Latest Android Update Breaks ‘kdeconnect’ on Debian Stretch (Already Resolved).

One of the apps which I have installed on my Android phone, is called ‘kdeconnect’, and I’ve blogged about it before. This is an app that allows a compatible Linux widget to sync certain data with the smart-phone.

Screenshot_20190216_130006

(Screen-Shot from some earlier version of this app, which did not constrain the available directories.)

The version which I have installed on the Debian / Stretch computer I name ‘Phosphene’, is 1.0.3~bpo9+0 . I actually needed to patch this package, so that for the following few months, it was able to browse the file-system of my phone, specifically, directories which I authorized on the phone app, from my Linux computer.

Well the Android companion to this app has just received an update through Google Play. This update broke the ability of my Linux computer to mount the remote file system – i.e., to browse any directories on the phone.

(Update at 18h25 : )

But what seems to have happened is that two updates were pushed to my phone in rapid succession, the second of which put the Android app version to 1.12.9 . The reason for which I’m inferring this, is the fact that this remote mounting of the phone’s chosen directories works now, with no actual intervention from me:

Screenshot_20190615_181900

The detail of this experience which puzzles me, is the thought that I had in fact been testing v1.12.9, when I first reported the app as broken… :-?

However, this ‘broken’ result can also occur, just because of faulty communication between the two devices.

(Update 7/6/2019, 21h25 : )

Continue reading Latest Android Update Breaks ‘kdeconnect’ on Debian Stretch (Already Resolved).

Issues in getting ‘kdeconnect’ to browse the Android phone’s file system.

One Plasma 5 application which I’m a fan of, is the ‘kdeconnect’ application, which, if The corresponding Android app is installed on a smart-phone or tablet, allows synchronizing data and use between the phone, and our desktop environment.

Here is what the widget looks like on the Linux computer’s desktop, when it’s showing no notifications from the phone:

Screenshot_20190216_125813

(It has the capacity to display notifications which appear on the phone, but on the Linux desktop, as well as numerous other capabilities.)

Under Plasma 5.8, most of what it has to offer works out-of-the-box. But there has been one issue, when trying to browse the phone or tablet’s file system, from the desktop, and which is a known issue. This does not work out-of-the-box. I know why it does not.

What ‘kdeconnect’ will do from the client, which is the desktop, is use an ssh-mount to mount the file system of the phone or tablet, virtually, on the Linux computer, after which that virtual file-system can be read from and written to, on the Linux computer. In order for this to work, the app on the phone needs to act as a file-server, and as far as I can tell, the Android app does so without flaw.

(Updated 3/27/2019, 7h40 … )

Continue reading Issues in getting ‘kdeconnect’ to browse the Android phone’s file system.

I’ve just tested the Android app(s) ‘LectureNotes’…

One of the subjects which has fascinated me for the past few days, is note-taking apps for the Android operating system.

As long as the tablet does not possess active stylus technology – which would need to pair a specific stylus-type with a specific tablet model, so that the hardware will be compatible – software that aims to allow input via a passive stylus works against a handicap. That handicap exists in the fact, that people will want to rest the side of their hand on the tablet’s screen, while also using the stylus to draw. In turn, the most basic touch-sensors, which are based on capacitance, cannot inherently distinguish between these two, simultaneous forms of contact.

In an earlier posting, and due to failure years ago to find suitable solutions, I had declared this aspiration to be a lost cause. But since that recent posting, I’ve found some software solutions that actually seem to work well, and that have hit the Google Play Store more recently. One such app is called “LectureNotes“, and is being published by the company named “Acadoid”.

This app requires configuration before use, which must at least state what level of hardware support the device provides. Thus, its higher modes will recognize a ‘Samsung S Pen’ – an active stylus, which will never create any ambiguity with the resting palm, and its lowest level of compatibility seems to use an API, to declare a rectangle on the screen to be a ‘Safe Zone’ , as the app calls it, or a ‘Palm Guard’ , as it might generically be called. This lowest compatibility level is likely to be supported by all the tablets, which have Android 4.1 or later running.

(Edit 1/03/2019, 16h10 : )

Actually, this Palm Guard feature does not use an API after all. Instead, it simply accepts points of contact from the O/S and ignores ones, which are either inside this safe zone, or outside an active input zone, if that has been configured in the app.

The reason I claim to know this is the observation that if the app is installed on a Samsung Tab S, First Generation, which can only input two points of contact, and if the palm already causes two points of contact to appear in the safe zone, then a third point of contact on the document, from the stylus, will simply fail to register…

(As of 12/28/2018 : )

I, personally, have just given this app a spin, on the Android 8 -based ‘Pixel C’ , using the lowest, passive stylus, compatibility mode, and discovered that the Palm Guard / Palm Rejection works well.

screenshot_20181227-191008

screenshot_20181227-192222

One detail about this app which I particularly like, is the fact that it offers many features in one place, that all have in common, suitability either for taking notes while a lecture is being given, or for giving lectures. And when one is using the app in sketching-mode, the rectangular Palm Guard region can either be resized and repositioned using two fingers, or have its position arrested, so that it doesn’t move around contrarily to the user’s intentions.

Additionally, as long as we may assume that the tablet has a working camera app, and a working file-manager app, photos can also be imported into the notebook. This means that a photo could be taken of a classroom whiteboard, or that an image from a file could be inserted, either of which may be resized and positioned, before the addition to the notebook is finalized…

At the same time, its tool-bar, which shows “action buttons”, defaults to displaying only a few of these buttons, the first few of which have fly-outs. But the tool-bar can be reconfigured in the settings, to display a much larger set of action buttons / icons.

Continue reading I’ve just tested the Android app(s) ‘LectureNotes’…

I have just tested the Android apps, ‘Mazec’ and ‘MetaMoJi Note’.

One of the tasks which some people might assign their Android tablets, is to use for note-taking. And as I wrote before, trying to do this becomes entangled in some hardware-related issues.

I’ve just downloaded two apps, which work best together, but which don’t absolutely need to be used together:

  1. MetaMoJi Note,
  2. Mazec.

To my great surprise I found that they work very well on my Google Pixel C tablet. One reason they do work, is because during normal use, only a narrow rectangle near the bottom of the display needs to be touched by the passive stylus, which means that the user’s palm is resting below the actual tablet.

When MetaMoJi Note is being used in line-drawing mode, in addition to allowing the user to jot his glyphs directly at their destination within the document, a very small icon in the bottom-right-hand corner expands a lower rectangle for drawing input, which remains associated with a movable, smaller rectangle within the document. The position of the destination-rectangle can be changed.

At the same time, when the doodling app is in text-mode, and if Mazec has been selected as the device’s input method – aka on-screen keyboard – again, the destination rectangle can be repositioned within the document, and writing proceeds in the text-entry rectangle at the bottom.

screenshot_20181226-004934

screenshot_20181226-005115

 

I do see three basic cons to this combination of software:

  1. It’s not obvious to me, that the maximum speed at which most users will be able to enter text drawn in this way, can be made to catch up with the speed at which, for example, University lectures will typically be given. I found that I needed to increase the Word Spacing from 0.4 to 0.7 because the initial need to write letters very close together, was slowing me down…
  2. These two apps together, amount to a price-tag, that may be higher than what most users are used to spending,
  3. Finally, Mazec is one of those input methods, which will not work to unlock the tablet, after a reboot. Depending on how the device’s version of Android is configured, directly after a reboot, there may be no icon displayed to change the input method. If that’s the case, then the additional risk exists, that the device might just crash while this keyboard is selected. And if that happens, to unlock the tablet may become difficult. Therefore, the total amount of time this IM is selected should be kept to a minimum, just to reduce this risk.

Dirk