Priorities for the next 12 months
As part of the BAC/TAC/staff meetings later this week, there has been a request for my/our top five priorities for the next year. The ask is for a one line per item + a short "why now" elucidation.
Please let me know what you want so I can present the committers' input to those present. Even if you are unsure, say something. Any discussion or input is far better than none.
@Neil Horman I mostly support your set of priorities. I have concern about 5 (zero global mutable data) though. IMO it is not achievable in time of the 4.0 release and as it will inevitably cause subtle API/ABI breakage it won't be suitable for 4.1.
If 4.0 it's going to take a year, will we have an other release in between?
@Kurt Roeckx 4.0 is April 2026 release - as per all releases now being on six months timelines.
Anyone have a clue as to what happened to the text I wrote to explain this?
It has disappeared for me.
From the history it looks like you deleted it yesterday at 11:33 :)
I edited it and it came up blank so I went back and edited again and got the text. I guess the blank must have stuck. It's not the first time I've gone to edit something here and been presented a blank comment.
@Paul Dale I re-added it as I had the text in my mailbox.
Thanks.
My priority is focused on DTLS 1.3. We need it for pqc encrypted voice and video calls and to be able to use providers.
@Frederik Wedel-Heinen We already have it as a priority for the Foundation as we have some funding to help with that (not for fully implementing it but we hope to finish it together with you and broader community).
Removing ENGINE support and probably some more deprecated stuff
Extending EVP_SKEY support
Mine is:
-
Support Distance-Preserving Encryption (DPE)
-
Since: it's very useful in LLM inference with RAG (Retrieval Augmented Generation) to protect the privacy
-
And I am glad to do a very brief introduction at the B/TAC meeting to RAG and related data security demanding in case not all of you are familiar with that realm.
To me, top priority is to get all PRs merged that change the API.
Next priority: reduce as far as possible the backlog of other open PRs. Many of those await reviews/approvals since long.
This is the suggestion I put forward for tonight's BAC/TAC/staff/directors call:
https://openssl-communities.org/d/6xLRj8rw/face-to-face-meeting-bac-tac/2
My thoughts alone, targeting the 4.0 development to take place over the next 12 months:
ENGINE removal (https://github.com/openssl/project/issues/1318)
selected deprecated function removal
discontinuation of atexit() use in OpenSSL (https://github.com/openssl/project/issues/271)
sslv3 removal (https://github.com/openssl/project/issues/1319)
zero global mutable data (https://github.com/openssl/openssl/issues/23451)