Dan Bricklin @danb@social.cleverlibre.org

@timbray @philnickinson That may be true now, but the corporation is signaling how they think about permanence. They should have erred on the side of retention. It is the opposite of S3 in the same company.

@ampressman @timbray @philnickinson They posted "The site will be available in read-only mode for a limited period afterwards." They posted you may be able to download your own content (forums) for limited time.

@timbray @philnickinson Deleting the site, not even in read-only mode, is a slap in the face to viewing web as a trusted repository. It signals that Amazon views all web content as ephemeral and not trustworthy for important works. Should we view S3 as something that could disappear in a few months?

@danhon Does it have to do with needing to know who the user is before you know who (which system, directory, service, Google, etc.) will validate them?

@donmelton Thanks for this! It's wonderful. I only have a few memories of interacting with Steve Jobs myself since I only saw him a few times other than with crowds at conferences. I did, though, have dinner alone with him once in his hotel room when the other attendee didn't show. I last spoke with him at his last WWDC closing a loop that started with VisiCalc (http://danbricklin.com/log/2011_03_09.htm#jobs2011). I got to tell him I was an iPad developer. He always treated me well, even when I said things I'm sure he didn't like. May his memory be a blessing.

@mimsical I also find that one-on-one with screen sharing for debugging, joint writing, etc., works better than sitting next to each other trying to view the same monitor or looking up at a distant screen.

39 years ago (in Sports time) we were 6 hours from the Apple Macintosh #1984 commercial. It signaled that the world was about to change.

(I already boosted the toot from @danhon which has a very interesting thread.)

Since he's an avid and successful developer for Apple Watch app, I wonder what @_Davidsmith thinks about @danhon's request for Activity Tracking acknowledging chronic or temporary needs to take it slow and not over-exert.

@Pat I took your advice and just pasted the whole thing in for my most recent toot. I hope that works. Maybe I should have started with something like a once sentence summary and then the CW.

@linkingdot Thanks for your comments. I appreciate it.

In addition to posting my Hallelujah essay on Mastodon yesterday, I also sent it to a mailing list of an organization I'm a member of. Almost immediately I heard from another member (a professional book author, no less) of a typo: "If it's not too late to make this change on Mastodon, the phrase you have in mind is a secret chord [not cord]." Oops! I thanked him, then added a short addendum to what I posted. Later, re-reading what I wrote, I found that I also wrote "songs to be song" instead of "sung". Oh, no! Two obvious mistakes.

When I used to write essays for my website (www.bricklin.com) a couple of decades ago, I often had a co-worker or friend proofread and give me edit suggestions. I miss that.

Microsoft Word had ensured that I caught my misspellings, but it did not catch obvious errors in the meaning.

Then I thought: Maybe with the new AI systems, we'll get that extra. I should try ChatGPT, which I had not been trying yet. Now it was time.

I played around a bit with "Find typos in this sentence: The quick brwn fix jumped over the lazy dogggies back." It found the misspellings but not the inappropriate "fix". "The rain in spain falls mainly in the plane." - "No typos found."

Maybe I should try a better prompt: "What's the next word after: The rain in spain falls mainly in the" - "plain". Good!

"Find the error: The Beatles wrote lots of songs. They wanted to hold my arm." It responded: "Hold my arm should be hold my hand." Wow! Maybe it could help me.

I then tried "Find the errors:" on one line, and then gave the context of the entire CORRECTED content of my Hallelujah piece. The response:

The reaction seems to be a thoughtful and well-articulated one, and I couldn't find any errors in it. The engineer provides their perspective on the first verse of the song "Hallelujah" by Leonard Cohen, drawing on their experience as a product creator and their love for music. The engineer seems to have a deep understanding of the song and its meaning, and the language used is clear and concise. Overall, the reaction seems to be a strong piece of writing.

Hmm. That was nice. I'll take the compliments.

Next, I tried with the original two mistakes in a new chat. Correctly spelled words, but obviously incorrect meaning even if you did not have access to the lyrics online (they are from before 2021 like the Beatles so they are). The response:

There are no errors in the text provided.

Missed the errors. Not as friendly. I tried "Are there any incorrect words? " - "No, there are no incorrect words in the text."

Oh, well. Not yet. (Of course, I assume what I write here is now a bug report and it will do better at some point in the near future. We'll see.)