Jay Levitt's dashboard

Jay Levitt shared an idea in reQall on October 18, 2009 15:52:

Privacy keyword

I want to use reQall to organize my task list, but I also want to use it to record personal thoughts.

Instead of an all-or-nothing flag on my account, why not let me opt out of human transcription for a specific message? That way, "buy new jeans" gets the full treatment, and won't accidentally become "buy new genes". But "private: my evil plans are about to come to fruition! muah-ha!" gets transcribed by computer, and if it spells muah-ha wrong, that's the price I pay.

Jay Levitt started following the idea "Hyperspaces should let you move the current application between spaces" in The CocoaBots.

Jay Levitt replied on July 19, 2009 11:15 to the problem "Webkit/Safari 4.0.2 still crashing?" in Dockland Software:

Also: could you give me the original radar # that Apple assigned? I want to report this, but I want to make sure they know this isn't the bug they already fixed.

A comment on the problem "Webkit/Safari 4.0.2 still crashing?" in Dockland Software:

Any chance of a way to ignore Webkit/Fluid/etc. the way you ignore Safari? I tried the top "filter" list, but it didn't seem to prevent the bug. – Jay Levitt, on July 18, 2009 19:46

Jay Levitt reported a problem in Dockland Software on July 18, 2009 18:24:

Webkit/Safari 4.0.2 still crashing?

According to the article linked from this topic, Webkit nightly builds and Safari 4.0.2 fix the accessibility crash. But I'm seeing both of them crash when Fantasktik is enabled. Worse, although Fantasktik will exclude Safari via the preference, it won't exclude Webkit, or other Webkit-based apps like Fluid.

Do you have a radar # for Apple? Are you able to reproduce this? I can get it to happen whenever Fantasktik's running by:

1. Go to Facebook
2. Find one of the "write a comment" textareas in your feed (e.g. someone's wall post, photo, etc.)
3. Type "a" <cr> <cr> "b"

Before you can type the "b", Safari crashes.

</cr></cr>

Jay Levitt shared an idea in Dockland Software on July 03, 2009 12:49:

Finer-grained control over which apps show up

There are too many display metaphors for background applications on the Mac, so even if you implemented a simple "show only apps with open windows" checkbox, the taskbar would still be full of apps I don't care about.

Right now, I hide the apps that I usually don't care about, but then I can't find their windows through the task bar when I do care! I also saw you mention that the applications pane is getting too crowded, and I agree.

How about this:

Instead of multiple list boxes, have a single, two-column list box, the way the Spaces preference pane does. For each application we add to that list box, let us select what Fantasktik should do about it:

* Display normally (default)
* Hide app
* Hide taskbar
* Enable taskbar auto-hide (what's the difference, anyway?)
* Hide if no windows are open
* Hide if only window titled "foo" is open (e.g., for Adium, which always has the Contact List available)

This makes it easy to change how an app is displayed, without having to delete it from one list and add it to another.

Then, I'd add one more global checkbox: Smart hiding. When smart hiding is turned on, any application that's been opened or in front during the last n minutes (say, 5 minutes) will appear in the taskbar, whether or not the rules above say to hide it!

This solves the problem of "I rarely want to see Skype in the launchbar, but if I was fiddling with its preferences, I do!". And I think the UI would be straightforward.

Jay Levitt replied on July 03, 2009 12:38 to the question "Auto-only-mostly-hide?" in Dockland Software:

I was thinking more about this.. I don't know anything about CoreAnimation, but the best way to do it is probably this:

1. Slide Fantasktik down the way you currently do.
2. Dissolve the remaining pixels into a plain gray bar, using a fairly short duration - kinda the way the iPhone 3.0 does when you go from the Home screen to the Search screen.

A comment on the question "Auto-only-mostly-hide?" in Dockland Software:

Ah, that makes sense. What if you optionally added a few "extra pixels" to the top - either all the time, or only as it hides? I tried changing the bar height under Appearance, but that changes the true size of the bar, not the padding. – Jay Levitt, on June 22, 2009 12:13

Jay Levitt asked a question in Dockland Software on June 22, 2009 01:23:

Auto-only-mostly-hide?

Is the taskbar supposed to still sorta-show-up when it's vertically hidden? It seems distracting; I preferred the way Windows does auto-hide, with a thin colored bar.

Jay Levitt shared an idea in Dockland Software on June 22, 2009 01:17:

Show only the current space

Fantasktik could be a terrific solution to the one big problem with Spaces: There's no way to keep your Spaces separate. How about an option to "Show only windows in the current space?" For apps that have no windows open, you could keep track of the last space they appeared in.

I know that Spaces APIs are limited, but I think apps like HyperSpaces and Warp have found ways around that...

Jay Levitt marked one of Lawrence Sinclair's replies in Pivotal Labs as useful. Lawrence Sinclair replied to the idea "Calculate velocity based on delivery, not acceptance".

Jay Levitt marked one of Marcus Ramberg's replies in Nordaaker Ltd. as useful. Marcus Ramberg replied to the question "Reporting version updates?".

Jay Levitt replied on February 28, 2009 21:40 to the question "Reporting version updates?" in Nordaaker Ltd.:

D'oh! I completely missed that link. I swear I looked for it. Clearly a Firefox bug of some sort. Yes, clearly.

Jay Levitt asked a question in Nordaaker Ltd. on February 25, 2009 21:16:

Reporting version updates?

Some apps will probably never be claimed by their "authors" - e.g. Adobe products, and so their version info is often out-of-date. Is there a way that we can easily report updated versions to iusethis.com?

Jay Levitt marked one of Lawrence Sinclair's replies in Pivotal Labs as useful. Lawrence Sinclair replied to the idea "Calculate velocity based on delivery, not acceptance". Jay Levitt and 4 other people think it's one of the best replies.

Jay Levitt started following the idea "Calculate velocity based on delivery, not acceptance" in Pivotal Labs.

Jay Levitt shared an idea in ExpanDrive on May 29, 2008 22:18:

Reconnect at Login: Not now!

I understand why most of the "edit" options aren't available while the drive is connected. However, the "Reconnect at Login" checkbox seems like a safe one to let us touch.

Otherwise, you have to remember that you WANT to connect to the drive at a time when you're not connected!