Appointment list - missing company names / status

The appointment list is missing names and company status (this applies in Beta and in the data returned from API) for some companies. Recent change? Example shows the names and numbers of the first few companies appointed to, but then is missing the names and status:

RM NOMINEES LIMITED
/officers/5Jcd7dBiPH80XOq7xhJpBd28XLU/appointments

e.g. can see

TENSOR PLC (03670909) (status active)
ORACA LTD (03750103) (status active)

But then:

(03962209) Company status
(04236472) Company status
(03657669) Company status

It seems that the difference is that the companies not shown correctly dissolved before the CH data limit for old companies e.g. around 7 years. (BTW could that be mentioned next time the documentation is updated?)

Checking a few e.g. via URI shows:

03962209 “OLD TIL LIMITED”,
“CompanyCategory” : “Private Limited Company”,
“CompanyStatus” : “Dissolved”,
“DissolutionDate” : “31/01/2010”

04089896 “MANAGECONSULT LIMITED”,
“CompanyCategory” : “Private Limited Company”,
“CompanyStatus” : “Dissolved”,
“DissolutionDate” : “02/03/2010”

04367196 “MOMENTUM MANAGEMENT LIMITED”,
“CompanyCategory” : “Private Limited Company”,
“CompanyStatus” : “Dissolved”,
“DissolutionDate” : “16/03/2010”

04637043 “LANKORA LIMITED”,
“CompanyCategory” : “Private Limited Company”,
“CompanyStatus” : “Dissolved”,
“DissolutionDate” : “13/04/2010”

We are currently (started a few days ago) in the process of deleting companies over 6 years dissolved, along with all the data associated with those companies.
This work is NOT complete yet.
When the deletions are complete, there will be a job run to update the lists and counts, so that these appointments missing data will no longer be displayed.
We apologise for the temporary data inconsistencies while the data is updated, as this is a relatively slow process.
If we correct the data any faster, it could impact the performance of the live service.

Thanks - thought it might be something like that.

Could this be more widely publicised e.g. what the API doesn’t have by design. (Did I miss it on the main page of docs / forums)?

How long data on dissolved companies is kept is one thing (given something about these companies can be found via WebCHeck / data URI). Personally I think this kind of information on relationships is useful. I’d rather know about a relationship and then turn up a blank when looking up a company - then I can fall back to the (more limited) info elsewhere. The alternative (as I understand) is there’s just no way of knowing that e.g. someone was a director.

Thanks for getting back anyway, keep up the good work!

Why are you deleting dissolved companies? This is like walking into the National Archives and trashing them. Just because they have been dissolved for 6 years does not make them uninteresting to journalists, historians and researchers of the economy and corporate finance. In Hong Kong, where I run Webb-site.com, the registry keeps all corporate records going back to the 19th century, and as a consequence I am able to do this kind of analysis on survival rates and incorporation/ dissolution rates, something that is currently impossible with your 6-year approach:

https://webb-site.com/dbpub/incHKsurvive.asp
https://webb-site.com/dbpub/incHKannual.asp

Please stop the electronic vandalism.

We are removing them from CHS only.
We keep all data, just not on CHS.

If you are removing the data from online public access under the principles of open data and open government, it amounts to the same thing. Why are you doing this?

If it is motivated by complaints from people who are embarrassed by their connection to dissolved companies, then that is not a valid reason. The European Court of Justice has recently ruled on this every point, that there is no “right to be forgotten” in personal data in the companies register:

Besides, I would note that:

  1. People who were directors of companies long ago are still in the dataset, as long as the company is still active; and
  2. Even for dissolved companies, there will be large fragments (or possibly complete copies) of the data available in private-sector websites (whether for free, ad-supported, or for a fee), so all that you are doing is giving them an information advantage, a reward for having collected the data from you while it was still available. Is that what Companies House really wants?