]

Thank goodness for that.

It's hard having to acknowledge to my friend whose a Active Directory admin the overall strength of AD when managing Windows, but it seems things are improving. I'm a bit out of the OpenLDAP loop, I catch bits and bobs but I generally stick to official Debian packages, of which the last update to OpenLDAP was in 2002.

Cheers for the update.

Open Source replacement for MS Active Directory

[tuxpert]

Finally ! this is the only Open Source Directory that can compare to the features that Active Directory has to offer, especially multi-master replication.

Unfortunately, this will probably mean OpenLDAP will fade into insignificance, but I may be wrong !

This is the 'stronger rope' I needed to hang the guys planning on making Linux authentication depend on MS AD where I work :)

Re:Open Source replacement for MS Active Directory

[diegocgteleline.es]

Unfortunately, this will probably mean OpenLDAP will fade into insignificance, but I may be wrong !

Why? KDE vs gnome? freebsd vs linux? It's not going to dissapear...

Re:Open Source replacement for MS Active Directory

[tuxpert]

well, in all those cases the alternatives have equivalent features (generally speaking) on offer. Whereas in this case, Fedora Directory Server is vastly superior to OpenLDAP in terms of features and I assume scalability as well going by what i've read/heard.

Re:Open Source replacement for MS Active Directory

[Kjella]

Why? KDE vs gnome? freebsd vs linux? It's not going to dissapear...

Yeah, the OSS community has a knack for keeping inferior products going.

*Watches as all the fanboys come flaming, thinking I meant them*

Seriously though, it depends on what you mean by better. Some products (apache,linux kernel) seem to be doing fine being just one, others are more of a preferance thing and usually have several. I think a high-end directory service is more of the type which only really needs one product though.

Kjella

Re:Open Source replacement for MS Active Directory

[sud_crow]

Unfortunately, this will probably mean OpenLDAP will fade into insignificance, but I may be wrong ! At first i thought indeed you may be wrong and actually OpenLDAP may have some benefit from this, but now, i think the most benefit will be for Netscape Directory, as this one is under GPL (most part at least) while the other seems to be under a BSD license. If im not wrong, BSD cant access GPL code, but GPL can use BSD code. Am i wrong?

Re:Open Source replacement for MS Active Directory

[LiTa03]

Unfortunately, this will probably mean OpenLDAP will fade into insignificance, but I may be wrong !

Why? KDE vs gnome? freebsd vs linux? It's not going to dissapear...

Nope, more like OpenGFS vs GFS. Check the opengfs webpage, they started work on improving global locking, etc etc, but once Redhat bought sistina and re-released GFS under GPL, opengfs had no reason to exist anymore. Shame for the devel effort going to waste...

Same happened to QT vs Harmony if anyone remembers (Harmony died when QT went for GPL if I remember)...

Re:Open Source replacement for MS Active Directory

[Anonymous Coward]

I wouldn't call it "vastly superior" in every respect. Ideally, the strongest parts of both projects would be merged together.

Re:Open Source replacement for MS Active Directory

[swv3752]

Perhaps you are forgetting the various BSD's? There are more webservers beyond Apache; try Xitami, Savant, or Boa.

About the only F/OSS software that stands alone is Tex, and instead of competing projects it has other software tha builds atop it like LaTex and Lyx. Heck, even core GNU utils like 'ls' have replacements in busybox.

Re:Open Source replacement for MS Active Directory

[Anonymous Coward]

Am i wrong?

Yes.

Re:Open Source replacement for MS Active Directory

[Hurderos]

Having the former Netscape directory product released as OSS is positive to the community. Unfortunately this doesn't imply a replacement for Active Directory. In fact there is very little available in the form of an Open-Source solution to compete with AD. I'm not sure that people fully understand the significance or importance of that.

AD is actually a fusion of a number of technologies. Most notably Kerberos for authentication, LDAP for identification and dynamic DNS/DHCP. As is common with Microsoft architectures all of these have been amalgamated into one large reasonably amorphous blob. So the presence of a directory server, even one as formidable as the Netscape product, does not imply an AD replacement.

Perhaps most significantly AD provides an architecture for service authorization. When AD issues network credentials they contain a Privilege Authentication Certificate (PAC) which is loaded into the optional authorization payload field of the Kerberos tickets. There has been a fair amount written about the 'openness' of this approach but Microsoft's use of the field is consistent with the specification and intended use of the field.

I've railed onward for a number of years about the hazards that AD poses to OSS solutions in the enterprise. Unfortunately the field of identity implementation and middleware is fairly exotic and not well understood. Its also not very sexy to work on and thus doesn't attract much of a developer following.

It is where the future of information delivery is going to get won and lost though. Ultimately the only thing that is relevent is who someone is and what they have access to. Microsoft has positioned itself to have a controlling position of this high ground.

Besides not having an integrated solution there is basically no open standard for answering the authorization question. This has basically lead to a situation where Microsoft is in a position of defining the standard for the industry. The ability to determine whether or not a user can run an application is not inconsequential when the majority of the desirable applications are controlled by the company setting the de-facto standard for answering the question. The implications for application/server and infra-structure tying are, of course, considerable.

The team working on Samba4 development are grappling with some of these issues now. One of the primary issues being argued about is whether or not Samba4 should contain its own implementation of Kerberos. It already contains a minimalistic version of a directory server. The arguement from the Samba camp is that the technical amalgamation model that AD is predicated on forces a similar architecture in a work-alike product.

I respect the Samba team and understand their motivations but I harbor significant reservations about efforts to 'clone' Active Directory. Its a pathway which is already leading to compromises in technical design. Beyond that the potential implications of legally defending this type of project is potentially troublesome in the current intellectual property climate.

So a directory server does not imply an Active Directory replacement. That will require tight integration of a number of technologies, not the least of which is an architecture for answering authorization questions.

Our Hurderos Project was conceived to integrate these technologies and answer the authorization question. Actually the fundamental tenant that Hurderos is based on is a solution to the problem of how to 'implement' identity, a question which is largely unanswered in both the OSS and commercial venues. Interestingly the authorization problem conveniently falls out from answering the identity implementation question.

Our latest release has been slowed up a bit but that was due to making arrangement to have the intellectual property protected in the name of OSS. Now that an acronym is in a position to do that we hope to spend more time on the technology.

Over the last year we have de

Re:Open Source replacement for MS Active Directory

[ImaLamer]

AD is actually a fusion of a number of technologies. Most notably Kerberos for authentication, LDAP for identification and dynamic DNS/DHCP. As is common with Microsoft architectures all of these have been amalgamated into one large reasonably amorphous blob. So the presence of a directory server, even one as formidable as the Netscape product, does not imply an AD replacement.

And...

So a directory server does not imply an Active Directory replacement. That will require tight integration of a number of technologies, not the least of which is an architecture for answering authorization questions.

Why does it seem that I'm a shill for Novell lately?

eDirectory is exactly that. It integrates authentication (multiple level, best in the business), software deployment, user rights, extended TCP/IP support, DNS/DHCP servers...

No it isn't open source, but replacing Active Directory can be done with an *open-like* or *open-friendly* solution is possible. (Sorry, after laughing at Novell's history for so long I've used eDirectory and found out why it is better than AD. I wish more people out there would use eDirectory, that's all...)

Re:Open Source replacement for MS Active Directory

[Luke-Jr]

No, actually, he isn't wrong. The BSD license permits GPL-type usage, but the GPL does not permit BSD-type usage.

Evil/Not Evil

[zerbot]

Depends on your point of view. I've quoted the "exception" below. It allows developers to distribute versions that are linked to non-GPL code as long as those links use approved interfaces. Developers who modify the GPL code are not required to continue the exception in the code they release. I'd put it in the decidedly non-evil camp, but GPL hardliners may view it as evil.

In addition, as a special exception, Red Hat, Inc. gives You the additional right to link the code of this Program with code not covered under the GNU General Public License ("Non-GPL Code") and to distribute linked combinations including the two, subject to the limitations in this paragraph. Non-GPL Code permitted under this exception must only link to the code of this Program through those well defined interfaces identified in the file named EXCEPTION found in the source code files (the "Approved Interfaces"). The files of Non-GPL Code may instantiate templates or use macros or inline functions from the Approved Interfaces without causing the resulting work to be covered by the GNU General Public License. Only Red Hat, Inc. may make changes or additions to the list of Approved Interfaces. You must obey the GNU General Public License in all respects for all of the Program code and other code used in conjunction with the Program except the Non-GPL Code covered by this exception. If you modify this file, you may extend this exception to your version of the file, but you are not obligated to do so. If you do not wish to provide this exception without modification, you must delete this exception statement from your version and license this file solely under the GPL without exception.

Re:Evil/Not Evil

[antrik]

As a GPL hardliner, I decidedly put it in the non-evil camp as well. But only in this context.

Generally, such exceptions for otherwise GPL'd software are not uncommon. They are always a compromise: The copyleft is somewhat weakened to be able to do something important that wouldn't be possible otherwise. (In this case, using the not yet free components.) By restricting the exception to the fixed set of interfaces, abuse can be prevented quite well however.

By leaving the option to change the list of allowed interfaces, they can reduce it by and by, as they replace the non-free components with free ones. Once they are down to none, they will remove the exception alltogether.

This way, they have the possibility to release the core under GPL already, while not fully done with freeing the whole application yet.

The important part is that you can use a perfectly free program, as long as you stick to the core. If you want to take a compromise and use one of the non-free components however, the license doesn't prevent that.

Of course, such a license could be abused theoretically, by putting all useful functionality into the proprietary modules, thus tricking all users into using the non-free parts. This would effectively make the application non-free. That's the risk with every license that doesn't consequently enforce copyleft.

Summing up: With a loosened copyleft, the license is still perfectly free, just like BSD or other lincenses without copyleft can be perfectly free. And just like BSD, it could be abused. (Though the potential is smaller due to the limited exception.)

One question:

[zenmojodaddy]

Billy G, is that you?

Re:Oh well....

[gst]

If you prefer to have an LDAP server under in the public domain then go and code one yourself. Did you pay ANYTHING to redhat to release their LDAP server? No? Then shut the fuck up and either use it or don't. But don't complain about things which others give away for free. As long as you neither contribute to the code nor pay anything to get it under another license NOBODY will care about your opinion of the GPL.

Boohhuuu

[Anonymous Coward]

The evil, evil GPL.

Requiring someone who uses free software to grant others the same rights he's taking advantage of is so, well, evil and unreasonable.

That said, this is about the dumbest discussion to start whining and trolling about the oh so evil GPL, as Redhat is, as mentioned in the article, releasing this under the GPL+Exceptions and these exceptions specifically allow to link non-GPL software to the code.

License

[tpv]

It's basically a half way point between the GPL and the LGPL.

For most Open Source developers the easiest thing to do is to just use the software under the GPL.

However, if you use the software as a library, and only make use of the specific APIs that Red Hat has listed, then it effectively becomes like the LGPL. You are not obliged to release your code under the GPL.

But unlike the LGPL, the set of allowed APIs is fixed, and defined by Red Hat. In a LGPL program you can open up new APIs and change existing APIs and as long as you release your changes to the library, you don't need to GPL your program. With the NS Directory Server, you can change the APIs all you want, but the new APIs you create can only be used under the GPL - i.e. you have to release your program under the GPL too.

Re:License

[Anonymous Coward]

Thanks for the info. It was worth a +1 underrated from me anyway.

I wish more groups would use this license. As much as I detest the GPL and wish people would use the LGPL (or my beloved BSD license) this one doesn't seem to be such an annoyance.

Of course, we'll find out it's GPL incompatible in about 2 months I'm sure ;)

Non-Evil

[Kjella]

I will let others decide on its evilness factor.

Basicly, it is like a dual license, GPL and "LGPLish". Anyone making modifications may do so under the GPL, or they may preserve the dual licensing. Same goes with every other piece of dual licensed software out there. You can make your own GPL-only fork of say Qt or MySQL - but don't expect Trolltech or MySQL AB to merge your changes.

Kjella

Stricter than LGPL

[Anonymous Coward]

Seems that it's stricter than the LGPL. The LGPL (as far as I remember) doesn't specify anything about approved interfaces, although it does require that it is possible to replace the library (dynamic linking or re-linking), but even this doesn't say anything about e.g. not touching deeply internal structures inside the library. As long as these internal structures are still accessable, it will still be possible to replace the library. This license does not allow that.

Re:Oh well....

[diegocgteleline.es]

they are stuck and are forced to have their code infected by this heavily viral licence.

Well, redhat is allowing you to use freely and for free a product that costs several millons of dollars. You should THANK them that they give you this possibility, if you don't like it don't use it or just shut up. If you are doing money from the product redhat opensourced but you don't want to give them back anything...well, I wouldn't really like to be your friend ;)

Re:Oh well....

[Anonymous Coward]

Excuse ME, but that's one of the stupidest posts I've seen on /. for a long time. (And that's saying something!)

If you prefer not to 'jump on the band wagon', fine - stick with existing closed source solutions. Nobody is holding a gun to your head demanding that you use this software.

Either way, whining about the fact that you actually have to do some work if you want to have this functionality - rather than just stealing their code outright - is something you probably should have posted anonymously. Or better still, not posted.

Released; Let the Benchmarks Begin

[LordMyren]

Not a dupe, its actually released now.

It'll be great to see the benchmarks to settle this;
SunONE
IBM's ldap thingy
OpenLDAP
Novell's eDirectory
maybe even AD for kicks.

Also, just a note, redhat's docs are actually pretty good. Even the web pages ~2500 word Architecture docs probably outweight the usefulness of everything else available on the web. One of the most frequenty Directory Service gripes is how bad the docs are; finding out how to build a good DS system is pretty much a black art. Part fo the reason OpenLDAP is so unacceptable as a solution is because you're at the mercy of whatever tools you can find; docs are MIA. RedHat's already done a decent job of making them accessible, which is good because I might need them to make this thing compile on Debian.

Way to go red-hat. Everytime red hat shows up on campus I always spend five-ten minutes asking about the Netscape DS. Thanks for the release; here's to long life.

Myren

Re:Released; Let the Benchmarks Begin

[Anonymous Coward]

This is not going to happen. The licenses for most commercial directories (and Netscape's directory in the past) forbid releasing benchmark information. Novell's clause is particularly nasty.

However, that doesn't stop you from performing the benchmarks for yourself to figure out which works best. Just make sure to carefully read all of the documentation to find all the tunable parameters that could have an impact on performance, as some of them may not be immediately obvious but could make significant differences.

License

[tpv]

Grrr, my previous reply was an accidental submit.. Why isn't the default action on the form to preview instead?

It's basically a half way point between the GPL and the LGPL.

For most Open Source developers the easiest thing to do is to just use the software under the GPL.

However, if you use the software as a library, and only make use of the specific APIs that Red Hat has listed, then it effectively becomes like the LGPL. You are not obliged to release your code under the GPL.

But unlike the LGPL, the set of allowed APIs is fixed, and defined by Red Hat. In a LGPL program you can open up new APIs and change existing APIs and as long as you release your changes to the library, you don't need to GPL your program. With the Fedora Directory Server, you can change the APIs all you want, but the new APIs you create can only be used under the GPL - i.e. you have to release your program under the GPL too.

It's covered at on the wiki.
As for evil/not evil, I'd say "not evil". It's seemingly designed to allow existing closed source software that used the NS directory to keep doing the same thing with FDS, but they've done a good job of preventing those users from getting a free run. They get to keep using the same APIs that they've always used, and now they have the freedom to fix bugs in the directory, but they don't get all the freedoms that GPL developers will get.

OpenLDAP vs this?

[Anonymous Coward]

Why would I want to switch away from OpenLDAP 2.2 for this? This is not to say it mightn't have advantages over OpenLDAP 2.2, but I've
been very happy with OpenLDAP recently, since it got really fine-grained permissioning and syncrepl.

Re:OpenLDAP vs this?

[Anonymous Coward]

Well, it does multimaster replication - openldap syncrepl is pretty orthogonal to that, you could syncrepl a local directory on your laptop (and have disconnected read-only operation when you're away from the corporate network) to a multimastered high-availability server farm, if you had a package that had both (i.e. some future parallel-universe merge of the two source trees :-) )

Also, as computer scientists have known for many years, hierarchy where there is no natural hierarchy is evil - netscape allows you to have a pretty flat database underneath and use "Virtual Views" to build up appropriate domain-specific hierarchies for different purposes, and maintain performance. It thus allows for LDAP directories that are less of a mess than is traditional, though I'd really like to see a LRAP (lighweight relational access protocol) that didn't have ODBC's intrinsic SQL-specifity and securability problems.

On the other hand, in my experience, recent openldap has much better schema validation than netscape server had when I last used it (yes, years ago openldap had very weak schema validation, that changed). Also,the use of openldap in linux distros has meant it's really been hammered hard on the security front, and is now rather secure.

Re:OpenLDAP vs this?

[podz]

Also,the use of openldap in linux distros has meant it's really been hammered hard on the security front, and is now rather secure.

Oh, you mean the same linux distro with openldap that is used in secure banking networks? Please. This has been the domain of Solaris/HP-UX and Netscape and Sun Directory Server for 8 years already.

Re:OpenLDAP vs this?

[Anonymous Coward]

Banking networks are laughably insecure, for the most part, actually. Linux users need good IT security because if our computer gets broken into, the police will laugh in our faces if we report it. Our only defence is IT. And the semtex wired to the chassis intrusion detect in my case, but I'm completely paranoid.

Banks can have you put away for a very long time because they 0wn society right now. They have a very big stick against people who are caught breaking their networks. Break a linux box and you'll probably be cheered on by Bush-loving corporatists.

Re:Oh well....

[Anonymous Coward]

What software are you using right now. What license is it under?

In other news, most people hate taxes, don't want to get old, and hate the taste of liver and onions.

If you care, why don't you stop what you are doing and write some BSD replacement?

In the old hunter/gatherer days, I can just imagine you complaining about the type of meat someone else just caught, ending with the day where you ARE the dinner. Jesus H cripes I can't believe people complain about free lunch because it's not their favorite dish.

MOST PEOPLE prefer the GPL, buddy. Or at least, most of the coders do. That's obvious.

Re:Oh well....

[frag thief]

I suggest you doing a google on:

kerberos + microsoft + [weasles | bastards | hijack | whatever]

Sure, the BSD license is more free than the GPL and would be a terrific license if everyone played nice.

But they don't.

typical slashdot

[mindwar]

here we have it ladys and gentlemen. the typical slashdot attitude. let some random guys tell you what you should be thinking. idiot.

Re:typical slashdot

[Creepy Crawler]

Our time is better suited at DOING THINGS rather than reading yet another useless, boring license.

Let the legalese geeks do their thing.

Uh-Oh.... binary only licenses from now

[HighOrbit]

From TFA at http://directory.fedora.redhat.com/wiki/FAQ#Genera l some componets are not yet open.

In order to make the Directory Server software available as soon as possible, some components will not be released as open source in the first release. Initially just the LDAP server itself is being released. The administration server and end-user console are not being released as open source at this time. However, the binaries will be available for those other components, so the full console, management, and web based applications will be available, just not the source code

Well... at least the core is Open. Maybe they have to write replacements for encumberered components (perhaps the Sun iPlanet parts??).

Re:Uh-Oh.... binary only licenses from now

[Anonymous Coward]

I am excited about this release. However, I will be much more interested if all components become open source and available on more platforms.

OpenLDAP is not without problems, but it does run on just about everything (including my little non-Red Hat/Solaris OS) so for now I will stick with it. I do appreciate the investment Red Hat has made, and I hope this works out.

Re:Uh-Oh.... binary only licenses from now

[Chris Blizzard]

It's not encumbered. It's just a question of time.

Re:Uh-Oh.... binary only licenses from now

[Anonymous Coward]

Read the comments at the beginning of the article. They haven't had time to get them out is workable OSS. They are coming but aren't done yet. Is that so hard to understand.

" Maybe they have to write replacements for encumberered components (perhaps the Sun iPlanet parts??)."

No, they just aren't ready yet. Let's not make a bigger deal of this then need be.

Re:Oh well....

[MynockGuano]

Amen, brother. Amen.

Re:Oh well....

[ChrisJones]

The phrase "viral license" is dumb. Viruses are alive and they reproduce by destroying the host.
GPL is a license you can choose to put on your work.

Nobody is forcing you to use GPL code, nobody is forcing you to modify GPL code, nobody is forcing you to write GPL code. If you choose to modify *AND DISTRIBUTE* a GPL program then you should be happy to put your modifications out too.
If you're not happy with that, go with some code that's under a different license and modify/distribute that instead.

It's not rocket science. Now stop with the FUD.

Re:Oh well....

[Anonymous Coward]

Manip is a known Microsoft fanboy. Check out his posting history at blogs.msdn.com: http://www.google.co.uk/search?q=manip+site%3Ablog s.msdn.com.

Re:Oh well....

[kylegordon]

Have you READ THE FUCKING ARTICLE?
In case you haven't, I'll save you the pain of clicking on 3 links and point you directly at http://directory.fedora.redhat.com/wiki/Annotated_ GPL_Exception_License#Special_Exception

Re:Oh well....

[LnxAddct]

Read the exception in the license, it specifically states that you can link against non-gpl code! Red Hat added a non-evil clause to the GPL to make it less restrictive like you were just complaining about.
Regards,
Steve

They had to do this or violate their own license

[HighOrbit]

If you look at their wiki, they have not released all the componets as Open Source yet. So, they had to have an exception or they themeselves would be linking non-free with free and in violation. With the exception, they can release at least part as GPL and get it out to the world. I'm betting the non-free parts are likely iPlanet componets that are still encumbered.

Because this introduces complexity (and possibly confusion), it sort of brings to question if the hollowed and sacred GPL was the most appropriate license in the first place or if they shouldn't have gone with either a Mozilla or MIT/BSD license. Redhat lives and breaths the GPL, but it is not the only free license around. Perhaps they should show some flexibility and consider others.

Formalizing exclusions, cif. Linux kernel

[Morgaine]

It's actually very clever wording by RedHat.

What they've done is to formalize a mechanism for exclusions similar to the one that we're used to in the Linux kernel, which also provides licensing exclusions to allow closed binary applications to run under it and closed binary drivers to run within it, despite the kernel itself being GPL'd.

Linux overcomes this issue because the GPL's explanatory notes describe a general exemption for closed applications that link to standard operating system programs and interfaces, without needing to reveal the application source code. It's rather imprecise though, hinging on the meaning of words like standard and distribution. Try defining standard for Gentoo, hehe. :-)

RedHat have improved on that by formalizing the exemption mechanism, ie. explicitly stating where you can link to without strings attached and mandating that linking anywhere else will place you under GPL rules. Clarification benefits everyone, lawyers excepted of course.

First realistic support statement

[Intron]

"To get in touch with us, you can try to reach us on IRC at #fedora-ds on irc.freenode.net or on one of our mailing lists.

At least they're honest.

Re:They had to do this or violate their own licens

[Anonymous Coward]

Because this introduces complexity (and possibly confusion), it sort of brings to question if the hollowed and sacred GPL was the most appropriate license in the first place or if they shouldn't have gone with either a Mozilla or MIT/BSD license. Redhat lives and breaths the GPL, but it is not the only free license around. Perhaps they should show some flexibility and consider others.

Red Hat holds the copyrights to this code. Whether the the GPL, this modified GPL with additional granted rights, or any other license is "appropriate" is solely their decision.

Side by side comparison?

[HermanAB]

Is there a side by side comparison of Network Information Service (Sun Yellow Pages), Open LDAP and Netscape Directory anywhere?

I'd love to know what exactly the differences are and I suspect that most people have no idea and simply use whatever it is that they are familiar with...

Re:Side by side comparison?

[Just Some Guy]

Here's most of what I decided I needed to know:

• NIS has essentially no security.
• I've never seen a non-Unix box support NIS.
• NIS is easy to set up and works great for what it's intended for.
• LDAP is extensible; adding new schemas is pretty trivial.

If you need to authenticate Unix logins on a secure network, NIS is easy to use and works out-of-the-box. If you need to store any other data (Windows logins, mailserver logins, IM server logins, address books, etc.) then LDAP is the clear winner.

Re:Side by side comparison?

[alexborges]

Is there a side by side comparison of Network Information Service (Sun Yellow Pages), Open LDAP and Netscape Directory anywhere?

NIS != LDAP

Just to clarify, NIS and LDAP really have no real point of comparission.

I mean yeah, you can use them both as authentication/authorization backend for pam, but then again, so can postgres.

NIS is a way to distribute some key file in an asyncronous way across a network. It works kinda nice when youve a full host of unixes and you need one authentication database for them all.

Let me put it this way, although its a flawed analogy:

NIS is to LDAP what Windows Network authentication services are to Active Directory.

Sun Java Directory Server meets some of those IMO

[WidescreenFreak]

Granted, this doesn't fit everything, but when it comes to Solaris environments, I've found that Sun Java System Directory Server fits the bill quite nicely. Don't get me wrong. I have not tried this outside of a Solaris environment, so I don't know how easy it is to implement into other operating systems like Slashdot's oh-so-precious Linux, but JSDS is available for Linux as well.

Your first requirement was not always the case. JSDS was a BITCH to install when I first started to work with it. It took me weeks to figure out, but after several test installs and documenting exactly what I did each time, I can now have a complete JSDS server for user authentication installed in about 15 minutes. (I plan on releasing my currently-64-page manual for installing and configuring it publically in the coming weeks.)

Don't know about your second requirement, as I haven't really tried it. There are only 4 admins for our entire company, so we all need equal access to the schema.

The rest are definitely there. JSDS contains a GUI client that the other admins and I use all of the time to look at data. It's not web based, but it provides access to almost every facet of the directory structure and configuration. Integration into other Solaris boxes is a breeze - run one command, fix the nsswitch.conf afterwards, and make some relatively simple pam.conf modifications. {Ding} Done.

As I said, and as I'm sure I will get flamed for, I have not tried it outside of a Solaris environment; however, I have found messages from others on the Net who currently use it with AIX and Linux as clients.

And don't complain about it not being open source. It comes with Solaris 10 and it's available as a free download. Although Sun claims that you need a license to use more than 100,000 entries, they don't do anything to enforce it. ;)

Re:Sun Java Directory Server meets some of those I

[parryFromIndia]

Except that Sun JSDS does not work on anything other than RedHat 7.2 or RHEL2 which is not what most people use today.

much happier with fedora-ds over openldap

[mrsbrisby]

it's been trivial to install, it's not eaten any of my data, and I must say I'm very happy with multimaster replication as provided.

it also seems much faster than openldap.

Re:much happier with fedora-ds over openldap

[lcs]

I wonder if fedora-ds is the answer to my never-ending

nss_ldap: reconnecting to LDAP server...
nss_ldap: reconnected to LDAP server after 1 attempt

problems ...

Re:Sun Java Directory Server meets some of those I

[Anonymous Coward]

Actually, the latest version also works on RHEL 3.

Not to mention other platforms like Solaris (SPARC and x86), HP-UX, AIX, and Windows. 64-bit versions are available for Solaris and HP-UX.

Useful for home networks?

[smartin]

I have a small network of Linux and Mac machines and would really like to set up one address book that is accessible to all my machines and all my accounts under both Thunderbird and Mac Mail.

Would this be useful for this application or is it overkill? What are the other alternatives, I played with openldap using something called abook once and it was unusable.

Re:Useful for home networks?

[Just Some Guy]

I've yet to find a mail client that I like that can write to LDAP address books. Read? Sure. Write? Nope. Apple's Address Book application is great for searching directories, but don't expect to use it to update entries (AFAIK - I'd love to be proven wrong!).

Re:Useful for home networks?

[rsax]

Here is an alternative: Novell eGuide.

eDirectory is a requirement so grab the 250,000 free licenses first and then read the documentation like there's no tomorrow.

Re:Useful for home networks?

[Anonymous Coward]

Oh god I wish.

I spent the past two weeks looking into having a unified addressbook schema available with LDAP as the backend.

At present it still cannot be done. The latest Thunderbird doesn't bother to read the MozillaOrgPerson schema during searches, so entering data using PHPLDAPAdmin with the MozillaOrgPerson schema buys you a broken address book. At least, all the address fields are empty no matter what you try. If all you care about is email addresses, you're fine.

I must be the only person that also wants _physical_ addresses to be displayed properly for contacts!

I finally gave up. I'll look again in six months perhaps.

If you're really curious, read all the posts about trying to find some kind of unified addressbook solution on MozillaZine in the Thunderbird forums. You'll see many, many people interested, but _no_ solutions. Period.

Very, very sad.

Enjoy!

Re:Useful for home networks?

[lcs]

Wasn't that one of the advertized features of Tiger?

Re:Useful for home networks?

[Just Some Guy]

The only LDAP-relevant feature I found on their feature list was:

Auto-Update LDAP Cards
Keep your server-based Address Book contacts synchronized with an LDAP based directory - perfect for mobile professionals.

OT: JuK blasted out "My Jesus I Love Thee" a couple of weeks ago when I left it on random play for a few hours. Thought you might like to know.

Re:Useful for home networks?

[Shadowlore]

I've yet to find a mail client that I like that can write to LDAP address books. Read? Sure. Write? Nope. Apple's Address Book application is great for searching directories, but don't expect to use it to update entries (AFAIK - I'd love to be proven wrong!).

Well we cant deal with the "like" aspect, but I've set up OpenLDAP to store my addressbook in using Evolution. I create new contacts there and they show up in the db.

Re:Useful for home networks?

[Just Some Guy]

Good to know - thanks!

Re:They had to do this or violate their own licens

[swillden]

If you look at their wiki, they have not released all the componets as Open Source yet. So, they had to have an exception or they themeselves would be linking non-free with free and in violation.

No, because as the owners of the copyright, they don't need a license to distribute the software.

The reason is the basis of the GPL's legal force: copyright. By law, only the copyright holder has the right to make copies or create derived works, or give permission to others to make copies or create derived works. So in order for you to copy a Linux CD you have to have permission from the copyright holders. That permission comes to you in the form of the GPL. If Linus owned all of the Linux source, he could make copies without relying on the GPL. He doesn't, so he has to comply, otherwise he'd be infringing on the copyrights of other Linux kernel contributors.

Red Hat owns the copyright to all of this code, so they don't need anyone's permission and can do whatever they like. They've decided to grant everyone else permission to use their copyrighted works in ways defined by this GPL+extra license, but that decision in no way constrains what Red Hat can do.

Re:They had to do this or violate their own licens

[killjoe]

I think the GPL probably works best for a company like Red Hat. Other companies can out spend them on development staff so they need to make sure they get back any enhancements made by oracle, sun, hp or what have you. Otherwise the project would fork and their version would be left behind.

Re:Sun Java Directory Server meets some of those I

[WidescreenFreak]

What happened to backwards compatibility in Linux? ;)

Seriously though, are you just looking at that from the software specifications or have you actually tried and failed to install it on a newer release? I'm not being sarcastic by that. There are a lot of things that were designed for Windows 95 that run fine in Windows XP. Similarly, there have been many times where I can run a Solaris 7 or even Solaris 6 binary in Solaris 9 without a problem. Since I don't work with Red Hat or Fedora Core, I cannot determine if what you say is true or not. But I must admit that I'd be disapponted if JSDS didn't run on newer versions of RH/FC.

Re:Sun Java Directory Server meets some of those I

[parryFromIndia]

I actually tried to install and work around the failures on FC3, Gentoo and Debian. Did not work. I hadn't tried RHEL3 so I shouldn't have commented on that but I thought RHEL3 == FC3, but maybe not.

How is this different than the LGPL?

[Anonymous Coward]

What's the deal? Isn't this pretty much the LGPL? Or are they really wanting to limit the number of places this thing can be linked to?

Hrm, fedora?

[EvilStein]

Ok, I read that *Red Hat* was going to release it. Then Red Hat spins off Fedora..

Now this is showing up under the Fedora project pages?

So, are they releasing it as an actual product that can/will be supported by Red Hat's support & support contracts, or are they just saying "Ok, Fedora project, you can have this. Have fun!" and letting it go at that?

Personally, I think that it'd be a little underwhelming to just release it to the world and say "here.." This product is something that could be quite useful to many organizations, but without that level of support available, it'll be seen as "more open source stuff with no support" by the PHBs and bean counters. :(

(yeah, I know, the "if you had competant admins you wouldn't need support" line, but if that was the way management types thought, Windows Server 2003 wouldnt be selling as well as it is..)

Re:Hrm, fedora?

[Chris Blizzard]

It's considered a Fedora project. Fedora is where we will take contributions and build a community around the code.

Red Hat has a supported version of the the Directory Server called "Red Hat Directory Server." Separate information about that product can be found on Red Hat's web site.

Re:Hrm, fedora?

[EvilStein]

Dude, best information I've seen on Slashdot all day. Thanks! :D

Re:Hrm, fedora?

[Shadowlore]

So, are they releasing it as an actual product that can/will be supported by Red Hat's support & support contracts, or are they just saying "Ok, Fedora project, you can have this. Have fun!" and letting it go at that?

Go read the docs, you'll find it is released as both Fedora-DS and RH-DS. I'll leave it to you to figure out which is a supported (as in pay and get a tech) and which is not.

yeah, I know, the "if you had competant admins you wouldn't need support" line,

Do you know the "yeah if you read the website you'd have known the answer" line? ;^)

http://www.redhat.com/en_us/USA/Home/Company/News% 2Band%2BEvents/Red%2BHat%2BPress%2BRelease%2BArchi ve/2005/Red%2BHat%2BAnnounces%2BDirectory%2BServer .html

They did both. That isn't difficult to understand.

Yes.. but what about *your* redistribution

[HighOrbit]

Ofcourse, if Redhat owns the copyright, they can do as they please and distribute it. But now you download it and you want to redistribute it. If the necessary build would violate the GPL (as this may because the necessary non-free componets link to to GPL covered componets), then *you* would be in violation if you re-distriubted it. So, now you can see the conundrum. To over-come it, they come up with this expection to the GPL. It is their right to do so, but perhaps not an optimal solution.

But they are work , their just not free yet

[HighOrbit]

No, they just aren't ready yet. Let's not make a bigger deal of this then need be.

Apparently, they have fully working binary-only versions, so this doesn't see to be a matter of it not being ready, but a matter of a *free* implementation not being ready.

go fuck yourself asshole

[Anonymous Coward]

you should just shut up and study for your mcse

Re:Sun Java Directory Server meets some of those I

[WidescreenFreak]

Fair enough. Then again, considering that Java 1.4 breaks a lot of stuff meant for Java 1.2.1, I shouldn't be surprised that Sun makes software that's not backwards compatible.

Re:Sun Java Directory Server meets some of those I

[Xoth]

Well I have been using this particular DS for many years and have tried it on a myriad of operating systems even windows. The platfrom support from Sun for RHEL and Windows has always been a version behind the current one. Only starting this year did Sun support RHEL 3 which in my book is pathetic. Remember Sun aquired the DS from Netscape, (nsswitch.conf notice the ns prefix) and dont really know what to do with it. This is a good DS and I am very pleased with Red Hats purchase. Its the best of both worlds. If you run Solaris you have a DS, and now if you run Red Hat you have a DS. And I think Windows shops have a DS too :)

Re:Yes.. but what about *your* redistribution

[swillden]

Agreed, assuming that you actually need one or more of the non-free components to build and distribute something useful. Also, even if you do need the non-free components, Red Hat would still not be violating the GPL, they would just be releasing software under the GPL that is not useful as-is. There's nothing wrong with that, actually, as the GPL contains no requirement that the software actually work.

Let's See, Any Way This News Could Be Older

[Master of Transhuman]

Maybe "Einstein joins Red Hat Fedora Foundation"?

Or "George Washington funds development of Red Hat Directory Server"?

How about "Jesus blesses newly released Red Hat Directory Server"?

And some moron says, "Well, this story includes the LINK!"

DUH!

eDirectory!

[ImaLamer]

Even runs on linux...

Overkill

[Anonymous Coward]

Overkill. When I worked at Netscape, I ran a bunch of the server products on my home O2 -- this (the ldap server), mail, news, http ... I forget if I ran any others. They worked like a charm, but they used up a lot of resources waiting for the other 99,999 members of my household to show up.

But it was cool having Netscape Communicator profile roaming between home/work/laptop. I miss that with Mozilla.

Re:Sun Java Directory Server meets some of those I

[wobblie]

nsswitch has nothing to do with netscape, it means "name service switch"

Re:Sun Java Directory Server meets some of those I

[Xoth]

Oops bad example... on meds for walking pneumonia :) My point is Sun bought the DS, they certainly did not develop it. Anybody who uses the Sun DS knows most of the config files and operational attributes are prefixed with "ns" which stands for netscape.

Re:Oh well....

[MaoTse]

It can be fetched from Sun for no cost for 4 years now ...

For "RH 7.3" too ;-)))

And you get all the Sun enhancements (secure replication, plugins. many more) too.

Re:Oh well....

[diegocgteleline.es]

It can be fetched from Sun for no cost for 4 years now ...

...and no source

Re:Oh well....

[MaoTse]

The point "for RH 7.3" might be more important for some people, I'm afraid.

Re:Sun Java Directory Server meets some of those I

[podz]

Although Sun claims that you need a license to use more than 100,000 entries, they don't do anything to enforce it. ;)

Not until the SBA comes a knocking with audit papers in hand.

Re:Oh well....

[Luke-Jr]

Any team can add features, so long as their changes remain Free. The GPL only says "if your screwed up mentality requires permission to use code, then if you want to modify this program, you need to let others modify it also"

No they didn't

[Shadowlore]

If you look at their wiki, they have not released all the componets as Open Source yet. So, they had to have an exception or they themeselves would be linking non-free with free and in violation.

The owner of the code is not bound to the license terms others have to agree to to use the code! If I write something and GPL it, I, as owner, can sell a proprietary app using said GPLed code w/o violating the GPL license you have to obey.

But they would, if it is to be re-distributable

[HighOrbit]

Yes, for Redhat themselves, they can do as they want as owners. But the whole point of OSS is to make it re-distributable. As I said in another post, now you download it and you want to redistribute it. If the necessary build would violate the GPL (as this would because critical non-free componets link to to GPL covered componets), then *you* would be in violation if you re-distriubted it. That is the problem. They want to have binary-only distribution of some componets that are linked to GPL components. So, in effect, they are putting other people in the position of violating the GPL (except for the exception they have provided). So the exception is necessary, otherwise Redhat would be causing other people to violate the GPL.