@Icelandair_HR I think the airline should consider finding a more family friendly, and by extension more delay-forgiving slot, if this is as common an occurrence as I and several (in fact every) other passengers I talked to seem to have experienced.
@Icelandair_HR Thanks. What are the delay statistics on these flights? Word among the passengers is that this flight is often delayed. The non-delayed landing time is bad enough (23:30). Now 10 minutes out from “departure “ and this is where we are:
Hey @Icelandair mikið óskaplega er Billund flugið lélegt hjá ykkur, lentum í ca 2ja tíma seinkun á leiðinni út, og núna aftur á heimleið. Þetta eru til að byrja með langsíðustu flugin hérna á Billund svo ekki er á það bætandi. Ótal bugaðar barnafjölskyldur á ferð. Fussumsvei!
I am enjoying this discussion with Claude. Meeting the demands of users, is the goal with most software. We've all heard the "build me a better horse" analogy, which is good, but listening to users is still important :-)
Having thoughtful conversations with AI... I'm not sure where this sits conceptually. It's like an exercise (although the overwhelming obsequiousness makes it a fairly easy one) in conversing. Like talking to myself, but because it's in text form, it's more structured.
..and I actually 100% understand why companies rarely want product feedback/ideas from users. It's so rarely actionable, and often suffers from skewed context. But when it is actionable, and the small stuff often is, I think it adds up.
Filtering out the non-actionable though..
HN thread about failed moon landing. 296 comments, about 30 of those focusing on typos and a whopping 68 on how the adjective "private" is used in "robotic private spacecraft".
I lost what little interest I had in software security probably 25 years ago. Building things is a lot more fun than trying to break things, IMHO.
But sometimes, when looking through my logs to debug things, I see stuff like this. Does shit like this ever work?
A few weeks ago I was listening to the @BG2Pod where @bgurley really nailed what I consider the major challenge for enterprise-wide LLM adoption: numbers.
We at GRID wholeheartedly agree, and we’re trying to do something about it. https://t.co/5n1WrhH0O5
I think the "ProductChecker" and crippling of the client libraries was the final straw for me however. I don't think developers like being coerced into technical decisions, so hopefully this is something we'll see less of, not more, in the future.
About a month ago we switched from ElasticSearch hosted by https://t.co/cOryax1Vnl to AWS OpenSearch, largely for the reasons detailed in the thread below. It took us close to a year because we had just prepaid a year's worth of usage when the events below unfolded.
Do I know anyone who switched from ElasticSearch to OpenSearch after the SSPL switch?
A few days ago we suffered an outage due to our clusters on https://t.co/cOryax1Vnl returning 5xx without any updates to the status page. 1/2
... at this point in time I was clearly aligned with Elastic against AWS, thinking that this must be a nightmare for any F/OSS company, and I expected them to change their license to prevent this from happening without them being compensated or having more control.