muppeth

Mon, 16 Sep 2019 18:02:34 +0200

forteller wrote the following post Mon, 16 Sep 2019 16:38:31 +0200

"Without a truly open standard for payments, open systems in general have had a hard time competing with their proprietary counterparts."

Coil, Mozilla, Creative Commons have set up a $100 million grant to fix this. Very interesting!

https://coil.com/p/coil/Coil-Mozilla-and-Creative-Commons-Launch-100-Million-Grant-for-the-Web-to-Advance-Web-Monetization-f/Bdp8Hgcf4

poVoq

Mon, 16 Sep 2019 20:58:47 +0200

I hope GNU-Taler will also see some boost from this, but it might be a case of too little too late, with both Facebook (Libra) and Telegram (Gram) pushing into this market with a full warchest very soon.

ActivityPub mentions

muppeth

Sun, 08 Sep 2019 14:00:50 +0200

!Hubzilla Support Forum
The annoying thing about mastodon is that it looks like you need to always mention people in the thread to apear in the conversation on the otherside. I find it very annoying and pointless, but that how they roll it seems like. I've been wondering if its possible to auto mention everyone inthe thread when replying to AP post so that its consistant with how at least mastodon does it.

show all 4 comments

Mike Macgirvin

Mon, 09 Sep 2019 02:16:10 +0200

I implemented this for the Osada project around a year ago.  But just collecting tags for people in the conversation isn't actually the right way to do this.

I'll tag @Indio  because he's been learning that codebase.

Look for the string 'authors' in Zotlabs/Lib/ThreadItem.php in Osada or Zap (the code still exists in  Zap, even though ActivityPub is disabled there). This is the current implementation - and is wrong; although in a pinch it's probably better than nothing.

I believe the correct implementation is to tag the author of the specific tweet/toot/dent/plerk/whatever you are replying to as well as anybody they tagged in their tweet/toot/dent/plerk/whatever. I believe you can ignore the rest of the conversation. You would need to verify this as I do not use ActivityPub personally and very little that Mastodon does makes any sense in a conversational (Facebook-style) social network.

Somebody should also verify if it is necessary to tag yourself if you reply to your own post/comment. This wasn't needed on the Zot side but it might be important to Mastodon.

There is an additional function called commentAuthors() in view/js/main.js which is responsible for inserting this mention list when you reply to something.

For Hubzilla, there are additional constraints/complexities. You probably do not want auto-mentioning if the ActivityPub Protocol isn't installed, and you also probably don't want it if all the members of the conversation are using a conversational network (Hubzilla, Zap, Diaspora or Friendica). Only if somebody involved in the conversation is using a Twitter-style service. It "should" also be a function of the pubcrawl plugin to implement this feature since it isn't needed or desired functionality in Hubzilla core.

poVoq

Mon, 09 Sep 2019 08:32:33 +0200 last edited: Mon, 09 Sep 2019 08:33:06 +0200

At least for the new comment reply button there is that functionality. IMHO to make the UI consistent I would hide the comment box by default and add the reply button even to top level posts.

muppeth

Mon, 09 Sep 2019 12:45:59 +0200

Thanks @Mike Macgirvin for the info. The mastodon way of doing things is very annoying from the perspective of normal conversation like threads. Dont see much sense in having to 'mention" everyone everytime.I guess if the mention everyone in the thread should definatelly be used only in AP threads and only with the plugin enabled.

@poVoq which new comment button?

Max Kostikov

Mon, 09 Sep 2019 17:19:44 +0200

@muppeth
To enable see the option in "Conversation Tools".
Also I just opened the MR with automatic quotation for highlighted text in reply (only works for text in the post you are replying to).
Add selected text quote on comment reply (!1724) · Merge Requests · hubzilla / core

Max Kostikov wrote the following post Tue, 30 Apr 2019 17:55:18 +0200

Add 'reply on comment' feature
!zotlabs.org | Hubzilla Development

I have added merge request with new "Reply on comment" feature.
https://framagit.org/hubzilla/core/merge_requests/1614
Now you can enable button to reply on specific comment.

When pressed on you'll be redirected to comment input field where post author which you reply will be mentioned. Also "Submit" button will be marked with arrow indicator and floating hint.
This also should resove issue with Mastodon threaded comments
https://framagit.org/hubzilla/core/issues/1357

muppeth

Sat, 07 Sep 2019 11:44:36 +0200

CryptPad wrote the following post Fri, 06 Sep 2019 18:47:29 +0200

CryptPad 3.0.0 featured a few minor server changes that made a big difference in memory usage.

Our next few releases should continue to improve performance, in case you needed more reason to update.

muppeth

Tue, 23 Jul 2019 02:18:49 +0200

!Disroot Community
Ok. I had a hard day. Just had a glass of wine. Time to try, and fix (break) diaspora support on hubzilla.

show all 16 comments

Mario Vavti

Tue, 23 Jul 2019 10:53:00 +0200

Time to try, and fix (break) diaspora support on hubzilla

I see you have enabled it again...

Did it get any better? I am really curious about your findings and where the bottleneck is...

muppeth

Tue, 23 Jul 2019 12:27:27 +0200

Its hard to tell. It seems to be working now. I have disabled php presistant postgres connection and tweaked hell outta postgres and php. Atm it seems to work but I have not yet checked the amount of db processes yet (busy at work). As far as I could observe it before closing my eyes last night, it was going ok for about 10 - 15 minutes (up to 5 active processes tops) and then it jumped to about 20. I suspect that the amount of diaspora missed posts to deliver in the queue must have been big so I left it like that.
In the past however this amount of active db processes would grow until it would reach php's max_children set on hubzilla server which would render hub offline of course. No matter how high i set max_children it would hit the roof and stay there until diaspora addon would be switched off. I guess the root cause is missconfiguration either on postgres or php since noone else seem to suffer from this, but I wouldn't exclude diaspora addon though. Our hub is aware of about 17k diaspora accounts and about 43k activitypub, yet its diaspora that the moment its on keeps on flodding with posts constantly so maybe some rate limit on diaspora addon could help the situation. Do you know if there is any other hub around that gets a lot of diaspora content? I wonder if they experience issues too (Guessing they most probbaly use mariadb which would be interesting to check and compare).

Mario Vavti

Tue, 23 Jul 2019 13:36:33 +0200

Have you tried the queueworker addon yet?

muppeth

Tue, 23 Jul 2019 18:18:05 +0200

Yes its enabled. Is there any options for it that could be tweaked?

muppeth

Tue, 23 Jul 2019 18:19:47 +0200

Seems like every few hours the amount of workers jumps over threshhold for few seconds. I will add check to zabbix to monitor amouunt of active processes to be able to pinpoint things in time and compare with the logs.
It is already much better as its usable even with diaspora addon on (just few seconds downtime every now and then).

Mario Vavti

Tue, 23 Jul 2019 20:06:46 +0200

I do not have much experience with the addon. Also it has been improved in dev branch recently.

The settings for the addon are in /queueworker

Also i am curious about your other delivery settings from the admin panel.

muppeth

Tue, 23 Jul 2019 23:38:43 +0200

🤔 getting page not found on admin account.

muppeth

Tue, 23 Jul 2019 23:45:14 +0200

Currently I set all deliver to the lowest numbers:

Mario Vavti

Wed, 24 Jul 2019 00:02:15 +0200 last edited: Wed, 24 Jul 2019 00:04:18 +0200

You should actually get to queueworker settings as admin. Do not see why that could be... It should only throw a not found if you are not logged in and not admin.

You should probably tweak the queue threshold a little. 30000 is quite high (that basically means your server will try to deliver 30000 items immediately before starting to queue them).

Maybe set it to 10000 for now and see how that works...

Also you can probably set deliveries per process to maybe 2 or 3. Should be no problem on bare metal with good internet connection i guess... Maybe even higher...

muppeth

Wed, 24 Jul 2019 00:08:32 +0200

Hmmm... weird indeed. i can see all other admin pages on the admin account.

Thanks for the explanation. I set the threshold long tiem ago when we had issue with delivery (you told me to put it higher to process gigantic queue we had). I should have put it lower after but I did not really understand the meaning of those values. BTW. is this documented somewhere? If not I will try to ask around and compile and publish some notes. Might help others.
As for other values I will adjust as your suggestion. Also because of lack of understanding what they exactly do I wasnt sure what to set it to.

muppeth

Wed, 24 Jul 2019 00:10:12 +0200

According to the admin page the queueworker is on:

muppeth

Thu, 08 Aug 2019 16:23:46 +0200

Update
Just small update for those who ask whats the state of diaspora federation. Nothing has changed. Everytime I switch it on our instance goes down within minutes. The amount of active db queries matches the amount of php children worker amount and continues like that until diaspora addon is off. I havent have much time to debug it last days.

Thinking maybe just dump the idea of federating to diaspora (so put is as low profile on the list of todo for proper hubilla launch).

poVoq

Thu, 08 Aug 2019 16:57:37 +0200

@muppeth IMHO, dropping Diaspora is probably the way to go (I never enabled it on our Hubzilla in the first place). The protocol used by Diaspora is just too different from AP or Zot, and bridging beween the latter two is already difficult.

Mario Vavti

Fri, 09 Aug 2019 23:27:43 +0200

Hm... Basically storing items is the same for every protocol. Processing diaspora messages will have some overhead compared to activitypub because we have to parse xml instead of json which is slightly harder. It should not make such a big difference though imho...

muppeth

Tue, 13 Aug 2019 16:40:15 +0200

I planning to put it back on the table for upcoming spint starting this week so in next two weeks time I will put some significant amount of time trying to either resolve the issue or make decission to drop diaspora support entirely and move on to other hubzilla related stuff instead.

Mario Vavti

Tue, 13 Aug 2019 21:42:29 +0200

I won't be much available in the next couple of weeks... Anyway, it would be good to find out where the overhead comes from. Outgoing messages? Incoming messages? Something else?

muppeth

Tue, 23 Jul 2019 02:17:12 +0200

Need to look into .mod .xm player plugins for hubzilla. Me has a plan. Me wants moar tracker music.

show all 4 comments

muppeth

Tue, 23 Jul 2019 12:36:29 +0200

You mean this one? https://www.stef.be/bassoontracker/ Yes I was already thinking of putting it as an app for hubzilla (would be cool to integrate it with files so you could compose and save/load from your hubzilla account).

But I looking for a simple player that does not require opening entire app. something like the current videoplayer or soundcoud's music player.

muppeth

Tue, 23 Jul 2019 12:39:21 +0200

Damn bassontracker has integration with modarchive and mnodules.pl. and you can define between xm and mod and decide on amount of tracks. this is cool.

rialtate

Thu, 25 Jul 2019 00:56:48 +0200

It looks really cool but is definitely not the one I found before. There's probably a few of them I'd imagine. What I liked about the one I found was it was written more like a library and you could build your own interface from components. It also had a very efficient HTML5 audio engine.

muppeth

Thu, 11 Jul 2019 00:28:17 +0200

test 123

G. Love

Thu, 11 Jul 2019 01:34:34 +0200 from ActivityPub

@muppeth test test test!

João Azevedo

Thu, 11 Jul 2019 19:42:36 +0200 from ActivityPub

@muppeth Teste!

muppeth

Wed, 10 Jul 2019 23:38:59 +0200

muppeth updated their profile photo

muppeth

Sun, 23 Jun 2019 01:11:24 +0200

!Hubzilla Support Forum
Going through logs I see lots of:
:Router.php:179:__construct: index.php: page not found: /receive/public ADDRESS: ... QUERY: q=/receive/public
Any hints? Tried checking Router.php but not sure if it means I should ignore it or indeed there is a missing module and if so which.

Mike Macgirvin

Sun, 23 Jun 2019 03:13:09 +0200

That's a remote server trying to deliver to Diaspora, which you've mentioned elsewhere it not currently enabled - hence your server can't find that endpoint.

muppeth

Sun, 23 Jun 2019 12:15:41 +0200

Thanks for clearing that one out.

muppeth

Sun, 23 Jun 2019 00:23:47 +0200

!Hubzilla Support Forum

What does the error in the logs mean
Finger.php:139:run: invalid signed token: https://hub.disroot.org/.well-known/zot-info?f=&address=user&token=... Does that means channel does not exist?

NetNut404

Sun, 23 Jun 2019 01:10:27 +0200

I haven't got any response from hub.disroot.org in a while so I suppose it can't reach it

muppeth

Sun, 23 Jun 2019 01:16:59 +0200

I'm posting from hub.disroot.org so obviously it's online. I do have some issues mainly apearing the moment I switch diaspora addon on (causing my db to choke because of concurrent processes). I dont know whether this is the root cause (the diaspora addon is currently off) but i'm now trying to go thought logs to resolve all possible warnings/issues before I move on to diaspora.

The fact you see my post and I see yours should indicate hub.disroot.org is online. That piece from logs is just one server. I see also hubzilla.eu;hub.g3l.org;gerzilla.de; start.hubzilla.org;hub.weho.st and few others.

Mike Macgirvin

Sun, 23 Jun 2019 01:25:48 +0200

Ultimately there are hundreds of things which could cause this, but it basically means that somebody on the other end signed a request and we couldn't verify the signature on your end. Assuming it isn't a blatant forgery, It is highly unlikely to be an issue caused by your server, and very likely to be an issue with the requestor's server. It could be as simple as the server not responding in a timely manner.

If it's causing an identifiable/recurring problem we might need logs on both sides to figure out the ultimate cause. But if it eases your mind a bit, we upgraded this particular verification step to use OWA probably around two years ago and still support the old mechanism, but it's pretty ancient. This tells me the other server is probably running really, really old code and if it's causing an issue the very first thing I would do is get them to update.

muppeth

Wed, 19 Jun 2019 14:09:31 +0200

I've been asked at work to edit a short video. I have a feeling it will be a dumpster fire on its own excluding the software choice (no real concept and the bad material to work with).
I havent use any video editing software for long time. And last time I used one was when @Antilopa and I were making apt-get shirt video

apt-get shirt --build
by apt-get shirt on Vimeo

Which we ended up doing in blender. I have close to zero skills in blender (all hard work was dne by antilopa) so I need to find other solution. I see quite some new kids on the video editing bloc, so I decided to test our Shotcut and Flowblade. Anyone with expirience in either. I hope the crash rate will be lower then all the other video editing software I worked with in the past.

#videoediting #videoeditingsoftware

projects

fuchsia

Wed, 19 Jun 2019 14:12:03 +0200 from ActivityPub

@muppeth @antilopa I use kdenlive for my quadcopter videos, which works alright. Crashes very rarely, so do keep saving your project

meaz

Fri, 21 Jun 2019 05:53:19 +0200

I also use Kdenlive. It works fine.

NullSpaceOddity

Fri, 21 Jun 2019 06:59:21 +0200

I've used Openshot for some projects. It was fairly intuitive for basic stuff, but not really stable.

Cinelerra GG is much more stable, but has a pretty big learning curve.
Cinelerra GG Infinity free video editing software for linux

Cinelerra GG Infinity is a free video editing software or program for Linux. Edit your videos like a pro for YouTube, Facebook and Instagram.

Missing and broken thumbnails

muppeth

Tue, 18 Jun 2019 03:20:45 +0200

!Hubzilla Support Forum
Since migration of thumbnails to fs I noticed quite few thumbnails missing, low quality or broken.

Gallery displays only couple of thumbnails. Missing thumbs are replaced with random ones (my pretty face in this case).

Once pressed to view the image in fullscreen the proper image is displayed, though I notice couple of them missing too. Also noticed my thumb in low quality.
I did run util/thumbrepair few times but with no success.

show all 6 comments

Xosé M

Tue, 18 Jun 2019 08:44:50 +0200

good morning! #TZAG

I'm now writting from cloned identity, dis-hub seems a little buggy this days. Hope all backs to normal.

Few days ago my avatar was changed to some default bird (eagle), I noticed some other users got this avatar though, so I guess was some missconfiguration.

Thanks anyway!

Mario Vavti

Tue, 18 Jun 2019 09:38:43 +0200

@muppeth there was another fix for the repair script shipped with 4.2.1
You might want to try util/thumbrepair once more after updating to 4.2.1

Max Kostikov

Tue, 18 Jun 2019 18:21:33 +0200

@muppeth
I just sent MR to master which probably should fix your issue with util/thumbrepair
https://framagit.org/hubzilla/core/merge_requests/1673

muppeth

Tue, 18 Jun 2019 19:10:39 +0200

@Mario Vavti I'm pretty sure I did it after the update again, I'll wait for @Max Kostikov merge and run it again. Thanks for looking into this.

Mario Vavti

Tue, 18 Jun 2019 19:24:32 +0200

Merged.

muppeth

Tue, 18 Jun 2019 23:40:04 +0200

Pulled and executed thumbrepair again. Looks like more of the thubnails is back. I see few white ones, but they do load when opening, so it might be my cache, generally something I can live with (hope that applies for others on my instance). Thanks a lot guys.

muppeth

Tue, 18 Jun 2019 02:18:29 +0200

Mike Macgirvin wrote the following post Tue, 18 Jun 2019 02:16:00 +0200

Share freedom – Get this Extension for 🦊 Firefox (en-US) Share freedom – Get this Extension for 🦊 Firefox (en-US)

Download Share freedom for Firefox. share the current tab on the fediverse

Diaspora addon temporarily disabled

muppeth

Thu, 13 Jun 2019 01:59:42 +0200

!Disroot Community
Due to stability issues caused by what looks to be Diaspora Addon, federation to d* has been disabled until the root cause is identified and fixed.

Lepe

Thu, 13 Jun 2019 18:50:23 +0200

Diaspora addon causes huge flood

muppeth

Thu, 13 Jun 2019 01:54:54 +0200

!Hubzilla Support Forum
For quite some time my instance is struggling with performance. Since last two days to the point it is essentially dead. No matter how many php processes I set, it will get filled up and fillup db with queries.

I see my hubzilla logs flodded with:


Receiver.php:1459:like: diaspora_like: parent item not found:
Receiver.php:748:comment: diaspora_comment: parent item not found: parent:

i have now disabled diaspora addon (i wonder what will happen now with all the accounts followed), and issue with db flooded with queries has stopped and I can finally write this post.
I would apreicate any suggestions, help or feedback on that issue. If there is anything you want me to run to provide more information in order to find solution to this, please let me know. I run latest hubzilla 4.2 (though the issue is there since about 2 months), php7.3 and postgres.

show all 17 comments

Eudaimon ꙮ

Thu, 13 Jun 2019 06:59:03 +0200

muppeth

Thu, 13 Jun 2019 23:06:31 +0200

I'm not sure this post has reached outside our instance. Can someone confirm they see it (outside hub.disroot.org)?

fridaysforfuture, but everyday

Thu, 13 Jun 2019 23:17:04 +0200 from ActivityPub

@muppeth I saw it from hub.librenet.de and todon.nl

There might be an issue with the server the support forum is hosted,..

paulfree14

Thu, 13 Jun 2019 23:22:40 +0200

jup, it says: " Apologies but this site is unavailable at the moment. Please try again later. "
see:
https://zotadel.net/channel/support

fuchsia

Thu, 13 Jun 2019 23:36:42 +0200 from ActivityPub

@muppeth confirming from pixie.town, no idea how to help tho :/

Iron Monkey

Fri, 14 Jun 2019 00:55:42 +0200

Can see it from hub.subzone.net @muppeth

Waitman Gobble

Fri, 14 Jun 2019 02:58:58 +0200

Something was killing my server too. I had a feeling it was diaspoora. Turning off all the plugins seemed to fix the issue i was having.

Mario Vavti

Fri, 14 Jun 2019 09:07:23 +0200

@muppeth we have been receiving malformed diaspora messages from friendica for some time. Not sure if that is still the case though. I have added some additional checks and loging to the diaspora addon. Please update with util/udall and see if the situation improves...

Alex

Coruscant, Fri, 14 Jun 2019 10:30:43 +0200

Diaspora Protocol is blocking my server zotadel.net almost completely at the moment.
I disabled all addons with "safemode off".

muppeth

Fri, 14 Jun 2019 12:17:27 +0200

@Mario Vavti Thanks. Updating addons, does the core need to be updated too (just to know if i need to merge to my fork).

Mario Vavti

Fri, 14 Jun 2019 12:28:50 +0200

There is a fix to the calendar in core. You should probably pull that too...

muppeth

Sat, 15 Jun 2019 13:11:24 +0200

hmmm.... The moment I re-enable diaspora addon the flood starts. At first I thought it might be due to lots of stuff missing because of the addon being off for few days, but it does look like it it just continues eg: Receiver.php:458:post: diaspora_post: Ignoring this author. (and the above. I will re-enable it this evening to post some queries that I see from db side as well as maybe more logs if needed.

Please let me know what is needed or what should I look at to find the root cause.

Mario Vavti

Sat, 15 Jun 2019 16:59:23 +0200

Ignoring this author

That is not an error message.

muppeth

Sat, 15 Jun 2019 21:49:55 +0200 last edited: Sat, 15 Jun 2019 21:52:42 +0200

Yeah I know. But the log is litterally flodded with those. I can barely see anything else. But yeah I will look around compile some logs and post some more observations.

Mario Vavti

Sat, 15 Jun 2019 23:01:19 +0200

I have just raised the loglevel for those messages to LOGGER_DEBUG. If you use a loglevel below, you should not see those messages anymore.

IIRC hub.disroot.org is subscribed to the social-relay with scope all. Maybe that just became overwhelming for your server?

muppeth

Sun, 16 Jun 2019 13:18:18 +0200

Possibly. Though we run it on baremetal with 24 cores and 32GB RAM and much heavier services we run on similar setup manage. It's not that I'm bothered by the messeges in the logs (its imo good to see them) I just wonder what exactly they mean and why it floods so much. I mean I'm subscribed to the same relay server with the same settings on diaspora and do not notice such a flood.
I would like to understand what is happening. Hmmm... I remember I have switch off fiorehose already months ago: config[diaspora][firehose] = 0

Mario Vavti

Sun, 16 Jun 2019 14:55:32 +0200

It's not that I'm bothered by the messeges in the logs (its imo good to see them) I just wonder what exactly they mean and why it floods so much.

It just tells you that a received messages is not allowed in a channels network stream.

Hubzilla updated to version 4.2

muppeth

Sat, 08 Jun 2019 01:33:16 +0200

!Disroot Community
Hi there.

As our hubzilla seems to be stable for last few days, which is great as it wasn't feeling well for past two months, its time to start working on it again. Starting from update released just few days ago:
Below you can see the link to the change log. If you are using Events or CalDAV app, it now has been merged and replaced by Calendar app. Make sure to enable it in
https://hub.disroot.org/apps/available

Here is the link to ChangeLOG: https://framagit.org/hubzilla/core/tags/4.2

muppeth

Thu, 06 Jun 2019 10:00:59 +0200

GNOME wrote the following post Thu, 06 Jun 2019 08:08:32 +0200

The @kde has done an absolutely incredible job creating a truly world class user experience. We at GNOME appreciate everything you have done to push the #FOSS desktop forward, and we continue to be inspired by your community and vision! 🤗

https://pointieststick.com/2019/06/05/kde-usability-and-productivity-are-we-there-yet/

muppeth

Wed, 05 Jun 2019 22:33:28 +0200

Havent posted much lately (if at all). Our hubzilla instance got a bit more stable these days and i hope this state will continue. I'm thus planning to be more active in the fediverse.

muppeth

Sun, 19 May 2019 17:13:19 +0200

!Disroot Community

Disroot State wrote the following post Sun, 19 May 2019 13:38:02 +0200

Nextcloud update

Monday, about midnight (CEST - Amsterdam Local Time), we will be upgrading nextcloud (https://cloud.disroot.org) to the latest version. This means for about 1 hour the service will not be available.

muppeth

Fri, 03 May 2019 22:51:46 +0200

!Hubzilla Support Forum

Hey. For weeks now my hub struggles with issue I did not manage to solve on my own. @Mario Vavti was trying to help out a bit, but I think bringing it to more people might spark some discussion resulting in finding out a root cause of the issue.

Since last updates (i think as my other post 0.4.0.1) my db is flodded by DEALLOCATE tasks (like DEALLOCATE pdo_stmt_0000559c). Those will grow to the amount of php workers setup (even if i crank it to like 200) which will result in the site unavailable until the tasks are finished. That in turn will work for short while until new deallocates build up again. I have not notice that prior the update and the instance was pretty snappy. Even currently once there is room for new php processes to do their job, the instance works well.

I run out of ideas here so thought someone might look at it with 'fresh eyes' and ask some questions that will bring the debugging on the right track. !Hubzilla Support Forum

Hans W

Fri, 03 May 2019 23:56:54 +0200 from Diaspora

Is it working with persistent connections?

Hans W

Sat, 04 May 2019 00:05:13 +0200 from Diaspora

A normal prepare would solve this but I think you might bump into another problem. If mysql is used then enable slow query logging. Based on that you might want to enable or disable some indexes. If a lot of indexes are used and updated then it will reindex after every insert, update or delete statement.

Logs flodded with missing column "channel_w_photos"

muppeth

Fri, 03 May 2019 22:15:35 +0200

!Hubzilla Support Forum

Sorry for double posting (added issue on git) but wasn't sure whether this is not a better place to post to.
For last weeks (possibly since update to 0.4.0.1) my server keeps spitting out SQL error of Undefined column: 7 ERROR: column "channel_r_photos" of relation "channel" does not exist

I did not find out which addon is causing it, and though someone here might know. I'm using postgres.

show all 10 comments

muppeth

Sat, 04 May 2019 16:00:22 +0200

@Mike Macgirvin ⚡ Looks like this did the trick. Thanks a lot for quck fix!

Max Kostikov

Sat, 04 May 2019 17:06:51 +0200

@muppeth Can we close issue report?
https://framagit.org/hubzilla/core/issues/1367

muppeth

Sat, 04 May 2019 22:09:26 +0200

@Max Kostikov Yes i think so. Sorry for double posting and thanks for responding on git really fast too.

muppeth

muppeth@hub.disroot.org