Slashdot Mirror
Can We Finally Ditch Exchange?
"With new releases on the way, like Mandrake 9.0 and the new Lycoris can we who try to use Free Software in business environments hope for any change? Do the commercial Linux distros have any plans to implement a free replacement for Exchange, including a Win32 client-side bridge? If not, why not? Do you feel it is too cost prohibitive to imitate Bynari in this case, or is it a decision more along the lines of 'we'd rather you used Evolution and Mandrake/Lycoris/Whatever, rather than OutLook and Win32'? If it's the latter I'd be severely disappointed, and I don't think I'm alone. Any discussion on this topic would be appreciated; but what I'd really love is a community push to get this done. Perhaps a running Web-A-Thon to raise the money to simply purchase the technology from Bynari? I personally think it would be a great move towards grabbing market share from some of the other distributions, some of which have the technology but choose to keep it closed, as well as from the Great Dragon. What do you think?"
Untill there is a standard calendar protocol, and that protocol is supported by exchange, you won't be able to get rid of it.
second society
1.) Companies are having difficulty implementing the calendar system that Exchange uses properly. 2.) Microsoft professional support, big business likes the idea of having someone to blame when things don't work. They sign contacts that make people have it fixed within a specific time period or they recieve massive compensation.
At the place I work at, we have both an exchange server and a POP3/IMAP server. We catch so much flak from our users over the exchange server than the POP3 server, Then again 90% of our users use OutLook even though we have a site license for Eudora,and offer Netscape and an IMAP web client. They get lost even though we have an excellent web-directory and one of the best calendar projects around. Everyone has a public folder that they can put stuff into to share with the rest of the network byt they still insist on using Exchange. Personally I can't find any feature that justifies all of the garbage we have to put up with to get it running. Outlook sucks!(there goes my karma) Outlook crashes more often then IE, Outlook is targeted my more virii then ever. If these people would change their mail client, they wouldn't have this problem. The exchange server is jacked up as well. We have to call and have it re-set every three days and I'd bet that the network "gurus"(/sarcasm) don't know how to admin it either!!!... argh!
First, most non-tech corporate types have heard of Exchange. Next, they like to have someone to sue. Even those projects with companies behind them don't have much to go after. Even though Microsoft has a EULA that supposedly frees them from any liability if the software screws up, it makes the corporate types feel better. Also, they can hire any MCSE off the streets to run the Exchange server. There aren't many standard certs that they can rely on when they need to hire your replacement after you've bundled together all this unfamiliar software on their servers. When you consider the hiring difficulties, lack of certifications, and lack of accountability of the authors of the software, the open source projects may, in fact, cost a good bit more than the $10,000 worth of Microsoft software. The entry costs of this software look enormous to individuals, but to corporations, it often doesn't appear to be much money. Corporations care much less about software politics than most of us do. The open source solution has the benefit of getting out of proprietary formats, but I don't think that's very high up on the list of priorities of the people making the decisions.
At work, we've been trying to switch over from using exchange but a lot of people have implemented some very neat features, like for instance. If I have an appointment to do a stress test on a patient, the nurses send an email so that it is loaded onto my palm pilot when i sync and a letter is automatically printed out letting the patient know when the test is scheduled.
The IT guys think they may have found an exchange server replacemetn with SUSE but for now exhcnage is very useful and would be very hard to replace.
Thanks for reading
Sigs are dangerous coy things
There has been plenty of call for an open source groupware application like Exchange and as yet there are still none. My appologies to the folks at PHP-Groupware but, even though this is often cited as a solution, it simply isn't an adequate solution especially for a medium or large enterprise.
Frankly, I had always thought that the Sendmail folks would be the one to deliver. They have certainly nailed down the mail side and I feel that they could do a great job integrating calendaring and other groupware features, most importantly a programming interface to make it an extensible solution like Exchange or Notes. Unfortunately, as of yet, they have not indicated that they are pursuing this.
OSS is still out in the cold when it comes to an OSS Groupware application that scales.
The problem with evolutions is that is doesn't run on windows, so it is fairly usless to mixed shops. Also, it only supports exchange AFAIK when it comes to calendar/scheduling. In theory it supports LDAP for the address book, but the gloss factor in the manual on that point is very high. (see for yourself: here)
I know that the small software company I work for would love to have Evolution on every desktop (windows and linux) using LDAP for a shared address book and calendar, but it just can't happen today. Oh well, here is hoping that the Kompany can get Aethera right sometime this decade...
apt-get install redhat please god - Me (take it easy, I love Debian)
The whole concept of Exchange, in my opinion, is flawed. Each Exchange server recreates a mini-internet within a lan, that connects to other mini-internets within other lans, tied together by wans (or mans--as the case may be) and also tied together by the real internet [a nebulous definition goes here].
Here, we have one Exchange server for 150 people. But then there are 9 locations, from San Francisco to san Diego. They all hit the same server through the wan.
Remote users (15+) also use outlook web access (i't really Exchange web access if you think about it) to access their mail. We have to allow that traffic through the firewall.
And every single one of our people have one or more other email addresses (AOL, Earthlink, RR, whatever).
I would say: have better addressing handling.
Email was first created by geeks for geeks (at univs. and gov.) and served its purpose well. When the move was made to the company, the whole transition was just done wrong.
I say the Exchange servers should be totally eliminiated in favor of a non-lan/wan centric solution (watch your step, marketing words all around), namely a true internet application, shared, replicable, and reliable.
As far as calendaring is concerned, we don't use it much. Our corporate values promote face-time and intelligent conversation more than lines on a spreadsheet, so meetings are more dynamic, more fluid, and less apt tp be "scheduled". Usually it's a phone call.
Anyway, I digress.
But this may be the reason no open-sourcer wants to tackle that issue. It may subconsciously feel flawed to recreate the Exchange architecture.
"Piter, too, is dead."
Hmmmm.. Looks like I got a bit carried away there, but you get my meaning.
Reality is defined by the maddest person in the room
I was approached by Bruce Perens at LWE and he stated that Debian needed better support for Open Office. I looked at him and told him as soon as one of us had a reason to care we would.
This is the fundamental problem with Open Source in business land -- you need a coder who has the time to code and actually cares about making it work. I see lots of sysadmin types complain about Exchange but no one seems to hate it enough to sit down and work on something better. Most of the businesses approaching Mandrake, RH, etc are looking to dump the Microsoft solutions entirely so Exchange is not a big deal there. Or they are only looking for server -> server solutions and not desktops.
Last but not least you have the problem that Exchange is 100% proprietary. Look at all of the "fun" Samba has had trying to get smb interoperability right. I also bet Microsoft would be VERY apt to sue a company that did this into the ground. Might as well paint a target on your head.
As with every other itch you just need to find someone to scratch it. You mentioned "clients", why not funnel some of that contracting cash to coders willing to work on the project.
OEone and Mozilla are working on an Open Source calendar server. Support it!
I know that mentioning Lotus Notes violates the Code of Slashdot Posting, but take a look at Notes sometime. The people who designed that system spent a long time thinking very hard about how to build a mobile, distributed, secure groupware system (note: you do not need to agree with the solution they built to acknowledge that they thought very deeply about the problem). Then - they spent a lot of time and money building what they had designed.
(Exchange is basically an imitation of the 45% of Notes' features that are most commonly used, without the thought, design, or security).
Who in the Free Software/Open Source world is going to spend that kind of time and effort? Particularly given that most Linuxians fall into the "don't like groupware" camp?
sPh
I've implemented several Exchange servers in a few large organizations in my day and to replicate what it does would be not be easy.
I think this is part of the problem with any existing attempts or lack thereof to replicate it...Exchange very elegantly handles messaging, calendaring and basic groupware with elegance.
For instance Exchange uses databases with transactional capability to provide extreme scaleability and reliability on the back-end. It has backup APIs that support amazing throughput for on-line hot backups. The database reclaims pages and defragments itself essentially in real time. Exchange supports every protocol in the book...but most customers implement it with their proprieatary MAPI protocol because it actually works a lot better than things like POP3, & IMAP.
Single instance storage allows Sally from marketing to send out her corporate spam to all internal unsuspecting users and the message will only be stored once in the database, there are semaphore links that track who has read the message or deleted it from their mailbox, disk consumption and server I/O load is dramatically reduced, especially when the message is 5 megabytes across 15,000 users!
I could easily come with a design document for a system that would essentially clone Exchange, the problem is around actually programming the system.
You would need a robust database back-end with excellent management support for things like hot backup and real-time database page reclamation, powerful & scaleable MTAs, an arm's length list of supported protocols and APIs, a user friendly cross platform client...
The ability to get all the developpers to agree on how to solve all of the above would be the biggest challenge.
Nobody ever got fired for buying Microsoft.
I know of three people who did get fired for buying Microsoft.
A friend of mine is now providing consulting to the companies in question. Two are running Twig on Linux servers, the other has their old non-ms, non-unix server back up and working (again) while they slowly transition to Linux.
Despite all the "I'll sound wise and neutral if I make out to be 'admitting' free software's flaws and giving Microsoft its due" commentary one sees here on slashdot as either an effort at karma whoring, or an effort at pro-Microsoft propoganda and astroturfing, the fact remains that there are really very few shops that cannot do without Microsoft, and many that actually benefit from running other platforms.
What is very interesting is the number of non-technical people who are coming to realize that, and while they don't necessarilly embrace free software in general, or GNU/Linux in particular, they are beginning to recognize just what a financial, technical, and time drain Microsoft and their products have become to their enterprises, and they are looking for ways out.
Even to the point where, now, people are starting to get fired for blindly purchasing Microsoft, and treating MS propoganda as a substitute for technical research and savvy.
Its a rather refreshing change, actually.
The Future of Human Evolution: Autonomy
Eww, sourceforge! Actually, there's not even a webpage- waste of time until there's something I'm willing to put up. ;)
But there are about seventy-five pages of analysis of Domino and Notes in real-world settings, some design documents and a few prototypes of critical components (probably about 200-300h of work so far). I like doing things the right way, which takes time.
Things like Domino and Exchange can be pretty effective if used well, but frankley they're not very smart. My personal research interest is managing the complexity of business and research processes, and I've found that Domino and Exchange don't really help the problems much: they don't help manage the complexities, they simply space-shift them. There's a lot of really interesting and hard problems when you start trying to solve the failings of these two systems. :)
Check out the Citadel project. This started as a BBS server, but it's gradually being built up into a groupware system. We've spent the last couple of years building up a solid messaging architecture and a fast, efficient server architecture. Right now it does IMAP, POP3, and SMTP natively (no tedious mucking about with Sendmail or Cyrus), and it's got a web interface, too. It has a single-instance, transactional data store. It has a pluggable, extensible architecture. And one of our design tenets is that it must be easy to install.
No calendaring yet, I'm afraid. We're still finishing up the server foundation. As soon as there are some decent calendar clients out there to test CAP (Calendar Access Protocol) with, we'll start building the calendar server.
I am absolutely serious about this project. This is not vaporware.
Tired of FB/Google censorship? Visit UNCENSORED!
It's made by Fujitsu, and runs on Linux, Solaris and NT. It has a really good web client, and fat (desktop resident) clients for WinX.
It does calendering, email, forums, file sharing and syncs across multiple sites. Directory services use X400/LDAP.
It's really cheap compared to Exchange and you can talk to it via IMAP, NNTP and, in version 6.0, webdav.
Check it out at www.teamware.com.
Chris.
-- I don't have a cool sig.
This has been an intensive issue in my firm (law firm, 120 users, two locations).
We avoided Exchange (there are all kinds of plusses and minuses there).
For email we use iPlanet's messaging server but are in the middle of switching over to Cyrus+postfix.
Our email clients are mostly netscape with some mozilla/outlook/outlook express/eudora. We plan wholesale conversion to Mozilla.
Many users prefer to use a web interface for email. Inital feedback on IMP has been good.
For a global addressbook we use Rolodap http://rolodap.sourceforge.net which is LDAP based. Rolodap has a web interface for searching, word processor use, data entry, etc. and also provides integrated autocompletion services to all the emailers we use. We also can extend its searches to include the local University (also LDAP) and Bigfoot if the sought address is not in our own database.
We use a central employee LDAP server (not the same one we use for Rolodap--rolodap has 35000 contacts in it) to manage email/mailboxes,web access, etc from a single interface. Though we are a touch "broken" now because of the conversion to Cyrus we will shortly be back to the goal of adding, changing, delting users, forwarding email, etc. through a single central interface. LDAP is our vehicle for that.
We like LDAP. A good open standard has all kinds of collateral benefits.
We use Steltor as our calendar -- calendaring is an important app for us and we are worried about its purchase by Oracle. LDAP controls Steltor as well.
Our documents are organized by client so the client network directories are what we use for shared files. We are about to add a means to email documents to the client file using Postfix and a little home grown app. Sharing files outside the client directories would cause us some concerns about whether we would meet our professional needs of maintaining a coherent client file.
A few observations:
First, we do not necessarily want a "one server does it all" solution. In our experience that compromises security, reliability, and redundancy. It kills flexibility, if for instance, to chnage your calendar you have to change your email server and addressbook. We much prefer to keep functions separate so long as we can have a rational presentation to the user.
Second, our users largely prefer web based interfaces if they have high performance. They find them simpler to use and more understandable. We find it simple to do "cheap" superficial integration by just giving them a master web page. Web based also provides an easy way for us to do remote access.
So for us, the key to replacing Steltor, if the Oracle acquisition doesn't work out, is a shared calendar that offers the same features, has good performance, integrates via LDAP. I do not know of an OSS app that does this.
I think both we and the original enquirer are looking for the same thing. A shared OSS calendaring server. We have email, contacts/addressbook, shared files, covered. We don't, outside of Steltor, have calendaring covered.
The original enquirer wanted calendaring that worked in Outlook. Browser based would be OK with us.
yes, I have looked at phpGroupware and its ilk and at SUSE Mail Server. All of them however provide me with too much integrated together or have deficiencies in the features/performance side of things.
If Lotus spent so much time thinking about Notes' design, why did they get it so horribly, horribly wrong?
News for Nerds. Stuff that Matters? Like hell.