Open-Xchange for the other 75% of us

[Martin Heiland]

Hi,

we're evaluating several options to get onto other platforms. One of the most important platforms in the market is Microsoft Windows Server. Since Open-Xchange started its work to provide a first class open-source groupware solution we focused to the Linux ecosystem. This was caused by historical facts since Open-Xchange was once part of SuSE/Novell before we started as an independent software vendor.

There are no exact numbers about the market share of Windows Server, but we believe that more than three quarter of all small and medium size organizations run this platform at their datacenters primarily. This also means administrators are skilled in maintaining a Windows Server environment. We believe that a organization should have the freedom of choice which platform they want to use, we already express this by providing packaged software for a dozen of Linux based distributions. On the other hand, it is always a venture to introduce a new Linux based solution. Administrators might need to switch to a new platform, processes might be changed and all the typical pain when introducing a whole new platform.

In short, we're evaluating Open-Xchange support for Microsoft Windows Server (2003, 2008, 2010...). This would include installers and documentation for this new platform. Since this project is really at the beginning right now, we'd like to get your thoughts about it and what would be important to you as an administrator. We already made some prototypes which worked pretty well. The command line tools have been ported to windows scripts and configuration files needed to be updated. One of the major obstacles would be support for IIS, so we ask for your thoughts if Apache on Windows would be okay for you as well or if IIS is a must-have for this platform. We also understand that dealing with configuration files is not the rule on Windows Server so an integration to MMC would be required. You may also vote what you're generally thinking about the approach.

/discuss

Note: This is *not* an announcement, just a proposal and a discussion.

[evildead]

Hallo,

ich bin zwar auch durchaus der englischen Sprache mächtig, jedoch da ihr deutsch sprecht möchte ich in meiner Muttersprache antworten.

Generell ist es eine gute Idee Windows Plattformen zu unterstützen, jedoch solltet ihr einige Dinge nicht vergessen.
Linuxbenutzer bzw. Admins verzeihen Fehler. In unserem tiefsten Herzen, wenn wir ehrlich sind, sind wir alle kleine Bastelkinder, die Spass daran haben komplexe System anzupassen. Ich denke auch das es sehr viele Windowsadministratoren gibt die in diese Kategorie eingeordnet werden können, allerdings wird meiner Meinung ein Softwaresystem von sehr vielen Windowsadministratoren nur als "working out of the box" akzeptiert werden.

Es gibt 3 kombinationen von Administratoren.
Linux (andere Unixartige Derivate einsetzbar) only.
Linux (andere Unixartige Derivate einsetzbar) + Windows
Windows only.
Die ersten beiden Kategorien könnt ihr jetzt schon erreichen. Die dritte erwartet ISS oder MicrosoftSQL Server oder Exchange support und eine voll konfigurierbare Benutzer/Gruppen/+x+/Rechteverwaltung. Sollte man denke ich in Active Direktory abbilden können. Oder alles zusammen.

Da ihr Java verwendet, denke ich euer kleinstes Problem wird sein die Software auf Windows lauffähig zu machen. Aber die Unterstützung der ganzen Windowswelt Standartformate wie sie von vielen vielen Windowsadmins eingesetzt wird wird sehr schwer und komplex werden.

Ich habe die Vermutung, das die eigentliche Unterstützung von Windows so viele Ressourcen auffressen wird, das die weiterentwicklung der eigentlichen Software hinten ansteht.
Es gibt einige Features, die dem System aktuell Fehlen um es so viel besser zu machen das ein Umstieg wirklich lohnenswert wäre.

Wenn ihr meine bescheidene Meinung wissen wollt und ihr Kunden gewinnen wollt im Bereich kleiner und Mittelständiger unternehmen, müsst Ihr Features anbieten die sonst kein System kann.
Es ist zwar schön das man aktuell mehrere Emailaccounts einrichten kann, aber man kann aktuell nicht mal vernünftig Termine ausdrucken (ausser in der Listenansicht). Für viele Unternehmen, die zb. Aussendienstler beschäftigen, ist sowas ein absolutes nogo.

Ich bin gerade dabei ein mittelständiges Unternehmen auf die aktuelle Version zu migirieren und letzte Woche habe ich einen kleinen Betrieb auf OXAE umgestellt. Dort treten Probleme auf, die niemals den Weg hier ins Forum finden werden, aber meiner Meinung nach den Einsatz stellenweise behindern.

Technisch betrachtet ist meiner Meinung nach das grösste Manko, das es keine einheitliche Administrationsoberfläche gibt. Damit ködert man auch keinen Windowsadmin. Konsolenbefehle gerade in Windows bereiten mir persönlich alleine beim dran denken Kopfschmerzen
Ihr dokumentiert eure APIs, könnt quasi innerhalb von 2 Wochen mit ein wenig Menpower eine Oberfläche zaubern und alles zb. in Mysql verwalten und in Univention wird LDAP benutzt. Unverständlich meiner Meinung nach.
Weider ein System mehr was angepasst werden muss. Nun soll Windows auch nochmal dazu kommen, womöglich alles in active Directory, das hört sich nach viel Arbeit an.

Grundsätzlich wüsste ich auf Anhieb 3 Features die kein anderes System unterstützt, die aber ein riesigen Benefit für viele Benutzer gerade im Unternehmensumfeld darstellen würden.

Wenn ihr möchtet stelle ich mal eine Liste zusammen über Anfragen von Benutzern was Sie am System nicht verstehen oder wo Probleme auftreten, die euch vielleicht helfen Schwachpunkte des aktuellen Systems zu beheben. Denn das Problem ist, das hier im Forum sowieso nur geeks und admins fragen stellen, das Problem sitzt aber meisstens bei den eigentlichen Anwendern und die lesen vor der Benutzung meisstens nicht das komplette Handbuch, sondern wollen mit dem System arbeiten.
Scheibt mir einfach ne PM oder eine Email. Die Infos gibts natürlich kostenlos

BTW. Trotz aller Kritik finde ich das System durchaus gut und benutzbar, sonst würde ich es nicht einsetzen. Aber ich bin auch kein normaler Anwender und kenne meine Vokabeln.

gruss

[Martin Heiland]

Hi,

i agree that Windows support needs more than just porting some scripts. This could be done by a partner and would not influence the core work so much. If i understand you correctly, IIS, AD and MSSQL is a must have. Using a MS Exchange Server als E-Mail backend only for OX works on the technical side - the question is if its reasonable on the business side (TCO etc.)

[pezi]

Andere Platformen sind ein gute Idee - insbesondere wenn man bedenkt das OX eine reine Java Applikation ist. Die Unterstützung der einzelnen Linux Distributionen (debian, etc.) ist sehr löblich. Aber für die Community Version wäre eine distributions unabhängige Version nicht uninteressant. Die Installation mag dann zwar aufwändiger sei - mehr Einzelschritte. Aber dann könnte das System unter anderen Linux-Distributionen, Unix Varianten installiert werden.

Was ich nicht sagen kann, ob dies möglich ist. Man müsste die diversen Bash-Scripte z.B nach Perl migrieren.

Weiters wäre es wünschenswert, dass OX auch mit Java 1.6 kompatibel ist - dass Java 1.5 nicht zwingend notwendig ist. Zumindest unter Centos habe ich Probleme wenn ich manche OX-Scripts unter Java 1.6 aufrufe.

[Martin Heiland]

Open-Xchange Server supports Java 1.6 (SUN and IBM). If you encounter any issues, please open bug reports. Since the code is public available, it's possible for third parties to create packages for other distributions.

[evildead]

Hi,

i agree that Windows support needs more than just porting some scripts. This could be done by a partner and would not influence the core work so much. If i understand you correctly, IIS, AD and MSSQL is a must have. Using a MS Exchange Server als E-Mail backend only for OX works on the technical side - the question is if its reasonable on the business side (TCO etc.)

So siehts aus. Ich bin ja nicht der Windows Fachmann, aber was für Mailserver gibt es denn noch für Windows? bzw. welche die auch vom Mainstream benutzt werden.
Sobald ja jmd irgendwo nen linux/unix mailserver hat, fällt er ja schon wieder in eine andere kategorie.
Zu MSSQL, das ist kein unbedingtes must have, mysql gibts ja auch für Windows, allerdings würde es bestimmt den hartgesottenen MSSQL Fan erfreuen wenn es unterstützt wird. Gleiches gilt im Prinzip auch für den ISS, es gibt tatsächlich einige Leute die benutzen den gerne und ausschliesslich

[mybulk]

I dont like the idea, and spent last 30 minutes thinking about some arguments why i don't wanna see a OpenXchange on windows. Several arguments came to my mind:

- Missing features of the Linux version requested by the Users for ages (like Admin Interface, Backup , and many more). Fix that first, then expand to other OS

- Minor Bugs in Openexchange GUI that set back OpenXchange clearly behind other solutions (e.g. the mentioned printing of dates)

- The weak documentation (FAQ, WIKI, Forum help, setup Guides) of OpenXchange compared to other solutions like Zimbra. Sorry, don't wanna upset anyone, just saying that OpenXchange clearly falls behind when it comes to uptodate documentation and forum activity. This makes me think that OpenXchange is not the top solution for most people atm, if it where more people would use it and forum activity would be way higher.
So exporting something barely breathing to a new OS does not make much sense to me.


But ........ all that above is just my personal opinion and falls behind the killer argument why OpenXchange should not be exported to Windows, because any effort doing so will be wasted time.

Microsoft Small Business Server

Windows users ALLREADY have it all they want,
- Mail server, Database, Active directory Support,
- Web frontend with web server, integrated OS with File and Printserver,
- Terminal services, easy installation and HUGHE community of free and
payed support,
- Free Activesync support, well tested software from the market leader
- fully support migrating other solutions to exchange (via 3th party software), regular Updates and Bugfixes from a huge company that will surely exist the next 10 years
and moreover a competitive pricing for small companies

So there is no need to invent the wheel a second time.


PS: No, i don't use SBS with my customers not do i like it. But it is the de facto standard for business messaging

[Martin Heiland]

Thanks for the feedback, let me comment some of the statements.

- Missing features of the Linux version requested by the Users for ages (like Admin Interface, Backup , and many more). Fix that first, then expand to other OS

An administration and Backup interface is availabe with the Appliance Edition, it's part of the administration backend from Univention. They are focusing on operating system management. We develop a groupware, not an operating system or backup system. There is a lot of sophisticated backup software out there, and i think it would be a waste of time trying to copy all their features. We're focusing on groupware. Administration is quite the same topic, we build software for integration. We offer various interfaces to let others administrate the groupware but we don't force a specific interface. There are free and open-source administration tools available for the pure groupware admin part (http://oxgui.wordpress.com/) Other solutions like the one from Yahoo have a nice administration backend, but did you ever tried to integrate it to an existing mail system?

- Minor Bugs in Openexchange GUI that set back OpenXchange clearly behind other solutions (e.g. the mentioned printing of dates)

Yes, totally correct and we're on it - we fix a large number of small and large bugs every release and add better usability. Calendar printing is scheduled for 6.14 but not finally there yet. The integration to a Windows Server system might not be done by the same guys that code the user interface or the groupware server, that might be a job for an external partner.

- The weak documentation (FAQ, WIKI, Forum help, setup Guides) of OpenXchange compared to other solutions like Zimbra. Sorry, don't wanna upset anyone, just saying that OpenXchange clearly falls behind when it comes to uptodate documentation and forum activity. This makes me think that OpenXchange is not the top solution for most people atm, if it where more people would use it and forum activity would be way higher.
So exporting something barely breathing to a new OS does not make much sense to me.

It's okay to drop some critic where it is appropriate. I don't think that the number of forum accounts is soley an indicator for a good or a bad solution. Zimbra might be more popular and people are more active at the community, but is this really an indicator? I mean Yahoo bought them two years ago and they did not manage to integrate the software to their offerings. Now rumors are around that they want to sell them again (for a fraction of the original price of $350m). No gloating but that does not put them into a good light on the technical side.
Regarding the other documentation, what is missing? Just open a thread if you're missing something. If there is demand for a specific kind of documentation we would be happy to support, but we have to know what you want

[mybulk]

Oh I don't think OpenXchange is BAD straight away (or i would not use it). But the pure fact that you can handle all the questions alone on forum( must i mention that u do it well ,beside those "ugly monotone" responses when it comes to Admin interface and backup) , indicates that OpenXchange is atm just for a small group of users compared to other solutions.

I expect u to know best why this is so, so no need to discuss about. Maybe you launch a pool next to collect ideas what can be done to change that and spread OpenXchange into the world, I just guess you would not like to hear it :-)


Anyway, imho OX is just not a product to be ported on another Platform with heavy competitor Windows SBS, OX is atm "just" a GUI not a whole Product as it needs lots other software to work (not counting the OS here)
So going "Windows" seems a very bad idea.

Instead go and become the f... (am i allowed to say fucking here ?) best enterprise mail software for Linux (you already have the best GUI, other features are just 3th class, or none existing at all)

And then, only then, and just maybe go windows.

[Martin Heiland]

We'd like to hear any kind of suggestions as long as they are not against the law
No need to open an extra poll for this, just open a thread yourself and post your proposals.

Of course it's always nice to have a big and most active community (some people are) but this is a result, not a precondition of a great product. However, many of our end users are customers of hosting providers around the world. Compared to "direct" users, those are not as active since they "only" use the software but do not administrate it.

[joan-open]

OX is a good product but it is too fragmented and difficult to configure

Hi Martin,

I think OX is a very good groupware, and I like most OX than Zimbra because I like having a standar back-end (postfix, cyrus, openldap) and not a propietary back-end with my mails related in a close-dark database. And it is much better that kolab, phpgroupware, opengroupware, egroupware, ....

BUT, I think that OX is too fragmented and it is very dificult to integrate this fragments:
- OX-LDAPsync in a fragment not included in the package: TODAY is essential integrate the new aplications in the existing LDAP or AD to avoid create users for each new application
- Funanbol (or similar) is not included
- Mozilla SunBird/Thunderbird conector is not included
- OX server admin by Peter is not included
and the documentation to integrate this fragments is not complete.
and when the versions changes it is possible that this fragments don't work any more.

I think that is good having a complete suite like Kolab with openldap+postfix+cyrus+open-xchange+ldapsync, to avoid incompatibility versions.

I spend 2 hours to have a Zimbra comunity ready to work and 1 week to have one server with openldap+postfix+cyrus+open-xchange+ldapsync almost ready (now I spend only 1 day).

Is is only a beginner opinion...

[Martin Heiland]

Hi,

i agree that OX is fragmented. This is by design since we use a plugin architecture and most parts of the groupware are already plugins that can be changed dynamically if required. It's also true that such an architecture might be confusing to starters. It might be difficult to integrate all those plugins, but it is *possible* at all - which separates the software from most others that are build monolithically.

You pointed out four components that are not maintained nor made by Open-Xchange. Since we cannot just take Version X of Funambol and package it with OX i don't see a real solution other than better technical documentation about such integrations. This documentation shall be available at the wiki. Right now it is not perfect, but anyone is free to make the documentation better. On the other hand there are partners that developed solutions like OXAE which come exactly with a ready-to-run integration.

Greetings

[joan-open]

Hi Martin,

About your original reason for open this Thread....

I think that Open-Xchange in Windows server is not a good idea. Open-Xchange works fine in Linux, like a front-end of a very powerful configuration (Cyrus, Postfix, Openldap), and with this configuration is better than one Exchange server. But only changing a Linux for a Windows server you lost performance of your global system, and if you change too the MTA for a windows server MTA and the cyrus IMAP (or other like Doevecot, Courier) for a Windows server IMAP , the OpenLDAP for a ActiveDirectory, you lost more performance amb close your system, then I prefer Exchange Server than this new hibrid-Open-Xchange.

I think that:

- Spend your effort working in the missing features of the Linux version requested by the Users for ages (Admin Interface, Backup, best LDAP integration, and many more).

- Work in minor bugs

- Work in documentation: FAQ, WIKI, Forum help, COMPLETE setup Guides, and you can arrive to more users

and then, if you have time to work in expand to another operating system choose plattaform to increase performance: FreeBSD, Solaris, not a plattaform to decrease performance.

Exchange server is a good application for a windows user: Mail server, Database, Active directory, pretty Web frontend, integrated with Fileserver, Free Activesync support, ALL installed with a couple of RETURN's

Joan