Mozilla and Google — Exchange Killers At Last? 336
phase_9 writes "The latest version of Mozilla Thunderbird may still only be in beta but already the user community have started creating an extensive set of viable Exchange killers. One such example is the latest mashup between Thunderbird and Google Calendars, providing bi-directional syncing of calendar information from both the client and internet. How long will it be before open-source software can provide a complete, accessible office suite for a fraction of the cost that Microsoft current imposes?"
$PRODUCT killers is a continuum (Score:1, Interesting)
Replaces important functionality of the product
The other end, I'll call "type 2:"
Replaces the product
Type 2 means among other things your users won't notice any functional differences and the new product can read and write the old product's files perfectly and/or there is a perfect two-way file-translator available.
Type 2 is rare unless the product is designed around an open specification. For example, some implementations of "gzip" or "cat" are type-2 replacements of each other.
nope (Score:5, Interesting)
Please don't flame me ... (Score:5, Interesting)
Re:Evolution??? (Score:2, Interesting)
Google is open source? (Score:5, Interesting)
Re:It's not going to happen (Score:2, Interesting)
This is all very clever and wonderful (Score:3, Interesting)
I'm not a huge fan of MS, but it's nice that external people can send you stuff (as they use Outlook) and it'll appear in your company outlook calendar.
Sooo if you want to defeat Outlook you've got to produce something that replicates outlook's functionality. I don't care what the other company is using, I just care it works with my outlook (or vica-versa).
Basically my point is we live in an Outlook eco-system. If you want to displace it, then you can't just ignore it and do your own thing (e.g. Mozilla+Google).
Notes (Score:4, Interesting)
And of course, Lotus Notes is what software would be like if it was written by Satan.
There ARE options, and it's not openXchange (Score:2, Interesting)
Alternative open-source solution (Score:4, Interesting)
Another opensource solution that has piqued my interest is zimbra [zimbra.com], which includes collaborative e-mail, scheduling and many other groupware functions. All the functions work through a web interface as well, but they're now developing zdesktop [zimbra.com] to allow on- and off-line sync/viewing of e-mail, scheduling as so on. It's in alpha, however. There are also programs to use on your mobile [zimbra.com] devices.
I haven't used this system myself, but I'd be interested in any thoughts from sys admins that have successfully (or unsuccessfully) implemented this.
Re:It's not going to happen (Score:3, Interesting)
There is no open source exchange killer in the offering here. As far as Outlook killers are concerned, Mozilla has been an Outlook killer for a very long time. Even with something as lame as courier Mozilla can easily work over 12G+ IMAP mail folders. Outlook (prior to 2003) caused massive corruption crashes and loss on anything above 2G (after the local cache exceeded 1G).
As far as the usual argument about "want it local", nope I do not. Provided that:
1. Google can offer compliance and logging features for a relatively strict regulatory framework including blanket logging of all emails per domain and retention as per user specified policy. I am aware that exchange does not do that, but there is enough third party software for it (as well as for exim, sendmail, etc). Before that google + mozilla are worthless for corporate use.
2. Gooogle can offer client side certificate based authentication and ssl-enabled protocols for everything. I do not mind google maintaining the CA for that as long as they can.
3. Backup requirements which correspond to 1.
And so on and so fourth. If all of these are complied to, is it local or remote is largely irrelevant. Google has enough datacenters to ensure that the latency is low. In fact, it better be remote. One less item to worry during disaster recovery.
Re:no bloody chance (Score:4, Interesting)
In the distant future there may be a commercial groupware solution based on open source, but it will almost certainly cost as much or more than Exchange.
Re:Notes (Score:3, Interesting)
On the other hand, at least the older versions of Notes did a number of things very well (I can't speak to newer ones), including security. However this required more skilled and educated administration. The MS pitch throughout the early to mid 90s on the server end was that you didn't need the kind of expertise you needed to run Notes to run Exchange, or to run Novell to run NT server. The rest, as they say, is history.
Re:Google is open source? (Score:3, Interesting)
Google is paranoid about internal security and leaks because what they really have is their own "special sauce", based on open source and commodity hardware, that they can't sell. Google is going to be "hosted only" for the forseeable future, and I for one would never consider an ASP or outside vendor for my groupware server. It's actually ILLEGAL in many US organizations due to Sarbanes-Oxley.
Re:your business E-mail is an open book anyway (Score:3, Interesting)
But let me ask you... Are you a Google competitor? If so, you don't really have reasons to host on their servers
Re:Careful now. Think this over carefully. (Score:5, Interesting)
Anything big is slow to move and is an easy target. Big things usually subtract the human element due to bureaucracy. I would say that big things are generally corrupt, and that would indicate Google too.
---Yes they're a corporation. Yes they're in it for the money. But they manage to do it by embracing technology and providing it to a wider base of users for FREE. They can data mine every second of my life if thats all they ask in return.
I dont know where you live, or what you do for a living, but I'm a 25 year old. At our local mall, there's a door with a company plate on it. It idnt spiffy looking, nor are there windows or anything else. They are a marketing firm. They are the ones that Coca-Cola, Pepsi-Cola and many other companies go to for aggregate and specialized data.
I have participated in a few of these studies (I cannot specify product names.. nda for company name I tested only). I usually am given 10$ worth of goods to test and then do a write up and phone interview for said products.
My average payout for these interviews is ~30$, along with free products, and getting a say on a new product. I KNOW that I'm in a database somewhere and I'm properly compensated for it. When companies come along and want "free information" for "free product", it tells me that what they offer isnt worth it, and my data is worthless.
Word to Google: Tell me how much my information is worth, and Ill pay for information if your product is worth what I deem it to be. Better yet, if they are willing to pay me, I'll list product names and prices and my personal writeups. Not all companies will like what I write.
WebCalendar + Thunderbird/Lightning (Score:3, Interesting)
I already run WebCalendar on my local server and it is an excellent program. But I would like to be able to tie it into lightning for calendar sharing. It doesn't work. First, the stable version of WebCalendar doesn't support publishing. The CVS version supposedly does, but while you can import a calendar into lighting, any changes you make there doesn't get published to WebCalendar. Lightning flashes a little bar, gives no errors but reloading the calendar or logging into webcalendar will show that the new changes were never uploaded.
I've never understood what is so difficult about combining email with a shared calendar. That solution alone would prevent the need to setup new exchange configurations. Most small and medium business only need integrated email and calendaring this leads them to Outlook, then they want to share calendars. That leads them to exchange.
As a developer I can't think of any great challenge involved in this (beyond not having time to write a solution myself). I have trouble believing that with (according to some EU state of FOSS paper) 2,000,000 OSS developers nobody has managed to come up with a solution for this basic fundamental and common need.
I won't even read the comments (Score:2, Interesting)
DKone
Re:Make a clone instead (Score:1, Interesting)
Yes. The only problem is it has been a "holy grail" only by mounth. You can see yourself by how many people even here doesn't distinguish Outlook from Exchange even with a handbook. All the pretended "exchange-killers" (scalix, openxchange...), are "closed sourced in disguise": they usually are mamoth-like apps with either no proper foundations nor/or proper open community procedures.
"yes, I know about Kolab and use Kontact myself, but how often do you actually see a Kolab server at a company"
Kolab is a very interesting example, and it's not the typical one. It has some good foundations but it has horrid problems too that prevent it to be implemented in real-world scenarios (that's why you haven't seen a single instance of it in production). While the Kolab starting point was a very good one and fullfiled its target (a fast hack for a single client) it is terribly hacker-unfriendly: you can't install but on a server on it's own (due to the horrid choose of a self-servant package database); it uses a heavily-nested template system that not only doesn't leverage but works against any knowledgeable sysadmin; it has quite a lot "hard-wired" constraints that impedes you to integrate it on your current structure (the most obvious example is it's LDAP schema), and I could go on an on. It's important to notice that the real problem with Kolab is not that it's feature-lacking (it's quite good on that respect) but that it is very early-adopter/hacker unfriendly (so you can't see it going anywhere).
"Exchange is very hard to clone and very hard to replace in an environment that uses it."
I don't see it being such a very important problem or at least a very critical one. The "exchange-killer" doesn't need to be an Exchange-clone for the very second reason: it's almost impossible for any non-microsoft service to cleanly cooperate with its microsoft-equivalent. That's valid even for the most perfect and matured "microsoft-clone" overthere which is Samba, so we can forget about it as a lost battle and move on. The "exchange-killer" should be functionaly-wise more or less equivalent to Exchange for a very valid reason: Exchange *does* offer a very valid and usefull feature-list, but it doesn't need to be a bug-by-bug Exchange-clone -it won't work. Even if exchange-dependant companies (that's valid for any other microsoft-based service) are "kipnapped" by that Microsoft-dependency it's their problem at worst: new companies are founded every day that could use new solutions provided they are useful and budget-savvy and gaining a competitive advantage out of it.
"In conclusion, then, a warning to anyone who does not now have Exchange and is thinking of it: don't. You'll be using it forever"
A very valid point; and you can say exactly the same about almost any other Microsoft software: that's the way they make bussiness.
"Look at other open source and closed source products first"
I've done. The problem is that there's nothing currently that can fulfill the Exchange gap nowadays. What's worse, it's very difficult for a truly open product to do it. Even while calendar, meetings, etc. are the most obvious advantages of the Exchange solution there are other benefits out of their office monopoly: with Exchange you can have workflow, templates, etc... only because they know exactly what the client-side is -Ms Office, that's it. Microsoft knows that and it's already exploding it even further with Sharepoint. You just can't have this kind of functionality/integration when you can have client-side Evolution -or Kmail, or Pine, or Squirrelmail; OpenOffice, -or KOffice, or Abiword, or Emacs or... you need a "vertical stack" where everything from client to server is known and developed to work together.
You currently have some "bricks"; some are good, some not as good, but they are probably good startin
Re:It's not going to happen (Score:5, Interesting)
False. You assume that Google's IT department and a corporate IT department have the same goals.
They don't.
Google's business model depends on providing access to their services to people outside of their network, while making sure those people outside of their network only get access to what they are supposed to access.
Corporate network admins, on the other hand, typically give first priority to doing something that Google fundamentally can't without interfering with their business model - prevent outsiders from obtaing ANY access whatsoever to the internal network. This is pretty easy with a proxying firewall. Optionally, after that begin providing access to authorized external users in a controlled and secure manner, such as an IPSec VPN using RSA SecurID tokens for authentication. Google simply can't force all users of their services to go get a SecurID token and VPN in, especially since such VPN systems usually force the client machine into connecting ONLY to the network it is being connected to via VPN.
Their next priority is usually controlling what internal users get access to what, but this is an easier job than "you vs. rest of world". You can usually ensure by methods already in place (Interviews of potential employees, locked doors with badge access and/or combo locks, etc.) that the likelihood of internal users being a skilled cracker is low, although IT departments should still assume that they are. Google can't place men with guns and network monitoring devices (IDS and other sniffers) at every potential user's home to say, "You may be doing something malicious. Stop now."