Since it isn't very obvious when superficially browsing the web pages just the remark that this is based on the GNUstep development framework according to https://www.sogo.nu/support/faq/how-do-i-compile-sogo.html .
This is propbably a testament for the maturity of GNUstep.
It's a bit larger then most groupware. It's more of an entire Google Workspace replacement. Installing and running it can also be a bit more of daunting task. We ran it for 10+ years on premise before we migrated to Google Workspace.
I think most people can safely rule-out Open-Xchange unless they've got deep pockets.
Sadly commercialism has taken over there, and their sticky hands have now reached Dovecot.
The open-source version of Dovecot 3 will have the vast majority of multi-server functionality stripped from it. OX have stated on record[1] that their plan is that open-source Dovecot 3 will supported and maintained for single-server use only. If you want multi-server you will need to pay $$$$$$$$$ for the commercial version.
Last thing I know about trying to purchase commercial Dovecot, you had to harass the sales team to get an answer, and a few thousand users with 10TB+ data was "too small" for them.
The website's not great IMO (the nav bar doesn't seem to work at all on my machine) but about half way down https://www.open-xchange.com/products/ox-app-suite-cloud is a grid showing the capabilities, incl calendar, doc editing & storage and tasks, on top of email.
OX provides alternatives for Gmail+contacts, Google Calendar, Google Drive, Google Docs/Sheets/etc.
I remember it having chat and calling capabilities too, but on github their messenger is marked as archived and the mobile app seems to have been taken down, so I'm not so sure about that.
I assume you mean from on-prem OX to Google Workspace?
Maintaining OX is not super much work but it is a complex piece of software. You'll have a lot of moving parts; Open-Xchange (OSGi / Jetty application), MySQL, Postfix, Dovecot, etc. All these parts need monitoring, disaster recovery, high availabilty setup in some cases, etc.
The "absolute amount of time" these things cost isn't even super much. Especially if you are already pretty proficient in the technologies. But, personally, the hardest part of maintaining a bunch of these opensource solutions on-prem is that you are people's "go to person" for it. Meaning everything what happens to it is "your problem". For example if people perceive the solution as slow, you will need to deep dive in it. If Google Workspace is slow it's a Google problem. Maintaining these solutions (at the time we also used Zulip on-prem, Asterisk and some other stuff) becomes a recurring distraction. Patching, investigating stuff like a failed backup, breaking updates, testing updates, user complaints, etc.
So for us (about 60 people at the time) it simply wasn't worth the effort anymore anno 2018 to self-host it if you could offload it for 15 Euro a month to Google. But when we started out in 2005, it was. Also keep in mind that in my role as software developer the on-prem stuff was a "secondary objective".
Thanks for the explanation. I think a common mistake is to not make the maintenance a part of someone's job description and actually pay them for this type of work. If keeping things running feels like a secondary objective, a distraction, it's quite hard to stay on top of things and be proactive, not reactive.
Over two decades ago I started worked on open source groupware projects. I moved on about 15 years ago. Back then it was all about building an open source alternative to MS Exchange (or Novell GroupWise). Today little has changed.
In the early days it was mostly PHP and Perl front ends for a MySQL database and various daemons. The mid 00s saw a new batch of solutions emerge such as OpenXchange, Zimbra and OpenGroupWare.org. SOGo traces its roots to OpenGroupWare.org. Google Apps, which went on to become G Suite and finally Google Workspace launched around the same time.
When looking at most of the products in this space today, including M365, Google Workspace and the various open source offerings, I don’t need much innovation in terms of features. The core product is still email, calendar, todos and notes. There have been compounding incremental improvements in these features, but haven’t been any big ground breaking changes.
Today Google and Microsoft bundles their “office” apps and “groupware” apps into a single integrated offering. I don’t see an open source product that comes close to either of them. The self hosted / on prem offerings are servicing a shrinking niche.
I keep watching this space, but I have given up expecting to see any significant changes.
The real innovation I think would be a traditional groupware application that behaves more like Notion or Atlassian or a CRM. I feel like Google started in this direction, but didn't go far enough.
Imagine if every component is tightly integrated and highly cross-linkable. Right click somebody's name in a document and get their contact card from which you can email them and find all of their public documents. Find all backlinks to a particular document. Rich global search across types of assets.
Our current technology seems so close, but the integrations are inconsistent and the UX is poor.
It feels more like social media/sharing app than something suited to a business environment. Email, scheduling, chat and document sharing are must haves in a modern business. Hubzilla lacks all but one of these.
I have bad memories of SOGo, from one company that insisted on self-hosting absolutely everything (including email, git, issues, Nextcloud, etc) on their own hardware. It wasn't even like a gov't institution, but a normal commercial webdev shop with strange whims.
All of it was poorly maintained: several versions behind, suffered from unplanned downtime, performance issues... the choices were also questionable. Rather than gitlab or gitea, we've been using something like gitosis - except it wasn't gitosis or gitolite (these can work for very small teams, which we weren't), it was something way worse (can't even remember the name). Of course I didn't stick around there for too long.
I guess this is a pretty big risk factor for self-hosted projects, that nobody seems to be talking about. I'm fairly certain SOGo is a great piece of software, but its name has been tarnished for me (and many of my ex-coworkers) by people that have nothing to do with the project.
Good point. I work in security and although I care about privacy and self-sufficiency, sadly enough most (usually smaller) companies that have a dogmatic self-host policy tend to have a complete mess as infrastructure. Old software, servers that haven't been updated in ages, known vulnerabilities, semi-configured services, non-existing documentation or contingency policies, the list goes on.
Companies should focus on their core business, and not underestimate the efforts involved in straying off the beaten path. You might be a master in gluing technological solutions together, but you might end up under a bus and woe upon those that have to figure out your handiwork. Even I myself have stopped self-hosting stuff like email for personal use because at some point, life happens, you have kids and jobs and deadlines and other responsibilities, and something might break just at that moment you really don't have time for it, and you just don't want to have to worry about stuff like if emails are getting through.
Luckily there are some more ethical alternatives to the big industry leaders, for whatever your interpretation of ethical might be. It's always possible to support companies that are contributing something back to open-source software by choosing their paid services.
Also maybe I'm wrong but business to business relationships tend to care more about privacy since 1. You're paying buckets 2. The loss or leak of info can have more serious fiscal harm.... I try to offload everything at my job
If you're going to self-host, you're going to need to dedicate people to keeping your infrastructure up to date, whether that's SOGo, Gitlab, or Microsoft AD.
Of course you'll still need to appoints someone to manage your hosted environments if you decide to go with external services. I'm not so sure if a company gimped by a badly managed SOGo environment would be able to keep their hosted Exchange working smoothly for long.
> Of course you'll still need to appoints someone to manage your hosted environments if you decide to go with external services.
Fair point! It's also fair to say that managing hosted Office365/Exchange is a very different skillset/background than a traditional Linux sysadmin role, and that they were making use of the resources they've had at hand (considering how finding a reliable IT partner can be challenging by itself; I've seen many shops struggle with that).
Still, my original post was how your branding/perception as a self-hosted project can be ruined by someone else's mistakes in setting up said self-hosting. I've been on the other side of that fence, trying to convince the management of a different company, to host a particular service - one of the execs had a horror story, and our team had a hard time working out a compromise. For open source projects, which rely on contributions, this external perception by decision-makers (or decision-makers-in-the-making) can have real impact.
Looks elegant and much nicer than eGroupware, but eGroupware offers a lot more functions, though. Visited the demo site of eGroupware, I can not shake off the feeling the UX and its GUI essentially remains the same in the last 15 years. eGroupware is functional but its GUI is unfortunately awkward.
I've been running this for a few years now, but for personal email I've found I rarely read it on a computer. Once in a blue moon I have to open SoGo to click a link or grab a code from an email if my phone isn't handy, but that's it.
It works fine, it's quite simple, the "material design" look is perhaps a little dated now but nothing like Roundcube.
I am using Roundcube with a modern theme called Elastic. It works like a desktop app. I have gmail through work too but I don't like using it. I think Roundcube is much better. But my colleagues prefer Afterlogic which has a much simpler interface.
I just upgraded my whole mailserver recently and the Elastic theme is growing on me. I had the usual gripes about everything moving around for no reason but the theme seems a lot nicer for modern displays and (my) aging eyeballs.
Oh interesting. I was looking for something like this about eight years ago when we were breaking off from our previous owners into a 30 person company
