I think the final step of moving my critical outlines to Drummer is complete with moving the
2022 books that I have read outline and updating
my master books outline index, which is an outline that includes other outlines, each one a list of the books that I've read during that year.
#Ok,
my now page is moved over to Drummer and the alias URL, frankm.info, forwards to the new public location. It does take a few minutes for the DNS change to occur to change the forwarding location. The cool thing is that because I am using an alias, all I need to do is update the DNS forwarder and everyone will be able to get to the new location.
#Fire drill time....
Dave says Twitter has provided notice that Twitter is cutting off access to their API that his apps use for authentication. Dave is not planning to make any changes to
Little Outliner, which I have been using for some time for things like
my now page. I've known that I should move these key outline files to Drummer, but haven't done so, but now I must. So, I've downloaded my outline files and I am going to import the key ones in to Drummer. This will be a good demonstration of how I been using DNS to redirect to the public version of these pages.
#Note to self: All of your public outlines that you have been editing using
Little Outliner are now moved to Drummer and being served via drummer.this.how. Should you need to, you can provision your own server running
pagePark to server these outlines via your own domain.
#
Good to be back here, there seemed to be some type of connectivity issue to Drummer when I tried yesterday and I never circled back to retry.
#
Over the weekend Dave posted a roadmap for FeedLand. As I understand the gist of the roadmap, there are two significant changes coming with what I will call
FeedLand Version 2. The identity model is moving from Twitter Oauth to Email, which I think will work the same as micro.blog. The other major change is that the backend for Version 2 will be released as open source so that users can deploy the back end to their own servers.
#I think Dave is doing the same with
FeedLand that he did with
Little Outliner and I think Drummer but I am more familiar with
Little Outliner.
Little Outliner can integrate with a Nodejs app called
nodeStorage.js. One can deploy nodeStorage to their own server and thus control where their data is stored, but the frontend remains proprietary and hosted by Dave. If Dave uses this same model, one will continue accessing
FeedLand via feedland.org, but they will be able to change what backend is used by executing a command via the browser console. I regularly switch
Little Outliner between the default backend that Dave hosts and my own that I host in Google Cloud.
#As an aside, nodeStorage also supports
1999.io, which is one of Dave's blogging platforms that preceded Drummer. Basically, nodeStorage handles the storage of files, in the case for
Little Outliner outline files in OPML and in the case of
1999.io blog posts in a JSON file, on the server it is hosted on or in AWS. For
1999.io I believe nodeStorage also handles the generation of each blog post into a separate HTML file and all of the navigation that is needed to produce a blog of static HTML pages. I think nodeStorage can also serve the blog from the same server the app is hosted on, or, as I do, it copies the generated HTML to an AWS S3 bucket, and if that bucket is configured be a public Internet site then it hosts the blog.
The blog that I wrote using 1999 is hosted in AWS and can exist independent of the server hosting nodeStorage but there are components (urlTwitterServer and urlChatLogSocket) that call back to originating server that causes a delay in the site loading if the originating server does not respond. (View source of
my 1999.io blog to see what I mean.) Because I've stopped using
1999.io I have considered decommissioning the server I have hosting nodeStorage but have not so do this impact on the blog site loading.
#I think that Dave is hopeful that other developers will write and support their own frontend applications that will use
FeedLand for feed management. It will be interesting to see whether any developers actually take the opportunity to produce their own RSS applications.
#
I am currently using
FeedLand to collect and review my RSS subscriptions and I am using Readwise's Reader app for reading items in my feed later. What I wish I had was a way to know from which sites do I read the most articles. With such data I could do a better job of pruning my RSS feeds. Over time I add sites to
FeedLand that I think might be interesting, but it's really only what I read, over time, that proves the actual value. I think the most likely source for this data would be Readwise. Even a narrower set of articles I read do I actually highlight, which are the real high value items.
#Note to self: Find out whether Readwise has an API for pulling the type of data I describe above.
#
Dear Google, why is it that the At A Glance widget on my Pixel Phone still is not capable of detecting the wallpaper color and adjust its font color accordingly? You provide a way to automatically change the wallpaper on the phone but when the wallpaper changes to a light color, like the sky in a photo, the white font used by At A Glance is unreadable. I expect that this is something Material You to fix.
#I hate how Google Chrome constantly changes the configuration of my Macbook by adding itself to the Open At Login setting. Somehow there has to be a way for me to prevent this from happening.
#
Put in the category of nostalgia computing:
DOSBox-X is a MS-DOS emulator for running DOS in modern operating systems.
#
Yesterday I successfully executed my plan to track my walks on the
Pixel Watch. I started walking before the tracking started and in each case the right distances were tracked.
#
- One of my projects beginning this week is leading a discussion of Jesus' Alternate Plan by Richard Rohr. I am going to write notes here to organize my thoughts in preparation for the discussion. HTML version of outline accessible here.#
How does one view HTML page source in Safari? I see a strange behavior in which the number list below that I entering as markdown in Drummer does not render as a numbered list in Safari. It appears as expected in Firefox and Chrome but when I view this page in Safari I just see the indented sentences without the proceeding numbers. Strangely, the other markdown for bold face and the bulleted list do get rendered.
#
- I track my walks in the Fitbit app on the Pixel Watch, and I've begun to experience what I think may be a bug that causes the distance, both miles and steps, from a prior walk be the starting amount for a new walk. The problem then is that if I don't see this happen the distance for the new walk is not properly recorded. Here are the steps to reproduce:#
Open the Fitibit Workouts tile and tap the walk icon to start a walk
The walk workout tracker starts, it shows miles as zero and it initiates a "count down" to start the walk. (You see the standard "play" icon with a circle being drawn around it for about ten seconds) During this time the GPS starts a connection and it needs several minutes to complete. Wait until the tracking starts and then move your arm to your side as you would when walking but do not start walking. Note that the key here is not actually be moving forward!
Lift you arm back up to look at the workout status. The expected result is that Miles would be zero, but instead what I see is the ending miles from the prior walk.
I end the walk, then start a new walk workout, but this time I don't wait until the "count down" completes, I just start walking and at this point, perhaps before the prior distance can be pulled, the new distance is added and thus is correct.
- I previously speculated that the covering of the watch face with my coat sleeve was causing the problem, but during more careful observation I could consistently cause the problem by simply moving my arm down to my side. The behavior suggests there is a problem in how the Fitbit app on the Pixel Watch is using the accelerometer sensor, if I am moving forward the distance tracking works as expected, but if it thinks the walk started and I move the arm so movement is detected BUT I am not moving forward, like on a walk, it then seems to pull prior miles and steps from the previous walk. #
- In the error the starting, "default," miles and steps has been consistently been the distance from the prior walk, it's easy to confirm this within the Fitbit app on my phone. Only the miles and steps is carried over, not the pace or avg heart beats per minute. The behavior suggests that these values are stored in a memory variable on the watch that is not being zeroed out after a a walk is ended, and that in some conditions that value is set for the new walk, as if the prior walk was paused, rather than being ended. I've been paying extra attention to be sure I manually end each walk.#
- I have verified that the GPS connection status is not the cause because it happens both when the GPS is connected or not connected. The problem appears to be triggered by how I start walk workout tracking, if I am already moving then the accelerometer sensor is used to calculate a distance. In fact, in this forum message, an explanation is provided for how distance is calculated if GPS is not connected. #
- Previous it seemed this behavior only for the first walk of the next day. However, in subsequent testing I was able to consistently reproduce the problem. #
- The consequence of this behavior is that I must immediately start walking once I start the walk tracking. It appears that if I am moving forward when the tracking starts the app properly reads the current values from the accelerometer rather than pull from the prior entry. I do not think there is a physical problem with the watch as the proper tracking data is valid in comparison to other watches and is consistent for each of my walks. I am convinced there is a bug related to the distance values, miles and steps, that expects a current non-zero forward movement value from the accelerometer at the time the walk starts, and it seems if it is zero for some reason it is using the prior walk miles and steps value. This behavior that should not be happening and thus is clearly to me a bug. #
- In a search for information about bugs with the Fitbit app on the Pixel Watch, I have found:#
I don't know if it is something I am doing or a real bug, but yesterday and today when I started the walk workout on my
Pixel Watch they workout seemed to start from the prior day's last walk. Going to restart my phone and watch.
#By the look of my feeds, I get the feeling CES 2023 is ramping up. A consequence of COVID is that there does not seem to be as much interest in CES as there used to be.
#
I took out the
Google Pixelbook to write that post below and that reminded me of a few things. One is just how nice the keyboard is on the Pixelbook, seriously, probably the best notebook keyboard that I have ever used. The other is just how much Google has improved the ChromeOS UI, it really looks nice and modern. A reminder to myself to use the Pixelbook more.
#Vintage computing is a thing. This morning I fell down the rabbit hole of watching a video demonstration of
C64OS, which is a new operating system for the Commodore 64. The Commodore 64 was first introduced in January, 1982.
#
The wording of the Twitter settings in Drummer imply either it posts to Twitter or it posts to Mastodon, and I don't know why that is the case. I would expect it be both/and. My guess is that the setting is really associated to the behavior of the Twitter icon in Drummer and what happens when one clicks it.
#I see that Dave is working to enable one to send Toots, which is a post to Mastodon, from Drummer. It is a work in progress. In the Twitter tab of Drummer settings there is a checkbox to post to Mastodon with a link to authorize that isn't working for me at the moment. I am curious to see whether Drummer requires authentication to a specific instance and I am not sure how that works when one has their account on another instance. I haven't posted to Twitter from Drummer often, but it is nice to have.
#
Surprised to find that I had not upgraded my
iPad Air to iPadOS 16, so starting to do that now having done so to my
iPad Mini 6.
#On April 13, 2001 gas cost $1.69 per gallon near me and I wrote of speculation of it reaching $2 or $3 dollars a gallon. Little did I know then.
#
Yesterday I upgraded the RAM my
home lab Proxmox host to 16 GB of RAM from 8 GB of RAM, which will enable me to run more virtual machines without a lot of memory swapping and thus gain better performance. The performance is still hindered by the slow disk I/O that causes long I/O waits.
#
Not too much of a surprise, if I create a post with a title then it renders nicely within Day One. It looks like IFTTT is the one stripping HTML and thus hyperlinks. If the link were created using markdown it way work as desired, but it would mean everything I want to send to Day One needs to be in markdown.#
- Day One is a journaling app and while it doesn't directly support RSS feeds, it does work with IFTTT and so one can create an applet that will send new RSS items to Day One. Unfortunately while the applet works, it's not perfect. The post below is the first one I sent and a couple of "bad" things happened. First, the hyperlink associated to Day One was removed. Next, the URL for the link to the item was populated in the title. #
I am considering resurrecting my use of
Day One for archiving some of my writing via IFTTT. I wonder how well titleless blog posts will look in the app.
#In my opinion, there is a fairly simply solution to
the "social network" problem, and that is to stop treating it as a place for publishing "content." Sure, post statuses, may be even post pictures, but these long threads and other articles should not be on Twitter or Facebook or anything other than the web.
#I've stopped cross posting from
my blog to Twitter as I am no longer interested in contributing content to that service.
#There is no such thing as the first amendment protecting "censorship" by anything other than the government. Those claiming that Twitter or Facebook or Apple are doing censorship need to understand that they are private corporations who make editorial decisions much like any other news organization does every day. The first amendment protects citizens from being censored by the government. IF the government forced Twitter to not allow something to be published, with a threat of penalty, then there would be a first amendment case, against the government, not Twitter!
#For some reason the Readwise Reader app is refusing to load for me in Fiirefox, so switched to Chrome for this morning.
#
Soundiiz syncs/transfers playlists between streaming music services.
#
- I am starting to check out passkey, which as I understand it is an alternative to passwords that uses platform biometrics and keystores. One of the sites that supports it is BestBuy.com, for which I have an account. I have a Macbook and I have iPads that support passkey, so I went ahead and set it up my account to use it. #
- The first problem is that I can only use Safari to set up and use passkey because it uses iCloud Keychain to sync across Apple devices. So, this appears to me that I either can only use my passkey configuration with the Apple platform or I have to set up separate passkeys for every platform that I use. Right off the bat I can see this is way too much friction. #
- Right now I use a password manager that works for every platform that I use, so I use the password manager to create one password and then I can use that password whether I am using a Mac, Android phone, iPad, or Windows. Passkey is dead on arrival if it is not as simple to use as passwords. I must be missing something because I cannot understand how this can be seriously promoted. Passkeys appears to be developed for people who ONLY use one platform vendor and I am not one of those people.#
- Another test I tried is creating a passkey for Passkeys.io using ChromeOS and my Pixel 4a. First I tried creating a passkey from the Chromebook and while it prompted for my phone (a Pixel 4a) it didn't store the passkey on the phone, so in following attempts it would not work. It wasn't until I first went to the site using my Pixel 4a, created the passkey on the phone AND then tried logging in from my Chromebook that it worked. Again, way too many hoops and way too complicated. #
- Finally, I went to a Windows computer running Chrome opened passkeys.io and clicked Sign in with a passkey. I select the Pixel 4a and then got prompted to provide my finger scan after which it logged in. #
- Thought of one more test, which is signing in to BestBuy.com from ChromeOS. First, I signed in to the site using Chrome on the Pixel 4a and set up a passkey. Next I went to the site in Chrome OS but it doesn't seem to know about using the Pixel 4a for the biometric authentication and since the Pixelbook does not have it, it says flat out I can't use Passkey. #
- In another test I opened BestBuy.com in Chrome on a Windows PC that has its own biometric capability and as such it also refused to use the Pixel 4a. Apparently, I would have to create yet another passkey to log in with one on the Windows PC. Mind you, to do any of this set up you must initially create an account with an ID and password, so tell me again how any of this is better than a password manager and passwords? #
- So, from my experimentation it appears that only Apple has figured out a way to create for users to create a passkey from one device, like the Macbook, and then be able to use that same passkey across multiple devices like the iPad. For Google it appears that passkeys are specific to the device so it has to be created on the "authenticating device," in my case the Pixel 4a, first, and then can be used across platforms on which Chrome is running so long as the computer I am using does not have its own biometrics. What a mess!#
I decided to revert the Readwise Reader app to pulling new items from Pocket rather than from my linkblog feed because it doesn't appear to poll my RSS feed frequently enough. I also notice some irregularities with it's RSS parsing that make each article appear as though it was published by me. So far I really like the Reader web app much better than Pocket, so it will be a worthwile move away from Pocket when it is finished.
#
I received an invite to the beta of the
Readwise Reader app, which appears to be intended to be an all-in-one read it later app. It combines the function of Pocket and a RSS reader like FeedLand. I triage my RSS feeds and send articles that I want to read later to Radio3 where they are added to my linkblog feed, which you see in the Links tab above. I have a IFTTT applet that monitors the linkblog RSS feed and adds new items to Pocket. I am attempting to do the same with Readwise Reader by taking IFTTT out of the loop and subscribing Reader to my linkblog RSS feed. So far it doesn't appear as though Reader polls from RSS as fast as I would like.
#Radio3 can now be used to
share links to Mastodon, but it requires an account on Dave's Mastodon instance, which feels problematic.
#
I've decided to try a Mastodon app, Tusky, on my
Microsoft Surface Duo, which is neither my primary tablet or phone. I now have two "social network" apps on the Duo, micro.blog (official) and Mastodon (Tusky).
#Pocket has fixed the keyboard short cuts (again) in their web app. I find it funny how this seems to happen frequently, obviously their programmers seem to forget to include that code in their builds.
#
Slowly working my way through building a new Active Directory domain controller. Done. I've created some users and groups, need to find information about organizational units.
#I am using
Proxmox for my home lab, but the computer that is hosting it, which is a HP Probook, only has 8 GB of RAM. I am finding that 8 GB is enough when running containers, it is not enough when running VMs as they don't appear to share memory. Not sure whether there is a ballooning issue, I have to research more, but I am thinking I really want a host with at least 16 GB of RAM.
#I think that the only
apps that I use that rely on Twitter as an identity provider are Dave's: Drummer,
Little Outliner, Radio3, and FeedLand. It's also my experience that once I am logged in to these apps from a browser I stayed logged in unless or until I wipe the browser cache. It looks like Dave is considering replacing Twitter's Oauth for Mastodon should he need to, but only will make the switch if he needs to. I am wondering whether a federated system like Mastodon's is a good solution for identity. My guess is that a single, reliable, provider is a better option. If it were me right now, I think I would be looking at Github as the replacement. Fortunately, I switched my 2FA for Twitter to Authy some time ago.
#
Once again the keyboard shortcuts in the Pocket web app has stopped functioning. It seems like every six months or so the keyboard shortcuts break.
#