Slashdot Mirror

Can you give examples of good Exchange replacements?

Yes, for that see DVL. Seriously, though you have to define what activities you need to do before you can ask for a replacement. MS Exchange is marketed in many niches and fails (on the surface) in most. The most spectacular is its failure as a mail server replacement, if you look at it as such. If you look at the wonderful cover of plausible deniability it gives executives by randomly losing and delaying mail, then that is a success.

Anyway, try looking these. Keep in mind that, unlike with M$ products, you can combine pieces of several packages.

• Kolab — http://www.kolab.org/
• Citadel — http://www.citadel.org/
• Dingo Calendar Server — http://andrew.triumf.ca/dingo/
• Darwin CalendarServer — http://trac.calendarserver.org/
• Bedework — http://www.bedework.org/
• Zimbra — http://www.zimbra.com/
• OpenGroupware — http://www.opengroupware.org/

If you are simply looking to improve reliability of e-mail they a plain Mail Transfer Agent (MTA) will do. Before it became too embarrassing for M$, it used to be recommended practice to put one of these in front of MS Exchange to improve reliability and security. Also look up ClamAV, Spamassassin and how to do greylisting.

• simta — http://rsug.itd.umich.edu/software/simta/
• Dovecot — http://www.dovecot.org/
• Postfix — http://www.postfix.org/
• Exim — http://www.exim.org/
• Sendmail — http://www.sendmail.org/
• qmail — http://www.qmail.org/

However, before you can think about "replacing" MS Exchange, you will have to get rid of the staff that selected and deployed it in the first place. They ignored all the licensing shortcomings, the bad reviews, high price and ongoing technical failure to instead push ideology over technology. People making decisions based on ideology are not going to accept any technical or economic arguments...

Can you give examples of good Exchange replacements?

Yes, for that see DVL. Seriously, though you have to define what activities you need to do before you can ask for a replacement. MS Exchange is marketed in many niches and fails (on the surface) in most. The most spectacular is its failure as a mail server replacement, if you look at it as such. If you look at the wonderful cover of plausible deniability it gives executives by randomly losing and delaying mail, then that is a success.

Anyway, try looking these. Keep in mind that, unlike with M$ products, you can combine pieces of several packages.

• Kolab — http://www.kolab.org/
• Citadel — http://www.citadel.org/
• Dingo Calendar Server — http://andrew.triumf.ca/dingo/
• Darwin CalendarServer — http://trac.calendarserver.org/
• Bedework — http://www.bedework.org/
• Zimbra — http://www.zimbra.com/
• OpenGroupware — http://www.opengroupware.org/

If you are simply looking to improve reliability of e-mail they a plain Mail Transfer Agent (MTA) will do. Before it became too embarrassing for M$, it used to be recommended practice to put one of these in front of MS Exchange to improve reliability and security. Also look up ClamAV, Spamassassin and how to do greylisting.

• simta — http://rsug.itd.umich.edu/software/simta/
• Dovecot — http://www.dovecot.org/
• Postfix — http://www.postfix.org/
• Exim — http://www.exim.org/
• Sendmail — http://www.sendmail.org/
• qmail — http://www.qmail.org/

However, before you can think about "replacing" MS Exchange, you will have to get rid of the staff that selected and deployed it in the first place. They ignored all the licensing shortcomings, the bad reviews, high price and ongoing technical failure to instead push ideology over technology. People making decisions based on ideology are not going to accept any technical or economic arguments...

Sadly this has been the situation for quite some time. I have been looking for a completely open standards based solution and have found little luck over time. OpenGroupware was able to technically fulfill most of the needs, but in reality was hard to use, configure, and manage when I last used. However, Recently I came across this project, http://www.bedework.org/bedework/ which looks to be promising. I plan to investigate it thoroughly when I next get a chance.

This question came up on the Fedora Dev list. Of the suggestions offered there, Bedework was not mentioned here.

http://www.bedework.org/bedework/

"Look Ma! The ferocious mountain lion is really a big kitten!"

Perhaps. What worries me is that Microsoft recently vowed to start buying open source companies. Most of the work on standards based collaboration software and related technology that I'm aware of (e.g. Chandler, Bedework, etc.) (exception: Apple's iCal server) is done by a few small tight groups. Are any of them going to withstand millions of dollars in cash for the sake of principle? I'm a die-hard F/OSS advocate myself, but if someone were to offer me a few million to abdicate my principles; money I could use toward my children's education and otherwise bettering my family's life - well, that would be a hard pill to swallow. I have other principles too, like maintaining my marriage for example.

I really hope a large corporation that buys the F/OSS vision (Are you listening SUN?) steps in to preempt such a hostile maneuver. If they do, everyone wins. I'd take a million dollars from SUN before taking 50 million from Microsoft any day; and I bet most F/OSS developers feel the same way. If the big players sit on their hands, we all lose, as there will never be an answer to Microsoft's dominant collaboration suite.

You can have a better OS, but you really need to hit all of the important application targets in order to present customers with a viable alternative to Microsoft hegemony.

It's not just the collaboration suites we should be worried about, BTW. Microsoft could buy UNIX IP as well (SCO is dead, but Novell is alive and well, and already getting cozy). The big F/OSS players better get their ducks in a row fast, or MS may very well soon act to back up their empty bluster with real products, real patents, and real lawsuits.

The *REAL* Exchange killer will be CalDAV. Yet another proprietary calendaring/scheduling back end from Zimbra means competing with MS on their turf - and we all know how that always turns out. But when F/OSS apps implement open standards, they kick ass. Think HTTP/HTML - Apache. Think SMTP/POP/IMAP - Sendmail, Postfix, Cyrus, etc. These are the most important protocols and applications around. It's time to do the same with calendaring.

Proprietary calendaring solutions only extend as far as your local implementation. I want to coordinate my schedule with more people than that. I want my coordinate my calendar with my family, with my friends, with my vendors, with my collegues at other institutions, with local town events, with my kid's school - NOT just my co-workers. I don't want to care if they use webmail, crackberries, or whatever. This requires open standards. The standards exist. They are pretty new. Now we just need good reference implementations. If Zimbra does this, I predict they will own the market.

Bedework already supports CalDAV. Bedework/Sakai integration is taking place. Lightning and Sunbird work on top of CalDAV. CalDAV is happening. So Zimbra: please please please get on CalDAV, so we don't have to continue suffering the interminable curse of incompatible proprietary calendering protocols. As we've seen, when open standards take off, even Microsoft has to play along. It's time.

Lightning supports CalDAV for sharing calendar information. Apple announced yesterday that Leopard iCal Server and the iCal application will both talk CalDAV, they released the server at http://trac.macosforge.org/projects/collaboration. Bedework is making a lot of progress as an institutional calendar server.

Oracle has a CalDAV stack. IBM has some stuff in the works as well.

It looks like exchange will have a fight on its hands very soon.

I've been helping on a CalDAV plugin for Outlook called Open Connector, which allows Outlook to take to CalDAV servers like Apple's and Bedework. We always need help, if you have a lot of experience developing COM apps in C++, come help out.