1. 22
  1.  

  2. 8

    The HPDA (called hipster more because it fits in one’s hip pocket, rather than because it’s beloved of node.js developers in Shoreditch).

    If you believe this is true, I have some artisinal mustache wax to sell you.

    1. 7

      Thoughts - perhaps there’s value in attempting something like the focus the original Palm and Newton devices had, despite their differences, and perhaps modern accoutrements like sync. They were both focused on trimming the cruft from pocket devices that accumulated, and focused on a few thing. Palm focused on the “core four” of contacts, calendar, tasks, and notes, and did them simply and did them well, while making a few additions. Newton focused on being able to write anywhere and make it useful anywhere - the closest to that today, I’d say, is OneNote.

      1. 3

        I used to be a Palm PDA owner for years and my sister used hers until few years back when empty batteries wiped her data.

        I am not sure I would describe Palm’s work as focusing on trimming the cruft. I think they correctly identified those 4 apps as the one that are most important and need to be polished, but otherwise was mainly a story of limited resources. Certainly platform itself let people add apps and there was a healthy market of different ones.

        1. 3

          I never made the connection before, but Plan 9 and acme come to mind as a similar model, in that everything is text and almost free-form and you can assign meaning to text while still being able to “just click and type anywhere”

          1. 7

            Acme ripped much off much of its metaphors from Oberon, and a little bit the Raskin UIs of SwyftCard, Canon Cat, and Archy. (MPW later was inspired by Acme and Oberon. Another little-known thing is that MPW and Plan 9 itself were also influenced by Apollo Domain.) Newtons focused on pen input for notetaking, which is why I suggested OneNote.

            1. 1

              Is there any way to run Aegis or Domain/OS under emulation, rather than trying to find and run aging Apollo DN* hardware, commercially or otherwise? Does anyone know the licensing situation with, I assume, HPe?

          2. 3

            I found value in them for privacy purposes. Simple, non-wireless devices are easier to secure. One designed to not trust anything outside the SOC boundary far as info storage and processing would be a step up over current smartphones or desktops. It could also run a lot of software (eg even Linux) based on prior work in CompSci. It could also be cheap as prior work used things like MIPS or SPARC processors with Leon3 being open-source. Now we have Rocket with its chip generator.

            There’s also potential to use them for improved speed, UI, or battery life while doing basic apps. No cruft = less stuff going on. Maybe program them to use modern components from smartphones like the battery or even touchscreen. Just more focused UX on top of that hardware.

            1. 3

              Even if they still had wireless functionality (required for any modern market), you could still do it 10x better than extant smartphones. Give much more fine-grained permissioning of peripheral access and stop the whole “wireless chips get DMA” thing. If you put all your wireless peripherals over untrusted ports (and, ideally, tried to make the peripherals themselves more trustworthy) that would be substantially better than the current situation. I know there have been several efforts to move the baseband in cell phones to USB, but as I understand these efforts have died due primarily to regulatory concerns.

              1. 4

                wireless chips get DMA

                This is a good thing, otherwise performance would be trash - what’s not a good thing is unrestricted DMA. This is why IOMMUs exist.

                1. 1

                  My original concept was a knockoff of the Nokia Communicator given the low-cost, secure components take up more space due to extra isolation. The original even had an Intel processor with segments that could’ve run GEMSOS security kernel assuming memory requirements didn’t prevent it. Today, we got lightweight kernels like Muen or OKL4 with embedded chips that have MMU’s and such for as low as $4 per 100 units. New version might be even smaller. :)

                  https://en.wikipedia.org/wiki/Nokia_Communicator

            2. 6

              I personally find another advantage of the pocket notebook–much faster data entry. Speed of data entry for me is [typing on a keyboard] —> [hand written notes] ———–> [typing on my phone]. So while my phone is the most conveniently at hand, it is the worst for input.

              I’ve had some success at using bullet journals for pen and paper note-taking. I think giving yourself a month using their process, then throwing away what you don’t use works well. For example, I find appointments and meetings best handled in Outlook, so I don’t bother recording them. But development goals are absolutely worth tracking and keeping at hand.

              1. 3

                I bought a small notepad that fits into my chest pocket of my collared shirt. With that, a pen, and a pencil, I’m always able to just jot down something quick and easy. It doesn’t even have to be reminders, it can be anything. There’s no limitations, and it feels very liberating.

                1. 3

                  The author calls out an interesting phenomenon w.r.t. to backpressure– I’ve found that at my new employer, my task queue tends to grow without bounds when using digital tools due to the lack of finite space available. At the same time, it’s extremely easy to pick up enough responsibility to overflow the backup and editing capabilities of setting tasks to the pen. Not to mention the potential to discard a list that’s still active and lurking beneath a pile of other documents!

                  Personally, I’ve transitioned to editing a list in a single plain-text ‘tasks’ file on my desktop (file length seems to provide a soft-cap to my queue size), but what systems do other crustaceans use to avoid taking on so many tasks that it’s impossible to complete them before the heat death of the universe?

                  1. 2

                    I’ve been a big fan of Trello, but it’s not immune to overflowing. I recently split my main todo list into “todo list” and “someday hell” to split all the wacky ideas and fear-based someday tasks , and other nerd add stuff somewhere else. Always be refactoring your todo list!

                    The other thing I do for indie software development in trello is move a list of finished tasks for a given month to a ‘myproject done board’ to remind me of times when I’ve been really productive.

                    When Joel / fog creek first started trello they mentioned wanting a “five things” board. Something where the design made sure having more than five things wouldn’t appear as easily.

                    One of the other things I’ve experimented with is keeping a “pervasive today” list on my pebble watch via the “agenda watchface”.

                    1. 6

                      I actually use a card system myself, but not a strict GTD system - my cards are notes and ideas and lists and brainstorming put on 2 ¼” x 4” blank Rolodex cards rather than index cards.

                      I have a three 66704s on my desk - one is for for “overflow”, one for “completed”, and one for “archive”. I generally don’t have a hard limit but I carry whatever is manageable and appropriate and when I get overwhelmed I move cards to the overflow file. I have this setup at home and work.

                      Each file holds 500 cards and when it’s full I look back on the oldest idea and decide if it’s a lost treasure worth revisiting - if it is, it goes right back to the active deck I carry and if not it goes into the archive file.

                      The “completed” file works the same - once it’s full with 500 cards they move from complete to archive.

                      Once “archive” is full the removed cards get put in a drawer and every few months I rubber band the cards in the drawer up, label with the date, and toss them in a box.

                      That box ends up like my journal or diary of sorts, with all my ideas and musings, good and bad, completed and never started projects, etc.

                      I end up looking at the cards in the “completed” Rolodex more than I’d expect, and the “overflow” less than I would like. On a day where I get am either bored - or maybe particularly energized, as the case may be - I will look through the “overflow” in search of a great but almost forgotten idea, and add it back to my active “carry” deck.

                      FYI - my coworkers all laugh at me and make punch card jokes and such, and ask how I access my cards in the cloud, or if I worry about the Russians hacking my pens. :(

                      However, my work-life balance has improved by having physical Rolodex’s separated between my house and my office. It enforces a rough “locality hygiene” in that I keep mostly focused on work at work and my personal grandiose projects stay mostly at home, besides what mixes in the active deck.

                  2. 1

                    I tried several time to implement GTD, but always failed… It seems too much to implement at once. Any advice/tips, improvement you made to stick to it?

                    1. 2

                      I mastered the “Capture” phase, but not the others. I haven’t found a proper implementation, and I’ve been looking for 10+ years.

                      I’m writing a blog post explaining how the future of UI will be exactly like GTD. I also intend to build an implementation.

                      1. 2

                        Same, but I’ve found the somewhat more flexible approach of a bullet journal works well for me.

                        1. 1

                          I’ve tried and failed at org-mode multiple times and keep meaning to give TaskWarrior a solid chance but I feel lost without my physical cards and their “immediacy”.