Question Hi, is there a way to hide this error message?

Hi, is there a way to hide this error message？



URL：
dev.webtrees.net/demo-stable/index.php ?route=%2Fdemo-stable%2Ftree%2Fdemo%2Fsearch-general&query=%F0%A1%8A%A8&search_families=1&search_individuals=1&search_locations=0&search_notes=0&search_repositories=0&search_sources=0&search_trees%5B0%5D=demo#

Looks like this issue: github.com/fisharebest/webtrees/issues/2950
in other words: the database of the demo server does not support 4-byte Unicode characters.

How to hide the error message? It can be prevented by using only Latin or Cyrillic characters, avoid using Chinese characters.

Or does this also occur on your server, and not just the demo server?

The amazing thing is: after logging in to your account, there will be no error when searching again, for example, logging in: "moderator"

My database is: MySQL5.6.24. If the field "Collation" is set to "utf8mb4_general_ci", special characters can be saved.

I tried to batch modify the collation of the table to "utf8mb4_general_ci", but it failed because of the foreign keys of some tables.
 


 

Try to do the tables first where those foreign keys are?

Note this comment in the issue:

fisharebest wrote: `
Using utf8mb4 on new installations is straightforward.
[...]
Updating existing databases has many difficulties. It may be impossible on some servers.

I am waiting for the doctrine/dbal project to merge my fix and create a new release.

See github.com/doctrine/dbal/pull/6370

When this happens, I will be able to create a fix for this issue.

I've been patiently waiting for this fix, hoping it can be merged into dbal sooner.

Thanks Greg for the status update. I also look forward to this. For those unfamiliar with the specifics: there are a series of Chinese (usually given) names which have no modern western translations and will crash the database currently used by webtrees when recorded in the program. See how the Peoples Republic of China 'deals' with this conundrum in a 2009 NYT's article: www.nytimes.com/2009/04/21/world/asia/21...=1&pagewanted=print#
Many westerners might see this when they use emojis copied from social media posts. For people documenting Chinese genealogical records, this runs into the core webtrees value of saving names. @xmif recommended recording homophones as a temporary substitute; he said this takes some time to figure our a correct word. My work-around (aggrevated by lack of language skills) is to record the unrecognized word into a unicoded text file and save it in the media folder. These names are stored in Gramps, which I assume use a different encoding in their database(?)