https://fedoraproject.org/wiki/Features/Zarafa
(AFAIK the reason why this was made public after feature submission freeze is that it was to be officially announced at FOSDEM last weekend on request by Zarafa's marketing).
This should only be considered for F-14, the feature owners could have come to fesco before feature freeze and taken steps to get things started. I don't feel we should make an exception for them. Everyone knows when they need features submitted by.
If you don't want to call it officially "feature" for Fedora 13, then simply don't do it. But as Fedora 13 will anyway ship the Zarafa packages (the server package is already approved, branched and submitted as updates, the webaccess is closed to be done as well), it would be just a rejected feature which is nearly completely done.
This stuff is already in Rawhide, so it makes no sense to postpone it to F14 just because of the submission deadline. It makes no sense whatsoever to advertise something as a F14 feature when it was already there in F13.
On the other hand, I wonder whether this is feature-worthy at all. We don't advertise every new package as a feature.
Replying to [comment:2 ausil]:
This should only be considered for F-14, the feature owners could have come to fesco before feature freeze and taken steps to get things started.
As soon as it's discussed by FESCo, it's in public.
I don't feel we should make an exception for them.
This sounds like you are rejecting this feature on principle and not from technical reasons. IMHO this is very bad advice.
Everyone knows when they need features submitted by.
If you read the mail from Robert that poelstra forwarded to fesco-list, then you will see that Robert did not know how to proceed in this case. I always felt like our feature process was too complicated and too bureaucratic, so I'm going to support this feature tomorrow even if it was late.
In addition (to my question whether this is feature-worthy), this is yet another case of the alarming trend of crippleware "community editions", see for yourself: http://www.zarafa.com/content/editions
They're reserving several features for the proprietary versions. This shows they do not really believe in Free Software, is quite angering (somehow they feel we do not deserve as much as their "enterprise" customers), does not make for a healthy community (they won't accept contributions without some kind of copyright assignment allowing them to use the code in their proprietary edition, and they'll probably reject patches which implement the intentionally missing features outright, not to mention that knowing those features exist, but are intentionally not being made available as Free Software is very demotivating to potential contributors).
IMHO it is a mistake to endorse this kind of business model by advertising these "community editions" as Fedora features (and I already voted against at least one other feature because of this).
So I'm against advertising this feature, but NOT on deadline grounds (which I agree are silly), but on grounds of noteworthiness and of political tactics.
(On a tangent, I wish we could get http://kolab.org/ packaged, which is much more in line with Fedora's ideals: reuses existing Free Software and is not crippleware (the only proprietary pieces are third-party plugins for Outlook which are obviously not required in a Free Software environment; Zarafa's Outlook integration is also proprietary). Of course it also means it's harder to package without bundled stuff (AFAIK, they normally bundle everything in the server), whereas Zarafa gets away with reinventing everything because they rewrote everything from scratch so they can sell proprietary "enterprise" versions. But enough ranting.)
Replying to [comment:5 cwickert]:
Right, I contacted John Poelstra because of his announcement e-mail, his offer to help, and as we are still in time for an exception.
Replying to [comment:6 kkofler]:
They're reserving several features for the proprietary versions.
Zarafa is Open Source for a short time now and they're rapidly making progress in the Open Source world. I'm sorry, but you can't seriouly expect, that a company is able to change their whole business model from one year to another.
The real differences between the Zarafa Open Source Edition and the best commercial one are the advanced multi user calendar, bricklevel backup (I never needed that even as a commercial user/admin) and multiserver support.
The Active Directory toolkit doesn't make sense without Active Directory (requires proprietary things), Blackberry Enterprise Server support doesn't make sense without an own proprietary Blackberry Server and the auto deployment tools are for Outlook users which doesn't make sense to Linux users as well (requires proprietary things). Finally, the "High Availability toolkit" is almost commercial support and some help for Zarafa with Heartbeat and DRBD as far as I know.
they won't accept contributions without some kind of copyright assignment allowing them to use the code in their proprietary edition, and they'll probably reject patches which implement the intentionally missing features outright, not to mention that knowing those features exist, but are intentionally not being made available as Free Software is very demotivating to potential contributors).
They're dual-licensing their code like other companies do with their products as well, e.g. MySQL to name an example included in Fedora. Anyway, you've interesting speculations regarding the way you're thinking they might handle patches, contributions and new features. IIRC you're right now not a Zarafa contributor so far to reason about that.
(On a tangent, I wish we could get http://kolab.org/ packaged, which is much more in line with Fedora's ideals: reuses existing Free Software and is not crippleware
Zarafa is not Zimbra or Scalix which are shipping its own MySQL, Java, MTA, ClamAV etc. Zarafa works and integrates well with standard openssl, libical, libvmime, pam, mysql, curl, perl, php, gettext, openldap and httpd from the regular distribution. And in times where Zarafa started, libmapi from OpenMAPI didn't exist as code, so they had to start with nothing. Even now OpenMAPI's libmapi is years behind from functionality and features compared to Zarafa's implementation.
Of course it also means it's harder to package without bundled stuff (AFAIK, they normally bundle everything in the server), whereas Zarafa gets away with reinventing everything because they rewrote everything from scratch so they can sell proprietary "enterprise" versions. But enough ranting.)
If you want to use Zarafa with Outlook you only have to install the zarafa-licensed package which includes a homonymous daemon/service doing it all. So you can use the Fedora packages and run a single zarafa-licensed on another box/virtual machine to satisfy your Outlook users. So no, the proprietary stuff is not part of the base package.
In addition (to my question whether this is feature-worthy), this is yet another case of the alarming trend of crippleware "community editions", see for yourself: http://www.zarafa.com/content/editions (...snip...)
(...snip...)
With this reasoning, please make a motion to throw out and expunge MySQL from the distribution. Please, do seriously consider creating a Special Interest Group in order to attack the problem of "crippleware" as you call and perceive it, and remove all of such from the distribution.
Of course there's some kind of professional value proposition, and there's always going to be unique features to such a professional edition that are not in the community edition, either in the software itself (crippleware), in another technical area (product add-ons and plugins), or in real life, such as support.
(...) This shows they do not really believe in Free Software (...)
That's an uncalled for blow below the belt.
For one, Zarafa has been going about Open Sourcing their products very hard this last year, with collaboration from Fedora, Red Hat and Open Source community members, open sourcing everything they had faster then anyone else out there has done / are doing. Make no mistake in arguing they don't care about it, judging by the first thing you see, which again is just an unfounded, carelessly made, unconsidered statement.
Additional References:
they won't accept contributions without some kind of copyright assignment allowing them to use > the code in their proprietary edition, and they'll probably reject patches which implement the > intentionally missing features outright, not to mention that knowing those features exist, but > are intentionally not being made available as Free Software is very demotivating to potential > contributors).
They won't have to and can't accept patches that implement the functionality that is now still in the proprietary product add-ons and plugins, because, well, they are separate product add-ons and plugins.
You seem to forget that Zarafa, including Zarafa webaccess, are standalone products which function perfectly fine and are Free Software (mind the capital F).
(On a tangent, I wish we could get http://kolab.org/ packaged (...)
I'll be working on that.
(...) I wish we could get http://kolab.org/ packaged, which is much more in line with Fedora's ideals: reuses existing Free Software and is not crippleware (...)
If you actually take a look at Zarafa's architecture, you would realize that this kind of statement is just pre-judgemental and unfounded.
This Feature was approved at the 2010-02-09 meeting.
Login to comment on this ticket.