Open Source Microsoft Exchange Replacements?

Carl Farrington asks: "Do you think you could try to raise public awareness of the importance for an open source replacement for Microsoft Exchange (Outlook/MAPI compatible for shared/public folders). Current offerings are SuSE Linux Groupware Server, Communigate Pro (Stalker Software), Samsung Contact (ex. HP OpenMail) all of which are not open source / free. Kroupware is in development, but there will be no Outlook Connector for it. otlkcon is in slow development as a possible connector for Kroupware. There is also OSER (Open Source Exchange Replacement) which again looks like it needs more help. Is there any chance of getting some people to back this stuff? It's so important and is probably the major problem facing Linux as viable replacements for Win2000 servers." While this seems to be a question that keeps popping up in one form or another, it's always worthwhile to come back and point out alternatives, in development, that might need your help to get off the ground and running. So, if you're looking for an alternative to Exchange, would you be willing to contribute some time to one of the projects listed above? If you've been using Unix as an Exchange replacement, what did you do and how well has it been working?

Ask Slashdot? Just ask the Magic 8-Ball.

"Outlook not so good"

Re:"Fall over" features

Guys, does it have to be that Outlook-compatible?

No, just shared calendars would work for me. Right now I have a RedHat server running UW IMAP, LDAP, qpopper, qmail, and squirrel mail. It works great, except all of my lusers want to share Outlook calendars. I've tried to get them to use the calendar features in squirrel mail, but they revolted and screamed like little children "Why can't I do this in Outlook. Other companies do it, why can't we." I even went as far as adding an IE shortcut in the outlook bar, so the squirrel mail pages would open up inside of Outlook, but they still screaned like little infants for their 3 am feeding. I wish I was anaccomplished programmer so I could contribute to a project, but unfortunately I am a lowly little SA that makes all of the shiny boxes talk to each other. I would gladly contribute documentation, money, or even be a beta tester. Hell I would give up my left nut if thats what it would take to be able to just share calendars for a reasonable price. I've looked into some of the replacement products, and for the price I might as well buy Exchange. Now that I'm done ranting, please somebody out there, please give me a way to just share multiple calendars. If someone could do that then I will worship you like a God!

Re:"Fall over" features

For some reason, I beleive you are earnest - and not a Troll!

That said, Exchange is a bloated, administrative nightmare. ANYTHING else is almost a complete privledge to manage by comparison. Yes, even Notes.

"Let's buy another 500 user licenses for this server!" is a good place to start bitching. I don't want to hear "$9 a user" from anyone ever again! Oh, and another $2 grand for software JUST TO BACKUP!?! This is the most basic and integral function of real server software - not an expensive, after-market opportunity.

Do you have multiple Exchange servers? Are they AD integrated? Do you need to retire the old hardware of the original box? Nightmares never end! The controls for EVERYTHING look identical, and there are eight separate plug-in control panels, each with less than 10% of the needed functionality to perform any moderately complex administrative task. "You are in a maze of twisty, little tree/pane browser widgets, all alike!" Exchange is so deeply, fundamentally flawed from an administrative perspective, that I am caused physical pain, just trying to think where to begin these descriptions! It was bad in 4/5.x, but to "Train Wreck" it's administration into the nightmare-that-never-ends of AD tools...

I'd rather be devoured by the Nameless Horror out of Time.

Communigate

You're right, Communigate isn't open source. It is, however one of the greatet things since sliced bread in terms of functionality/ease of use/stability. It runs on open source, isn't from Microsoft, works wonderfully, and isn't all that expensive.

Good enough for me.

Re:Communigate

It runs on open source, isn't from Microsoft, works wonderfully, and isn't all that expensive.

All else the same, why is "isn't from Microsoft" on that list? If MS put out something that: ran on open source, worked wonderfully, and wasn't all that expensive, why would you let the name brand discourage you?

Re:Communigate

All else the same, why is "isn't from Microsoft" on that list?

Because Microsoft designs their software to be as incompatible to anybody else's as possible and often even to their own. Microsoft technologies only run on Microsoft software and Microsoft software with some rare exceptions only runs on Microsoft Windows which runs only on x86. (No, don't try to play the Itanium card) Unix software on the other hand runs on many different OSes from tens of different vendors on many hardware architectures.

Choosing Microsoft is the final decision, because after that there won't be any easy choices anymore.

Therefore, any non-Microsoft product is usually a lot safer investment because you are not completely dependent on the whims of a single organization.

Mod me down all you want, but you know it's true.

Re:Communigate

Lets slip in a caveat, shall we? Microsoft will run right under other systems - until they find out about it, then all bets are off. That is why they changed the file formats - to thwart conversion programs - not because of any ineptitude on the part of Microsoft employees (I guarantee you if Bill said, "make an open source OS that is bug free", it would be done - but that does not make money - and hence is 'bad' in Bill's world view).

That is the key point behind all of this: Microsoft is morally bankrupt. The company will do whatever it can to ensure total domination. Any words to the contrary are just so much balderdash.

If you still aren't convinced, here are some examples that may shed light on this problem:

Sincerity: Programmer extends a recognized standard for the benefit of everyone; his enhancement is completely backwards compatible with the existing standard.

Insincerity: Microsoft extends a recognized standard, saying its for the benefit of everyone. Then they change their applications to not use the standard correctly - or use loopholes in the standard to prevent other applications from running with the new standard on machines running Microsoft software.

Sincerity: Open Source, and GNU allow users to view and modify the source code of all applications.

Insincerity: Microsoft creates hope in the development community by announcing its shared source initiative. Unfortunately, it limits what is shared, what is not, and by whom.

To put it even more simply: "Don't mind that man behind the curtain..." - The Wizard of OZ. His name is probably Bill Gates.

Re:Open Source is something more

On email? never. Its the lifeblood of our orginization, and there's absolutely no way I'd be willing to put in place an inferior product. And the managemnt knows this. Our IT department can comprimise on plenty of things, but email can't be one of them.

Re:Open Source is something more

Second that.

I'd lose my job if someone found out that I'd picked inferior software on a moral issue (unless of course it was hand coded by a 3 year old kid in a sweat shop).

Management couldn't give a monkeys about the license. They just want to know that when they click Send & Receive, it will indeed Send & Receive. Every time.

Re:Open Source is something more

Management couldn't give a monkeys about the license. They just want to know that when they click Send & Receive, it will indeed Send & Receive. Every time.

Well, at least that eliminates Outlook.

Groupware? MAPI?

Indeed. The feature list [stalker.com] is pretty impressive. The one that catches my eye is:

The CommuniGate Pro MAPI Connector acts as a "MAPI provider". It accepts Messaging API requests from Microsoft Outlook (Outlook 98, Outlook 2000, Outlook 2002, Outlook XP and later) running in the "groupware" mode, and from other Windows applications. The MAPI Connector converts these requests into extended IMAP commands and sends them to the CommuniGate Pro Server.

Which leaves me with two questions: (1) Does CommuniGate really have all the groupware functionality of Exchange? (2) Are there extended IMAP clients that you can use to access this functionality, so you can get away from Outlook/Virusmaker and MAPI/Crashmaker?

Bynari Connector + Cyrus IMAP

I am using Cyrus IMAP as an IMAP server, with the Bynari Connector to do Contacts and Calendars for outlook. This is less than ideal because storing contacts and calenders in a mail system encoded with tnef is plain ugly, but it works. For the windows desktops at least. We don't plan on Linux desktops just yet, but servers, almost totally converted. Samba + OpenLDAP + Cyrus IMAP + Postfix. It's working amazingly. Nothing to patch, no crashes, fast, secure. It's a match made in heaven. Outlook works 100%. I'd like to find a calendaring/contact system that didn't use Outlook though... perhaps something that stored in LDAP, and was very flexible. I dont know what to do with Calendars though.

Re:Bynari Connector + Cyrus IMAP

I'm evaluating the Bynari software now and looking for people with real-world experience with the server. I've had problems getting the client to work reliably with a mirapoint IMAP server but the company claims, fairly enough, that they can only guarantee the connector will work if you use their server. Anyone out there use it? Any other good experiences with the client and other IMAP servers?

All told the bynari people seem eager and their product has some great promise. Yea, I know it's not open source but right now I'd take ANY non-exchange solution for calendaring/contact management in an Outlook client environment. My god exchange is a horror.

Re:Bynari Connector + Cyrus IMAP

I'm evaluating the Bynari software now and looking for people with real-world experience with the server.

I evaluated it (Bynari Insight Server/Insight Connector) over a year ago while looking for suitable Exchange replacements. After my initial eagerness I was ultimately disappointed. At that time my perception was that Insight Connector was an inelegant, unreliable kludge. Very clever, but still nothing I'd consider putting on a desktop. It installed as an client extension to Outlook and behaved and looked like an external, intermediate mail process. It certainly wasn't transparent to the user and added delay and additional onscreen windows and messages that gave it a feel of a "bolt-on" solution. And several experiences of extended pauses while trying to retrieve mail fom the server (on the same LAN) and times when the connector software simply wouldn't do anything certainly wasn't confidence-inspiring.

Also, their Insight Server mail server component was little more than a collection of common open source IMAP/POP/LDAP software with an installer. I felt there was scant value added in additional functionality or ease of use.

Re:Bynari Connector + Cyrus IMAP

We had real problems with Outlook and Insght Connector. Some users seemed to experience failures with synchronising which on two occassions resulted in lost mail. That was using cyrus imapd running on SuSE as the server. Needless to say any lost mail was unacceptable and thus the project ended. That said, the plugin was plain ugly. It changed the way Outlook works from the users perspective - they need to synchronise their mailbox to get new messages, or schedule that every few minutes, when they're used to receiving mail the second it was sent from the desk across the office. The client has to meet the user's expectations if it is to be successfuly integrated into an existing office. That's why there's a need for something that really works with Outlook, as that's what the users are used to, sad as it may be.

OSER need helps... ya' think?

There is also OSER (Open Source Exchange Replacement) which again looks like it needs more help.

I think that's an understatement... from the front page of their site, go to If you would like to help out with the OSER project, please see this page [sourceforge.net] and then click on If you want to contribute code, please see Writing code [sourceforge.net] and then you get...

TODO

Yeah, they might need some help... =)

Honestly, sounds like a great project, but for the love, people...

It runs on Linux, right?

they should name it Linux Open Source Exchange Replacement

You are asking for a lot for a little...

I think part of the problem is that what people are looking for requires a lot of work to create. Exchange does have a lot of features that, while they may not work as well an OS equivalent, work adequately well, are (somewhat) easy to administer and are integrated together. Could a good alternative be put together, definitely, but the amount of work may be more than some are willing to put forth without monetary compensation.

Re:You are asking for a lot for a little...

Part of the problem is that people are looking at writing this from scratch, which is a lot of work.

However, in April 2003 the OOo Groupware team [openoffice.org] and a few Apache James [apache.org] developers discussed building groupware functionality into AJ.

Apache James is already a production ready POP, NNTP and SMTP server, and has partial IMAP support. It is highly componentised, being based upon the Avalon Framework.

Basically, it was determined by OOogw and a few Apache James developers that it was more than pheasible to complete the IMAP support and add iCal and iCAP, plus the necessary authentication modules (LDAP is partly there iirc, and others). This is not a difficult task because most of the foundation work is already done. It's just a matter of implementing the few protocols that are missing.

Sadly it has not been followed up by the OOogw or AJ developers because nobody really has the time - ever the problem with OSS and volunteers. If I were a Java programmer, I would make an attempt, but I'm not.

I guess this post is a feeble attempt to lure some actual Java developers to the cause.

Re:Exchange exchange for unix

I'd go as far as to say that Exchange has been their key product with regards Microsoft's domination of the Enterprise and the establishment of their monopoly. Exchange's role in Microsoft's success is often massively underrated.

They're office suite has only recently become the best. They're operating system has always been technically behind others. Every other Microsoft product has had arguably superior alternatives. Everything but Exchange.

But until recently nobody, other than maybe Lotus Notes, offered worthwhile groupware solutions. The Exchange/Outlook combination has been superior to anything else and is idiotically easy to administer.

If you ask businesses why they use Microsoft (and I'm talking about the tech guys here), the vast majority will list Exchange as a primary reason.

Re:Exchange exchange for unix

You've obviously never had an opportunity to recover one of Exchange's JET-based (that's right, MS-Access) message stores and manually clean the mail queues. And then explain to the CEO why his perfect MS solution ate his email.

Go web based.

Many people have become familiar with using services such as hotmail [microsoft.com] or yahoo [slashdot.org] as their main form of email.

You could take this oppertunity to use something like http://www.phpgroupware.org/ [phpgroupware.org] which will replicate all the mail/collabaration/task/meeting scheduling functions of Outlook.

Also its free and open sores software, take a look at some of these screenshots [phpgroupware.org] or try out the live demo [phpgroupware.org] and see for your self how great it is.
I'd like to mention that I have no affiliation except having a linux server hidden somewhere at work running this and allowing many people who get stupid outlook viruses an account on it too see if they like it, so far I'm getting a great response.

Re:Go web based.

Tried it. Our staff practially revolted at the idea of solely using web-based email instead of having a client. Its a great thing to have, but not as a replacement for a client.

Re:Go web based.

Tried it. Our staff practially revolted at the idea of solely using web-based email instead of having a client.

That's easy to fix: Install Firebird and set ther home page to your groupware server. Then rename the Firebird desktop icon to "Groupware XP Professional 2004."

Chances are the'll never catch on...

Re:Go web based.

Also its free and open sores software

I don't know if that was caused by a spell-checker run amok, English as a second language, or an intentional editorial stance. And I don't care. A new and highly useful descriptive phrase has entered my vocabulary.

SuSE Open Exchange

We started using the Open Exchange groupware where I work, and I must say, it is a very capable and professional package. Beyond the usual email, adressbook, and calendar functionality, I have used it to track jobs and projects, maintain document revisions, and it has all worked very well. I have even become a fan of the web interface, because it really is convenient to be able to access all of the above from any given computer.

Definately a contender to keep in mind...

Bill workgroup server

There's something called BILL workgroup server, and it acts as an exchange replacement.
Here is the url www.billworkgroup.org [billworkgroup.org]

We have gotten to the point...

..where a good majority of the features provided by Outlook can be incorporated into a web based application, thereby reducing the threats created by using Outlook, and allowing portability:

- scheduling, contact management : easy ....
- Attachments : easier....
- calendar sharing : easy...

Give me the man hours, a good development team, a solid web sever and database server, and you could have a semi-decent web based, accesible from anywhere, email solution. Email is such a simple application, and its so feasible to do the same work as a client, via server to browser interaction....

if none of this makes sense, its cause im running on about 20 cups of coffeee...

Re:We have gotten to the point...

Give me the man hours, a good development team, a solid web sever and database server, and you could have a semi-decent web based, accesible from anywhere, email solution. Email is such a simple application, and its so feasible to do the same work as a client, via server to browser interaction....

Been done.

It's called Outlook Web Access [microsoft.com]; it's got all of Outlook's features in a web client connecting to Exchange Servers.

MAPI is Wrong Choice - use Standards

Sometimes you can only replace one part of a system at a time, so you're stuck with some proprietary vendor's proprietary protocol, but whenever possible, you should use standards-based protocols so you have a choice of products.

• SMTP - Outlook Express and Netscape/Mozilla and most other email clients can send mail using SMTP.
• POP3 - Older standard for email retrieval, which Outlook Express and Netscape/Mozilla can use.
• IMAP - Newer standard for email retrieval, which can manage group and folder types of functions. Many email clients use it; not sure if Outlook does.
• NNTP - Usenet standard for groups - works Just Fine, and there are lots of clients, including Netscape / Mozilla's mail clients and newsreaders.
• Web Conference Boards - There are *so* many of these out there, and they're often a much better choice than shared folders or similar groupware. Depending on how many messages you're trying to handle, your users will often find simple dumb systems friendlier than powerful complex systems.
• HTTP and/or FTP - If you're trying to publish files to people, these are much better standards than email. Some of the web conference board things have convenient uploading interfaces, or otherwise you'll need to do permissions of some sort.
• Shared File Systems - SAMBA, etc. - If you're enough of a Microsoft shop to be running Exchange, surely you're also running a file server network of some sort. Set aside a directory for people to drag files into, and tell them to mount it as their "G Drive" or whatever.
  
• Calendar Systems - This is the other hard one to replace, but I've seen a number of calendar systems out there, typically web-based, and you can email people URLs to click on if you want to integrate with email. The one thing MS seems to have done well is encourage Palm and Nokia and other PDA makers to develop tools for syncing their PDAs with Outlook Calendar. I think some of the Linux-based systems have probably done that.

MS Outlook lumps a whole bunch of functions into one program, so if your people get used to using any two of them they tend to be hooked for life. It's not a very good choice, and if you're going to do something like that, it's much cleaner to use a browser as the one big tool you're hooked on.

Re:WTF???

What utter bollocks.

Email is email - SMTP/POP3 provide a perfectly good delivery service.

The mail and the folders should reside on the server. The status of the email (read/unread) should reside on the server. Using POP3, it ends up on the client. Even if you configure the mail reader to leave mail on the server, you don't have server based folders, nor do you have the server keeping track of which messages have been read.

Equally, a mailserver doesn't typically corrupt its own data or require frequent reboots.

Look, if you want people to believe your arguments about whether Linux is better, quit spewing FUD. You complain when Microsoft spews FUD about Linux, don't do the same. Many, many organizations run Exchange with no corruption of the data store and no need for reboots. Get over it - Exchange is a good product.

VB programmers really, really should keep it to themselves.

Does that make you feel better? Not everyone codes in C/C++/Java. Big whoop. I bet I get a lot more done for the company I work for than a C programmer would. Right tool for the job.

Re:MAPI is Wrong Choice - use Standards

And Outlook/Exchange works with *all* of those standards above (except the web forums, the public folders portion of exchange is the answer, and it does have a "dumb" mode). Here is the point that everyone seems to be missing... Outlook integrates it all into a really easy to use, and (suprisingly) intuitive interface. I have been using Outlook 2002 (aka XP) for almost a year now, and I am surpised almost every day how well they have tied my calendar, journal, e-mail, and contacts all together. If you really haven't seen and worked with it, they have really cleaned it up from the last version (2000).

It can be a real timesaver when I need to create a meeting for all the people on an e-mail string, even those on outside e-mail systems (iCal), or have to look up someone in the Global Address List (works much better in Exchange than LDAP mode).

Granted, to use most of these really cool features you have to be running Exchange, However, most features are functional on IMAP and LDAP servers. It just doesn't look and work as pretty as a native exchange install. Once you start pulling these functions apart into different programs, you really start losing functionality. I am not saying everything on your computer should be in one huge mega-application, but these are all related functions that give you a one-stop shop with a clean consistent interface.

Like most people here, if there was an OSS replacement, I would consider it, but we are part of a HUGE Exchange site (US Navy), and we have to have replication and so on. Interoperability is a must, and to be honest, there isn't a package out there that even comes close to matching the feature set and manageability of Exchange/Outlook.

Other side notes..... changing permissions on folders you own (such as calendars and what not) is really easy for users. They just right click, choose Properties, and choose who can see, change, add, etc. I haven't seen anything like that in the OSS world, and is a MAJOR thing, at least in my corner of the world.

Excryption, using PKI certs is a piece of cake, public keys are stored in the GAL, so I don't even need to get it ahead of time. Outlook checks every message, warns of bad certs and sigs, the whole deal. User can be brain-dead, but still send mostly secure e-mail.

I can choose the format of my e-mails (plain, RTF, HTML) and base that on the destination, so that I send plain out on the internet, RTF within the exchange site, and HTML to local addressees, etc.

Ties in with Windows messenger and NetMeeting, so I can click the name on en e-mail and talk to and see someone, using all internal servers, no MSN or any of that crap. Shared whiteboard? No problem. Shared Desktop? Ditto.

Exchange hosts IRC conferences, that can be scheduled via Outlook, and accessed by any IRC client out there.

Those are just off the top of my head. IMO Outlook/Exchange is the best software MS has, especially the latest versions. We haven't had a server crash or DB corruption (with 7,000 users and 2 TB info store) in over a year and a half, and when we did, it was because the SAN died, not exchange. If you have people that know what they are doing running exchange on good non-bargain-basement hardware, it works well and just runs. It's managed by *one* MMC snap-in tool to control all the protocols, stores, folders, etc. That's my $.02......

Re:MAPI is Wrong Choice - use Standards

You're missing the value-add of Exchange (and therefore the hard part of replacing it) entirely.

The complete integration of calendar, contacts, public folders, and email in Outlook (well copied in Evolution) is not client-side only -- it extends into the server. The two most useful and hard to replace parts are:

free/busy scheduler. Calendar, new appointment, select a few co-workers, pick a time. You can see if they're busy at that time or not. Timezone synchronization is automatic. Select some resources as well, like a conference bridge or a video projector -- you can see if it's in use at that time. This is the killer app of Exchange.

global address book. LDAP is great, but few Unixy solutions let you use it from the email composer address field, the calendar address field, and the contact editor. Evolution is pretty close, Mozilla does better but lacks the calendar part.

Public folders which include non-file content. Shared filesystem is okay if I want to share a spreadsheet, but a public folder can include an addressbook of people that you don't need in your everyday book, a calendar showing training schedules and the resources committed, all sorts of goodies like that. VB macros too -- workflow and virii to your heart's content :-)

why must it be OSS

measure a product on it's ease of use, stability, security, cost, etc. whether or not it is OSS or not shouldn't be an issue. it seems that exchange is a rather nasty program to admin, but it also seems that groupwise from novell is quite good. my school district uses it, and it is overkill for most teachers, we just need mostly simple email, but all the collaborative features are good. i think our problems have been on the admin side, since school districts aren't known for paying top dollar. if there was an OSS replacement fine. but it isn't the be all, end all. sorry. unless you're RMS or something, everything isn't about software philosophy. there are tons of good middleware apps for linux, and more to come. whether they are oracle, notes, db2, etc. just let the best program win.

Evolution connector

Maybe you should start by asking the Ximian guys why they can't release their Exchange Connector's source code... There is probably something in there that they don't own, so they had to license it. And if they had to license it, it is probably something so complex technically or legally that the average OpenSource developer should definitely not put his nose in it.
I am assuming of course that the poster/submiter of this story wants compatibility with existing Exchange clients, right?

Re:Why does everyone want to copy MS products in O

Shouldn't OSS be about solving problems that people want to work on rather than trying to be a cloning engine for Microsoft software?

Bingo. Sometimes I shake my head at the lengths people go to bash M$ at every chance they get, then spend tons of effort to clone them. The first blatent one was when RH shipped thier default windowing system to be FVWM95 [plig.org]. I still havn't gotten over that one. KDE and to an extent GNOME are not too far behind either. For example. Why in the world do they put the start thingy/taskbar/icon collector at the bottom of the screen? Because M$ put it there first. Take a look at your browser. See all the menus up top there? See the titlebar to move the window and close it etc? Shouldn't the taskbar be up there too?

Look at StarOffice and OpenOffice. They seem familiar. And there are plenty of others, but I think you get the point.

Another thing that M$ gets bashed on here is because they "embrace and extend". Many, many open source projects do exactly this.

Don't get me wrong. I like OS and there are beautiful examples of its success, like Apache, Linux, Galeon/Mozilla. The last one is an excellent example. I never thought of what I would want out of a browser, I just knew they all sucked a few years ago. However, Galeon is exactly what I want out of a browser.

So, what software do I use on a daily basis? Linux [kernel.org] for an OS, WindowMaker [windowmaker.org] for a window manager, mutt [mutt.org] for email, vim [vim.org] for an editor, and lord forbid a closed source calendar called corporatetime. I believe that Oracle bought this, its difficult to find info about it anymore.

So what is my point? I get along just fine without M$ nor do I use any software that really has a M$ equivalent. Why do these topics come up all the time? Maybe we should be cloning M$'s slogan too. "Where do you want to go today?" It is a fitting question, right now the answer seems to be "Wherever M$ was yesterday?"

What you'll need

I think what most people forget is that in order to replicate Exchange's functionality, or even come close to offering a fraction of the features of Exchange, you're going to need to put in close to the same amount of work that Microsoft did. If I remember correctly, Microsoft had a team of no less than a hundred programmers working full-time for two years to produce Exchange server 2000. Logic would dictate that the Open Source community would need to do the same, with the same amount of resources. A considerable undertaking. I believe it would make more sense to enlist in a corporation like Red Hat (who doesn't have the same amount of resources as Microsoft, but they do have the talent and organization) to begin development on a project such as this.

Re:What you'll need

and if you've ever admin'd an exchange box it would become clear that those 100 developers were from 50 countries, spoke 70 different languages and, since they hadn't developed their collaboration server yet, had no way to communicate. Plus 10 of them were from a country at war with 5 others and 1/3rd didn't have any computers to test with so they wrote code on paper and mailed it to MS. I'm pretty sure the guys who wrote the brick backup did it on paper, certain they never tested it.

And and and OH jesus don't get me started. Exchange=evil-come-to-earth.

Been waiting years for this type of alternative.

It seems that, for years, I swore our IT department would not convert off of Groupwise until we had an open-standards alternative that gave us the same integrated mailbox, public information store, and calendar solution. That was back in '97. When nothing prevailed to grasp as an integrated standard, the pressure finally caved when we had to make the choice between upgrading Groupwise or migrate to Exchange.
As we reviewed the options, we noticed that the only reason we were still using Novell servers was to support Groupwise. It was at this point that we did a cost-of-ownership study and found that supporting aging Novell servers was going to cost us more over time than a single platform solution from M$. The choice was made to convert.
Our conversion was very successful, and recieved much praise from the end users. Why? Because they all wanted to use Outlook. No one really cared that we were using Exchange, what they really wanted was Outlook. (Btw, the Groupwise plug-in to Outlook sucked at the time, maybe better now, but back then it was terrible)
As an Outlook user myself, I have to say that it is a great application. It works well, provides many options, and integrates with everything.
With that said, I believe our IT team would readily accept an opensource alternative, particularly if we could cut down on the cost for licenses. Not only that, but many of our partners and clients would convert too if they didn't lose Outlook. Honestly, I think fewer and fewer people outside of IT even know what Exchange is. All they want is outlook.
I can't offer much to the development of an Open Source Exchange replacement, but I sure would love to see one sprout up.

Not just another rollup

Here's a project worth checking out: Citadel/UX [citadel.org]. Admittedly it's only about 80 percent of the way there, but the thing that makes Citadel stand out from its open source brethren is that it's not just another Cyrus/Postfix/OpenLDAP/etc. rollup with some loose stiches put in to make them act like a single system.

We're actually taking the time to build something good from scratch. We've got a true journalling database oriented message store (thanks to Berkeley DB) including single-instance store (a message sent to 100 users doesn't get saved 100 times). Built-in IMAP, POP, SMTP protocols. A nice calendar service, and a Web interface. It's even got its own instant messenger.

The thing that's important, though, is that it's designed to be easy to install. One of the very few things that Exchange 5.5 had going in its favor was that it was relatively easy to install. Citadel aims for that as well -- plug in the RPM's or tarball, run the setup program, and you've got a basic server up and running. Inexperienced admins might be scared by editing /etc/mail/complicated.cf and /etc/init.d/S90scary.sh, but they don't mind running a "setup" program and then customizing with a web browser.

Where we really need the extra development work right now is to start writing some connectors for popular client software. Currently we are aiming for 100 percent compatibility with the Kroupware project (so you can use the Kontact client without having to install the clunky Kolab server) and eventually Evolution (which has a 'connector' architecture). Eventually we'd prefer to do everything in Mozilla (using Mozilla Mail and Mozilla Calendar), since it's cross-platform.

Again, it's not a drop-in Exchange replacement today, but it's a project worth watching, or better yet, helping out on.

What an assine statement

"It's so important and is probably the major problem facing Linux as viable replacements for Win2000 servers."

Right, because Lotus Notes [lotus.com] has the majority share of corporate e-mail solutions or because Bynari [bynari.com] offers an Exchange replacement that runs on Unix.

This is such a stupid statement. Active Directory is a much bigger problem in replacing Win2k servers since your Linux servers would more or less be stranded on the network as is.

Exchange versus POP, a sad story

I worked at a company which sustained most of the raw network services(DNS, mail) we needed on a single ancient Sun pizza-box single-processor system, maybe 200MB of ram, and one or two rather old SCSI disks. Clients used POP or IMAP to get their mail, and all was good. It almost never crashed(maybe once every 6 months), people liked the speed, etc. This was with 50 employees. All was good.

About a year after I joined the company, we got bought by a company which was thoroughly impressed with itself IT-wise; they were geniuses, we didn't know shit, supposedly.

They DEMANDED we switch to Exchange, because goddammit, we needed to be able to click the "Yes, I'll be there" button when they sent a meeting announcement. So we threw a Quad 500mhz Xeon box with 2 or 4GB(I forget which) of ram, 6+ SCSI drives with a high-end raid controller, etc. at the 'problem' and hoped for the best.

It crashed constantly. It corrupted its database incessantly. It had to be rebooted every week, sometimes more often. People were always having problems with the Exchange client; disconnects from the server, crashes, weird error messages, hosed mailboxes(which meant you lost all your mail). It took forever for the client to launch in the morning when you first opened it. All in all, we went from having to spend maybe an hour or two a month supporting mail services, to a full-time employee spending several hours a week feeding the damn thing. Rarely did people use the meeting scheduling stuff, or any of Exchange's other groupware features. The whole thing was collosally stupid.

Isn't it really fucking sad when a software package barely running on a $30,000 system is worse than a software package running nicely on a system you could buy off ebay for $100, and you did it all to give people features they never used anyway?

A friend worked at a company where someone suggested they move to Exchange off of POP/IMAP services. The CTO intervened VERY quickly and shot the whole idea down, saying it would be a terrible idea.

If someone at your company makes a similar suggestion and tries to get Exchange through the door, tell the execs to find another company that switched to Exchange, and ask them about reliability, TCO, and whether anyone is actually using the few things Exchange gets you over "just a mail client".

Typical FUD/Lies (was some BS story)

Why must people resort to lies to promote their holy cause? ANYONE who's really used Exchange (and has even half a brain) knows that this story is complete horseshit.

Look - here is a real one for ya all. Dual PIII-1000 system, 1 gig of RAM, mirrored pair of 72 Gig 10K SCSI drives in a 2U SuperMicro chassis connected to a 100 mb/s burstable circuit at level 3. That's what my company uses to host our exchange users; our own use plus those we host for.

Setup? Lesse, a basic load of W2K, hit windows update and did'em all. Single vendor provided driver was for the SCSI 0-channel RAID card. Time? About an hour.

Loading Exchange 2000? First, run dcpromo to turn this box into an Active directory domain controller. This process also automatically installed and configured the DNS. Then stuck Exchange 2000 CD in drive, followed the next next next, finish clicks and sat back. About 30 minutes later Exchange was running.

Configuration? Added domain name, added a user and left the checkbox for "Create Exchange mailbox" checked. Bingo, new user with automatically assigned e-mail address based on policy we wanted to use.

Full web access. Done. Full shared calendars and public folders. Done. Delegate access with full ACLs. Done. Offline support. Done. POP3 support. Done. IMAP support? It's in there. NNTP? All set. Instant Messenging? It's in there. IRC (chat) - It's in there. x.400 and SMTP, of course. No open relays by default. S/MIME? Digital certificates? Yep and yep. The list goes on, I won't bother with any more.

Total time to get up and running, a single afternoon.

OK, so it's up - now what? Well... nothing. Every night we do a backup, using built-in APIs that allow backing up without taking the information store offline. Virus scanning runs automatically and updates itself daily automatically. Antispam is fully automatic using statistical and phrase filtering. Nothing to do but look at the cute charts of spam blocked by user. Every so often there might be an applicable windows update to do - ok, so, hit windows update, download and (the ONLY part that sucks, I'll admit it) reboot.

That's it. Our uptime is 100.00% The only reboots are planned. Period. The hardware is not esoteric. The loads are easily managable on a simple dual PIII.

Client performance is flawless, and very fast. Database corruption? What's that? Never seen it. During preproduction testing we regularly would pull both power cables simultaneously while the machine was doing an full-text indexing crawl across our 60 gigabyte stores. Upon restoring power the entire server came up without a single hitch and without any delay whatsoever; the failed crawl was detected, and restarted. Transaction logs were played back and 0% loss sustained. We did this at least 30 times without ever suffering a single corruption or anything more than a few red Xs (something needs fixing) in the event log (followed by a few yellows (we're fixing it) then pretty blue I's to tell us "it's fixed.")

Anyone that thinks Exchange is just a POP/SMTP/IMAP server hasn't a clue. Anyone who would like to tell you that Exchange crashes is either lying or can't run a server. Period. With over 75 Exchange boxes in production and never a single chance to test our off-line disaster recovery plan -- we could not be more pleased.

Perhaps we need

an Outlook alternative.

Most of the problems seem to be with MAPI and Microsoft COntrol what Outlook does. However, on Linux we aheva hugely ca[able email program in Ximian's Evolution. If it were to exist on Windows and have a server based company wide contacts calendar sharing and task managment Microsoft would be under pressure even on their home turf.

qmail/vpopmail/qs/clamav/squirrelmail/openldap/etc

In my office we completely remove exchange and put up a complete system without shedding any money (FREE) except for the hardware of course. We used it for both local and internet mails.
FREE software:
qmail - mail
vpopmail - pop3/multidomain
courier-imap - imap3
qmail-scanner - email filter
spamassassin - spam filter
squirrelmail - web-based mail
openldap - email directory
clamav - antivirus
ezmlm-idx - mailist
apache - webserver
qmailadmin - email administration

With this u can use clients eg outlook, mozilla mail, evolution, eudora, etc

Features
SMTP Mail Server with SMTP-AUTH (Plain, CRAM-MD5), TLS (SSL) support, and SPAM/Virus Scanner.
POP3 Server with APOP and SSL support
IMAP Server with TLS (SSL) support
WebMail Server
Quota Support (usage viewable by webmail)
Autoresponder
Mailing Lists
Web-Based Email Administration

Why clone a bad idea?

The notion of tightly integrated servers and clients strikes me as stupid. I'd much rather use a high-quality web-based groupware suite. If you really must have a GUI for some operations (e.g., calendar maintenance), it can be implemented as Java applets or through SOAP, but with the web based interface being the primary interface.

What about windows versions?

While I find the idea of an Exchange replacement under Linux nice, it's also worthy to note that a lot of 2K/2K03 IT admins would probably like an exchange replacement running on Windows as well. It's not because you can afford a windows liscence that you can necessarely afford (or actually want to) shell out extra money for everything that could be replacable and potentially more stable/easier to manage.

What I hate about MS's licensing isn't the fact that it costs about 50$ per CAL seat after paying for the OS itself, that I can live with it. What I don't like is all those CAL seats for ALL the software after... it's nuts, CAL for SQL after buying SQL server, CAL (client access licenses) for MS Projects after shelling 1000$ for it, CAL for this CAL for that, in the end, your server for 50 users costs a fortune, and forget it if you want to run it in cluster mode; there's no rebate, you have to shell out exactly 2X for the licenses, plus Win2k costs you more for Advanced server (because win2k server cannot cluster). I think you can make 2 nodes with the standard 2003 server though, but still... you need 2x of everything.

At work I simply ditched Exchange and used a standard POP3/MAPI E-Mail server (merak) which came cheaper. For the contacts management and exchange of information, we run this through a local intranet that does its job pretty well. Of course having something like exchange would be really nice, but the horror stories I heard about it and the fact that I would have to shell out another few grands out from my budget simply made me back off.

If there's anything replacing Exchange and/or having some solid functionnality for outlook running under Windows out there, I'm sure there would be a lot of people willing to at least evaluate it.

Recent Article on this topic

There was a recent article in the April 2003 edition of Linux Magazine [linux-magazine.com]

They discussed and tested the following

1. SuSE OpenExchange Server 4 [suse.co.uk]
2. Samsung Connect [samsungcontact.com]
3. Stalker CommuniGate [stalker.com]
4. Easygate Workgroup [n-h.de]
5. Bynari Insight Server [bynari.com]

Only Easygate and Samsung had full Outlook MAPI support, whilst Communigate and Bynari Insight Groupware had partial support.

The April archive is online and link is here [linux-magazine.com]. There are a number of PDF files with the article details in them.

Use Emacs

As an operating system it supports everything.

Chronos as a replacement

LQT Systems [www.lqt.ca] has been selling Chronos [sourceforge.net], a system I developed when I was working there, to many clients. Numerous enterprises have replaced the calendaring part of Exchange with Chronos successfuly. The tools are out there. You just have to find them.

A list of candidates

There tends to be confusion in these discussions because of lack of agreement on what the term "Exchange replacement" means. At one extreme, something qualfies only if it accepts Microsoft-proprietary RPC connections from MS-Outlook for MAPI transactions providing 100% of the functions the Outlook / Exchange Server combination du jour supports. At the other extreme, Web-based access (e.g., Sherpath) and glorified BBSes (First Class, Citadel/UX) are deemed worthy of consideration. Anyhow, here's a list I maintain as part of http://linuxmafia.com/~rick/linux-info/groupware [linuxmafia.com]:

• MS Exchange Server (server end; NT only), MS Outlook (client end; Win32, MacOS). Very limited support of open-protocol clients (IMAP, webmail?). Microsoft Corp. wants to sell you Exchange 2000, these days, but Exchange 5.5 is still very common.

• Lotus Notes / Domino (server end, Linux supported), Lotus Notes (client end; Win32, MacOS). Limited webmail access (iNotes).

• Novell Groupwise. http://www.novell.com/products/groupwise/ [novell.com] Server end runs on either Novell NetWare 5/6 or WinNT. Client end is proprietary Win32 client or webmail. A native Linux client is under development.

• SuSE Linux Openexchange Server (formerly SuSE Linux eMail Server). Standard, good open-source components (Postfix, Apache, Cyrus IMAP, OpenLDAP, OpenSSL) preconfigured to work well with one another, plus a couple of proprietary components: YaST2 for graphical administration, and SkyrixGreen for integrated scheduling and group discussions (shared folders). Client access from any OS, including but not limited to webmail. A full-functional trial version (lacking only "maintenance") is available for US $20 at http://www.suse.com/openexchange/slox_eval_form.ht ml [suse.com] . Sites are known to scale well to at least 1,000 users per site. The largest deployment yet known (March 2003) is 1,900 users.

• Bynari Insight Server, http://www.bynari.net/ [bynari.net] . Server end is Linux-based. Intended as a plug-compatible replacement for MS-Exchange Server, based on POP3, IMPA, SMTP, and LDAP, but also with full support for all the special, proprietary MS-Exchange Server RPC-based protocols for group discussion, scheduling, contact management, task lists, etc., when used with MS-Outlook clients. Review: http://linuxjournal.com/article.php?sid=6734 [linuxjournal.com]

• Bynari InsightConnector, http://www.bynari.net/ [bynari.net] . Extensions that load into MS-Outlook clients to let them perform MS-Exchange-type functions (scheduling, contact-management, public folders) without needing an MS-Exchange server, using only open-standard IMAP, SMTP, and LDAP servers, instead.

• Samsung Contact (formerly HP Openmail), http://samsungcontact.com/en/ [samsungcontact.com] . Server end can be Linux-based (or Solaris/AIX). Based on SMTP, IMAP, POP3, LDAP. Supports proprietary protocols for e-mail, scheduling, etc. native to Samsung's Contact client (which is available on Linux and Win32). Webmail access. Implements Microsoft's (documented, for a change) MAPI protocol for scheduling, public folders, offline folders.

• Oracle Collaboration Suite, http://www.oracle.com/ip/deploy/cs/ [oracle.com] . Formerly Steltor CorporateTime, http://www.steltor.com/ [steltor.com], until that firm's recent acquisition by Oracle. (That product is said to have emerged from Netscape Calendar.) Does IMAP, POP3, SMTP, E-mail, real-time conferences, voicemail, scheduling. Apparently implements all of the special, proprietary MS-Exchange Server RPC-based protocols for group discussion, scheduling, contact management,

Kroupware?

Kroupware? There is an Open Source product whose that is going head to head against major proprietary mail server packages, and someone actually thought to call it 'Kroupware'?

Is that like 'HackingCoughWare' or, perhaps, the more subtle 'ScreamingInfantWare'? Ok, perhaps this is a troll, but I've historically had a hard enough time selling open source stuff into various enterprises. ("MySQL? Aww, what a cute name. Now go get us something that sounds professional." I've heard that. Literally. Twice.) I realize we're all smart enough to know better.

Selling a product is as much (if not more) selling an image than it is selling features, reliability, etc. At least for the PHBs I've had to sell to in the past. Trying to bring a mission critical piece of software in that's named after an anoying childhood malady will, before anything else, elicit a bunch of laughs from the powers that be, and then there's that much more of a hole to dig out of.

Oh, well, there goes what little karma I had, but I had to say it.

Our solution to the Exchange & Outlook problem

The users in our company are heavily dependent upon Public Folders and the Calendar in Outlook. Yet, we were being eaten up by Spam and the odd virus that would get through our filtering on the Exchange server. It got to the point where we had daily downtime and two scheduled daily reboots of the Exchange server.

Our solution was to remove the load of incoming email from the Exchange server, moving over to a FreeBSD/SendMail/SpamAssassin POP server. Internally, the Exchange server is still available for Public Folder, Calendars and in-house email, but all outgoing and incoming email never hits the Exchange server.

We didn't remove Exchange from our organization, but we did remove it's biggest liability: MS-specific virii and Spam.

Re:What does Outlook do besides carry viruses

Outlook Express does little more than email.

Email is the least of the features of its big brother, Outlook, however. Outlook handles: task lists (very important...our comapny uses hierarchies of these task lists for all bug tracking in development , because it's stupid enough to be flexible with regards to input), global contacts (as in, for an entire organization), group management, Sticky notes, alerts, a "journal" which tracks changes on all your office docs (fucking awesome), syncing with pocketpc and I THINK palm, publishable schedules, and this is jsut the stuff I actually USE.

Best of all, Outlook is pretty stable, unobtrusive, and surprisingly easy to use. And since our smtp server cleans viruses before they even GET to Outlook, the second biggest downfall is eliminated for us. The biggest, of course, is price, and our license came "free" with the MSDE subscription we get anyway to do our work.

I prefer Squirrelmail for email, and use iCal at home for the killer rendezvous support. But for doing all the sundry business crap I gotta do on Windows, Outlook is second only to a personal assistant (insert secretary shagging joke here).

Re:What does Outlook do besides carry viruses

It really comes down to people being cheap, nothing more.

Exchange 2000 (I've tested 2003 and it's going to be impresssive) with the current SP is quite nice. During the initial launch there was a lot of bugs, noting the first SP was bigger than the cd install.

Now, it's pretty rock solid. We use a sendmail front end under solaris to do our initial queuing, aliases and distribution. Being skeptical over migrating to Exchange we decided to keep our sendmail frontend in case of a catastrophy. From there we have 4 exchange servers, 3 in the US, one in Europe. It could be done with 1 but we cater to remote offices to make their lives more pleasant.
We previously had a Netscape + pop3 implementation about 3 years ago.

We have 1 exchange and domain administrator for nearly 600 employees. One. The amount of problems and headaches we go through is quite minimal now.

For the price you pay Exchange just works now. You can have a functional server up OS + exchange install in about 3 hours if you know what you are doing.

Oh and screw the smtp gateway for antivirus scanning. That won't do you any good if an internal user sends an email to another. We've been using Antigen from Sybari. It does real-time scanning with 3 different engines, incoming and outgoing. It will also scan any message you move between folders or grab from a personal folder you just attached. We've never seen a single virus, Not 1, get through in nearly 3 years.

I know I've heard many horror stories of Exchange 2000, Outlook and viruses. I truly believe if you take the time to sit down and plan the installation (most people just jump into shit blindly) you can have a very competent mail system running on a Microsoft product. The problem is most Microsoft admins are guilty of being next next next admins and give MS a bad name.

Re:What does Outlook do besides carry viruses

Listen. I moved to this system from TestDirector, and I must say it kicks the SHIT out of the latter. I've also used a couple other solutions I won't mention because even though they suck, I'm friends with their developers.

1) Outlook's native, not activex/java-inna-window, so it never crashes. TD (and one of the other apps) has the tendency to do that unless you use their client, and they make you pay per license. Outlook's client is basically "free," since it comes with MS Word, Excel, and all the other crap you "have to have" at a business.

2) Most of them FORCE you to enter information. This can take a long time. Sometimes, I just want to add a task to remind me to find a faster way to execute an algorithm. It is much quicker and much easier to use Outlook.

3) Generally (at least 90% of the time), even WITH all the extended information, I needed to meet with the tester who found a problem to watch them replicate it. It's nearly impossible to codify some of the more complicated activities we perform, and many testers, sadly, aren't technical writers. They're clever sadistic people who get their jollys off in proving you wrong (j/k guys, I love you all! Beta Forever!)

4) There's nothing by way of completeness or exactness that you get in a bug tracking system that you CAN'T get with Tasks. Need to know what version they're running? Say, "hey guys, when you enter a task, include the version." Done. Need to include a screen shot, patch file, etc? Done. Need to SEARCH on these things? Done. Maybe not as nicely as you'd like, but you can do it...and it's already here.

But then again, I *like* post-it notes on a whiteboard. And I used to work with this guy [softwarereality.com].

Re:Not wishing to appear ignorant but..

Exchange handles Active Directory integration (you need to add Samba for that), IMAP, POP3, and shared collaboration folders. You can cobble together replacements for most of it except the requirement to handle MAPI for Outlook integration.