We do have a fairly good idea, but certainly not enough to describe it's exact relationship to exactly how long we can except vaccines to last, to OP's concern :)

I don't think the media has sensatialized any of the strains, and I am equally unconcerned as you. I expect I may need a booster in a year, probably in reality two, but one is safer under uncertainty.

Do you have anything on the Astra-Zenaca vaccine and how well it works against the do called "South-African" variant. There was a story that they'd stopped a trial because the results were so bad. Do we have any harder data?

Yes, Astra-Zeneca has been shown to be repeatedly poor against variants, a big story in the UK right now is whether the increase in cases and sense of being behind the US is due to gov't overly committing to a shoddy vaccine, or if it's variants burning through unvaccinated populations.

The picture isn't nearly as rosey as OP makes it sound, unfortunately: sort of a gish gallop about knowing the rate at which we can expect buffer overflows in C, but in reply to a comment noting we don't have an ETA for ending security vulnerabilities, so they unintentionally implied the picture is good.

Astra-Zeneca is not an mRNA vaccine.

The mRNA vaccines (Pfizer and Moderna) have both shown excellent (though in some cases reduced) efficacy against all strains [1], and the protein regions they target have not shown substantial alteration against antibody binding affinity (fun fact: this is "easy" to test - you give people the vaccine, isolate the antibodies in their blood, then test them against COVID in a petridish - so monitoring and change discovery is quick because you're not dependent on finding infected patients). The linked paper is notable, because while efficacy is effected, in tests the antibodies still ultimately inactivate the entire viral sample: no survivors means no natural selection criteria.

Given the widespread of COVID-19, the emergence of strains which do not successfully alter targetable regions for vaccines means that those regions are not easy (read: lead to reduction in fitness) to change. Similarly, the spike protein displays suboptimal binding to ACE receptors (tested by producing modified ACE receptors which have better binding) but there's been little field emergence towards a superior binding motif in any of the faster spreading strains [2], which implies again an evolutionary blockade or dead-end to improvements in that dimension.

Finally, antibody studies have found that in people who have had COVID, long term antibody production sets up in the bone marrow - this was recent research, but thats important because the basis of long term immunity is the development of immunological complexes which keep some antibody circulating. This is a hallmark of most long-lived immunities, and is for example one of the desired (though difficult to induce) end states of a potential mRNA HIV vaccine. [1]

Taken together (and again: you can also just go read what epidemiologists and microbiologists are publishing, plenty is out there) - the consensus opinion at the moment is that COVID-19 immunity is long-lived based on established science from decades of previous experience with vaccinating against viral diseases.

[1] https://asm.org/Articles/2021/February/SARS-CoV-2-Variants-v...

[2] https://www.nature.com/articles/s41598-020-70812-6

[3] https://www.nih.gov/news-events/nih-research-matters/lasting...

You gish gallop'd me through the floor by answering, yet again, as if I claimed something I hadn't. You should be ashamed for yourself, the very intent you had to better inform people was defeated by you.

You know most of the world isn't getting mRNA vaccines, and even if we quibble over 'most', it has nothing to do with your original nonsense, that somehow we know exactly when boosters for mRNA will be needed.

I meant some actual links to some actual data...

I don't know why I have to also do the Googling of the terms you mentioned as well as telling you what the results say.

You don't have to do anything at all that might assist anyone else. But without support it's just like "your opinion, man." ;-) And difficult for anyone to act on in any meaningful way.