Recent searches
No recent searches
Do NOT Truncate Comments Without Warning
Posted Apr 09, 2017
When you post a comment to a ticket it may be truncated without singe notice from Zendesk if it will find it too long. So you (and your end-users and agents) may lose a part of your comment text you submitted and be in a confidence the message is fully delivered.
At the moment, "too long" means 64Kb. In case of UTF your safe harbor may be up to 20K characters.
But since the limitation is not documented (see this and this API doc), the limit may be changed by Zendesk without any prior notice.
Official reason for such limitation is "performance issues", as per ZD Support answer.
As per my request to Zendesk, at this moment they do not find that type of our data loss is a problem. They even do not want to show a warning message for such cases.
So i added this feedback to let you know of that potential problem. And in hope to gather enough vote-ups and comments to make Zendesk change their mind.
As of possible solutions, i see 2 ways to work this limitation around in a proper way:
- Show "remaining characters" countdown somewhere near the comment text area (as decent people do).
- Save the full comment in a separate object or make it avaible by a click (e.g. "too long" emails are hidden under "Show full email" link, but not lost).
- ... what's yours?
Thank you for your support.
36
26 comments
Official
Amisha Sharma
Hello All,
We heard you!
We have released the Email Truncation Alert feature (only on Agent Workspace). This feature will consist of two changes:
1. Counter will show at the bottom right of the composer when you 5Kb away from hitting the limit. The counter will keep changing as you enter more content and will go in negative at some point.
2. An error message will pop up if you try to submit a ticket that is over the 64Kb limit. You will need to edit your content and then try submitting again.
We are hopeful that this initial version will help. We will continue to collect feedback and improve this even further. Thanks!
0
Michael Goldman
I vote for solution #1.
2
Nicole Saunders
Michael, can you tell us more about your use case?
2
Michael Goldman
Sure. My agent recently copy/pastes a SQL output with lots of whitespace and formatting. It got butchered upon copy/paste since it was too many characters. She tried it again 2 more times and kept getting the same result thus the customer got the same comment 3 times. I had to go in and make 2 of them "internal" to fix the customer view which leads to a bad customer experience and a poor agent experience.
The only reason we found this out was due to her telling me that the system was broken and I did some research.
3
Amjad Talib
I have a use case where tickets are created via email with tables and if the tables exceed a certain number of rows, the comment is truncated in the body, plain_text, and html_body fields..
I posted here about it as well in a similar request:
https://support.zendesk.com/hc/en-us/community/posts/115000714147-Retrieve-Original-Email-Content-Using-Core-API
I know we can retrieve the html, but it isn't very useful if the message is truncated without any ability to retrieve the remainder.
2
Christopher Stock
Another vote here for truncation warning - we've been caught out by this limit recently.
4
Chayce O'Neal
We've hit the limit. Upvoting this! :)
3
Monica
Would love #1 at least and #2 would be a nice addition.
3
Robert Welch
We just ran into an issue where we had an automated alert get truncated that had customer data on it. As a result, we missed notifying some of our customers and were put in a very uncomfortable position. Any type of visual indicator that a message is truncated is critical to us.
2
Chris Swinney
Annoyingly, we have just seen the same thing happen. We can have incredibly long and complex tickets. I personally make things worse by operating almost exclusively via HTML email, which is a requirement for me personally as the application interacts with my screen reader in a much more compatible fashion. Several new comments in the thread were simply blank or only contain a partial line, making responses between the customer and ourselves almost impossible.
If you are going to truncate, it would seem sensible to truncate the historical messages not the new ones.
3
Natalie Cherwin
We have a special ticket type that is triggered to notify email targets. The recipients of the email targets cannot view the HTML link to view the full message. This is a problem we are missing parts of orders. The people receiving emails through target extension emails need a way to view the full message, and they end up looking silly because they have to reply to the end user for the full email, or submit a request to help desk crew for the full email.
My solution would be to display the HTML link in the ticket itself so this can be forwarded to cc's or email targets.
Thanks for considering.
2
Roman Sheydvasser
This appears to have happened to us with a zendesk chat transcript that was too long. Either a warning or a link to a full transcript would be great.
2
RealConnections Support
Feedback for the product team
2
Bruno Cruz
Would we awesome if we could just get the warning right now, improving on this in the future. We were actually unaware of this char limit, an agent lost part of its internal comment due to that.
4
Montel G.
Hi team,
Any updates on this? It appears that this request has been out there for a while and I just ran into a loss of an hour's worth of research in an internal note as a result of unexpected truncation and I am beyond frustrated.
3
Monica
Adding another vote for number one and a warning pop-up if a person exceeds the limit.
1
Brook
My team uses internal notes for recording important information regarding customer billing issues and NCMEC/LE reporting. These notes are often lengthy and when they are cut off without warning, we end up with incomplete information that is vital to both completing appropriate refunding/product distribution for customers, and providing adequate information when reporting important things like CSAM and real life threats to appropriate agencies.
1
Amisha Sharma
Hey everyone, Thank you all so much for taking the time to share your feedback with us in regards to comments being truncated without warning. Our team would like to investigate this feature request in the future and find the best way to address this limitation. Currently, we are focusing on composer stability so we won't be able to prioritize this to our roadmap for at least the next 6 months. In the meantime, please keep sharing your use-cases with us as this will help us to prioritize this functionality in our future roadmap planning.
0
Amisha Sharma
Hey everyone, Thank you all so much for taking the time to share your feedback with us in regards to comments being truncated without warning. Our team would like to investigate this feature request in the future and find the best way to address this limitation. Currently, we are focusing on composer stability so we won't be able to prioritize this to our roadmap for at least the next 6 months. In the meantime, please keep sharing your use-cases with us as this will help us to prioritize this functionality in our future roadmap planning.
0
David Rose
@... just been caught out with this giving the customer some quite complex SQL to fix a database problem. Thankfully the truncation resulted in a syntax error so the code wouldn't run and we were able to identify the problem, but it's quite possible that the truncated code could have been legal, but wrong - this could be very dangerous.
e.g if the truncation had occurred after the first of these two lines of SQL, the code would run but the result would be very different than that intended, deleting all records rather than just the one.
DELETE FROM table
WHERE record = 1
BTW, Slack manage it very well, offering the user the option to create a snippet (attached text file) on the fly.

1
Kyle Pinkley
This has been reported on a couple of occasions by our agents. In some cases the limit is not high enough depending on how much output from logs is sent. The customer needs to receive all of the response the agent is sending. Having a warning show when the limit is reached and when the customer submits a response would at least help.
2
Tim Grimshaw
Hi everyone! This is a ‘plus one’ on this one from us - we sometimes run into a rare, but super frustrating issue where an agent will compose an internal comment which is larger than the allowed limit for comments. They submit the internal comment, and it silently truncates it, so the agent then loses anything they've written which was outside of the max limit.
I came across this feature request, and got super excited when i saw that the top comment says that there is now a feature “Counter will show at the bottom right of the composer when you 5Kb away from hitting the limit” … but I can't see it anywhere, and can't find any other references in documentation to see where to enable this.
Am I misunderstanding that comment, or am I just completely missing where to enable that feature please? Thank youuu!
0
Shawna James
0
Tim Grimshaw
Thanks for the update Shawna James 🙏🏻 appreciate it! Maybe I'm just missing the counter in the interface in that case, cos I'm still a bit confused.
Re the post at the top of this post, the pinned comment mentions:
“Counter will show at the bottom right of the composer when you 5Kb away from hitting the limit. The counter will keep changing as you enter more content and will go in negative at some point.”
Do you have a screenshot of that counter in the compose box please? Maybe that'll help me work out why I'm not seeing it?
Thanks again!
0
Tim Grimshaw
Ah… i see the issue now. Just done some testing with a large comment that got lost, and the issue looks to be:
That means that while the protections are in place for public replies - for internal comments we're still finding that people's comments are getting silently truncated 😭
Is that a bug with the current feature (it's only enabled on public replies instead of both public replies and internal comments?) or should I raise a feature request to get the same behavior applied to internal comments too? Or… was there a reason that it wasn't implemented on both?
Thanks again :)
0
Shawna James
1