I solved this by installing SSL as that allowed the ws connection to work properly through the ISP in question.

Hi,

I have a few users (including myself) that are experiencing issues with the image upload.

The images start to upload then stop like...

Images do however successfully arrive on imgur.

This started about 2 months ago.

NodeBB version 1.4.2

Any one else experiencing this issue?

Any suggestions of what I can try?

Thanks

Ted

Users are still receiving the invalid buffer error as well, it appears to be (but I am not sure) posts with a lot of images. Is that likely related?

Does the buffer array limit/length include the image content? I wouldn't think so.

Note: Our images are currently sent to imgur.

Thanks

Ted

@baris it can be any at random times, no consistency at all, a refresh of the page solves the issue

Thanks again for the reply

Still get the INVALID DATA when upvoting/downvoting all the time though

@PitaJ I don't think it is a memory issue, I have about 10 concurrent users as a maximum at the minute. I have just installed the new graph metrics provided by digital ocean memory is around 30%, will keep an eye on that.

I have also increased the post length in the ACP.

Will report back if it continues

Thanks

Ted

@baris it happens randomly, sometimes not at all with huge posts. I use redis as the db.

However I can write a post, press submit, get this error. Copy the contents, refresh the page, paste the contents and then it posts fine.

Thanks for your reply

@Pilvinen said in How to improve Nodebb - few suggestions from the top of my head:

Thread tools: Edit poster/adopt post
The admin should be able to switch the owner of the post. (right now I have orphan posts left over from import and the only way to change their owner is to use the redis-cli). Probably not a very popular/useful feature in general, but should be simple enough to implement because why not?

Is this possible yet?

Thanks

Ted

Hi,

I myself have received this error when posting large posts and now some of my users are reporting it. I was wondering if anyone else has experienced it?

Also, people get this one all the time when up or down voting posts...

Refreshing the page usually resolves these issues but that is not ideal as for the first one you lose your post and most of my members don't have the ability to copy, refresh, paste the contents and retry.

Any ideas greatly appreciated.

Thanks

Ted,

@etakmit isn't this the 'Site Title' under Settings > General, you can either change the text or turn it off

@jarey No probs. I would have figured it out sooner if i could get my local install to behave itself, had to use the server one and figure out some sort of logging using winston

You can correct this by changing the line (about 42) in the library.js in the plugin folder

from...

if (p.deleted && !(privi[0].isAdminOrMod || p.selfPost)) {

to...

if (p.deleted === 1 && !(privi[0].isAdminOrMod || p.selfPost)) {

Hope that helps

Thanks

Ted

@metalkramp there is a plugin called nodebb-plugin-assign-newuser-to-group that is in the plugin list that will allow you to select a group for users to be assigned to upon registration.

There is also the setting 'Make user info private'...

under settings > users that keeps the email and other personal information private, for just email there is a further default setting for users - 'Show email'

on the same page

Currently it looks like the groups only load 10 posts and that is it, is there anyway to turn on auto scroll or at least increase this limit?

Thanks

Ted

Upgraded to 1.4.2, same issue, just so you know.

Cheers

Ted

@PitaJ sorry I should have looked a little harder first I just noticed other people putting issues in here. That post fixed my issue. Thanks ever so much!

@PitaJ Thanks again, I use nginx for the URL, however I just checked and if i access through the direct IP and port the problem is not present. I assume that would mean there is some missing configuration but I am unsure of how to fix it, I am still learning node and nginx so any pointers appreciated

@PitaJ No, no difference between en-US and en-GB, this is only in admin (ACP) though. The front-end is fine.

In dev mode I receive no errors in the terminal console or in the error log in Advanced in the ACP.

In the Chrome console when running in dev mode I get

Which would explain the issues. I can see these files in the /public/language/en-US/admin and en-GB folder

Thanks

Ted

@PitaJ Thanks for your reply, I currently have en-US. I did have en-GB but I changed it earlier to see if it was that but no difference

Hi,

I have upgraded from v1.3 to 1.4, then 1.4.1, 1.4.2, it all appears to have gone successfully from the front end side. Awesome job, very smooth from a setup side

However admin, seems to have lost a lot of it's language localization strings....

Note this is only a few examples, there are loads.

Anyone else experience this issue? Did I do something wrong? Any way to fix it?

Thanks

Ted