Remove old (v0) titleindex (title pointer list)

[kelson42]

At the time of "removal" of namespaces, the "old" titleindex has been superseeded by the new "titleindex which is available as a normal item via X/listing/titleordered/v1.

We should considering removing the old pointer list at . This should not be used anymore because deprecated 2 years ago and use storage "for nothing".

The move is easy just put 00 in the header in place on the title pointer list pointer and stop writing the old version of this pointer list.

But before doing that we should better check wit third party ZIM readers to verify that they don't use anymore:

• The old index
• The header pointer to the index

[Jaifroid]

Given the discussions on the complexities of supporting X/lisiting/titleOrdered/v1, and in particular on the fallbacks that are necessary if we don't find data in that listing (see kiwix/kiwix-js#597 and kiwix/kiwix-js#708), I would just urge a very cautious approach towards deprecating the titlePtrList, and a relatively long period of testing given the range of scrapers involved (i.e. different ZIM varieties).

[kelson42]

@Jaifroid I don't understand your comment: do we have any problem with the current ZIM spec and/or implementation on any part of our software portfolio?

[Jaifroid]

@kelson42 No, but this is a low-level potentially breaking change, and because we still use a custom JS backend in Kiwix JS, I would like to be sure to have enough time to make any low-level adjustments if needed.

[kelson42]

@Jaifroid It's not breaking anything regarding current ZIM specification... if implementation of reader is OK. You are IMO the only one in position to check if kiwix JS implementation is OK and such a ticket won't break anything there. We need to be cautious with FUD effect.

[kelson42]

@Jaifroid Should I assume all good on your side and we could consider us ready for this issue version 10 of the libzim?

[Jaifroid]

@kelson42 Apologies, I'm on holiday so not responding as fast as normally. I can't say for sure that there would be no consequences without a ZIM to test. From memory, we use the urlPtrPos to find the listing for X/titleOrdered/v1, and fall back to X/titleOrdered/v0, then fall back to titlePtrPos in that order, so we should be OK. However, what I can say for sure is that if there are any issues, we can easily fix those with a test ZIM. So do go ahead.