Exchange Comes To Linux As OpenChange

joesmart writes to tell us that new work on OpenChange and KDE seeks to bridge the gap between groupware compatibility and open source. KDE developer Brad Hards spoke at the Linux.conf.au conference and said the goal of OpenChange is to implement the Microsoft Exchange protocols as they are used by Outlook. "OpenChange has client and server-side libraries for Exchange integration and relies heavily on code developed for Samba 4. It is open source software licensed under the GPL version 3. Hards said more work is being done on the client side and 'we have code for the server,' but estimates another 12 months of development is required to produce an OpenChange server ready for production."

Comments

Here we go again.....

[wintermute000]

The goal is laudable but strategically speaking: do we really want to focus more OSS efforts to replicate MS protocols and methods?

Whilst a million enterprises out there shrug their shoulders and think 'why would I want to wrestle with this when I could just go along with the AD stack that I know, trust and my MSCE admins love'

Of course they may come out with a fantastic 100% interoperable and virtually bug free product and I'll have to eat my words. But history is not on their side.... also will this have to plug into openldap/kerebos/samba nightmare?

Re:Here we go again.....

[devman]

Also worth noting this will be nice for people like me who work in windows shop but would like to run a Linux and actually use exchange functionality from a native client.

Re:Here we go again.....

[techno-vampire]

The goal is laudable but strategically speaking: do we really want to focus more OSS efforts to replicate MS protocols and methods?

If you want to telecommute, you need to be able to access your work email. If your company is one of the many who use Exchange, you have to use a client that can talk to it. Having a native Linux client that can do this would mean that you wouldn't have to run Windows, even in a VM box if you didn't want to, just to get your work email.

Re:Here we go again.....

[Raul Acevedo]

The goal is laudable but strategically speaking: do we really want to focus more OSS efforts to replicate MS protocols and methods?

Yes, we do.

Why do you think Microsoft has such a stranglehold on the corporate desktop? Outlook and Exchange are the cornerstone of that lock. It's brilliant if you can produce a true Outlook replacement; that means everybody's email and calendars can stay the same. If you try to introduce a brand new calendering/email system, you have to deal with migration, and that is a ridiculously huge headache affecting the entire organization. Not to mention all the retraining and retooling (and likely re-hiring) you have to do with a new server architecture...

No wonder nobody does it.

If you can replace the client, you are much more likely to have clients that can talk to multiple back ends (e.g. Exchange or an open source alternative). Then you have the real possibility of replacing the back end much more transparently at a later date.

Unfortunately this two step solution is, for the next few years, the only real way it could possibly happen in most companies.

Re:Here we go again.....

[wintermute000]

Well at least if its OSS then its zero cost to try it out in the lab, except for time of course.

I'd be interested to see how well it plugs into an otherwise stock MS active directory domain. If it wants to take on MS in their home turf it must get this bit absolutely right.

Also note as MS's embrace extend extinguish approach has brought us all sorts of 3rd party apps that plug into exchange e.g. voicemail to email for VOIP stacks like Cisco CCM, I can only foresee lots of pain

Another point, sure us IT types are more open to this kind of change. We are also (at least those of us in Dilbert corporate land) very wary of the consequences of messing with core systems that are working fine. Despite what Cisco QoS teaches you, email is regarded by your users as tatamount to electricity and plumbing. Until this project gets to a critical mass here like say apache or mysql its an easy sell to management, you will find it hard to justify ripping exchange out for this unknown quantity

Re:Here we go again.....

[wintermute000]

good points, I must admit I glossed over the client side and was thinking primarily on the server side.

Having said that though I find exchange web interface perfectly adequate, although of course its tied to IE for full functionality (shakes fist at MS)

On the client side, I ask another (possibly stupid) question: how is this different from say Evolution's exchange plugin (which I have used via https and from what I could tell, it did what it said on the tin, if slow as molasses)

Re:Here we go again.....

[realmolo]

Well, Exchange is *part* of the reason people get locked into MS products. But the bigger reason, by far, is Active Directory.

AD *works*. It's easy. It integrates seamlessly with Windows. The management tools are good, and easy to use. There are tons of third-party products that integrate with it. Seamlessly.

The current LDAP/Kerberos/Samba situation is a fucking MESS. It's unusable in a production environment. It's hard to manage. It doesn't have GROUP POLICIES, for Christ's sake.

Samba 4 supposedly fixes some of these problems, but I doubt it comes even CLOSE to providing all the functionality of a genuine Windows Server OS.

THAT is why people are locked into MS products. They simply work better than the alternatives in many cases, especially on a corporate LAN.

Re:Here we go again.....

[rabbit994]

It's a mess to get it all working properly and to get Windows clients to swallow it.

While I applaud their laudable goals, I don't see this making it very far. In 12 months, Exchange 2010 will probably be out and they will continue to play catch up. Also, it needs to drop into Active Directory without Windows AD servers not complaining and Outlook clients not noticing a change. For most businesses, no email world stops and price of Exchange is worth it to many businesses.

Re:Here we go again.....

[TheRealFixer]

What's nightmarish about OpenLDAP, Kerberos and Samba? I run this combination on my home LAN. Couldn't be easier.

Key words being home LAN. In a corporate LAN, or even a mid-size company network, management of these alternatives quickly becomes a nightmare. Stuff just doesn't work quite right with the Windows clients, and you don't have key components of Windows management available, like Group Policy Objects. Might be good enough for your hobby network at home, but multiply that across a couple thousand clients and it's not exactly fun.

I'm all about cutting costs by going open source wherever I can, but Active Directory, when you're dealing with a Windows environment, just works. The headaches and time I'd waste trying to get the current LDAP/Kerberos/Samba "alternative" working well enough that we wouldn't be getting flooded with calls about stuff not working how the users expect, greatly exceeds the cost of just implementing and maintaining Active Directory.

There's some hope that Samba4 will fix a lot of that, and after it's released I'll look at it again.

Re:Here we go again.....

[Antique Geekmeister]

Sadly, it's unlikely to work well past t he next Exchange or MS Office upgrade. You _cannot_ maintain compatibility when the primary authors of a product are determined to break your compatibility, and it certainly fits Microsoft's history to do so.

Re:Here we go again.....

[TooMuchToDo]

Wow, I guess you haven't been around long enough to pay Microsoft huge yearly amounts for support contracts to have them say to one of your problems "We have no idea. Good luck." Give me OSS anyday.

Re:Here we go again.....

[evilviper]

AD *works*. It's easy. It integrates seamlessly with Windows. The management tools are good, and easy to use.

Like ALL Microsoft products and technologies... Active Directory is pretty easy to get into a minimally working state if you like all the defaults. And isn't too difficult to get it to do some of the lowest-common-denominator simple tasks that everybody wants, like single sign-on, roaming profiles, and a few policy restrictions.

AD isn't really "easy" unless your time is worthless, and you don't mind insane problems cropping up. You're going to be browsing around context menus, sub-sub-sub-sub options with utterly insane names and absolutely no comprehensible scheme, to find the one option you want to toggle.

God help you if you want some slight variation of how Microsoft thinks it should work, because you've just gone from "easy" to "practically impossible" and are going to be delving into the darkest realms of the registry, and deeply hidden configuration menus and files.

I know plenty of companies who think Windows servers are easy, and work well... Plenty of them have hired me to get them to stop "working" the way they do.

Whatever time and money you think you've saved by going with Windows servers goes out the window the first time you try to copy a very big file to a Windows Share, only to have it fail at 2GBs... Yes, Windows quietly decides your gigabit LAN is a dial-up link, and decides to go for the slow, high-delay, 2GB filesize limit variation of SMB. Samba never does.

The current LDAP/Kerberos/Samba situation is a fucking MESS. It's unusable in a production environment. It's hard to manage. It doesn't have GROUP POLICIES, for Christ's sake.

I have no idea what you are talking about. You can manage group policies on a Samba server with some of Microsoft's own management tools (ie. from a Windows workstation that logs-on to the domain).

And once you've got Samba setup, it will silently work, exactly how you configure it to do so, forever. A Windows server will require CONSTANT attention, as weird one-off bugs continually spring up, performance suddenly drops dramatically one day, and slowly starts recovering over the next week, but never quite gets back where it was. Never mind the standard Windows practice of quietly disabling/corrupting one driver or another for no particular reason. And did I mention the utterly useless error messages, and logs with lots of useless information and NONE of the HELPFUL information you could possibly use.

THAT is why people are locked into MS products. They simply work better than the alternatives in many cases, especially on a corporate LAN.

No. They just sound better when you're reading the spec sheet, and trying to get a basic server minimally working...

The fact that Windows is popular with numerous companies is actually a sad commentary on corporations, who go for the quick way to save a buck, and ignore the vast amount Microsoft costs them over time.

Re:Here we go again.....

[scamper_22]

here we go again... missing the key point.

"Well at least if its OSS then its zero cost to try it out in the lab, except for time of course."

Yes, it is the time and labor cost that is the move expensive. What kind of staff do you think it is going to take to truly evaluate and support this kind of project? Let's not even get to the training the staff, installing new software on servers... You're looking at several hundred thousand dollars...

or you can just pay microsoft their regular fee and be done with it.
Think about it this way. An OEM copy of Windows costs 50 dollars.

Assuming a tech support person costs 25 bucks an hour.
All it takes is an extra 2 hours of support/training for a transition to linux to cost as much as simply installing windows. This does not even take into account lost user time dealing with new things.

Re:Here we go again.....

[Bonobo_Unknown]

Exchange is Microsoft's last fortress protecting the enterprise. If we could run an Exchange clone on Linux it would be so much easier to ditch all the rest of the Microsoft suite.

Re:Here we go again.....

[rolfwind]

The goal is laudable but strategically speaking: do we really want to focus more OSS efforts to replicate MS protocols and methods?

Something like Wine will be really helpful to the linux movement when some boxed software has in it's requirements list: XP, Vista, 7, and Wine 1.x compatible. If linux gain more, it may come! And it doesn't have to gain as much as if the software makers were forced to do a total rewrite. Once that happens, Linux has its foot in the door. And microsoft cannot change the API too much without breaking backwards compatibility and pissing off a ton of customers.

The end goal isn't to run Windows compatible apps but to make the transition to Linux easier. If Openexchange achieves the same thing, more power to them.

In an ideal world, Microsoft would conform to Open standards. But since this isn't an ideal world, and Microsoft has majority market share, open standards can, from time to time, conform to it.

Re:Here we go again.....

[mlts]

There is also Sarbanes-Oxley and other issues. Part of the costs of keeping "Due diligence" valid by doing Exchange is that Exchange comes with a lot of the features needed for compliance built in. For example, with E2007, it is almost a no brainer to set up archiving and retention so incoming and outgoing E-mail is retained as per laws... laws that are a bad thing to break.

An OSS product is going to have to not just grok the Exchange 2007 protocol, but be able to support features that Exchange offers, from OWA, to replication and clustering (larger installations have one Exchange server on their DMZ and a cluster for their mailboxes.) Most importantly, companies will need to rely on the solution to be able to archive and audit. If a solution can't produce logs when auditors come by, people go to prison, as per HIPAA, Sarbanes-Oxley, or CALEA.

Maybe RedHat could do something like this and get it FIPS/Common Criteria/whatever certified so people have an alternative to Exchange, but until then, a lot of companies will remain tied to it and Active Directory.

Re:Here we go again.....

[Gordo_1]

I'm not saying MS doesn't have incentive to break the protocols, but they do have to maintain some sort of compatibility between versions of Exchange. That's because corporations typically update Outlook software across the organization in a continuous fashion and asynchronously from Exchange server upgrades. IT departments would raise bloody hell if MS didn't provide a mostly seamless transition.

Re:Here we go again.....

[Mista2]

What would be more likely to fly would be a feature complete client to exchange. Email - no problems but it is still a headache to get calendar and contact information. Where Exchange and Outlook rule is integrating this all into one place, and now Comms server brings in voice and chat/IM, yet more systems MS has tied into a Windows client by extending a set of open protocols so that noone else is compatible 100%

Re:Here we go again.....

[GigaplexNZ]

There is a big difference between not attempting to maintain compatibility and actively going out of your way to break the compatibility.

Re:Here we go again.....

[TheLink]

"Especially as it's their job to learn a new system if/when it is introduced."

But for most, that's _far_ from their main job. They may be good at sales/marketing/purchasing/managing people or projects/etc but not as good at learning new software. So change is disruptive and costly.

If the change is perceived as being useless or pointless it is no surprise when end users protest.

Another thing - there are valid reasons to use windows.

For one, 5 years ago Desktop Linux was crap. Alternatives to Microsoft Office were abysmal. Things have improved a bit but still the OSS alternatives are behind in many ways (even OSX is doing better than Desktop Linux).

Thus Microsoft Software may have been the best choice back then. And if there is no need to change, why change? So no surprise if 5 years later companies are still using the same stuff.

Thing is, now Microsoft is forcing the issue with huge changes like Microsoft Office 2007, Vista, Windows 7 and so on - all these involve extra training and cost.

So there is a big window of opportunity for OSS stuff - since either way the corporation has to retrain staff and spend $$$. Whereas previously sticking with Microsoft was a known, acceptable cost.

If there was a painless and cheaper way to migrate off Microsoft products, many companies would go for it.

Re:Here we go again.....

[bit01]

There is also Sarbanes-Oxley and other issues.

Sarbanes-Oxley applies to the USA only. 95% of the world's population don't give a damn about Sarbanes-Oxley.

IN any case archiving is trivial and there is no need to duplicate system functionality in yet another application. Email logging is built into almost all email systems. Clustering is available in all major OS'. Setting up country applicable audit trails is trivial.

You're just FUD'ing.

---

Adopt an astroturfer. Make their life hell.

Re:Here we go again.....

[Antique Geekmeister]

Yes, they do have to maintain some compatibility. But that compatibility is to be able to read _old_ versions, not to keep new files from new versions of MS Office compatible or even legible to old versions of MS Office. Upgrades that work for MS owned software seamlessly, in particular, but by default save old files in new formats, are absolute hell to keep interoperating with third-party tools.

The result is that features can be added to Outlook that are not compatible with _any_ third party software, and even directly violate third-party API's, and they can and will say "gee, you should have used Exchange/Outlook/Word/Excel! That works!!!!"

Re:Here we go again.....

[EvilRyry]

How many exchange admins do you have that it would cost "several hundred thousand dollars"? We have tens of thousands of users around the world and a single exchange admin that keeps up just fine (plus a few on/off folks that pitch in when he's out).

Would this transition require anymore testing than upgrading to a new version of Exchange? For us, upgrading exchange meant setting up new VMs in a test environment, testing everything, new VMs in production, and then make it start seamlessly migrating users to the new cluster.

Assuming that OpenChange works as well out of the box as a stable version of Samba, the process should be pretty much the same as an Exchange upgrade.

Re:Here we go again.....

[Kjella]

Sarbanes-Oxley applies to the USA only.

And to any company that is on a US stock exchange, or is linked up through corporate ownerships to a US listed company or is a company that is considering the possibility of being bought by or merged with a company or is a joint venture with a US company or...

Let's just say that I'm far from the US and I've heard SOX mentioned quite a few times anyway. Those that don't need to comply are usually thankful for that but even they will easily want "SOX-compliance" on any product featuresheet. It tells you that you have the capability even if you haven't put the routines and staff in place to actually be compliant. The last thing they want is to find that they must be compliant with something SOXish and recieve a huge bill of required upgrades. The difference between chose to and forced to upgrade can be quite huge, particularly for the manager in question "The lousy system you picked last year now forces us to do $5m worth of upgrades!" is a much bigger career killer than the typical bad choice of software.

Kontact is cool.

[SanityInAnarchy]

If by "KDE integration" they mean Kontact, I'm all for that.

Mostly because of the design -- Kontact looks and feels like a monolithic, Outlook-esque application. Instead, it merely combines pieces you already have as standalone programs -- KMail, Akregator, KOrganizer, and so on.

Re:Kontact is cool.

[sskang]

KParts is a non-unique concept implemented pragmatically, leading to KDE devs actually using it.

The entire framework, from querying, instantiating and integrating KParts is optimised for the common case, ie shared libraries used in-process on the local machine, which means it's easy to learn and use.

Other attempts such as Bonobo and the erstwhile KOM/OpenParts were designed for maximum flexibility but didn't catch on because they made developers' lives difficult for these common cases.