I read it, but am unclear on if OnlyOffice Document Server CE is the same thing as Nextcloud Hubās version.
it is not the same, as far as I know and understand in the livestream. It ist called as a small, simple, easy to install part from the server, not a full server.
I seem to recall that when Nextcloud forked from Owncloud, one of the reasons given was to get away from the opensource crippleware model that Owncloud embraced - where the community edition of software offers reduced functionality compared to the āenterpriseā versions. The primary reason I made the switch was on that understanding.
This business of integrating OnlyOffice (which also embraces that crippleware model) as a core feature, seems to have the effect of introducing the same nonsense via the back door.
I have researched both thoroughly.
There are only 2 players. Onlyoffice and Collabora
Collabora is Libre Office compliant. Libre is mostly compliant with MS Office but starts to falter when any complexities are introduced from an MS created doc.
Onlyoffice on the other hand dovetails seemingly perfectly with the Microsoft document formatting even as complexities in the document increase. This MS compatibility is extremely important when a business / workplace is using it. Clients will not want it when MS formatting starts to fall apart.
Otherwise who is going to recreate this functionality from scratch?
Onlyoffice at least has a 20 user freebie entry use before your breaking any rules.
All that considered, I agree with the Nextcloud decision makers. Onlyoffice is the only way to go here.
Jay
CompuMatter
As I recall, though, it was a very similar situation at the time of forking from Owncloud (when it was widely agreed to be a problem.) Owncloud was attempting to restrict access to the online office integration to the enterprise edition.
With the integration of Onlyoffice in Nextcloud restricting the community edition to 20 concurrent documents (which any but the smallest of deployments will hit in no time), weāre into murky territory.
I do think that not locking up functionality in the paid version is a pretty fundamental consideration, and at one point it was pretty clear that this was not the intended direction of Nextcloud.
You can unlock the 20 concurrent connections with LibreOffice Online (not collabora) for free by compiling it yourself.
Itās a pain in the *** but i made it once. And a pain to update it too.
You can do the same with OnlyOffice but the documentation isnāt updated so for the moment you canāt compile it by following the documentationā¦
Again if you donāt want or canāt afford paying the license, you can unlock the limitation by compiling it.
OFF TOPIC : it could by great if someone can make a docker image of unlimited LibreOffice Online (LooL) or Onlyoffice Document Server.
Yes, I understand the possible solutions (and their added burden).
The point for me is that there are several pieces of software out there that follow the opensource crippleware strategy, and then there are other projects that are truly opensource. Many of us avoid the former and prefer the latter. Originally, this concern was stated as part of the reason for the fork.
I just think that this default of including a crippleware solution for part of the functionality isā¦ questionable.
Please change your angle of view. How to offer users of Nextcloud an office integration without limitations?
What are you talking about? Are you sure that there is such a limitation? The new integration seems to be a rewrite / re implementation of the onlyoffice community server in PHP. The code is here: GitHub - nextcloud/documentserver_community: Document server for onlyoffice
Please tell me that you already tested it with 20 users and found out that there is such a limitation or are you just guessing?
Well, thatās a different question. There are many software solutions out there that donāt have a good open-licensed alternative. But, that doesnāt mean that we should bundle restrictively-licensed software with our opensource projects. The answer may well be that there is no good option for office integration at present. In that case, we either accept that we donāt have that functionality yet, or we offer a separate opt-in integration with the restrictive-licensed options, which is essentially what weāve had until now.
With this level of integration, and the software restrictions removed by buying the enterprise versions, I think that weāre getting into problematic territory. I realise that some will see it just as a convenience, but I think itās a bit more complicated than that.
Ok, I understand that the limitations of the OnlyOffice integration need to be outlined.
Then everyone can evaluate whether the use of OnlyOffice is worth considering.
Does that hit the nail on the head?
I think making the limitations clear is certainly important.
For me, that doesnāt quite address the fundamental problem, concerning the values of the project, which arises from the fact that a default part of Nextcloud Hub will be feature-restricted compared to the the paid versions.
Understood. Without office integration Nextcloud feels feature incomplete (my personal opinion after seeing all the apps the last years). So how to close this gap? Adding an optional limited solution? Works for me (private, 10 users, family).
What is your use case?
Weāve been using in a mid-sized organisation for years.
Integrated office functionality would be a bonus, but we donāt require enterprise support, and donāt wish to go down that road.
On the other points, I donāt want to go round in circles. I think that introducing paid feature restriction is a fundamental problem in itself (and a change from the original position of the project.) How to provide an alternative open-licensed solution for a new feature is a different question, but true open-licenced projects donāt typically solve that problem by integrating restrictive software. To my mind, that rather undermines the whole point.
I personally donāt understand the hype of these online-applications. Itās nice for certain documents to work on them at the same time, however there is always a certain delay and working with larger amount of data, itās not very efficient.
There are a lot of great ideas with these new features, but I see many current problems are completely neglected. Big announcement of client-side encryption. That feature is not documented, no idea if it works decently on any of the clients (android perhaps?). The desktop sync client, conflict files popping up from files that havenāt been changed for years, performance on many small files is not so good (one could perhaps bundle man small files), now the virtual drive, ā¦ and for webdav as virtual drive or client-side sync there are already solutions (cryptomator for client-side encryption, for webdav-mapping itās not free but there are several), there are perhaps ways to partner with someone.
nextcloud/server on github has 2160 open issues and 2021 with no:assignee
nextcloud/desktop on github has 768 open issues and 699 with no:assignee
@tflidd do you mean this?
btw: there are also good news. my two āhandsfree self updating(*)ā docker based installations just survived the update from 17.0.2 to 18. yeah!
(*) i know. for real admins thatās a total no go. but iām brave.
@Reiner_Nippes
The raw numbers may be sobering, but donāt trust statistics you havenāt falsified yourselfā¦
There may be duplicates, issues that are already resolved but not closed and so onā¦maybe it lacks just someone who does some issue tracker maintenance, maybe itās really that bad as the numbers say.
@Semjel
TL;DR: There more than just āblackā and āwhiteā regarding āopen-sourceā.
Your point may be valid from an ethical point of view and iām definetely not that happy with some decisions as wellā¦but i think itās about one thing that almost every open source project that grew past ājust for fun/learning/cause i can do itā to āneed to make a living out of itā: If even āmid-sized enterprise organisationā expect to get everything āfor freeā, how would Nextcloud GmbH who provides a core team of developers for the project keep things going?
Iām not sure if a project of this size would exist without one (or more) companys backing it in regard of ressources and devs.
Iām contributing to some open source projects myself - just for the fun of it - and i like to see open-source as an idealistic approach too, but the reasons i can do this is that i spend only a small amount of me spare time for it and i have a paid job to earn a living.
OnlyOffice is not crippleware per se. It is fully open source to the core and can be used without limitations when recompiling. What it is lacking is end user documentation and support for ARM servers. And I agree that the company behind isnāt very supportive of non enterprise customers.
But the tech itself is great and there is a lot of possibility for it to become a more community driven project. Might at some point a Owncloud/OpenOffice like fork be necessary? Maybe yes, but so far I think not, as the company behind OnlyOffice has not actively sabotaged any open source efforts.
Oh and with the new 0.1.2 update the new Community Document server now seems to work fine with the 18.0.0-Apache Docker image.
There are a lot of great ideas with these new features, but I see many current problems are completely neglected.
@tflidd I agree with you. I often wonder what happens to those announcements after the media hype. Take the projects for example. There is no real way to manage those and not even some kind of documentation how to add support for projects (collections codewise) to apps. Iām using Nextcloud as Dropbox replacement but still some basic file synchronization features are missing or not working
nextcloud/server on github has 2160 open issues and 2021 with
no:assignee
nextcloud/desktop on github has 768 open issues and 699 withno:assignee
As @anon99283430 already said. Itās not that easy You can use the labels to filter the data.
is:open is:issue label:bug
: 1163 open
is:open is:issue label:enhancement
: 885 open
is:open is:issue label:"1. to develop" label:bug
: 191 open and confirmed bugs
is:open is:issue label:"1. to develop" label:enhancement
: 420 open and accepted enhancement.
ā1. to developā means a issue is reproducible or a enhancement is acceptable. A bug/enhancement is not planned at this stage.
Just some general remarks about issues on GitHub: Iām doing a lot of housekeeping but itās hard. People often ignore the issue template, refuse to provide additional information or they donāt respond anymore. Also those search thing seems to be ā¦
We probably need a better way to handle issues or enhancements on GitHub. It make sense to keep the app code into different repositories but for issues/enhacements itās hard not to lose track.
Some examples
- NC 18.0.0.10 (from download after Nextcloud HUB annouce). Ā· Issue #18974 Ā· nextcloud/server Ā· GitHub Actually I should close this. Itās not a bug report. But Iām sure he will complain then. So just leaving it open until the time beeing.
- Nextcloud and Onlyoffice: Error when trying to connect (cURL error 60: SSL certificate problem: unable to get local issuer certificate (see http://curl.haxx.se/libcurl/c/libcurl-errors.html)) Ā· Issue #18899 Ā· nextcloud/server Ā· GitHub Configuration issue I guess. Not easy to spot of course but nothing we can fix.
- Cant upload files Ā· Issue #18656 Ā· nextcloud/server Ā· GitHub / https://github.com/nextcloud/server/issues/18671 ā¦ at least this one is closed
- Feature suggestion: Database backup function Ā· Issue #17328 Ā· nextcloud/server Ā· GitHub People hate it if we close their enhancement requests. What to do with this one? As a feature it makes sense. But there are 100000000 tools out their (bullet proof with proven history) to make backups.
- Show admin warning if number of outstanding jobs > 200 Ā· Issue #18582 Ā· nextcloud/server Ā· GitHub was a bug first and enhacement now. Sure the issue is with Nextcloud and the ajax cron. He basically ask to reinvent cron.
- Thatās just the last weeks and only nextcloud/server ā¦
At some point people started to write hate emails to me because I closed their issue / rejected the enhancement request. I removed my primary email and name from GitHub because of that.
Please remember this post if the next time someone complains about the number of open issues on GitHub. Thanks